progress and paradigms in the search for the neural correlates of consciousness progress and paradigms in the search for the neural correlates of consciousness editorial introduction sascha benjamin fink (philosophy departement, university of magdeburg; center for behavioral brain sciences (cbbs), university of magdeburg) ying-tung lin (institute of philosophy of mind and cognition, national yang ming chiao tung university) how to cite: fink, s. b. , & lin, y.-t. (2021). progress and paradigms in the search for the neural correlates of consciousness: editorial introduction. philosophy and the mind sciences, 2, 3. https://doi.org/10.33735/phimisci.2021.103 © the author(s). https://philosophymindscience.org   (issn: 2699-0369) twenty years ago, thomas metzinger attested that the empirical science of consciousness is in a pre-paradigmatic phase. some still see the field in this stage. while there are numerous theories of consciousness, none of them can claim to be the paradigm of the field (even though some – like the global workspace theory, higher order thought theory, predictive processing – are more prominently name checked than others). pre-paradigmatic stages of science are rare. according to thomas s. kuhn (1962), sciences usually progress by going from paradigm to paradigm via scientific revolutions. before the first paradigm, fields of research are marked by disagreeing writings that try to “build the field anew from its foundations,” because scientists are not able to take any “common body of belief for granted” (kuhn, 1962, p. 13). how do sciences arrive at a paradigm? one possibility is that they inherit a paradigm from the science they develop out of: cognitive science established its paradigms after its split from a psychology dominated by behaviourism. but this was not a proper, all encompassing revolution, in two ways. first, some paradigms in cognitive science can still be seen as left-overs from behaviouristic psychology. second, behaviouristic psychology was not overcome: some psychologists still remain true to behaviourism. the danger of such incomplete revolutions lies in their possible instability: as long as the not-fully-overcome paradigm has its prominent adherents, it may simply take over again. this may not always be a bad thing. but in order to stabilise an incomplete revolution, secession is needed: cognitive scientists needed to establish a field of their own and defended it against behaviouristic psychology by searching for allies in different fields (linguistics, robotics, cybernetics, computer sciences, neuroscience) and by doing “boundary works” (gieryn, 1983). some see the contemporary science of consciousness as a development out of cognitive science and thereby simply transpose paradigms of cognitive science onto this new field of research. others resist this conservative effort and propose alternative paradigms. in such a climate of balkanisation, researchers need to find common ground among the disparate fractions in order to lay the foundations for a unified field of research. one way to find common ground is in the abstract: meta-theoretic reflection and reasoning independently of a paradigm (also known as philosophy), may lead to shared starting points. for example, most researchers accept that consciousness exists and needs to be accounted for. another way to find common ground is in the concrete: the data all theorists need to account for. data can provide common ground if it can be framed in a way acceptable to several fractions. in the science of consciousness, one kind of datum to account for might be how we describe our subjective experience of ourselves and the world, i.e., our phenomenology. but how do we arrive at a phenomenology? due to the subjectivity and privacy of our experiences, the call for first-person methods can be heard in large parts of the field. however, there is reasonable scepticism vis-à-vis first-person methods, stemming in part from the idea that science must uphold objectivity and allow for intersubjective checks. first-person methods might be hardly compatible with that. but at least something is commonly accepted: that neural processes correlate with our experiences. the search for neural correlates of consciousness (nccs) therefore plays a fundamental role for a pre-paradigmatic science like the empirical science of consciousness: it is one of the few sets of largely uncontested data that would-be theoryand paradigm-builders have to take into account. why does this neural activation correlate with that experience? the second fundamental reason why neural correlates have proven to be so central to this field is their metaphysical promiscuity: they are not committed to one specific theory on how consciousness and the brain actually relate. still, whichever metaphysics philosophers may converge on – identity theory? epiphenomenalism? supervenience? grounding? neutral monism? pre-established harmony? interactionist dualism? –, as long as consciousness and the brain exist, we can expect some statistical coupling between neural and phenomenal variables, i.e., a correlation. james ward (1911, p. 600), who already used the term “neural correlates of consciousness” in the entry on psychology in the 11th edition of the encyclopedia britannica, called this methodological dualism, where we accept at the beginning that we grasp the two “correlated” entities with different methods; but he already speculated that this may lead to a metaphysical monism in the long run. this commitment to a purely statistical and largely measurable relation between the neural and the phenomenal – correlation – was crick and koch’s (1990) elegant way of bracketing ongoing metaphysical turf wars thirty years ago: let them fight, we’ll start our science. the discussion around the ncc therefore combines the abstract and the concrete to foster common ground: the meta-theoretical reflection on a metaphysically neutral but empirically accessible relation between neural and phenomenal variables gave rise to a suitable way to capture data and bracket discussion that could have divided the field prematurely. the search for nccs therefore was central at the dawn of the science of consciousness. the question is whether it still is. maybe the science of consciousness has progressed far enough to overcome nccs. maybe the problems associated with nccs are an indicator of a fundamental crisis this field will face if it binds itself to this concept. then, we should embrace alternatives, like the neural difference makers of consciousness which klein, hohwy, and bayne (2020) introduced in the first part of this special issue. but maybe we should shy away from calls to leave the search for nccs behind us. if we abandon the search for nccs in favour of more elaborate or theory-bound approaches, we may risk fragmenting the young field yet again – simply because it may lead into the usual cycle of paradigms, revolutions, and secessions. but maybe, a convergence onto one and only one paradigm is not even needed in the sciences. an alternative view, one we find quite plausible, is that the science of consciousness does not mirror the natural sciences in its use of paradigms; rather, it is more like fields such as economics. in the natural sciences, one paradigm governs all – and is largely replaced after a revolution. the usual business of scientists is nearly always done completely under one theoretical roof. in fields like economics, however, paradigms like the chicago school are not used to establish a normal scientific practice of solving problems arising from that paradigm. instead, they are reference points used to orient scientists in the field. as paul hoyningen-huene and nicholas wüthrich (hoyningen-huene, 2020; hoyningen-huene & wüthrich, 2017) point out, knowing orientational paradigms allows communicating how one’s approach differs from them. one needs to know them to work in this field, but one needs not accept them. in the science of consciousness, theories might fulfil the role of orientational paradigms. however, no theory can yet claim to be such an orientational paradigm, even though some may have the potential. but when there were no empirically adequate theories during the first years of the neuroscience of consciousness, the concept of an ncc fulfilled the role of an orientational paradigm, one that continues to overlap and build common ground for a broad range of theories in the field. the ncc thereby enabled the progress we have seen in this science since the 1990s. here, we continue to celebrate the progress made in the science of consciousness, specifically on the ncc, and for a good reason: it has been 30 years since francis crick and christof koch put the search for nccs in their towards a neurobiological theory of consciousness (1990) back on the table and 20 years since the conceptual contribution by david chalmers on the question what is a neural correlate of consciousness?, which was published in thomas metzinger’s seminal anthology on the ncc (2000; see also the introduction by fink, 2020, to the first instalment of this issue). a lot has happened since these works shaped the field we work in today, reflected in the papers collected here. most of the authors in this issue specifically address the tension mentioned above, namely between metaphysics, specific theories of consciousness, and the goal of facilitating scientific progress. alex lepauvre and lucia melloni (2021) address the methodological challenge ncc research is facing and present three strategies that will accelerate progress in the field. the currently prevalent contrastive method, in which neural activities correlating with conscious states are contrasted with those correlating with non-conscious states, have encountered difficulties in dissociating the proper ncc from neural precursors or successors. lepauvre and melloni suggest moving beyond the contrastive method. first, they call for adopting methodologies that capture the phenomenological dimension by incorporating a systematic description of phenomenal consciousness. second, large-scale coordinations and aggregation of data across paradigms, stimuli, and research groups in common databases must be facilitated. third, the field needs to adopt pre-registered adversarial collaborations, which directly test contradicting predictions. all of these, they argue, will move the field forward. morten overgaard and asger kirkeby-hinrup (2021) caution against overblowing results from ncc research: finding the ncc will not answer all of our questions because it is unclear how our theories of consciousness map onto data gathered by ncc research. we are missing an isomorphism between our conceptual frameworks for theories of consciousness and the brain. so even if we knew the ncc, it would be unclear what this says about our theories of consciousness. this situation is worsened by the fact that one of the correlates, phenomenal consciousness, is not directly accessible. the development of theories should take the missing isomorphism between conceptual frameworks as well as theories about consciousness on the one hand and neuroscientific data about nccs on the other hand into account. they thereby emphasise the distance between ncc research and metaphysical theorising that already motivated crick and koch 30 years ago, but which numerous researchers have since attempted to bridge. chen song (2021) underlines the importance of neural structures for the science of consciousness. her presentation of the structures of the central nervous system illustrates the various ways in which neural structures shape the way in which we experience. while differentiation and integration are considered to be essential to conscious experience, the brain must be able to facilitate these properties. song argues that the key features of the brain are the structural diversity between neurons and the modular topology of neuronal connections which allow the brain to generate differentiated yet integrated activity patterns. this view is empirically supported by studies that look into interand intra-individual changes: structural diversity is positively correlated with the diversity in conscious experience. sascha benjamin fink, lukas kob, and holger lyre (2021) also argue for the view that structures are important in the search for neural correlates of consciousness – so much so that they constrain what can count as a proper neural correlate of an experience (in contrast to a mere statistical correlate). they advocate a neurophenomenal structuralism and argue that (i) what type-individuates any experience in a domain (i.e., what makes it specifically of that type) depends on its relative similarity and difference to all other experiences in that domain and (ii) that these individuating structures in the phenomenal domain must be mirrored in their neural substrate in a surjective homomorphism. this structural similarity constraint allows progress in so far as it allows for distinguishing merely statistical correlates from those neural events in which consciousness has its physical foothold. tobias schlicht and krysztof dolega (2021) doubt that predictive processing (pp), a paradigm currently in high regard, has the power to become a theory of consciousness in its own right or can work as an overarching framework in the neuroscience of consciousness, as suggested by jakob hohwy and anil seth (2020) in the first part of this special issue. hohwy and seth argued that pp might be a unifying framework for the science of consciousness, out of which it may develop into a theory of consciousness in its own right. schlicht and dolega strongly disagree: not only is the evidence in support of pp still equivocal with regard to empirically measurable neural activations, mechanisms, and architectures, but even if this were not the case, the framework's conceptual attachment to consciousness is also too loose. their article cautions against adopting pp and its allies (e.g., friston’s free energy account, as presented in the contribution of wiese & friston in this issue) as overarching models, frameworks, or theories for the neuroscience of consciousness. wanja wiese and karl j. friston (2021) argue for an expansion of the search for correlates: not only neural correlates should be on the agenda, but also computational correlates of consciousness (ccc). finding such cccs would allow for a much easier transfer of our empirical findings established in human experimentation to more exotic instances, from non-human animals to artificial systems. wiese and friston locate this endeavour within the larger framework of the free energy principle and leverage its mathematics to address not only whether there can be “islands of awareness” (and if so, whether they are marked by complexity reduction or not), but also how to distinguish a system that truly is conscious from one that only simulates consciousness. progress in the field can then also come from accepting a minimal model of consciousness or from transgressing from a neural to a computational perspective. michael pauen (2021) advocates for a bolder approach: the search for nccs is not as metaphysically neutral as commonly expected. he considers what it means for the search of nccs to be “theoretically neutral.” drawing from the history and philosophy of science, in particular the development of the identity between water and h2o, he addresses the importance of the dependency between theoretical assumptions and empirical facts. the author argues that theory dependence is crucial for ncc research and is already present in pain studies. editorial note: due to covid-19, numerous steps in the publication of both parts of this special issue were delayed. this foreword has therefore been updated over the course of 2021 to reflect the new additions to this special issue. for their support, we are grateful to all contributors, to all submitters, and to all reviewers. we also want to thank the editorial board, and specifically wanja wiese, as well as fabian fuchs and anouk kinzel, for their support during the typesetting process. references chalmers, d. j. (2000). what is a neural correlate of consciousness? in t. metzinger (ed.), neural correlates of consciousness: empirical and conceptual questions (pp. 17–39). mit press. crick, f., & koch, c. (1990). towards a neurobiological theory of consciousness. seminars in the neurosciences, 2, 263–275. fink, s. b. (2020). a double anniversary for the neural correlates of consciousness: editorial introduction. philosophy and the mind sciences, 1(ii). https://doi.org/10.33735/phimisci.2020.ii.85 fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness. philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 gieryn, t. f. (1983). boundary-work and the demarcation of science from non-science: strains and interests in professional ideologies of scientists. american sociological review, 48(6), 781–795. http://www.jstor.org/stable/2095325 hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 hoyningen-huene, p. (2020). orientational paradigms. unpublished manuscript. hoyningen-huene, p., & wüthrich, n. (2017). why is general philosophy of science as yet irrelevant for economics and what can be done against this? in eipe 20th anniversary conference book of abstract. https://www.eur.nl/sites/corporate/files/book_of_abstracts_.pdf klein, c., hohwy, j., & bayne, t. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii). https://doi.org/10.33735/phimisci.2020.ii.60 kuhn, t. s. (1962). the structure of scientific revolutions. university of chicago press. lepauvre, a., & melloni, l. (2021). the search for the neural correlate of consciousness: progress and challenges. philosophy and the mind sciences, 2, 4. https://doi.org/10.33735/phimisci.2021.87 overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 schlicht, t., & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 song, c. (2021). brain structural complexity and consciousness. philosophy and the mind sciences, 2, 6. https://doi.org/10.33735/phimisci.2021.9185 ward, j. (1911). psychology. in encyclopedia britannica: vol. xxii (pp. 547–604). cambridge university press. wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 open access this article is distributed under the terms of the creative commons attribution 4.0 international license, which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. models and mechanisms in philosophy of psychiatry. models and mechanisms in philosophy of psychiatry editorial introduction lena kästnera (lena.kaestner@uni-bayreuth.de) henrik walterb(henrik.walter@charite.de) philosophy of psychiatry is becoming increasingly popular. one of the reasons is its scope and interdisciplinarity. it not only relates to philosophy of mind, philosophy of medicine, philosophy of science, and phenomenology, but also to ethics and social and political philosophy. a central question in the field is how mental disorders can be best conceptualized, modelled, and explained. within cognitive neuroscience, philosophy of science, and philosophy of mind, a range of new approaches has recently been proposed to serve this end. these approaches include, among others, the new mechanist philosophy, artificial (deep) neural networks, symptom network theory, predictive processing, computational psychiatry, connectomics, and 4e-cognition. although these accounts exhibit promising features for a modern scientific approach to psychiatry, they leave unanswered important questions regarding the theoretical and conceptual foundations of psychopathology as well as the practical limitations arising for clinical practice. this special issue sets out to address some of these questions. the background for this special issue is the multidisciplinary workshop “minds, models and mechanisms: current trends in philosophy of psychiatry” which was held at saarland university in april 2021. though we had to switch to an online format due to the pandemic, the discussions at the event have been extremely inspiring. it brought together experts from diverse disciplines, like clinical psychiatry, neuroscience, computational modelling, philosophy of mind and cognition, auniversity of bayreuth. bcharité-universitätsmedizin berlin, corporate member of freie universität berlin and humboldtuniversität zu berlin, department of psychiatry and psychotherapy. kästner, l., & walter, h. (2023). models and mechanisms in philosophy of psychiatry: editorial introduction. philosophy and the mind sciences, 4, 1. https://doi.org/10.33735/phimisci.2023.10398 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://doi.org/10.33735/phimisci.2023.10398 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org lena kästner & henrik walter 2 and philosophy of science. joining forces, researchers from these disciplines presented their ideas on the subject. in the aftermath of the workshop, the organizers issued an open call for contributions. after a rigorous review process five papers made it into the final issue, each making an original contribution to the field. in her paper “mental disorder: an ability-based view” sanja dembic (2023) puts forward an ambitious, general theory of what mental disorders are. the central elements of her theory are (in)abilities, reason, and harm. while dembic builds on well-known approaches of harmful dysfunctions, she aims to avoid some of the problems that these approaches have by relying on abilities rather than on functions. abilities as well as reasons are personal level features, so dembic’s account is intelligible not just to scientists and philosophers, but also to patients and lay people as well. in their paper “the skill of mental health” garson leder and tadeusz zawidzki (2023) put forward a related, but still different approach which also operates on the personal level. according to them, mental health is nothing over and above the ability for skilled action, viz. a metacognitive skill. importantly, this includes the ability to self-regulate one’s cognitive abilities in order to adapt to a changing environment. accordingly, leder and zawidzki conceptualize mental disorders as a breakdown of this metacognitive skill. the next two papers focus on a recently very influential approach in scientific psychiatry, namely the use of computational approaches based on machine learning and modern developments in the field of artificial intelligence. this approach comes with the aspiration or at least hope that gathering big amounts of data and analyzing them in automated ways will help psychiatrists find new regularities in patients with mental disorders. thus, proponents suggest, computational approaches will eventually contribute to a better understanding, classification, and treatment of mental illnesses. in their paper “machine learning and its impact on psychiatric nosology” georg starke, bernice simone elger, and eva de clercq (2023) present data from a qualitative study using semi-structured interviews with 15 german and swiss experts in computational psychiatry. the interviews took place between april 2020 and june 2021. in the results presented here, the authors focused on the questions (i) whether computational approaches will be able to contribute to an improved nosology and (ii) whether this would be desirable. the results were mixed, showing there were optimists and sceptics with respect to both questions. the authors discuss the various reasons for this optimism and skepticism, respectively, and highlight the relevance of these for philosophical theories of mental disorders. in his paper “understanding as a bottleneck for the data-driven approach to psychiatric science” barnaby crook (2023) gives an excellent review on the role of computational approaches, machine learning, and ai within scientific psychiatry. the main point of his paper is to show that despite its powers in finding new patterns in data, machine learning neglects an important aspect of psychiatry that is indispensable if psychiatry wants to improve patients’ treatment: patient kästner, l., & walter, h. (2023). models and mechanisms in philosophy of psychiatry: editorial introduction. philosophy and the mind sciences, 4, 1. https://doi.org/10.33735/phimisci.2023.10398 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10398 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org models and mechanisms in philosophy of psychiatry 3 understanding. crook argues that patient understanding, i.e. the understanding of mental disorders by those who are afflicted by them, has necessarily to be included into such approaches if they want to achieve substantial progress in the treatment of mental illnesses. the final paper in this issue is “how does the psychiatrist know?” by adrian kind (2023). in this paper, kind takes a closer look at how the diagnostic process in psychiatry can and should best be understood. in doing so, he confronts phenomenological approaches with standard clinical psychiatric examination and argues that diagnostics procedures depend on modeling mental disorders within the diagnostic procedures. he explains his model-based account in detail and argues why it is superior to the phenomenological approach. a common theme emerging from these papers is that personal level accounts do or at least should play an indispensable role in modern psychiatry. this is true for the concept of what a mental disorder is, how mental disorders are diagnosed, how they should be investigated, and how, or to what extent, computational approaches might be utilized to improve treatments of mental illnesses. with all of this, we hope that the current collection of articles will contribute to advance the philosophy of psychiatry. acknowledgments we would like to take this opportunity to thank the editors-in-chief, especially sascha fink, for their support and assistance throughout the editorial process. we further thank all anonymous reviewers for their work, which helped to greatly improve the published manuscripts. finally, we would also like to thank lea nickel and the team behind philosophy and the mind sciences for their help with preparing the printer-ready version of this issue. work on this edition has been supported by the deutsche forschungsgemeinschaft (dfg, german research foundation) grant 446794119. in addition, hw has been supported by dfg grant 337619223 / rtg2386 and lk has been supported by volkswagen foundation grant 9b830 (formerly 98 509). references crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 dembic, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735 /phimisci.2023.9630 kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci. 2023.9391 leder, g., & zawidzki, t. (2023). the skill of mental health: towards a new theory of mental health and disorder. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology: findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/p himisci.2023.9435 kästner, l., & walter, h. (2023). models and mechanisms in philosophy of psychiatry: editorial introduction. philosophy and the mind sciences, 4, 1. https://doi.org/10.33735/phimisci.2023.10398 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://doi.org/10.33735/phimisci.2023.9630 https://doi.org/10.33735/phimisci.2023.9630 https://doi.org/10.33735/phimisci.2023.9391 https://doi.org/10.33735/phimisci.2023.9391 https://doi.org/10.33735/phimisci.2023.9684 https://doi.org/10.33735/phimisci.2023.9435 https://doi.org/10.33735/phimisci.2023.9435 https://doi.org/10.33735/phimisci.2023.10398 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org lena kästner & henrik walter 4 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. kästner, l., & walter, h. (2023). models and mechanisms in philosophy of psychiatry: editorial introduction. philosophy and the mind sciences, 4, 1. https://doi.org/10.33735/phimisci.2023.10398 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10398 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org truth, success, and epistemology: a response to stokes's thinking and perceiving. truth, success, and epistemology a response to stokes’s thinking and perceiving zoe draysona(zdrayson@ucdavis.edu) abstract in thinking and perceiving, stokes challenges “the pernicious cognitive effects assumption”: the assumption that it would be epistemically problematic if our thoughts were to directly influence our perceptual experience. in doing so, stokes takes himself to be supplementing the epistemological claims of philosophers like siegel and lyons with descriptive claims about human psychology. i argue that his conclusions are more radical than they first appear, to the extent that stokes’s project is at odds with the standard epistemological discussions of cognitive penetration. keywords cognitive penetration ∙ epistemology ∙ perception this article is part of a symposium on dustin stokes’s book “thinking and perceiving” (routledge 2021), edited by regina fabry and sascha benjamin fink. if our thoughts directly influence our perceptual experience, epistemological worries seem to arise. philosophers have tended to assume that such a scenario, in which what we perceive is causally dependent on our beliefs and expectations, would result in epistemic concerns about circularity, unreliability, or doublecounting (vance, 2015). stokes (2021) calls this “the pernicious cognitive effects assumption”, and one main aim of his book thinking and perceiving is to challenge the assumption: he argues that “there is no principled reason to assume that cognitive penetration, should it occur, is bound to produce epistemic problems” (p. 137).1 stokes’ challenge echoes recent claims by jack lyons and susanna siegel, who also argue that the cognitive penetration of perceptual experience doesn’t necessarily undermine its epistemic role. lyons (2011), for example, argues that the epistemic impact of cognitive penetration will depend on the mode of penetration: auniversity of california, davis. 1all page references are to stokes (2021) unless otherwise specified. drayson, z. (2023). truth, success, and epistemology: a response to stokes’s thinking and perceiving. philosophy and the mind sciences, 4, 12. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-1665-3319 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org zoe drayson 2 while some top-down influences could negatively affect the epistemic role of perception, others could potentially improve it. and siegel (2017) has argued that while the justificatory force of a perceptual experience can be reduced by its etiology (‘epistemic downgrade’), the converse is also possible: the etiology of an experience can increase its power to justify certain judgments (‘epistemic upgrade’). similarly to lyons and siegel, stokes proposes that some top-down influences on perception could be “epistemically boosting, rather than downgrading” (p. 137) and could thus lead perceivers to “see better in a variety of ways” (p. 223). stokes’ proposal, however, is notably bolder than those of lyons and siegel. lyons and siegel are mainly concerned with challenging the conditional claim at the heart of the pernicious cognitive effects assumption: they propose that if thinking were to affect perceiving, then we wouldn’t necessarily run into epistemic worries. in thinking and perceiving, stokes appeals to empirical evidence to argue for the truth of the conditional’s antecedent. he proposes the thesis that ‘thinking affects perceiving’ (tap), where this is a descriptive claim about cognitive architecture. he then draws on further empirical data about the performance of experts in specific perceptual domains to make the claim that ‘thinking improves perceiving’ (tip), where this is a normative claim about the epistemic consequences of tap. stokes describes his project as building on the conceptual work of philosophers like siegel to make empirical claims about human perception: “what the architectural analysis given here does is supplement siegel’s epistemological model (which is defended as a conceptual possibility, not as a descriptive claim about human psychology) with a range of empirically grounded cases.” (p. 178) i will suggest that stokes’ position in thinking and perceiving is more radical that it first appears. where previous debates about the epistemic consequences of cognitive penetration have assumed that epistemology is governed by norms of truth, stokes’ conclusions seem to require the adoption of a pragmatist epistemology which is governed by norms of success. as a result, i propose that stokes’ view of perceptual epistemology fundamentally contrasts with the views of philosophers like siegel and lyons, and thus his analysis cannot be understood merely as a supplement to existing epistemological models. traditionally, epistemological constraints have been defined in terms of truthevaluable propositions. this is what goldman calls “the familiar assumption” that epistemology centers on belief, and that belief aims at truth and error-avoidance (goldman, 1995, p. 171). these assumptions are shared by internalists and externalists alike. siegel’s focus, for example, is on rationally evaluable inferences, where these are the epistemically appraisable transitions between beliefs. she is interested in how phenomenally identical experiences can provide different levels of justification for the belief that p, depending on how the experience has been modulated by cognitive inferences (siegel, 2017). lyons replaces talk of rational inference with talk of reliably truth-conducive processes, and explicitly states that drayson, z. (2023). truth, success, and epistemology: a response to stokes’s thinking and perceiving. philosophy and the mind sciences, 4, 12. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org truth, success, and epistemology 3 our evaluations of perception as epistemically better or worse rely on “truth-linked notions” like reliability, safety, sensitivity, and power (lyons, 2011, p. 300). when stokes introduces the idea of perceptual experts as epistemically better perceivers, he seems initially to share the same traditional epistemological assumptions as siegel and lyons. like siegel, he proposes that the “experiences of perceptual experts contribute in a positive way to their rational standing as epistemic agents” (p. 178, my italics). like lyons, he acknowledges that perceiving better or worse can be a matter of “producing accurate representations of the environment with more or less frequency” (p. 189, my italics). but stokes goes on to propose that many of the improvements which make perception epistemically better “are not improvements in accuracy or veridicality in this straightforward sense” (p. 200): “improved sensitivity to behaviourally relevant patterns, organizational features, categoryand diagnostic-specific information, less distraction, speed […] are instead improvements in seeing what is relevant to a task and with increased speed and efficiency.” (p. 194) stokes is not merely making the (relatively uncontroversial) claim that our perceptual performance can improve in non-epistemic ways. we can all agree that there is some practical sense in which faster perceptual identification skills improve the success of our performance at certain tasks, for example, but traditional epistemology would not generally classify this as an epistemic improvement. this is where stokes parts company with epistemologists like lyons and siegel: he proposes that such improvements in performance amount to a perceptual epistemic good. “what the data show, if they show nothing else, is that experts perform more accurately, more rapidly, with less cognitive effort, and in ways that present advantages for working memory. expertise is therefore an epistemic good simply because the expert is moving closer to an optimal cognitive stance on the world (or a part of it), where she can better acquire behaviourally relevant category and diagnostic information.” (p. 175) for stokes, epistemic agents aim at optimality rather than truth, where optimality is understood broadly in terms of success “along a variety of measures that align with aims for action, performance, and worldly engagement” (p. 229). improvement along these sorts of measures would not, i take it, qualify as epistemic improvement for most traditional epistemologists, including siegel and lyons. how does stokes justify characterizing epistemic aims in terms of success rather than truth? he proposes that we take a virtue-based approach to epistemology. virtue epistemologists address standard epistemological questions about knowledge, justification, and scepticism by appealing to the intellectual virtues of the agent, where these virtues can be understood as cognitive faculties (by reliabilist virtue epistemologists such as sosa and greco) or as character traits drayson, z. (2023). truth, success, and epistemology: a response to stokes’s thinking and perceiving. philosophy and the mind sciences, 4, 12. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org zoe drayson 4 (by responsibilist virtue epistemologists such as zagzebski). stokes suggests that where top-down effects on perception are epistemically enhancing, it is because they have increased the intellectual virtues of the perceiver: “perceptual experts achieve skills that are best described as intellectual virtues” (pp. 149-150). stokes’ adoption of virtue epistemology, however, does not seem to be sufficient to justify his characterization of epistemic constraints in terms of success rather than truth. notice that the virtue epistemologists quoted by stokes are as truth-focused as other traditional epistemologists. for sosa, an intellectual virtue is a reliably truth-conducive faculty, “a quality bound to help maximize one’s surplus of truth over error” (sosa, 1991, p. 225); for greco, intellectual virtues are faculties which “enable a person to arrive at truth and avoid error in some relevant field” (greco, 2002, p. 287); and for zagzebski, acts of intellectual virtue have the epistemic end of reaching the truth (zagzebski, 1996, p. 271). none of these virtue epistemologists frame epistemic improvement in terms of (non-truth-governed) success, and so they cannot help stokes to make the case that improvements in the speed or practicality of perceptual performance are epistemic improvements. what stokes needs is an approach to virtue epistemology which shifts the scope of epistemology away from truth-apt belief and towards more generally successful action and performance. such a variety of virtue epistemology may exist, but stokes does not give us reason to adopt it. and if he did adopt such an approach, he would no longer be engaged with the existing debates over the epistemic consequences of cognitive penetration, which characterize epistemic constraints in terms of truth. my aim here is not to argue against stokes’ claims in thinking and perceiving. it is rather to highlight that his conclusions are more radical than they first appear, making few points of contact with the epistemological literature on cognitive penetration which stokes takes himself to be supplementing. we might, for example, understand stokes’ approach as a move from ‘mentalism’ to ‘pragmatism’ (fodor (2008)). according to the mentalist standpoint, the paradigmatic goal of a cognitive agent is to ascertain truth: behaviors are only rational insofar as they are produced by a valid (truth-preserving) reasoning process. according to the pragmatist standpoint, the paradigmatic goal of the cognitive agent is to act successfully, where the success of a behavior is measured by its consequences (e.g. its adapative benefits) rather than by the validity of the processes which produced it. some proponents of the pragmatist approach go so far as to claim that we can understand rationality in terms of non-truth-linked success. some philosophers have characterized rationality as “getting ahead, cognitively, in the world” (dennett, 1987, p. 97), and rational norms as constrained by “functionality, not veridicality” (chater et al., 2018, p. 800).2 if stokes were to take this pragmatist approach to rationality, he could retain his claim that experiences of perceptual 2the project of explaining rationality, on these approaches, becomes “the problem of explaining the production, in social, environmental, and technological context, of broadly appropriate adaptive response” (clark, 2001, p. 143). drayson, z. (2023). truth, success, and epistemology: a response to stokes’s thinking and perceiving. philosophy and the mind sciences, 4, 12. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org truth, success, and epistemology 5 experts “contribute in a positive way to their rational standing as epistemic agents” (p. 178, my italics), but he would not longer be engaging with lyon’s or siegel’s concept of theoretical rationality. if stokes is proposing to adopt a pragmatic conception of epistemology, then interesting questions arise about the relationship between stokes’ descriptive claims about cognitive architecture (tap) and his normative claims about the architecture’s epistemic consequences. much of the philosophical interest in the cognitive architecture of perception and cognition is grounded in the assumption that the epistemic status of a cognitive capacity is a matter of the sorts of processes which produced it: their rationality, or reliability, for example. if we switch a successbased version of epistemology, then claims about the architecture seem less relevant. the more we focus on success rather than truth as our epistemic goal, the less reason we have to suppose that cognitive architecture will provide the answers to our epistemological questions. references n. chater, t. felin, d. funder, g. gigerenzer, j. koenderink, j. krueger, d. noble, s. nordli, m. oaksford, b. schwartz, k. stanovich, & p. todd (2018). mind, rationality, and cognition: an interdisciplinary debate. psychonomic bulletin & review (vol. 25, pp. 793–826). https://doi.org/10.3758/s13423-017-1333-5 clark, a. (2001). reasons, robots and the extended mind. mind and language, 16(2), 121–145. https://doi.org/10.1111/14680017.00162 dennett, d. c. (1987). the intentional stance. mit press. fodor, j. a. (2008). lot 2: the language of thought revisited. oxford university press. goldman, a. i. (1995). social epistemology, interests, and truth. philosophical topics, 23(1), 171–187. https://philpapers.org /rec/golsei greco, j. (2002). virtues in epistemology. in p. k. moser (ed.), the oxford handbook of epistemology (pp. 287–312). oxford university press. https://doi.org/10.1093/0195130057.003.0010 lyons, j. (2011). circularity, reliability, and the cognitive penetrability of perception. philosophical issues, 21(1), 289–311. http://www.jstor.org/stable/41329425 siegel, s. (2017). the rationality of perception. cambridge university press. sosa, e. (1991). knowledge in perspective: selected essays in epistemology. cambridge university press. https://doi.org/10.1 017/cbo9780511625299 stokes, d. (2021). thinking and perceiving: on the malleability of the mind. routledge. vance, j. (2015). cognitive penetration and the tribunal of experience (vol. 6, pp. 641–663). https://philpapers.org/rec/mic cpa-4 zagzebski, l. t. (1996). virtues of the mind: an inquiry into the nature of virtue and the ethical foundations of knowledge. cambridge university press. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. drayson, z. (2023). truth, success, and epistemology: a response to stokes’s thinking and perceiving. philosophy and the mind sciences, 4, 12. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.3758/s13423-017-1333-5 https://doi.org/10.1111/1468-0017.00162 https://doi.org/10.1111/1468-0017.00162 https://philpapers.org/rec/golsei https://philpapers.org/rec/golsei https://doi.org/10.1093/0195130057.003.0010 http://www.jstor.org/stable/41329425 https://doi.org/10.1017/cbo9780511625299 https://doi.org/10.1017/cbo9780511625299 https://philpapers.org/rec/miccpa-4 https://philpapers.org/rec/miccpa-4 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the agency-first epistemology of psychedelics. the agency-first epistemology of psychedelics lisa bortolottia (l.bortolotti@bham.ac.uk) kathleen murphy-holliesa (klm276@student.bham.ac.uk) abstract letheby’s book is an engaging and crystal-clear exploration of the philosophical issues raised by the use of psychedelic drugs. in this paper, we focus on the epistemological issues letheby examines in chapter 8 and argue that his analysis requires an agency-first approach to epistemic evaluation. on an agency-first approach, epistemic evaluation is about identifying the skills agents needs to acquire in order to pursue and fulfil their epistemic goals. keywords epistemology ∙ agency ∙ psychedelics∙ self-know-how ∙ skills ∙ self-regulation this article is part of a symposium on chris letheby’s book “philosophy of psychedelics” (oup 2021), edited by chiara caporuscio and sascha benjamin fink. 1 introduction letheby’s book is an engaging and crystal-clear exploration of the philosophical issues raised by the use of psychedelic drugs. while reviewing the conceptual and empirical literature bearing on the topic, letheby also succeeds in shifting the way we think about fundamental aspects of our existence, from knowledge to spirituality, from the structure of the self to aim of therapy. we were genuinely intrigued and challenged by the claims letheby powerfully defends throughout the book. in the limited space we have here, we will focus on the epistemological issues letheby examines in chapter 8. just like letheby, we have also drawn attention in recent work to the positive contributions of cognitions that misrepresent reality or violate norms of rationality (bortolotti, 2020; murphy-hollies, forthcoming). letheby compels the reader aphilosophy department, university of birmingham (uk) bortolotti, l., & murphy-hollies, k. (2022). the agency-first epistemology of psychedelics. philosophy and the mind sciences, 3, 5. https://doi.org/10.33735/phimisci.2022.9283 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-0507-4650 https://orcid.org/0000-0001-5061-4674 https://doi.org/10.33735/phimisci.2022.9283 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org lisa bortolotti & kathleen murphy-hollies 2 to re-examine many common-sense assumptions about what it means to hallucinate, asking what epistemic benefits the psychedelic experience can have and whether there are any reasons to believe that its beneficial effects are long lasting. one temptation in this area is to see psychedelics as a source of hallucination and false beliefs that make us feel better in some way but compromise the accuracy of our representations of reality. letheby challenges this trade-off view, arguing that psychedelic experiences have epistemic benefits of their own. letheby’s arguments for the epistemic benefits of psychedelics won’t be persuasive unless the traditional aims and nature of epistemic evaluation are revisited, departing from the received view that epistemic evaluation is confined to the project of examining a belief for its truth and well-groundedness. this is not a criticism of letheby’s project but a wholehearted endorsement of a broader understanding of the role of epistemic evaluation. this broader understanding can be cashed out in several ways (see henderson & greco, 2015, for some examples), but here we will explore the idea that the purpose of epistemic evaluation is to identify epistemic practices that support successful agency, and that successful agents have skills that enable them to pursue and fulfil their epistemic goals. 2 agency first in recent work in epistemology, an important shift can be identified. we move from asking whether the agent’s belief is true and well-grounded to asking whether the belief enables the agent to succeed in their epistemic practices. suppose that mary, as a result of her psychedelic experience, comes to believe that she is spiritually connected with nature. a traditional epistemologist may ask whether mary has any good evidence for that belief or whether the belief accurately represents how things are. but there is another important question to ask, whether for mary adopting a belief about being spiritually connected with nature means that she can be a more successful epistemic agent overall. what is going to change in mary’s behaviour due to that belief? are any of the changes relevant to her epistemic performance? in this framework, the neat distinction between epistemic and psychological effects of the agent’s practices comes under pressure. letheby challenges the assumption that false or ill-grounded beliefs act merely as comforting illusions, enhancing the agent’s wellbeing at the expense of their grasp of reality. in some contexts, comforting illusions may enhance an agent’s wellbeing and also be the most effective means for that agent to maintain a firm grasp on reality. traditionally, epistemic evaluation concerns outputs of reasoning tasks, beliefs, theories, and it measures their accuracy, well-groundedness, coherence. the outcome of such evaluations is an important guide to decision-making. if we realise that a belief is insufficiently supported by evidence, we may choose to abandon it, or suspend judgement until more evidence is available. bortolotti, l., & murphy-hollies, k. (2022). the agency-first epistemology of psychedelics. philosophy and the mind sciences, 3, 5. https://doi.org/10.33735/phimisci.2022.9283 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9283 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the agency-first epistemology of psychedelics 3 but what can we say about the agent who engages in those epistemic practices, the agent who produces inaccurate reasoning outputs, adopts ill-grounded beliefs, endorses incoherent theories? does their local failure suggest that they are unsuccessful in the pursuit of their epistemic goals overall? when epistemic evaluation concerns not merely the single output, the belief or the theory, but the interaction of the agent with their environment, a more complex picture emerges. we realise that the agent’s psychological make-up and social context affect what the agent can achieve epistemically, not just by revealing unsurmountable limitations but also by bringing to the fore the often-neglected role of mood and motivation for the pursuit of epistemic goals (bortolotti, 2020). let’s see how this can apply to the epistemology of psychedelics. letheby identifies a number of epistemic benefits that can be attributed, either directly or indirectly, to the experiences mary has when she takes psychedelic drugs: 1. a capacity to access emotions and feelings that she may not have been fully aware of before using psychedelic drugs (e.g., mary can gain access to the irrational jealousy sometimes driving her behaviour and missing from her previous model of herself); 2. a capacity to adopt a more open attitude towards the uncertain and the unknown, becoming more curious, accepting, and open-minded (e.g., mary becomes more skilled at letting go of her thoughts prior to meditation practice); 3. an enhanced realisation of what is possible (e.g., mary now realises that her mind has a vast unrealised potential and that her sense of self is not set in stone and can be changed); 4. an appreciation of previously known facts under a new mode of presentation (e.g., mary already believed that she was interconnected with nature, but she now appreciates that connection in a different way); 5. a preservation of psychosocial functionality leading her to exchange information with others and extend her appreciation of possibilities for change (e.g., the reduction in mary’s anxiety enables her to function better socially and her experience of alternative possibilities enhances her capacity to imagine a different future for herself). none of these benefits can be captured by analysing the single perceptions or beliefs that are due to mary’s psychedelic experience. the effects as letheby describes them concern epistemic agency as a whole and the capacity to adapt to new situations and evolve; preserve or restore engagement with the surrounding environment; and gain knowledge of the self by realising that the self is malleable. the epistemology of psychedelics is an agency-first epistemology, and that is why it has something important to say about what other things agents care about beyond knowledge, such as psychological wellbeing, mental health, and the meaningfulness of their lives. bortolotti, l., & murphy-hollies, k. (2022). the agency-first epistemology of psychedelics. philosophy and the mind sciences, 3, 5. https://doi.org/10.33735/phimisci.2022.9283 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9283 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org lisa bortolotti & kathleen murphy-hollies 4 3 self-know-how letheby argues that many of the epistemic benefits of psychedelics can be applied to learning about, and making changes to, oneself. these possibilities for selfinsights and self-change are exciting. letheby invites us to imagine that mary sees herself as a generous and selfless person, and yet sometimes she harbours irrational jealousy and selfish motivations. given the moral and normative consequences of jealousy and selfishness, what tools are available to mary so that she can fix these failings? here, letheby’s views about psychedelics can provide us with new ideas, which complement strategies which we discussed elsewhere (murphy-hollies, forthcoming). when agents live up to the good traits which they like to regard themselves as having, this may be incentivised by imperfect cognitions, such as post-hoc rationalisation, confabulation, and self-deception. imperfect cognitions may explain why mary doesn’t realise that she can be irrationally jealous and selfish. yet, such imperfect cognitions are not merely inhibitory—yes, they hide the agent’s own shortcomings, but there is more to them than that. they can contribute to the process of self-change. having ‘self-know-how’ enables agents to more robustly live up to their selfconcept. self-know-how is a skilful form of self-regulation: agents can bridge the gap between how they see themselves and how they really behave. the idea comes from de bruin and strijbos (2020) who discuss self-know-how in relation to the agent preserving first-person authority over their self-ascriptions even when they make inaccurate self-ascriptions, as in instances of confabulation. in essence, although agents may self-ascribe traits which they do not have—such as mary thinking of herself as selfless while having selfish motivations—they can make the illusory self-ascriptions a reality. what does this self-know-how involve? agents need to be open-minded and flexible with regards to their interpretation of themselves and of the situation they find themselves in. they need to be receptive to how others see them, as in the process that wilson calls ‘self-revelation’ (wilson, 2002). they need to be curious about, and attentive to, their own thoughts and feelings. they need to be cautiously confident in their own appraisals without being arrogant or paralysed by self-doubt. thanks to such attitudes, agents can regulate their own behaviour, and close the gap between their self-descriptions and the reality of their behaviour. in other words, they can make changes to themselves in the direction they desire. letheby’s views on the effects of psychedelic drugs on human experience nicely complement considerations about self-know-how, suggesting that psychedelics such as psilocybin can enable people to see themselves as mutable and in control of the stories they tell about themselves. letheby suggests that psychedelics can provide ‘knowledge-how’: mary gains know-how when she starts to pay attention to her own mental states in a more open, decentred, and flexible way. bortolotti, l., & murphy-hollies, k. (2022). the agency-first epistemology of psychedelics. philosophy and the mind sciences, 3, 5. https://doi.org/10.33735/phimisci.2022.9283 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9283 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the agency-first epistemology of psychedelics 5 furthermore, the other forms of knowledge discussed by letheby can also be seen as instances of agents acquiring new insights into themselves, seeing themselves as changeable, and making changes to the self: • psilocybin can relax and inhibit self-related priors, which ordinarily work defensively to protect mary’s more egoistic and flattering conceptions of herself. hence, she can gain more accurate insights into herself and her motivations. • mary can experience herself as a mutable and contingent being, and realise that her sense of herself is a story which she has the power to rewrite. • psilocybin expands the boundaries of what mary can access phenomenally, so that she can experience differently certain beliefs which she already had about herself. • indirectly, preserving psychosocial functioning and facilitating interactions with others can empower mary to imagine other ways of being and fitting into the world, bypassing some of the cognitive biases which pathologies such as depression and anxiety can bring. on de bruin and strijbos’s account of self-know-how and letheby’s account of the role of psychedelics, agents eventually accept and embody new self-concepts. via self-know-how, agents might draw on their own reflections and the input of others to consider new self-related ideas and behaviours. via psychedelics, agents can directly experience new conceptions of themselves which they would not otherwise be open to—experiences with psychedelics might present new selfrelated ideas and behaviours directly to agents. both approaches invite us to think of self-knowledge as an activity as opposed to a special type of unaided introspection or sheer willpower. as letheby writes, coming to know ourselves as constructed and mutable enables us to change our story. references bortolotti, l. (2020). the epistemic innocence of irrational beliefs. oxford university press. de bruin, l. & strijbos, d. (2020). does confabulation pose a threat to first-person authority? mindshaping, self-regulation and the importance of self-know-how. topoi 39 (1):151-161. henderson, d. k., & greco, j. (2015). epistemic evaluation: purposeful epistemology. oxford university press. murphy-hollies, k. (forthcoming). political confabulation and self-regulation. in a. jefferson, o. palermo, & j. webber (eds.), values and virtues for a changing world. cambridge university press. wilson, t. (2002). strangers to ourselves. harvard university press. bortolotti, l., & murphy-hollies, k. (2022). the agency-first epistemology of psychedelics. philosophy and the mind sciences, 3, 5. https://doi.org/10.33735/phimisci.2022.9283 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9283 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org lisa bortolotti & kathleen murphy-hollies 6 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. bortolotti, l., & murphy-hollies, k. (2022). the agency-first epistemology of psychedelics. philosophy and the mind sciences, 3, 5. https://doi.org/10.33735/phimisci.2022.9283 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9283 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction agency first self-know-how varieties of perceptual improvement. varieties of perceptual improvement zed adamsa(zedadams@gmail.com) abstract in this short comment on dustin stokes’s thinking and perceiving, i attempt to clarify what is at stake in the debate between fodorian modularists and stokean anti-modularists. keywords modularity ∙ perceptual learning ∙ expertise this article is part of a symposium on dustin stokes’s book “thinking and perceiving” (routledge 2021), edited by regina fabry and sascha benjamin fink. distinctions are the lifeblood of philosophy. one of the many virtues of dustin stokes’s thinking and perceiving is the wide array of distinctions it introduces for understanding the nature of the relationship between thinking and perceiving. these distinctions help to clarify what, exactly, is at stake in debates between fodorian modularists (fms) and stokean anti-modularists (sams). in this short comment, my goal is to encourage stokes to refine a distinction he introduces between two types of perceptual improvement, in order to achieve greater clarity about the precise point at which fms and sams disagree. my hope is that this will better direct our attention towards the sorts of evidence that would help decide between these views. to this end, i will introduce two accounts of the debate between fms and sams—a coarse, overly simplistic account, and a more nuanced, fine-grained account—as a way of illustrating the progress stokes has made in articulating what is really at stake in this debate. i will then introduce two problem cases for stokes, cases that don’t fit neatly on either side of the distinction he draws between two types of perceptual improvement, as a way of raising the question of whether he has drawn this distinction in the right place. on a first, overly simplistic, account of the debate between fms and sams, it is a debate between modularity and malleability. fms hold that perception and cognition are distinct mental faculties and that perception, unlike cognition, is fixed and anew school for social research. adams, z. (2023). varieties of perceptual improvement. philosophy and the mind sciences, 4, 11. https://doi.org/10.33735/phimisci.2023.10264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-0820-0106 https://doi.org/10.33735/phimisci.2023.10264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org zed adams 2 immutable. sams, by contrast, hold that perception is akin to cognition in allowing for improved performance as the result of experience. on this first construal of the debate, diachronic changes to perception that allow us to better discriminate different items, group similar items, and focus our attention all support sams over fms (connolly, 2019). the problem with this first account of the debate, as stokes himself recognizes, is that fms readily acknowledge the existence of diachronic improvements to perception as the result of experience (pylyshyn, 1999). they merely hold that that these changes are wholly intra-perceptual: they are developmental changes to the perceptual module itself that are triggered by exposure to the environment. for example, in the case of color vision, there is evidence of “improvement in color discrimination from infancy up until late adolescence, with saturation thresholds approximately halving (i.e., sensitivity improving) with every doubling of age until adolescence” (maule et al., 2023, pp. 91–92). this perceptual improvement might simply be the result of the slow maturation of the neural pathways underlying color discrimination (crognale, 2002). on a second, more nuanced, account of the debate between fms and sams, one endorsed by stokes, it is a debate between acquisition and learning. it rests on a question about the factors that lead to perceptual improvement: are they exclusively the result of developmental changes (i.e., acquisition) or can they result from “agent-driven, accuracy enhancing training” (stokes, 2021, p. 182) (i.e., learning)? fms hold that the informational encapsulation of perceptual processing implies that it cannot be influenced by central cognition and, as such can only be improved by developmental changes. sams, by contrast, hold that in addition to developmental changes, perceptual improvement can also result from “cognitive or conceptual learning” (stokes, 2021, p. 151). sams, but not fms, can claim that “perceptual expertise is a genuinely cognitive phenomenon, dependent on the conceptual information [in] the domain of expertise” (stokes, 2021, p. 165). in short, for sams but not fms, “thinking improves perceiving” (stokes, 2021, p. 202). the clarity with which stokes states this conclusion is admirable. but there is an aspect of his view that i wish he had said more about: namely, what, exactly, is involved in “cognitive or conceptual learning” and why is it required for learning? some sense of the difficulty here becomes evident when one notes that stokes does not think that (what he calls) “low-level perceptual learning” suffices: repeated exposure to colour chips or flavour testers as one finds in a sommelier kit may suffice, without any category or concept learning, to enhance one’s perceptual capacities to recognize relevant differences in those stimuli tokens. (stokes, 2021, p. 109) expert training, in and outside of laboratory circumstances, requires more than mere exposure to stimulus types (in contrast to many examples of low-level perceptual learning, such as colour discrimination). (stokes, 2021, p. 165) adams, z. (2023). varieties of perceptual improvement. philosophy and the mind sciences, 4, 11. https://doi.org/10.33735/phimisci.2023.10264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org varieties of perceptual improvement 3 as these quotations make clear, stokes thinks that in order for “conceptual learning” (i.e., learning) to occur, “expert training” must be involved. this immediately raises the question: what exactly is “expert training” and why is it required for learning? as a way of pressing this question, it will help to focus on two examples that stokes himself returns to multiple times: color vision and language learning. first, lewis forder and gary lupyan have recently argued that color words have a “warping effect” on the perception of colors (2019, p. 1109). they found that after hearing a color word, people more accurately discriminated objects of that color from objects of nearby color categories. they concluded that color words lead to enhanced categorical perception, to an increased sensitivity to the differences between objects of different categories (goldstone & hendrickson, 2010). this is relevant because it seems to be an example that cuts across stokes’s characterization of the distinction between acquisition and learning; it is an example of an improved perceptual ability to categorize objects that is neither merely the result of a developmental process nor one that requires “expert training.” one response open to stokes would be revise his characterization of learning to abandon the requirement that it must involve “expert training.” a better response, i think, would be to introduce a threefold distinction between acquisition and two different forms of “conceptual learning”: the sort studied by forder and lupyan (which involves enhanced categorical perception without “expert training”) and the sort discussed by stokes (which requires “expert training”). if this is on the right track, i am curious how stokes himself would characterize this new threefold distinction. second, as many have noted, learning a new language brings with it a significant perceptual change with regard to utterances in that language: simply put, a language that you understand sounds different from a language that you don’t. one response you could have to this example of perceptual change would be a version of stokes’s own “thinking improves perceiving” view: you could hold that the perceptual change results from being able to think about the meanings of utterances in the new language (cf. brogaard, 2018). an alternative response, however, might involve distinguishing between the perceptual and cognitive components involved in learning a new language. you might think that the perceptual change is limited to learning how to parse the phonemes in the new language, and that the cognitive change involves becoming able to retrieve lexemes from long-term memory and to use them to generate the meanings of utterances (cf. gross, 2022). the problem is that it is hard to know where to locate this alternative response in stokes’s characterization of the debate between fms and sams. it doesn’t fit neatly on either side of his distinction between acquisition and learning. this problem becomes especially acute when one realizes that fodor himself seems to have held a version of this sort of alternative view. fodor held that linguistic comprehension involves the combination of two capacities: a perceptual capacity to parse an utterance into its phonemic constituents and a cognitive capacity to understand what that combination of phonemes means (fodor, 1983, pp. 88–90). this raises the question of why stokes thinks that the debate between fms and sams can be adequately characterized in terms of the adams, z. (2023). varieties of perceptual improvement. philosophy and the mind sciences, 4, 11. https://doi.org/10.33735/phimisci.2023.10264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org zed adams 4 distinction between acquisition and learning, and not in terms of a choice between some more complicated combinations of both. my hope is that if stokes can refine his account of different types of perceptual learning in a way that better accommodates these two problem cases, we will be left with a characterization of the debate between fms and sams that focuses on what exactly is at stake in this debate, as well as what sorts of evidence would decide between them. acknowledgments i am indebted to conversation about the topics of this paper with greg bartels, jake browning, adam gies, nat hansen, dan harris, joe lemelin, chauncey maher, eric mandelbaum, and leo sedlmayer. an anonymous reader also gave very helpful comments. references brogaard, b. (2018). in defense of hearing meanings. synthese, 195(7), 2967–2983. https://doi.org/10.1007/s11229-016-1178-x connolly, k. (2019). perceptual learning: the flexibility of the senses. oxford university press. https://doi.org/10.1093/oso/ 9780190662899.001.0001 crognale, m. (2002). development, maturation, and aging of chromatic visual pathways: vep results. journal of vision, 2, 438–450. https://doi.org/10.1167/2.6.2 fodor, j. a. (1983). the modularity of mind. the mit press. https://doi.org/10.7551/mitpress/4737.001.0001 forder, l., & lupyan, g. (2019). hearing words changes color perception: facilitation of color discrimination by verbal and visual cues. journal of experimental psychology: general, 148, 1105–1123. https://doi.org/10.1037/xge0000560 goldstone, r. l., & hendrickson, a. t. (2010). categorical perception. wiley interdisciplinary reviews: cognitive science, 1(1), 69–78. https://doi.org/https://doi.org/10.1002/wcs.26 gross, s. (2022). is there an empirical case for semantic perception? inquiry, 0(0), 1–26. https://doi.org/10.1080/0020174x.2 022.2074099 maule, j., skelton, a. e., & franklin, a. (2023). the development of color perception and cognition. annual review of psychology, 74(1), 87–111. https://doi.org/10.1146/annurev-psych-032720-040512 pylyshyn, z. (1999). is vision continuous with cognition?: the case for cognitive impenetrability of visual perception. behavioral and brain sciences, 22(3), 341–365. https://doi.org/10.1017/s0140525x99002022 stokes, d. (2021). thinking and perceiving: on the malleability of the mind. london: routledge. https://doi.org/https: //doi.org/10.4324/9781315189895 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. adams, z. (2023). varieties of perceptual improvement. philosophy and the mind sciences, 4, 11. https://doi.org/10.33735/phimisci.2023.10264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11229-016-1178-x https://doi.org/10.1093/oso/9780190662899.001.0001 https://doi.org/10.1093/oso/9780190662899.001.0001 https://doi.org/10.1167/2.6.2 https://doi.org/10.7551/mitpress/4737.001.0001 https://doi.org/10.1037/xge0000560 https://doi.org/10.1002/wcs.26 https://doi.org/10.1080/0020174x.2022.2074099 https://doi.org/10.1080/0020174x.2022.2074099 https://doi.org/10.1146/annurev-psych-032720-040512 https://doi.org/10.1017/s0140525x99002022 https://doi.org/10.4324/9781315189895 https://doi.org/10.4324/9781315189895 https://doi.org/10.33735/phimisci.2023.10264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org 404 not found perceptual expertise and object recognition: an explanatory task for modularists and antimodularists. perceptual expertise and object recognition an explanatory task for modularists and antimodularists aleksandra mroczko-wąsowicza(aleksandra.mroczko-wasowicz@uw.edu.pl) abstract dustin stokes’s book contributes to one of the continuing debates in empirically informed philosophy of mind and cognitive sciences which concerns the relation between thought and perception. the book sheds new light on such questions as: whether vision is modular, informationally encapsulated, and thus cognitively impenetrable or rather the opposite – whether it is malleable and sensitive to further improvements by cognitive states. stokes supports the latter by referring to empirical evidence on perceptual expertise. proponents of the modular and malleable architectures of the mind offer different explanations of the phenomena involved in perceptual expertise, viz. object identification and categorization. interestingly, both views assume some kind of automaticity of the recognitional capacities for identifying and categorizing objects. in this article, i examine the influence of perceptual expertise on object recognition and how the seeming automaticity of object recognition may be approached from the modularist and antimodularist (malleabilist) perspectives. keywords perceptual expertise ∙ object recognition ∙ automaticity ∙ modularity this article is part of a symposium on dustin stokes’s book “thinking and perceiving” (routledge 2021), edited by regina fabry and sascha benjamin fink. dustin stokes’s book sheds new light on processes that constitute the human mind, the way they interact, and enable us to make contact with the world. he approaches the relation between thought and perception by dealing with such questions as: whether vision is modular, informationally encapsulated, and thus cognitively impenetrable or rather the opposite – whether it is malleable and sensitive to further improvements by cognitive states. stokes supports the latter by appealing to empirical research on perceptual expertise (bukach et al., 2006; drew et al., 2013; kundel et al., 2007; scott, 2011) and his own investigations of this phenomenon (ivy et al., 2021, 2023; stokes, 2021a). auniversity of warsaw, faculty of philosophy. mroczko-wąsowicz, a. (2023). perceptual expertise and object recognition: an explanatory task for modularists and antimodularists. philosophy and the mind sciences, 4, 14. https://doi.org/10.33735/phimisci.2023.10247 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-3511-1355 https://doi.org/10.33735/phimisci.2023.10247 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org aleksandra mroczko-wąsowicz 2 expertise involves experience and training specific to a relevant domain such as radiology, ornithology, or fingerprint examining. proponents of the modular and malleable architectures of the mind offer different explanations of phenomena involved in perceptual expertise, viz. object categorization and identification. modularists interpret it as the capacity for visual object recognition – a postperceptual cognitive process of late vision. for stokes, perceptual expertise, including expert recognition, is a genuinely perceptual phenomenon, which can be sensitive to cognitive influences and as such exhibits both a perceptual and cognitive achievement (stokes, 2021a, 2021b). capacities for object recognition are capacities to either categorize a perceived object as belonging to a conceptual category or to identify it as being a specific individual (abid, 2021). the two processes can be seen as different stages of the recognition of an item, with categorization being temporally more efficient and occurring before identification. there is wide agreement that at least some of these processes operate automatically (dell’acqua & job, 1998; mroczko-wąsowicz, & anaya 2022; serre et al., 2007). a process is automatic, as opposed to voluntary or intentional, when it is not under the subject’s conscious control (moors & de houwer, 2006; papineau, 2013, p. 177). empirical studies have determined that recognition can occur as quickly as within 200-300 ms after stimulus detection (dicarlo et al., 2012; mohr et al., 2018). this short time frame excludes any possibility for subjects to exercise conscious control over the unfolding of the process. stokes claims that recognition by experts differs from that by novices. this is because expert perception is cognitively enhanced as a result of their domain-specific concept-rich cognitive learning and unconscious impact of perceptual learning (visual memories) on attention. as a result, they do not only know better but see better. the empirical findings stokes (2021b, p. 193) refers to suggest that in expert recognition, the memorized visual information affects saccadic patterns so that experts are less distracted by features of a stimulus that have no consequence for recognizing a specific category of objects. stokes discusses eye-tracking studies which have shown that experts such as radiologists make fewer but longer saccades and fixate less on features of a radiogram that are diagnostically irrelevant for identifying an abnormality (drew et al., 2013; kundel et al., 2007). in consequence, decreased visual distraction leads to enhanced sensitivity to category-specific information. experts develop visual object representations of the relevant kind more rapidly and see relevant objects more accurately than laypersons. this is why it is concluded that experts display differences in eye movement patterns, enjoy an advantage for a specific category of objects in visual short-term memory, and exhibit automatic successful performance (stokes, 2023). in a similar vein, stokes and colleagues argue in their own experimental work that visual expertise is more than meets the eye (ivy et al., 2023). they examined holistic visual processing (hvp) – a behavioral marker of a visual-expert search strategy, which shows that experts are able to process information from a larger mroczko-wąsowicz, a. (2023). perceptual expertise and object recognition: an explanatory task for modularists and antimodularists. philosophy and the mind sciences, 4, 14. https://doi.org/10.33735/phimisci.2023.10247 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10247 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perceptual expertise and object recognition 3 region of space with a more focused gaze pattern. hvp turns out to be transferable across domains but supporting reduced search time and greater accuracy only within an expert’s particular domain of expertise. this led the researchers to conclude that visual search success does not depend exclusively on the occurrence of hvp, but also on the explicit knowledge of an expert’s domain, including their knowledge how to search and where to search. the seeming automaticity of object recognition may be an outcome of the interplay between these factors, because recognition is a complex phenomenon consisting of perception, concepts, and associated perceptual memories of categories or particular items. commenting on behavioral and phenomenological aspects of expert recognitional capacity, stokes emphasizes its instantaneousness: the first strand of behavioural evidence concerns “automaticity”. not only do experts more rapidly perform categorizations or other forms of recognition (…), but they do so in ways that they often cannot carefully describe. (…) expert radiologists often report a sense that there is something anomalous in a medical image before they can point to the anomaly. (…) [they] report that the relevant object or feature is “highly salient” or just “pops out”. (…) these reports and the speed of performance suggest that the expert expends little or no deliberate cognitive effort and that her performance is non-inferential. (stokes, 2021b, p. 152) on the other side, theorists aligning toward the modularist view (fodor, 1983; marr, 1982; pylyshyn, 1999) may explain visual object recognition differently. this explanation also involves some aspects of recognition’s automaticity, but this is limited to the early sensory component of recognition. accordingly, visual recognition includes automatic early sensory processing delivering new visual representations of basic sensory properties for purposes of later categorization and identification. this means that the new visual representations are compared with stored visual memories. in consequence, feedback from visual memories helps to search the newly formed sensory representations for visual matches. even quick comparisons, which lead to finding recognitional matches between novel visual representations and those stored in memory, are large enough to accommodate attention allocation and to be classified by modularists as late vision or post-perceptual processes (raftopoulos, 2011). modularists acknowledge that late vision is cognitively penetrated and involves modulation of processing by cognitively driven attention (pylyshyn, 2003). by classifying visual object recognition as a late vision phenomenon, they do not mean that object recognition enhanced by perceptual expertise is an additional indicator of the malleability of the mind. for modularists, perceptual expertise does not make object recognition more sensitive to cognitive influences than it is already the case for regular (non-expert) recognition. on their view, perceptual expertise makes the low-level component of object recognition – early mroczko-wąsowicz, a. (2023). perceptual expertise and object recognition: an explanatory task for modularists and antimodularists. philosophy and the mind sciences, 4, 14. https://doi.org/10.33735/phimisci.2023.10247 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10247 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org aleksandra mroczko-wąsowicz 4 visual sensory processing – more fluent and facilitates perceptual discriminations of fine-grained subordinate categories. modularists emphasize that outputs of early vision result from unconscious, mandatory, fast, and automatic modular processes. as such, early visual representations do not alone determine late vision phenomena like object recognition. interestingly, although automaticity in the modularist view is meant in the strong sense of operating pre-attentively, it can be considered to be a graded matter, similarly to modularity itself (deroy, 2014; drayson, 2017; fodor, 1983; mroczko-wąsowicz, 2022). in this commentary, i have examined how object recognition and its automaticity may be approached from the modularist and malleabilist perspectives. although some relevant questions regarding the status of late vision remain open for further research, it seems that non-modular approaches to object recognition may enjoy more advantages. this is because such approaches are capable of accommodating twofold explanations of the rapidity of expert recognition: (1) bottom-up explanation compatible with the modularist suggestion concerning the extraordinary fluency of discriminatory processes in the early sensory component of object recognition, and (2) the malleabilist proposal concerning the top-down penetrating impact of expertise-related higher cognitive states on perceptual processing and perceptual phenomenology. even if one agrees with modularists that the influence of expertise on object recognition can be interpreted as cognitive effects on post-perceptual processes, this does not have to be disruptive to stokes’s proposal. it seems his proposal is capable of acknowledging that cognitive contact with the world can have an effect not only on perceptual contact with the world but also on related hybrid phenomena such as recognition (stokes, 2021b, pp. 3–7). the main thesis of the book, namely the cognitive improvement of perception in cases of top-down effects of expertise on recognition, remains untouched by the pertinent warning by firestone & scholl (2016, pp. 15–17) who suggest not to confuse two integral but separable constituents of recognition: perception and memory. carefully distinguishing these constituents is essential, because cognitive effects on back-end memory have no implications for front-end perception. thinking and perceiving avoids this pitfall. why does explaining expert recognition matter? providing a satisfactory account of the smoothness of perceptual recognition of objects for which we serve as experts would not only inform debates in philosophy of mind/perception and the cognitive sciences but also furnish us with a better understanding of the process that is ubiquitous in ordinary life and that is central in efficient interactions with objects in our environment (see mroczko-wąsowicz & grush, 2023). the reason is that we all are experts in perceiving and recognizing some kinds of objects (mroczko-wąsowicz et al., 2023). acknowledgments the work was supported by the national science centre, poland (grant 2019/35/b/hs1/04386). mroczko-wąsowicz, a. (2023). perceptual expertise and object recognition: an explanatory task for modularists and antimodularists. philosophy and the mind sciences, 4, 14. https://doi.org/10.33735/phimisci.2023.10247 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10247 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perceptual expertise and object recognition 5 references abid, g. (2021). recognition and the perception–cognition divide. mind and language, 37((5)), 770–789. https://doi.org/10 .1111/mila.12362 bukach, c. m., gauthier, i., & tarr, m. j. (2006). beyond faces and modularity: the power of an expertise framework. trends in cognitive sciences, 10(4), 159–166. https://doi.org/10.1016/j.tics.2006.02.004 dell’acqua, r., & job, r. (1998). is object recognition automatic? psychonomic bulletin & review, 5, 496–503. https://doi.or g/10.3758/bf03208828 deroy, o. (2014). modularity of perception. in m. matthen (ed.), oxford handbook of philosophy of perception (pp. 755–778). oxford university press. https://doi.org/10.1148/rg.331125023 dicarlo, j. j., zoccolan, d., & rust, n. c. (2012). how does the brain solve visual object recognition. neuron, 73, 415–434. https://doi.org/10.1016/j.neuron.2012.01.010 drayson, z. (2017). modularity and the predictive mind. in t. m. & w. wiese (ed.), philosophy and predictive processing (pp. 1–12). frankfurt am main: mind group. https://doi.org/doi: 10.15502/9783958573130 drew, t., evans, k., vő, m. l.-h., jacobson, f. l., & wolfe, j. m. (2013). informatics in radiology: what can you see in a single glance and how might this guide visual search in medical images? radiographics, 33, 263–274. https://doi.org/ 10.1148/rg.331125023 firestone, c., & scholl, b. (2016). cognition does not affect perception: evaluating the evidence for “top-down” effects. behavioral and brain sciences, 39e229, e229. https://doi.org/10.1017/s0140525x15000965 fodor, j. (1983). the modularity of mind: an essay on faculty psychology. cambridge, ma: mit press. ivy, s., rohovit, t., lavelle, m., padilla, l., stefanucci, j., stokes, d., & drew, t. (2021). through the eyes of the expert: evaluating holistic processing in architects through gaze-contingent viewing. psychonomic bulletin & review, 28(3), 870–878. https://doi.org/10.3758/s13423-020-01858-w ivy, s., rohovit, t., stefanucci, j., stokes, d., mills, m., & drew, t. (2023). visual expertise is more than meets the eye: an examination of holistic visual processing in radiologists and architects. journal of medical imaging, 10(1), 1–15. https://doi.org/10.1117/1.jmi.10.1.015501 kundel, h. l., nodine, c. f., conant, e. f., & weinstein, s. p. (2007). holistic componentof image perception in mammogram interpretation: gaze-tracking study. radiology, 242, 396–402. https://doi.org/10.1148/radiol.2422051997 marr, d. (1982). vision: a computational investigation into human representation and processing of visual information. san francisco: freeman. mohr, s., wang, a., & engell, a. d. (2018). early identity recognition of familiar faces is not dependent on holistic processing. social cognitive and affective neuroscience, 13(10), 1019–1027. https://doi.org/10.1093/scan/nsy079 moors, a., & de houwer, j. (2006). automaticity: a theoretical and conceptual analysis. psychological bulletin, 132(2), 297–326. https://doi.org/10.1037/0033-2909.132.2.297 mroczko-wąsowicz, a. (2022). modularity. in b. d. young & c. dicey jennings (eds.), mind, cognition, and neuroscience: a philosophical introduction (pp. 149–163). new york: routledge press. mroczko-wąsowicz, a., & anaya, a (2022). how to explain the automaticity of object recognition. in j. culbertson, a. perfors, h. rabagliati, & v. ramenzoni. proceedings of the 44th annual conference of the cognitive science society. mroczko-wąsowicz, a., & grush, r. (eds.). (2023). sensory individuals: unimodal and multimodal perspectives. oxford: oxford university press. mroczko-wąsowicz, a., o’callaghan, c., cohen, j., scholl, b., & kellman, p. (2023). advances in the study of visual and multisensory objects. proceedings of the 45th annual conference of the cognitive science society. papineau, d. (2013). in the zone. royal institute of philosophy supplements, 73, 175–196. https://doi.org/10.1017/s135824611 3000325 pylyshyn, z. (1999). is vision continuous with cognition? the case for cognitive impenetrability of visual perception. behavioral and brain sciences, 22(3), 341–365. https://doi.org/10.1017/s0140525x99002022 pylyshyn, z. (2003). seeing and visualizing: it’s not what you think. cambridge, ma: mit press. raftopoulos, a. (2011). late vision: processes and epistemic status. frontiers in psychology, 2(382), 382. https://doi.org/10.3 389/fpsyg.2011.00382 scott, l. s. (2011). face perception and perceptual expertise in adult and developmental populations. in g. r. rhodes, j. haxby, m. johnson, & a. calder (eds.), oxford handbook of face perception (pp. 195–214). oxford: oxford university press. https://doi.org/10.1093/oxfordhb/9780199559053.013.0011 serre, t., oliva, a., & poggio, t. (2007). a feedforward architecture accounts for rapid categorization. proceedings of the national academy of sciences of the united states of america, 104(15), 6424–6429. https://doi.org/10.1073/pnas.07006 22104 mroczko-wąsowicz, a. (2023). perceptual expertise and object recognition: an explanatory task for modularists and antimodularists. philosophy and the mind sciences, 4, 14. https://doi.org/10.33735/phimisci.2023.10247 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1111/mila.12362 https://doi.org/10.1111/mila.12362 https://doi.org/10.1016/j.tics.2006.02.004 https://doi.org/10.3758/bf03208828 https://doi.org/10.3758/bf03208828 https://doi.org/10.1148/rg.331125023 https://doi.org/10.1016/j.neuron.2012.01.010 https://doi.org/doi:%2010.15502/9783958573130 https://doi.org/10.1148/rg.331125023 https://doi.org/10.1148/rg.331125023 https://doi.org/10.1017/s0140525x15000965 https://doi.org/10.3758/s13423-020-01858-w https://doi.org/10.1117/1.jmi.10.1.015501 https://doi.org/10.1148/radiol.2422051997 https://doi.org/10.1093/scan/nsy079 https://doi.org/10.1037/0033-2909.132.2.297 https://doi.org/10.1017/s1358246113000325 https://doi.org/10.1017/s1358246113000325 https://doi.org/10.1017/s0140525x99002022 https://doi.org/10.3389/fpsyg.2011.00382 https://doi.org/10.3389/fpsyg.2011.00382 https://doi.org/10.1093/oxfordhb/9780199559053.013.0011 https://doi.org/10.1073/pnas.0700622104 https://doi.org/10.1073/pnas.0700622104 https://doi.org/10.33735/phimisci.2023.10247 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org aleksandra mroczko-wąsowicz 6 stokes, d. (2021a). on perceptual expertise. mind and language, 36(2), 241–263. https://doi.org/10.1111/mila.12270 stokes, d. (2021b). thinking and perceiving: on the malleability of the mind. routledge/taylor & francis group. stokes, d. (2023). précis of thinking and perceiving. philosophy and the mind sciences, 10. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. mroczko-wąsowicz, a. (2023). perceptual expertise and object recognition: an explanatory task for modularists and antimodularists. philosophy and the mind sciences, 4, 14. https://doi.org/10.33735/phimisci.2023.10247 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1111/mila.12270 https://doi.org/10.33735/phimisci.2023.10247 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org stokes's malleability thesis and the normative grounding of propositional attitudes. stokes’s malleability thesis and the normative grounding of propositional attitudes christopher molea(cmole@mail.ubc.ca) abstract the position that stokes’s thinking and perceiving aims to overthrow is committed to the idea that the facts about one’s propositional attitudes and the facts about one’s perceptual experiences are alike grounded in facts about representations (in various formats) that are being held in a short or long term memory store, so that computations can be performed upon them. claims about modularity are claims about the distinctness of these memory stores, and of these representations. one way in which to reject those claims is to deny only that distinctness. a more radical way would be to reject the underlying idea that facts about perception and facts about propositional attitudes are alike grounded in facts about stored representations. although the more radical approach might seem to face a problem concerning causal efficacy, i suggest that the way is open for stokes to take it. keywords modularity ∙ perception ∙ expertise this article is part of a symposium on dustin stokes’s book “thinking and perceiving” (routledge 2021), edited by regina fabry and sascha benjamin fink. 1 rejecting the fodorian picture if you were a cognitive scientist in the last decades of the twentieth century then you may have been attracted to an exclusively feed-forward picture of the processes by which beliefs are acquired from experience. that picture is likely to have been one in which the processes of perception were taken to implement an algorithm computing a function from proximal stimuli to representations of one’s environment (along the lines theorized by marr, 1982). the representations generated by this algorithm would then serve as the inputs for a subsequent process, with the role of generating and regulating our doxastic commitments. the most auniversity of british columbia. mole, c. (2023). stokes’s malleability thesis and the normative grounding of propositional attitudes. philosophy and the mind sciences, 4, 13. https://doi.org/10.33735/phimisci.2023.10226 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-6144-2089 https://doi.org/10.33735/phimisci.2023.10226 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org christopher mole 2 philosophically complete account of this picture was given by fodor, in his 1983 (fodor, 1983). working memory was implicated in the second, and more doxastic, of these processes. so too were consciousness and abductive rationality. explaining this second process was therefore a daunting prospect, but the hope was that, before that prospect needed to be faced, explanatory questions about the first, merely perceptual phase of belief acquisition could be broached more or less independently. such hopes were founded on the idea that this perceptual processing would turn out to be the work of a module, which would be encapsulated enough to be innocent of any explanatory entanglement with the more philosophically puzzling aspects of the mind. if this fodorian picture had been accurate then perception would have been relatively easy to study with the methods of cognitive neuroscience, and relatively easy to model computationally. the epistemologists’ task might also have been a more straightforward one, since it would have been relatively easy to credit perceivers with uncontaminated objectivity. those implications would have depended on the picture being one in which cognitive states do not serve as arguments for the function that the perceptual process computes (although facts about one’s environment might nonetheless figure in the explanation of the efficacy of the algorithm that that computation implements: this algorithm’s success in arriving at accurate representations of depth might, for example, be explained by the fact that light comes typically from above, but the algorithm would need to depend on this fact without making use of any representation that grounds a personal-level attitude to that fact.) those who adopted this picture often did so as a working hypothesis of convenience, rather than as a matter of conviction. no element of the picture was ever free from controversy. philosophers over the last twenty years have taken issue with every part of it, launching attacks on various fronts, from various positions. the picture nonetheless retains its cachet. it continues to serves as a point of departure, with new approaches defining themselves in contrast to it. in the first half of thinking and perceiving, dustin stokes marshals arguments suggesting that, when this picture’s rationale is examined in the light of more recent empirical findings, it is no longer credible as an account of the way in which perception is actually accomplished, at least in the human case. stokes also thinks that it would be a mistake to regard this as a partial picture, from which reality deviates only in exceptional cases. he suggests that our current theorizing has been unhelpfully skewed by this picture’s ongoing role, as a foil for more recent views. summarizing the results of the discussion in his book’s first half, stokes writes: [m]odularity does not deserve the default theoretical position: neither its arguments nor its explanations are sufficient to justify such a default. accordingly, novel proposals in the philosophy and cognitive science of perception need not attach their success conditions to undermining or disproving modularity; they need not make the case mole, c. (2023). stokes’s malleability thesis and the normative grounding of propositional attitudes. philosophy and the mind sciences, 4, 13. https://doi.org/10.33735/phimisci.2023.10226 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10226 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org stokes’s malleability thesis and the normative grounding of propositional attitudes 3 that an interesting cognitive-perceptual phenomenon ‘counts’ as cognitive penetration. to do so is to continue to make that very default assumption and suffer the theoretical restrictions that come with it. (stokes, 2021, p. 141 f.) in the second half of his book, stokes therefore sets out on a course that attempts to leave this old picture behind. while doing so, he makes a plausible case for a claim (‘the tip thesis’), according to which thinking sometimes exerts an epistemically beneficial influence on perception. that claim is contrary to the spirit of the old picture, and undermines some of the considerations that were supposed to speak in favour of it. in advocating it, stokes hopes to exemplify the benefits of working with a rival picture of perception’s relationship to cognition. the reader is shown enough of this rival picture for us to know that the malleability of perception features prominently in it. but there are various different pictures within which stokes’s tip thesis could be accommodated, and malleability might be prominent in several of them. to effectively exclude the old picture from its role as the default view, one would like to instate some particular rival in its stead. one would therefore like to know which of the various malleabilityaccommodating views to favour. the most radical of those views would reject a fundamental part of the old picture, by taking perception or cognition to be something other than a computational process that generates representations. less radical alternatives would retain the idea that both cognition and perception are constituted by computational processes that generate representations, whilst rejecting only the idea that these processes are encapsulated from one another. the lion’s share of stokes’s discussion takes the less-radical approach, treating perception as a representationgenerating computation. there are some remarks (especially in the book’s epilogue) suggesting that this might be a simplification that has been made for the purposes of exposition, and that stokes’s sympathies might tend in the direction of something more radical. but one who departs from the old picture only on the point concerning modularity might accept the arguments given in all of his book’s more precisely regimented parts. i would like to take the present opportunity to draw stokes out on the point of just how radical he thinks our departure from the fodorian picture should be. to this end i want, below, to sketch one way in which the traditional picture might be rejected. having sketched this view, we will see that its radicalism is somewhat circumscribed: it breaks from the old picture’s conception of belief, but does so while retaining some of the explanatory advantages that had made that old picture’s account of perception appealing. mole, c. (2023). stokes’s malleability thesis and the normative grounding of propositional attitudes. philosophy and the mind sciences, 4, 13. https://doi.org/10.33735/phimisci.2023.10226 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10226 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org christopher mole 4 2 the old picture’s metaphysics the to-be-rejected picture was primarily a hypothesis about processing architecture, but it was imbued with ideas about the metaphysics of the psychological entities that this architecture was supposed to explain. more specifically, it was a picture in which facts about perception and facts about cognition are both grounded, more or less directly, in facts about the representations that are handled by computational processes taking place in the brain (and perhaps also by processes taking place in some more extended system of which the brain is a proper part, but extendedness is not the point that is currently at issue). the advocates of this to-be-rejected picture took one’s perceiving that p to be grounded in the fact that one has some representation of p among the set of representations that one’s perceptual processes are now outputting. they took one’s believing that p to be grounded somewhat similarly, in the fact that one has some representation of p in a memory store from which it can be summoned for use in reasoning. the representations in these two cases might have very different formats, and the processes in which they participate might be computationally very different, but, in each case, there would be a personal-level fact about psychology – that some thinker is having an experience as of it being the case that p, or that the thinker believes that p – and this fact would be metaphysically grounded in a sub-personal fact about a representation being held in a store of the right sort. advocates of this view were therefore willing to speak as if the brain contained, inter alia, a ‘belief box’ and a ‘perception box’. when deciding how radically to depart from this picture, a crucial choice-point is whether to retain these two claims about the metaphysical relationship between the personal and subpersonal levels. i would like to use this opportunity to encourage stokes to be explicit about his own choice at this point. both of the old picture’s metaphysical claims can be retained in a theory that follows stokes in rejecting the encapsulation of perception from cognition, just by allowing the contents of the belief box to serve as arguments for the function that is computed by the process that generates the contents of the perception box. but some of the remarks in stokes’s epilogue suggest that, when arguing for the rejection of encapsulation, he is hoping to point us towards a picture in which the relationship between facts about beliefs and facts about representations in the brain is metaphysically different in kind from the relationship between facts about perceptual experiences and such facts about the brain. perhaps the facts about perception are indeed grounded in facts about the right sort of representation being present in the right sort of box, whereas the facts about propositional attitudes stand in a metaphysically different sort relationship to those facts. perhaps those attitudes are instead grounded in facts about the most rational interpretation of a person’s overall behaviour (in the style of dennett, 1987). perhaps they are grounded in facts about that person’s position in the interpersonal game of giving and asking for reasons (in the style of brandom, 1994). or perhaps they emerge in some very mole, c. (2023). stokes’s malleability thesis and the normative grounding of propositional attitudes. philosophy and the mind sciences, 4, 13. https://doi.org/10.33735/phimisci.2023.10226 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10226 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org stokes’s malleability thesis and the normative grounding of propositional attitudes 5 different way. the brain might still contain representations of some of the propositions to which these attitudes are taken, and these representations might make a notable contribution to the behaviour that is to be interpreted, or to the way in which reasons are given and asked for, but the role of those representations, in providing a metaphysical grounding for facts about a person’s attitudes, would be different in kind from their role in providing a grounding for the facts about that person’s perceptual experiences. the picture of perception that stokes is rejecting was advanced as part of a larger fodorian programme in the philosophy of mind, which had views like dennett’s and brandom’s as its opponents. all of these theorists face the challenge of explaining how attitudes can be causally efficacious. if stokes’s malleability claim is right then this is a challenge that needs to be faced sooner rather than later, since attitudes already have causal work to do in bringing the thinker into perceptual contact with her surroundings, even before that thinker starts to act on those surroundings. answering this challenge might seem to be relatively straightforward work for a theorist who takes the less radical route of retaining fodor’s idea that attitudes and perceptions are alike grounded in facts about the presence of representations in boxes: representations in one box can exert a causal influence on the process that delivers representations to the other box, merely by serving as inputs for the computation that this process implements. that much requires nothing metaphysically recherché. but this challenge looks more severe on a dennettian or brandomian account of the way in which the attitudes are grounded since, for these theorists, the influence of the attitudes on perception will already be a case of interaction between entities from metaphysically distinct categories. if facts about belief are grounded in normative facts (such as facts about one’s entitlement to make moves in the game of giving and asking for reasons) then stokes’s malleability claim will require such normative facts to exert a causal influence on the facts about representation that ground perception. something metaphysically recherché does seem to be required here. and so it can seem like metaphysical complications are avoided by taking the less radical path, at the choice-point that was identified above, by retaining the fodorian conception of the way in which attitudes are grounded. i nonetheless want to suggest that the way is open for stokes to take the more radical path, without thereby running into metaphysical difficulties. 3 the perceptual benefits of expertise are not effects of conscious beliefs as such one takes this more radical path by adopting a normative account of the facts in which attitudes are grounded. these facts about normative standing will not map directly onto any merely descriptive facts about the brain, but those facts about the brain will still make some difference to that standing, by making a difference mole, c. (2023). stokes’s malleability thesis and the normative grounding of propositional attitudes. philosophy and the mind sciences, 4, 13. https://doi.org/10.33735/phimisci.2023.10226 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10226 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org christopher mole 6 to the way in which a thinker plays the game of giving and asking for reasons (or by making difference to the way in which that thinker behaves, and thereby making a difference to the way in which their behaviour is best interpreted). on either of these views, the attitude-grounding facts will include facts that are in the right place to do the causal work that stokes’s malleability claim requires. but the literature on mental causation has made us familiar with the idea that being in the right place is not enough: having a rook on one’s chess board may play a role in grounding the normative status of having checkmated one’s opponent, and it might also put this rook in the right place to exert a causal influence outside of the game, without the normative standing that it grounds being relevant to those extra-ludic effects. the dog that chokes on a checkmating rook has not been incapacitated by a winning-move as such. to explain the efficacy of the mental as such, it is not enough for the states that ground the attitudes to enter into causal relations. we want them to do so ‘qua mental’. and this seems to make a familiar sort of trouble for a dennettian or brandomian version of stokes’s picture. the point that i want to bring to light here is that, whether or not that trouble is soluble, for stokes it is trouble that can be avoided. to see why, notice that the empirically-demonstrated effects that provide stokes with his reasons for claiming that thinking exerts a virtuous influence on perception are derived from studies of the perceptual benefits that come from the acquisition of expertise. the latter half of his book examines the ways in which an expert’s thinking (about fingerprints, or radiological scans, or greebles) is found to have a virtuous influence on the speed and accuracy of their perception. that thinking will include the formation of propositional attitudes concerning these various topics of expertise. but it is not obvious that the virtuous influence of this thinking is as an influence of these attitudes as such. a counterfactual test for causal relevance suggests otherwise: suppose that a fingerprint expert has been so thoroughly bamboozled by some global sceptic that she no longer believes any of the propositions that she learned when acquiring her expertise. perhaps she even acquired this expertise while remaining resolutely agnostic about the theories on which it was based. if she nonetheless remained motivated and attentive then the lack of a believing attitude seems unlikely to deprive this expert of the perceptual abilities that her expertise explains. and this suggests that, even for the expert who avoids falling into scepticism, it was never the attitude of belief as such that was explanatory of their perceptual performance. it was instead some subdoxastic body of stored information that figured in the grounding of such an attitude, when accompanied by an appropriate background. but if it is not belief as such that explains the expert’s perceptual abilities then the qua-problem is toothless as an objection to the idea that that expert’s attitudes have a different metaphysical grounding from the perceptual states that show the influence of her expertise. a similar point can be made about the expert’s consciousness of the contents of the attitudes that are concomitant to the acquisition of her expertise. a 2002 study by fahle and daum (2002) suggests that at least some patients with profound mole, c. (2023). stokes’s malleability thesis and the normative grounding of propositional attitudes. philosophy and the mind sciences, 4, 13. https://doi.org/10.33735/phimisci.2023.10226 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10226 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org stokes’s malleability thesis and the normative grounding of propositional attitudes 7 anterograde amnesia can improve their performance in a visual discrimination task at the limits of their acuity, and that they can retain this benefit for some time. the perceptual benefits of such learning must involve some stored information, but they seem not to depend on that information being accessible to working memory, since these benefits are present even when amnesia prevents such access. perhaps, in the normal case, the states that represent this information might play a role in grounding some conscious attitude to the propositions represented, but the effects of these representations on perceptual performance seems not to be an effect that they have qua grounders of a subject’s conscious propositional attitudes. the effects of expertise that stokes cites can therefore be taken as support for his claim that thinking has a beneficial effect on perception, but they should not be taken as establishing that the processes of perception are influenced by one’s conscious beliefs as such. the way is therefore open for him to adopt a position in which conscious attitudes stand in a relationship to the brain that is metaphysically different in kind from the relationship that is seen in the case of experiences. rather than departing from the fodorian picture merely by allowing the contents of the belief box to exert an influence on the processes that output representations to the perception box, stokes can make the more radical move of rejecting the idea that facts about belief are grounded in facts about the appropriate boxing of representations. such a move is unlikely to be welcomed by the diehard fodorian. it nonetheless enjoys some of the advantages of the fodorian picture with which we began. by allowing that the states influencing perception are not conscious beliefs as such (even if they are representations that do typically play some role in the grounding of such beliefs), the version of stokes’s picture that i have sketched allows the explanation of perception to be disentangled from the explanations of consciousness, and of the abductive inferences in which beliefs characteristically participate. it therefore warrants some of the same explanatory optimism that accompanied the picture with which we started. references brandom, r. (1994). making it explicit: reasoning, representing, and discursive commitment. harvard university press. dennett, d. c. (1987). the intentional stance. mit press. fahle, m., & daum, i. (2002). perceptual learning in amnesia. neuropsychologia, 40(8), 1167–1172. https://doi.org/10.1016/ s0028-3932(01)00231-7 fodor, j. a. (1983). the modularity of mind: an essay on faculty psychology. mit press. marr, d. (1982). vision: a computational investigation into the human representation and processing of visual information. w.h. freeman. stokes, d. (2021). thinking and perceiving: on the malleability of the mind. routledge. mole, c. (2023). stokes’s malleability thesis and the normative grounding of propositional attitudes. philosophy and the mind sciences, 4, 13. https://doi.org/10.33735/phimisci.2023.10226 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/s0028-3932(01)00231-7 https://doi.org/10.1016/s0028-3932(01)00231-7 https://doi.org/10.33735/phimisci.2023.10226 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org christopher mole 8 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. mole, c. (2023). stokes’s malleability thesis and the normative grounding of propositional attitudes. philosophy and the mind sciences, 4, 13. https://doi.org/10.33735/phimisci.2023.10226 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10226 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org rejecting the fodorian picture the old picture's metaphysics the perceptual benefits of expertise are not effects of conscious beliefs as such a double anniversary for the neural correlates of consciousness: editorial introduction. a double anniversary for the neural correlates of consciousness editorial introduction sascha benjamin finka (sfink@ovgu.de) this article is part of a special issue on “the neural correlates of consciousness”, edited by sascha benjamin fink. at the heart of the contemporary science of consciousness is the search for neural correlates of consciousness (nccs). nccs and the research programme associated with it rose to prominence 30 years ago with the article towards a neurobiological theory of consciousness by francis crick and christof koch (1990). the notion of a neural correlate of consciousness itself was around before their seminal paper. crick and koch, for example, cite logothethis and schall’s neuronal correlates of subjective visual perception (1989), which arguably points in the same direction. that paper reports of single neuron recordings of the subjective perception of movement. it is therefore far less ambitious than crick and koch’s, which asks generally about the where, when, how, and why of neural activation correlating with consciousness. but during the reign of cognitivism, where consciousness was largely ignored, its overall usage was sparse and mainly in passing, e.g. when the term is also used by c. l. hardin (1988, p. 130) in color for philosophers: unweaving the rainbow where he talks of the chromatic cells in the visual cortex as “the hypothetical neural correlate of ‘consciousness’ ” – arguably shying away from such a thorny subject as consciousness by putting it in scare quotes. earlier, we find the term in j. s. griffith’s a view of the brain (1967, p. 25) as part of a more mathematical model of the brain’s workings. however, before the advent of behaviorism, nccs where popular enough to be mentioned in the entry on psychology in the encyclopedia britannica, written by james ward (1911). aphilosophy-neurosciences-cognition otto-von-guericke universität magdeburg fink, s. b. (2020). a double anniversary for the neural correlates of consciousness: editorial introduction. philosophy and the mind sciences, 1(ii), 1. https://doi.org/10.33735/phimisci.2020.ii.85 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-7535-1882 https://doi.org/10.33735/phimisci.2020.ii.85 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 2 this indicated that the idea that conscious processes have neural correlates was not uncommon in 19th century psychology, e.g. when edmund gurney (1881, p. 156) writes that “the movements of mind are correlated with nervous processes” or when mary whiton calkins (1892, p. 399) speaks of “cerebral correlates of […] intellectual processes”. so while crick and koch may not have originated the term “neural correlate of consciousness”, their article was crucial in pushing it onto the scientific agenda by sketching a general research programme guided by the question “where in the brain are the neural correlates of consciousness?” – which still occupies researchers today, 30 years later. the paper therefore deserves to be the most cited source and the central root of our contemporary neuroscience of consciousness. spelling the goal of this investigation out in terms of neural correlates (rather than something else) had several advantages and allowed researchers to get an empirical investigation off the ground. first, methods for capturing correlations (unlike more philosophical notions like supervenience, grounding, co-extensionality, identity and so on) is something already within the standard toolbox of empirical science. at its core, it measures the strength of dependence between two variables. bernard baars (1997) made this point explicit: in the ncc-framework, scientists treat consciousness as a variable in experimental set-ups. the goal is to see which neural variables co-vary with it. second, and more importantly, while correlation may not be fully metaphysically neutral, it is not too restrictive either. unlike neural substrates, constituents, reduction bases, or other notions that are more metaphysically loaded, neural correlates are compatible with a bunch of different but mutually excluding metaphysical views on how our neural goings-on relate to consciousness. even in leibnizian parallelistic dualism, there can be a neural correlate of a conscious experience, as ward (1911, p. 600ff) already pointed out. but the same holds for monistic identity theory, where we might think of the correlates as ways in which one and the same thing is given to us or picked out, introspectively or neuroscientifically. only a denial of the existence of matter or consciousness (i.e., eliminativism or irrealism) precludes a search for correlates because there is nothing to correlate with. or one must claim that conscious goings-on lack even a statistical relation to neural goings-on, which is rare even among proponents of enactive, embodied, or extended approaches. beyond such extreme positions, most metaphysics will allow for something neural correlating with consciousness. this metaphysical promiscuity allowed researchers to bracket ancient but ongoing philosophical debates on the mind-body problem and simply start their empirical work. if coined in terms of correlation, the results will be compatible with most of what philosophers (eventually or never) converge on. empirical researchers could also bracket their own varying metaphysical temperaments – some like popper & eccles (1977) have dualistic, others more reductionistic tendencies. such divergences need not hinder fruitful collaboration in a coordinated attempt to find out: “where in the brain are the neural correlates of consciousness?” while elegant in how it opens up avenues for empirical work, the ncc research programme also raised conceptual, methodological, and empirical questions, some fink, s. b. (2020). a double anniversary for the neural correlates of consciousness: editorial introduction. philosophy and the mind sciences, 1(ii), 1. https://doi.org/10.33735/phimisci.2020.ii.85 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.85 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a double anniversary for the neural correlates of consciousness 3 of which are still hotly debated. consider, for example, that correlation is relatively cheap as any two variables correlate to some degree. in order to prevent the search for nccs from becoming trivial, we must ask: which further constraints on correlation should be imposed to distinguish mere statistical correlates from core correlates that hint at a foothold of consciousness in the material world? how can we screen off core correlates from enabling conditions or causal factors? how can we best access, categorise or measure the conscious side of the correlation, which seems elusively subjective – especially if it might be tied to observable behaviour in only a loose sense? how do neural correlates relate to theories of consciousness? how can candidates for core nccs be tested? how do correlates relate to theories of consciousness? how do correlates of specific experiences (like seeing vermillion or smelling burnt toast) relate to correlates of broader conscious states like (like dreaming or waking)? how many different core nccs should we expect for a given conscious state? how does this research programme relate to embodiment and extended views of the mind or to research on artificial consciousness? the list continues. in approaching questions like these, the anthology the neural correlates of consciousness: empirical and conceptual questions (metzinger, 2000), edited by thomas metzinger, stands out. metzinger, who later also served as president for the association for the scientific study of consciousness ten years ago (2009/2010), brought together a wide range of empirical researchers and philosophers and thereby illustrated that the self-conception of the budding science of consciousness includes a continuous reflection about the philosophical foundations of its empirical side – but also a demand for deriving testable implications from philosophical views on consciousness in order to keep them accountable to empirical findings. this anthology also featured a detailed analysis by david chalmers (2000) of what a neural correlate of consciousness is supposed to be: an ncc of an experience 𝐶 is the neural state 𝑁 that is minimally sufficient for that experience 𝐶. being merely sufficient allows for several neural systems to give rise to a conscious experience. for example, before a stroke, neural area 𝑁1 may have brought about experiences as of seeing red, feeling one’s hand, or comprehending a word. after recovering from the stroke, these experiences may be brought about by 𝑁2. neither 𝑁1 nor 𝑁2 is therefore necessary for that experience. but searching for only sufficient criteria might be trivial: the state of my whole brain is sufficient for my seeing red now. thanks to the minimality criterion, we can reject such trivial answers and ask again: “where in the brain are the neural correlates of consciousness?” chalmers’ definition has become the gold standard, the one most often proclaimed to be what ncc-researchers aim to find. this special issue of philosophy and the mind sciences celebrates this double anniversary: 30 years since the research programme for finding neural correlates of consciousness was brought back onto the scientific agenda and 20 years since the standard definition of an ncc was published in a seminal anthology. here, we fink, s. b. (2020). a double anniversary for the neural correlates of consciousness: editorial introduction. philosophy and the mind sciences, 1(ii), 1. https://doi.org/10.33735/phimisci.2020.ii.85 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.85 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 4 want to take up some of the conceptual, methodological, and empirical questions raised and take stock of how the field has developed since then. in this issue, seven original peer-reviewed research papers take up these threads. talis bachmann, mototaka suzuki, and jaan aru in dendritic integration theory: a thalamocortical theory of state and content of consciousness (2020) improve on the prominent view of thalamo-cortical loops as nccs. they start from ten observations motivated by phenomenology and empirical data. a neuroscientific account of consciousness must explain all ten. their dendritic integration theory focuses specifically on l5p neurones that receive input from both cortex and thalamus and thereby can mitigate between the important cortico-cortical processing (relevant for the specificity of what one is conscious of) and thalamo-cortical processing (relevant for the integration of different conscious features into a unified experience). this allows for an interplay between contents of consciousness and larger background states being thalamically modulated. they argue that this also accounts for the ability for unconscious cortical processing. jakob hohwy and anil seth present predictive processing as a framework for progressing the search for nccs in their predictive processing as a systematic basis for identifying the neural correlates of consciousness (2020). the problem they raise is that there is a discrepancy between theories of consciousness (which are not always cashed out in neural terms), and empirical findings about neural correlates of consciousness. they stress that we must distinguish between arbitrary correlates – statistical coincidences not predicted by any hypothesis or theory – and systematic correlates, which are directly implicated by a theory. the goal of the neuroscience of consciousness is to understand, explain and predict mechanisms facilitating consciousness. for this, we require a theory. but so far, there is no convergence onto one. we may facilitate convergence by comparing different theories neuroscientifically. in order to cash out theories of consciousness in neural terms and make them comparable with each other and neuroscientific data, they suggest predictive processing as a suitable mitigator. predictive processing as a framework therefore promises a unification of the field without a commitment to a specific theory. this go-between, however, also has the potential to become a theory of consciousness in its own rights. colin klein, tim bayne, and jakob hohwy question that we should search for neural correlates of consciousness at all in their explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (2020). they make the point that the search for nccs rests on some controversial assumptions, which culminate in the lite-brite-model: (i) that a neural event brings about a specific experience autonomously and independently of what else goes on in consciousness or the brain, (ii) that an experience is uniquely dependent on a single neural element, and (iii) that global changes in consciousness uniformly affect all occurrent experiences. they contrast this with the difference maker framework. here, we ask which neural events make a difference to consciousness. not only is this research programme logically more complex, it fink, s. b. (2020). a double anniversary for the neural correlates of consciousness: editorial introduction. philosophy and the mind sciences, 1(ii), 1. https://doi.org/10.33735/phimisci.2020.ii.85 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.85 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a double anniversary for the neural correlates of consciousness 5 is also tied to a rich tradition of analysing causal interactions, experimental setups, and has facilitated great progress in other fields like genetics. the days of searching for nccs might be numbered. tomáš marvan and michal polák in their generality and content-specificity in the study of the neural correlates of perceptual consciousness (2020) pick up the question of what makes any content conscious. this is to them the pivotal question. according to them, just searching for nccs is then too broad: the notions must be split up into neural correlates of content (ncc, with a lower case 𝑐) and a general neural correlate of consciousness (gncc), which is the mechanism shared by all conscious contents. they assume that when we compare the neural correlates of conscious events, we will find the general neural correlate of consciousness as a stable part, but combined with varying neural correlates of content (which can, in principle, occur outside the ncc as well). so it is this broad mechanism that, in interaction, turns the content of a neural representation conscious. they map out which theories do or don’t allow for such a distinction – which in turn lends itself to evaluating such theories if the distinction is accepted. matan mazor and stephen m. fleming in their distinguishing absence of awareness from awareness of absence (2020) focus on one of the most important paradigms in the search for nccs, namely the contrastive analysis of trials where a subject is aware of a stimulus and those where a subject is unaware of a stimulus. they point out that this falls short of what is actually needed: subjects are not only aware of stimuli, they sometimes are also aware of the absence of a stimulus. or, vice versa, they can be unaware of the presence of a stimulus but also unaware of the absence of a stimulus. in both cases, the awareness of absence and the awareness of a stimulus, the neural correlate of awareness might be present. similarly, in both the unawareness of a stimulus and the awareness of a stimulus, a neural correlate of the stimulus may be present. if we do not capture this difference, we might either discard or fail to see relevant neural data. hence, they argue for a twodimensional report scheme. in concordance with marvan and polák, they suggest that this allows for distinguishing between content-specific and content-invariant nccs. andy mckilliam (2020) asks the question what is a global state of consciousness? this is a central question to ncc research, specifically regarding the distinction between content-specific and content-invariant correlates. a general mechanism for consciousness might also vary when global states vary from wakefulness to drowsiness or to dreaming and so on. still, such global states might be content-invariant. in order to research global states of consciousness, we need to know what it is. mckilliam finds all explications currently at play wanting, be they as levels, phenomenal states in their own right, total states constituted by all occurrent experiences, phenomenal signatures, or as phenomenal structures. as a contrast, he presents the capacities account of global states, where global states simply govern what a creature is capable of experiencing and how it is capable of accessing these experiences while in that global state. but i need not occurrently exercise my capacities in order to have them, e.g. i have the capacity to ride a bike fink, s. b. (2020). a double anniversary for the neural correlates of consciousness: editorial introduction. philosophy and the mind sciences, 1(ii), 1. https://doi.org/10.33735/phimisci.2020.ii.85 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.85 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 6 even though i am not currently riding a bike. the controversial upshot is that we may miss nccs for global states if we only look at neural data correlating with occurrent experiences. this suggests a totally new approach for investigating the neural correlates of waking, dreaming, being under anaesthesia and so on. matthias michel and hakwan lau report on the dangers of conflating strong and weak versions of a theory of consciousness (2020). they emphasise that we have to distinguish markers from constituents and both from correlates of consciousness. neural markers are events that we can exploit and use epistemically as indicators of consciousness. in that role, they differ from someone telling us that they are conscious mainly in that they are neural, not verbal. constituents, in contrast, lend themselves to reduction because they are what conscious events can actually be identified with. both are neural correlates, but markers are correlates only in a statistical sense, while constituents go beyond chalmers’ conception because they are also necessary in some sense. in that, neural constitution is more metaphysically demanding than correlation – and much harder to prove empirically. failing to make the trifold distinction between neural markers, constituents and correlates of consciousness might lead to overestimating the scope of a theory. they illustrate this with a critique of the integrated information theory: if integration is understood as a marker of consciousness, then the theory is empirically testable and maybe even supported – but the theory becomes untestable if integration is understood as a constituent of consciousness. integrated information theory in this fundamental reading goes far beyond finding correlates. instead, it detaches itself from empirical findings by claiming something untestable. these seven original research papers clearly illustrate on the one hand the impact which nccs still have on the science of consciousness and on the other hand the controversies this notion still raises. we still make progress and amend both the research programme inaugurated by crick and koch 30 years ago as well as the definition of an ncc presented by chalmers 20 years ago, all in the interdisciplinary, empirically grounded and philosophically open spirit exhibited in the anthology by metzinger. philosophy and the mind sciences will continue this discussion in a second issue in 2021 with several more articles. these, like this issue as a whole, have been severely affected and hindered by the covid-19 pandemic. so i am especially thankful to all the authors and reviewers for their work during these hard times. my gratitude goes to them as well as our supporters, but also to wanja wiese and jennifer windt as well as fabian fuchs for their help in getting this first part of the special issue out in 2020. only thanks to this communal effort was it possible to celebrate this double anniversary for the science of consciousness this year. fink, s. b. (2020). a double anniversary for the neural correlates of consciousness: editorial introduction. philosophy and the mind sciences, 1(ii), 1. https://doi.org/10.33735/phimisci.2020.ii.85 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.85 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a double anniversary for the neural correlates of consciousness 7 references baars, b. j. (1997). in the theater of consciousness. oxford university press. bachmann, t., aru, j., & suzuki, m. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii: special issue on the neural correlates of consciousness). https://doi.org/10.33735/phimisci.2020.ii.52 calkins, m. w. (1892). a suggested classification of cases of association. the philosophical review, 1(4), 389–402. http: //www.jstor.org/stable/2175939 chalmers, d. j. (2000). what is a neural correlate of consciousness? in t. metzinger (ed.), neural correlates of consciousness: empirical and conceptual questions (pp. 17–39). mit press. crick, f., & koch, c. (1990). towards a neurobiological theory of consciousness. seminars in the neuroscience, 2, 263–275. griffith, j. s. (1967). a view of the brain. oxford university press. gurney, e. (1881). monism. mind, vi(22), 153–173. https://doi.org/10.1093/mind/os-vi.22.153 hardin, c. l. (1988). color for philosophers: unweaving the rainbow. hackett publishing company. http://gen.lib.rus.ec/boo k/index.php?md5=3b87964d6381925dc1468be87b5b077e hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii: special issue on the neural correlates of consciousness). https: //doi.org/10.33735/phimisci.2020.ii.64 klein, c., bayne, t., & hohwy, j. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii: special issue on the neural correlates of consciousness). https://doi.org/10.33735/phimisci.2020.ii.60 logothetis, n. k., & schall, j. d. (1989). neuronal correlates of subjective visual perception. science, 245(4919), 761–763. https://doi.org/10.1126/science.2772635 marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii: special issue on the neural correlates of consciousness). https://doi.org/10.33735/phimisci.2020.ii.61 mazor, m., & fleming, s. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii: special issue on the neural correlates of consciousness). https://doi.org/10.33735/phimisci.2020.ii.69 mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii: special issue on the neural correlates of consciousness). https://doi.org/10.33735/phimisci.2020.ii.58 metzinger, t. (ed.). (2000). neural correlates of consciousness: empirical and conceptual questions. mit press. michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii: special issue on the neural correlates of consciousness). https://doi.org/10.33735/phimi sci.2020.ii.54 popper, k. r., & eccles, j. (1977). the self and its brain. springer. ward, j. (1911). psychology. in encyclopedia britannica: vol. xxii (11th ed., pp. 547–604). cambridge university press. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. fink, s. b. (2020). a double anniversary for the neural correlates of consciousness: editorial introduction. philosophy and the mind sciences, 1(ii), 1. https://doi.org/10.33735/phimisci.2020.ii.85 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 http://www.jstor.org/stable/2175939 http://www.jstor.org/stable/2175939 https://doi.org/10.1093/mind/os-vi.22.153 http://gen.lib.rus.ec/book/index.php?md5=3b87964d6381925dc1468be87b5b077e http://gen.lib.rus.ec/book/index.php?md5=3b87964d6381925dc1468be87b5b077e https://doi.org/10.33735/phimisci.2020.ii.64 https://doi.org/10.33735/phimisci.2020.ii.64 https://doi.org/10.33735/phimisci.2020.ii.60 https://doi.org/10.1126/science.2772635 https://doi.org/10.33735/phimisci.2020.ii.61 https://doi.org/10.33735/phimisci.2020.ii.69 https://doi.org/10.33735/phimisci.2020.ii.58 https://doi.org/10.33735/phimisci.2020.ii.54 https://doi.org/10.33735/phimisci.2020.ii.54 https://doi.org/10.33735/phimisci.2020.ii.85 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org belief now, true belief later: the epistemic advantage of self-related insights in psychedelic-assisted therapy belief now, true belief later: the epistemic advantage of self-related insights in psychedelic-assisted therapy chiara caporuscioa,b,c,d (caporusciochiara1@gmail.com) abstract chris letheby’s defence of psychedelic therapy hinges on the premise that psychedelic-facilitated insights about the self are in a better epistemic position than those about the external world. in this commentary, i argue that such a claim is not sufficiently defended. more precisely, i argue that one element is underexplored in letheby’s otherwise compelling picture: namely, that unlike new beliefs about the external world, beliefs about oneself have the capacity to turn into self-fulfilling prophecies. recognising the psychedelic experience and the subsequent integration process as opportunities not only to apprehend certain facts about the self but also to actively shape and redetermine those facts is key to understanding the epistemic differences between insights patients have about themselves and about the external world. keywords first-person authority ∙ insight ∙ integration∙ psychedelics ∙ psychedelic therapy ∙ self-shaping this article is part of a symposium on chris letheby’s book “philosophy of psychedelics” (oup 2021), edited by chiara caporuscio and sascha benjamin fink. according to the comforting delusion objection to psychedelic therapy (henceforth cdo), psychedelic therapy should not be recommended even though its psychological effects are beneficial. the justification is that psychedelics produce positive psychological effects only because they induce comforting beliefs in a joyous cosmology, a divine consciousness, or an ultimate reality. because such beliefs are incompatible with naturalism and therefore most likely false, defenders of the cdo argue that the therapeutic potential of psychedelics aotto-von-guericke-universität magdeburg, germany. bdepartment of psychiatry and psychotherapy, charité–universitätsmedizin berlin, germany. cberlin school of mind and brain, faculty of philosophy, humboldt-universität zu berlin, germany. dresearch training group 2386 “extrospection”, humboldt-universität zu berlin, germany. caporuscio, c. (2022). belief now, true belief later: the epistemic advantage of self-related insights in psychedelic-assisted therapy. philosophy and the mind sciences, 3, 6. https://doi.org/10.33735/phimisci.2022.9310 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-3267-6251 https://doi.org/10.33735/phimisci.2022.9310 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chiara caporuscio 2 comes with large epistemic costs that outweigh their psychological benefits (letheby, 2021, p. 2). chris letheby’s book the philosophy of psychedelics is an elaborate and largely convincing rebuttal of cdo. letheby accepts that the ethical status of psychedelic therapy hinges on its epistemic status. however, he contends that mental wellbeing is promoted not by changes in metaphysical beliefs but by changes in beliefs about the self: for example, “i am in touch with my emotions” (watts et al., 2017), “my identity is not tied to being a smoker” (noorani et al., 2018). according to letheby, these self-related beliefs are less likely to be delusional than metaphysical beliefs, making the mechanism of psychedelic therapy epistemically (and therefore ethically) innocent. when assessing letheby’s argument, we are immediately faced with the following question: if psychedelic-induced beliefs about external reality are probably false, why should the same not be true of psychedelic-induced beliefs about ourselves? unlike the metaphysical beliefs questioned by the cdo, beliefs about the self are compatible with naturalism; however, this does not necessarily make them more likely to be true. letheby makes a good case for the non-propositional epistemic benefits of psychedelic self-related insights, such as the acquisition of knowledge-how and knowledge by acquaintance, and for the indirect epistemic benefits gained through increased psychological well-being (see chapters 8.4–8.8). however, his argument about the direct acquisition of knowledge-that about the self through psychedelics, discussed in chapter 8.3, is less compelling. in what follows, i present a three-part response to this chapter. 1. i reconstruct letheby’s argument for the acquisition of knowledge-that through psychedelics, and suggest that it leaves self-related insights in a very similar epistemic position to that of beliefs about the external world. 2. i argue that, if psychedelics are equally likely to bring about false beliefs about the self and about the world, letheby’s reply to the cdo is not very convincing. 3. i offer an alternative argument, underexplored in letheby’s book, as to why self-related insights might after all be in a better epistemic position than insights about the external world. first, letheby’s argument. chapter 8.3 defends that we have grounds to believe that at least some of the self-related insights facilitated by the psychedelic experience are genuine and can promote the acquisition of new propositional knowledge, or knowledge-that. the argument for the possibility of obtaining knowledge-that about oneself from psychedelic administration goes like this: 1. decreasing the weighting of self-related priors can increase the probability of accurately apprehending certain facts about oneself (from general predictive processing theory) caporuscio, c. (2022). belief now, true belief later: the epistemic advantage of self-related insights in psychedelic-assisted therapy. philosophy and the mind sciences, 3, 6. https://doi.org/10.33735/phimisci.2022.9310 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9310 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org belief now, true belief later 3 2. psychedelic administration temporarily decreases the weighting of selfrelated priors (from the rebus model and the predictive self-binding theory) 3. therefore, psychedelic administration can increase the probability of accurately apprehending certain facts about oneself. (letheby, 2021, p. 169) there is a problem with this, which letheby himself points out: the fact that certain insights are likely to be accurate does not mean that all insights will be. as andy clark (2016, p. 288) puts it, priors are always both constraining and enabling. by filtering out evidence that contradicts them, they can impair access to certain facts, but they can also prevent implausible hypotheses from being considered and accepted. a temporary loss of confidence in the brain’s prior knowledge does not imply that all resulting beliefs will be veridical, nor that they will all be false: “[…] when it comes to propositional knowledge about our own mind, psychedelics facilitate both genuine insights and placebo insights, and there is no general formula for telling the two apart” (letheby, 2021, p. 171). according to letheby, the only way to assess the accuracy of these insights is through sober integration after the psychedelic session. letheby’s emphasis is on beliefs about the self. however, the same argument can be applied to self-independent beliefs. an example of priors constraining our knowledge about the external world is provided by popular explanation of the hollow mask illusion, according to which a strong reliance on the assumption “faces are convex” causes our brain to ignore evidence to the contrary and perceive a rotating concave mask as popping out. according to this explanation, a loss of confidence in the brain’s top-down priors causes patients with schizophrenia (dima et al., 2010) or people on psychedelics (millière et al., 2018) to perceive the mask more accurately than controls. hence, letheby’s argument for the possibility of obtaining knowledge-that about the self through psychedelic use can be generalised as follows: 1. decreasing the weighting of priors can increase the probability of accurately apprehending certain facts (from general predictive processing theory). 2. psychedelic administration temporarily decreases the weighting of priors (from the rebus model and the predictive self-binding theory). 3. therefore, psychedelic administration can increase the probability of accurately apprehending certain facts. if the argument generalises, the special status reserved by letheby for self-related beliefs seems unjustified. of all new insights acquired after a psychedelic session, some of those about the external world may be veridical, and some of those about ourselves may be false. imagine your friend maria (letheby, 2021, p. 162), who caporuscio, c. (2022). belief now, true belief later: the epistemic advantage of self-related insights in psychedelic-assisted therapy. philosophy and the mind sciences, 3, 6. https://doi.org/10.33735/phimisci.2022.9310 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9310 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chiara caporuscio 4 claims to have gained propositional knowledge from a psychedelic experience. when you ask her what she has learned, she responds by listing three new beliefs she has gained. first, she has discovered some aspects of herself that were hidden from her before, including a deep desire for human connection. secondly, she has discovered something she feels is a deep truth about another person’s mind: she realises that the actions of a family member that she had always thought stemmed from selfishness and greed are actually motivated by anxiety and insecurity. thirdly, she has gained a metaphysical insight, namely that all existence in time is equally real. all of these insights are plausible, and all of them have the potential to cause lasting psychological benefits to maria. but is one of these insights more likely to bring epistemic benefits, and in particular new propositional knowledge? in other words, are there epistemic differences between psychedelicmediated insights about the self, about another person, and about the external world? i argue that relevant epistemic differences between self-, other-, and worldrelated insights cannot be found if we treat the epistemic status of the psychedelic experience as purely dependent on acquiring new knowledge of pre-existing facts. maria’s insight about herself and her two insights about self-independent objects are all likely to have been caused by a weakening of her prior beliefs, which has allowed her to see herself and others in a new light. but this is no guarantee that any of the newly acquired beliefs is true. thus, letheby’s optimism about the possibility of gaining propositional knowledge about oneself rather than the external world seems unmotivated. psychedelics put maria in an epistemically promising but uncertain position, where she is likely to have acquired false beliefs about the world, other people and herself that she can only tentatively differentiate from the true ones by carefully scanning them for plausibility after her session. if it is the case that self-related insights have no firmer epistemic grounding than self-independent ones, letheby’s rejection of the cdo is considerably weakened. at least some of the self-related insights driving psychological improvement are probably still comforting delusions, and the epistemic status of psychedelic therapy is only partially rescued by indirect epistemic benefits and post-session evaluation. there are obvious epistemic faults in a therapeutic method that will sometimes work by convincing a person who lies continually that they are honest and dependable, for example. and because letheby accepts the cdo’s premise that epistemically bad means ethically bad, the possibility of comforting delusions about the self is a reason to refrain from recommending psychedelic therapy. i propose an alternative reason why maria might be in a better epistemic position regarding her insights and newly acquired beliefs about herself than regarding those about the external world. more precisely, i will argue that one element is underexplored in letheby’s otherwise compelling picture: namely, the recognition of the psychedelic experience and the subsequent integration process as opportunities not only to apprehend certain facts about oneself but also to actively shape and redetermine those facts through exploratory thinking and behaviour. caporuscio, c. (2022). belief now, true belief later: the epistemic advantage of self-related insights in psychedelic-assisted therapy. philosophy and the mind sciences, 3, 6. https://doi.org/10.33735/phimisci.2022.9310 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9310 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org belief now, true belief later 5 moran (2001) and mcgeer (2008, 1996; mcgeer & pettit, 2002) both talk about the the power of first-person authority to shape the self. in their view, the authority of self-knowledge derives not from a passive, error-free ability to detect our mental states, but from our capacity to regulate our thoughts and actions in accordance with the claims we make about ourselves. for example, the thought “i hate laundromats” might contribute to creating and sustaining the declared hate for laundromats, thus turning into a self-fulfilling prophecy (schwitzgebel, 2011). deciding as a child that your favourite colour is blue might causally influence your choice of outfits, objects, and self-expression, feeding into a growing appreciation for the colour blue. according to mcgeer (2008), self-shaping is not only a capacity, but a moral responsibility: in order to be intelligible as rational agents, we owe it to ourselves and others to behave and think in the ways we declare we do. our core beliefs about ourselves can (and should) turn into self-fulfilling prophecies, providing us with familiar patterns of expression and behaviour, allowing us to act as predictable agents, and fulfilling the expectations that we have created in ourselves and others. however, this also means we might end up stuck in our core beliefs about ourselves. in pathological cases, this is extremely problematic. a depressive patient who believes they are unable to find pleasure in going outside will behave in accordance with this belief—or, in terms borrowed from the active inference framework, they will sample their environment for evidence that will confirm the belief and avoid evidence that will disconfirm it (ramstead et al., 2020). not only will this behaviour reinforce the belief in an endless loop, but it may also make it true: by committing to act in line with the belief that they are incapable of getting out of bed and having a nice day, they will make it impossible for themselves to enjoy being outside. in his book, letheby argues that the psychedelic experience allows for a relaxing and rewiring of self-related priors, thus allowing patients with negative self models to access evidence about themselves that was previously hidden because it conflicted with those models. this is likely to be true, but not the whole story: if damaging self-related beliefs are (at least to some extent) self-fulfilling, losing confidence in them allows psychedelic users not only to access previously hidden evidence but also to create new evidence by acting and thinking in new, unconstrained ways. imagine, in line with moran and mcgeer’s examples, that by thinking of herself and presenting to others as self-sufficient, independent, and emotionally distant, maria has been committing for most of her adult life to act and think in a way that would fulfil that expectation. because of this, at the time of her psychedelic experience, evidence of a deep desire for human connection is not only hidden by her self model, but scarce. she has been leading a life of voluntary isolation, keeping distance from her family and friends, and she has rarely been imagining or wishing for a different lifestyle. caporuscio, c. (2022). belief now, true belief later: the epistemic advantage of self-related insights in psychedelic-assisted therapy. philosophy and the mind sciences, 3, 6. https://doi.org/10.33735/phimisci.2022.9310 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9310 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chiara caporuscio 6 however, by relaxing her belief about herself through psychedelic use, she is able to temporarily break free from her commitment to act and think in line with it. she can exploratorily entertain new thoughts, imagine new modal truths about how her life could be (letheby, 2015; 2021), and test out different behaviours, like opening up and expressing closeness to her trip sitters or companions. these new thoughts and behaviours then contribute to the evidence for her new, emerging insight that at her core is a deep desire for human connection. if maria comes out of her trip with strong confidence in this belief then this confidence will, in turn, exercise its power to shape the self. during the sober integration period, maria will consolidate her belief by maintaining and incorporating into her life the thoughts and behaviours she tentatively explored during the psychedelic experience. contrast this with the knowledge-acquisition story described by letheby in chapter 8 of his book. maria is not only detecting previously unnoticed patterns of her prior behaviour that indicate her deep desire for human connection; she is (during the psychedelic experience) exploring new patterns of thought and action in line with her newly formed belief, and (during the subsequent integration period) sticking with them. let us return to the epistemic status of self-related insights mediated by psychedelics. two questions might be asked: 1) has maria acquired new propositional knowledge? 2) is she in a better epistemic position regarding her insights about herself than regarding those about the external world? her epistemic position regarding her new self-characterisation is an interesting one. prior to the psychedelic experience, her position would probably have been inaccurate: she had a strong belief that she was happy by herself and never had thoughts or carried out behaviours indicating that she was longing for connection. however, months after the experience, maria finds herself enjoying the company of others and letting barriers down with her loved ones, as she has learned to do during her trip and has consolidated the habit of doing during the integration period. belief first, true belief later: she might have ended her trip with an illusory insight, but months later this has turned into new propositional knowledge. does the same apply to her self-unrelated belief that the actions of her family member were motivated by anxiety and insecurity? imagine this second insight was also, prior to the experience, false: this person’s actions were in fact guided by selfishness and greed. does this inaccurate belief have any self-fulfilling capacity? maybe, very indirectly and to a much lesser extent: by reconnecting with the other person thanks to her favourable disposition, maria might be able to partially influence their ways of acting and thinking in line with her belief. however, this is likely to be much harder and only successful in specific circumstances (for one, the other person must be receptive and willing to connect). finally, beliefs whose object is completely mind-independent, like maria’s metaphysical insight about the nature of time, are even more clearly not self-fulfilling. maria’s capacity to shape and influence self-independent objects and people is not comparable to her capacity to shape and influence her own mental states and behaviour. caporuscio, c. (2022). belief now, true belief later: the epistemic advantage of self-related insights in psychedelic-assisted therapy. philosophy and the mind sciences, 3, 6. https://doi.org/10.33735/phimisci.2022.9310 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9310 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org belief now, true belief later 7 the self-fulfilling character of these beliefs has some noteworthy implications. first, my account suggests a more prominent role for the integration period and psychological support following the psychedelic experience, which aligns with existing evidence that these elements are important predictors of successful outcomes (johnson et al., 2008; teixeira et al., 2022). in letheby’s account, follow-up sessions are important for epistemic purposes because patients can soberly scan insights for plausibility and distinguish accurate ones from placebos with the help of a trained therapist. while this is likely true, i argue that direct truth-testing is not the main mechanism that renders the months following the experience crucial in determining epistemic benefits. the epistemic success of psychedelic therapy happens in two steps: the trip is for discovery, and its aftermath is for consolidation and commitment. during the trip itself, self-related beliefs are relaxed, allowing patients to momentarily escape the self-fulfilling effects of damaging self models, explore new and healthier ways of thinking and acting, and consequently acquire new insights about themselves. however, the process of self-shaping involves changing one’s behaviour and thinking over much longer timescales than six or twelve hours. this is why most of it happens while sober. with the help of trained therapists, patients can consider which beliefs about themselves that have been tested out during the psychedelic experience are worth committing to, and plan for ways to change their behaviour to integrate and fulfil these new (or newly appraised) beliefs. if they do not adapt their lifestyle, thoughts, and actions to conform to their trip-induced insights, their newly formed beliefs are unlikely to turn out to be true and will probably be abandoned. the second consequence of the account i have presented is that while the self-fulfilling nature of self-characterisations helps mitigate the epistemic risks of psychedelic therapy, it also carries serious psychological risks. not only is there nothing intrinsically true about psychedelic-mediated insights, there is also nothing intrinsically positive: while some people report feelings of bliss and newfound self-acceptance, others experience anxious spiralling, negative emotions, and selfdeprecating thoughts. if positive insights can actively shape users’ future mental states and behaviour, so can psychologically damaging ones. in this account, a bad trip triggering the new belief that one is deeply incapable of being happy carries more than just epistemic risks. in order to prevent damaging self models from being accepted and incorporated as a result of a negative psychedelic experience, it is extremely important to take the danger seriously and carefully assess whether and to what extent it can be reduced. i suggest that, in order to mitigate the risk, we should focus on the following two aspects. first, appropriate preparation before the psychedelic experience, moderate dosage and controlled set and setting can significantly reduce the chances of a bad trip. secondly, psychological and behavioural support during the integration period can help an individual not only to integrate psychologically and epistemically beneficial insights into their life but also to recognise and discard damaging ones. caporuscio, c. (2022). belief now, true belief later: the epistemic advantage of self-related insights in psychedelic-assisted therapy. philosophy and the mind sciences, 3, 6. https://doi.org/10.33735/phimisci.2022.9310 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9310 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chiara caporuscio 8 psychedelics can facilitate a state of temporary flexibility where damaging self models can be discarded and new ones can take their place, be consolidated and powerfully shape one’s future self. old beliefs are abandoned, and new beliefs are accepted and turned into reality. however, this increased flexibility is neutral in itself: it is what happens before, during and after the session that really determines whether the newly adopted models will be harmful or therapeutic. acknowledgments funded by the german research foundation (dfg) 337619223 / rtg2386. references clark, a. (2016). surfing uncertainty: prediction, action, and the embodied mind. oxford university press. https://doi.org/ 10.1093/acprof:oso/9780190217013.001.0001 dima, d., dietrich, d. e., dillo, w., & emrich, h. m. (2010). impaired top-down processes in schizophrenia: a dcm study of erps. neuroimage, 52(3), 824–832. https://doi.org/10.1016/j.neuroimage.2009.12.086 johnson, m., richards, w., & griffiths, r. (2008). human hallucinogen research: guidelines for safety. journal of psychopharmacology, 22(6), 603–620. https://doi.org/10.1177/0269881108093587 letheby, c. (2015). the philosophy of psychedelic transformation. journal of consciousness studies, 22(9-10), 170–193. letheby, c. (2021). philosophy of psychedelics. oxford university press. mcgeer, v. (2008). the moral development of first-person authority. european journal of philosophy, 16(1), 81–108. https: //doi.org/10.1111/j.1468-0378.2007.00266.x mcgeer, v. (1996). is “self-knowledge” an empirical problem? renegotiating the space of philosophical explanation. the journal of philosophy, 93(10), 483–515. https://doi.org/10.2307/2940837 mcgeer, v., & pettit, p. (2002). the self-regulating mind. language & communication, 22(3), 281–299. https://doi.org/10.1 016/s0271-5309(02)00008-3 millière, r., carhart-harris, r. l., roseman, l., trautwein, f.-m., & berkovich-ohana, a. (2018). psychedelics, meditation, and self-consciousness. frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018.01475 moran, r. (2001). authority and estrangement: an essay on self-knowledge. nj: princeton university press. noorani, t., garcia-romeu, a., swift, t. c., griffiths, r. r., & johnson, m. w. (2018). psychedelic therapy for smoking cessation: qualitative analysis of participant accounts. journal of psychopharmacology, 32(7), 756–769. https://doi.or g/10.1177/0269881118780612 ramstead, m. j. d., wiese, w., miller, m., & friston, k. j. (2020). deep neurophenomenology: an active inference account of some features of conscious experience and of their disturbance in major depressive disorder. [preprint]. http://philsciarchive.pitt.edu/18377/. schwitzgebel, e. (2011). introspection, what? in smithies, d., & stoljar, d. (eds.) introspection and consciousness (pp. 29–48). oxford university press. teixeira, p. j., johnson, m. w., timmermann, c., watts, r., erritzoe, d., douglass, h., kettner, h., & carhart-harris, r. l. (2022). psychedelics and health behaviour change. journal of psychopharmacology, 36(1), 12–19. https://doi.org/10.1 177/02698811211008554 watts, r., day, c., krzanowski, j., nutt, d., & carhart-harris, r. (2017). patients’ accounts of increased “connectedness” and “acceptance” after psilocybin for treatment-resistant depression. journal of humanistic psychology, 57(5), 520–564. https://doi.org/10.1177/0022167817709585 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. caporuscio, c. (2022). belief now, true belief later: the epistemic advantage of self-related insights in psychedelic-assisted therapy. philosophy and the mind sciences, 3, 6. https://doi.org/10.33735/phimisci.2022.9310 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1093/acprof:oso/9780190217013.001.0001 https://doi.org/10.1093/acprof:oso/9780190217013.001.0001 https://doi.org/10.1016/j.neuroimage.2009.12.086 https://doi.org/10.1177/0269881108093587 https://doi.org/10.1111/j.1468-0378.2007.00266.x https://doi.org/10.1111/j.1468-0378.2007.00266.x https://doi.org/10.2307/2940837 https://doi.org/10.1016/s0271-5309(02)00008-3 https://doi.org/10.1016/s0271-5309(02)00008-3 https://doi.org/10.3389/fpsyg.2018.01475 https://doi.org/10.1177/0269881118780612 https://doi.org/10.1177/0269881118780612 https://doi.org/10.1177/02698811211008554 https://doi.org/10.1177/02698811211008554 https://doi.org/10.1177/0022167817709585 https://doi.org/10.33735/phimisci.2022.9310 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org serotonin, predictive processing, and psychedelics serotonin, predictive processing, and psychedelics matteo colomboa� (m.colombo@uvt.nl) abstract letheby’s philosophy of psychedelics relies on predictive processing to try and find unifying explanations relevant to understanding how serotonergic psychedelics work in psychiatric therapy, what subjective experiences are associated with their use and whether such experiences are epistemically defective. but if predictive processing lacks genuinely explanatory unifying power, letheby’s account of psychedelic therapy risks being unwarranted. in this commentary, i motivate this worry and sketch an alternative interpretation of psychedelic therapy within the reinforcement learning framework. keywords predictive processing ∙ reinforcement learning ∙ reinforcement learning ∙ unification this article is part of a symposium on chris letheby’s book “philosophy of psychedelics” (oup 2021), edited by chiara caporuscio and sascha benjamin fink. chris letheby’s (2021b) philosophy of psychedelics relies on predictive processing (pp) to try to find unifying explanations that could bridge computational function, neural mechanism and phenomenological analysis relevant to understanding how serotonergic psychedelics work in psychiatric therapy, which subjective experiences are associated with their use and whether these experiences are epistemically defective. but if pp lacks genuinely explanatory (counterfactual supporting) unifying power, then letheby’s account of psychedelics will rest on fragile bridges between computation, mechanism and phenomenology; and with such fragile bridges, one might worry that letheby’s interpretation of psychedelic therapy in terms of pp is unwarranted. in this commentary, i motivate this worry and sketch an alternative interpretation of psychedelic therapy within a computational framework different from pp. atilburg university colombo, m. (2022). serotonin, predictive processing, and psychedelics. philosophy and the mind sciences, 3, 7. https://doi.org/10.33735/phimisci.2022.9320 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-2251-2248 https://doi.org/10.33735/phimisci.2022.9320 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org matteo colombo 2 virtually all researchers working within the pp framework agree that some psychiatric conditions originate in and are maintained by the malfunctioning of precision-weighted prediction error signalling (colombo & fabry, 2021; see e.g., corlett et al., 2009). this disruption would result in too much (or too little) weight being given to sensory evidence compared with prior “beliefs”; and due to this misallocation, people would form delusional beliefs that could explain away overly salient perceptions (or would tend to perceive only what they are expecting to perceive). in particular, pp models such as carhart-harris & friston’s (2019) “relaxed beliefs under psychedelics” and letheby & gerrans’s (2017) self-binding model of psychedelic ego dissolution share the idea that serotonergic psychedelics disrupt “predictive models implemented in high-level cortical networks… and this disruption creates an opportunity to revise these models in beneficial ways” (letheby, 2021a, p. 9). for letheby, psychedelics weaken the precision of maladaptive selfrepresentations (i.e., mental representations of one’s self), which are often associated with anxiety disorders, depression and addiction; and this weakening would provide psychiatrists with an opportunity for therapeutic interventions aimed at “changing how patients see themselves and relate to their own minds and lives” (letheby, 2021a, p. 12). one set of questions relevant to evaluate the empirical adequacy of this suggestion, as well as the extent to which pp provides us with a genuinely unifying explanatory bridge for understanding the workings of psychedelic therapy is this: what’s the function of serotonin, which is the primary target of psychedelics? what kind of information does serotonergic neuromodulation carry? does serotonin implement precision estimates, precision-weighted high-level priors or some other aspects of precision-weighted computing? researchers relying on pp have made different suggestions about these questions. some posit that the computation of precision estimates is implemented by the activity of dopaminergic neurons (fitzgerald et al., 2015; friston et al., 2014). others suggest that increased acetylcholine activity enhances sensory precision by amplifying bottom-up signals when external stimuli are predictable (moran et al., 2013); other researchers that both noradrenaline and dopamine are implicated in implementing estimates of the precision of sensory evidence and in tracking the volatility of the environment (yon & frith, 2021); and others yet hypothesise that precision estimates are computed and manipulated by the orchestrated activity of dopamine, acetylcholine and serotonin neurons (adams et al., 2013; clark, 2013). these different suggestions within the pp literature could be taken to indicate that computational posits like precision-weighted high-level priors are somehow illdefined. but this conclusion is too quick, as the notions of precision and high-level prior are precisely and explicitly defined within many particular pp models (aka active inference models) of particular psychiatric phenomena (see e.g., lawson et al., 2017). more plausibly, the plurality of suggestions within the pp literature could be taken to highlight that there is a plurality of pathways and mechanisms implecolombo, m. (2022). serotonin, predictive processing, and psychedelics. philosophy and the mind sciences, 3, 7. https://doi.org/10.33735/phimisci.2022.9320 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9320 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org serotonin, predictive processing, and psychedelics 3 menting precision-weighted computing in the brain. this plurality of pathways and mechanisms for precision-weighted computing is consistent with the evidence from the neuroimaging studies that letheby (2021b, chapter 5) and letheby & gerrans (2017) review, which indicates the existence of an association between psychedelic therapy and changes in the default mode and salience networks in the brain. it is consistent also with the molecular neuroplasticity theory (letheby, 2021b, chapter 4) discusses, whereby serotonin and particularly 5-ht2a receptors (one among a dozen known subtypes of serotonin receptors) have many distinct effects at several spatial and temporal scales—for example, serotonin activity influences both excitatory and inhibitory neurotransmission in neural circuits in the forebrain and midbrain, and modulates the release of other neurotransmitters like dopamine and acetylcholine, which impact neural plasticity and play various roles in learning (hoyer et al., 2002; jacobs & fornal, 1999). while precision-weighted high-level priors might be implemented across different structures of the nervous system, the current lack of knowledge of how they are actually implemented should not suggest that a computational model positing precision-weighted representations is untestable or that serotonin cannot contribute to implement those representations. but the lack of consistency within the pp literature about the computational function of serotonin and the putative physical realizers of precision estimates and precision-weighted high-level priors could also indicate that the bridge between computation, mechanism and phenomenology offered by pp is gappy and with limited explanatory power, which would call into question letheby’s choice to use pp to work out a unified, explanatory account of psychedelic therapy. the worry here is that pp offers no more than re-descriptions of mental and neural phenomena. while these re-descriptions might be consistent with known empirical data, they would fall short as explanatory accounts that are experimentally tested and empirically evaluated against alternative hypotheses grounded in different theoretical frameworks (colombo & wright, 2017; litwin & miłkowski, 2020). one such alternative framework is reinforcement learning (rl), which is increasingly popular in psychiatry (colombo & heinz, 2019; maia & frank, 2011). compared to pp models, rl models have been submitted to experimental testing more frequently and more stringently, and they have proved themselves to enjoy a substantial degree of genuinely explanatory unifying power in the sciences of mind and brain (colombo, 2014). like pp models, any rl model of a mental phenomenon is incomplete, abstracts away from mechanistic details and its supporting evidence is mixed. unlike pp models, however, rl modelling is grounded in the notions of value and reward, which pp models eliminate reinterpreting planning and decision-making as forms of inference (colombo, 2016). now, the posits of value and reward in some rl models have been mapped onto various aspects of serotonin activity in algorithmic hypotheses about how the brain generates affective dynamics that drive learning and decision making colombo, m. (2022). serotonin, predictive processing, and psychedelics. philosophy and the mind sciences, 3, 7. https://doi.org/10.33735/phimisci.2022.9320 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9320 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org matteo colombo 4 (dayan, 2012; doya, 2002). specifically, according to one of these hypotheses, serotonin represents expectations about negative outcomes (i.e., punishments or negative rewards) and, on the basis of these representations, it would compute appropriate inhibitory motor responses (dayan & huys, 2009). if this hypothesis is true, increased serotonin activity will promote the downregulation of impulsive emotional reactions to expected punishment or provocation, boosting patience and coping with anxiety. while this hypothesis has received some empirical corroboration (crockett et al., 2009), more recent theoretical and empirical work suggests that serotonin plays more complicated functions, some in opposition and some in synergy with the functions played by dopamine in learning and decision making, depending on the agent’s expectations about the statistical structure and controllability of the environment (boureau & dayan, 2011; liu et al., 2020). according to another hypothesis developed within the rl framework, serotonin computes “the time and resources available for action, learning and development” (doya et al., 2021, p. 121). if this hypothesis is true, then serotonin activity is involved in computing multi-dimensional representations of the temporal horizon and the value of different outcomes at different time scales in different cognitive and biological tasks. higher serotonin activity in a decision-making task would signal that there is enough time available for deeper deliberation, more explorative choices and longer-term predictions of the likely outcomes of different choices (doya et al., 2021, table 1). both rl hypotheses just sketched dovetail computational function, algorithmic processes and neural mechanism, while they also integrate affective and temporal dimensions that are central to the phenomenology of psychedelic experiences (letheby, 2021b, chapter 3). if we exclude the ecumenical possibility that separate networks of serotonin neurons and different subtypes of serotonin receptors perform different kinds of computational functions, then those rl hypotheses are not obviously consistent with the hypothesis that serotonin implements precision estimates and is involved in the weakening of the precision of high-level priors. but how would these hypotheses, if true, reconfigure letheby’s key idea that “psychedelic therapy works mainly by changing self-representation” (letheby, 2021b, p. 6)? one possible answer is that the truth of either of these rl hypotheses would not change anything in letheby’s key idea, because time, value and affect are core aspects of self-representations and self-representations might be implemented in high-level cortical networks encoding “the most fundamental assumptions concerning self, [but also] space, time, causality, and so forth: the basic parameters of our phenomenal worlds” (letheby, 2021a, p. 8). yet, if different computational hypotheses about serotonin are all compatible with letheby’s idea that psychedelic therapy mainly works by changing self-representations, then one may conclude that this idea is too generic to advance existing understanding of psychedelic therapy, or of serotonin function in relation to self-representations. a different answer is that the truth of either of these rl hypotheses indicates that psychedelic therapy mainly works not by changing self-representations, but colombo, m. (2022). serotonin, predictive processing, and psychedelics. philosophy and the mind sciences, 3, 7. https://doi.org/10.33735/phimisci.2022.9320 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9320 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org serotonin, predictive processing, and psychedelics 5 by changing one’s representations of time or value, where such changes would increase or lower, for example, one’s levels of impulsivity and modulate responses to threat—e.g., if the available time is plenty, then one may cope or avoid threats, rather than freeze or panic (seo et al., 2019)—or would enhance or lower aversive affective reactions to anticipated negative outcomes for self or others (crockett et al., 2010). so, if either of the two rl hypotheses i have sketched is correct, then psychedelic therapies may not mainly “induce different forms of self-modelling” (letheby, 2021a, p. 10), but would mainly change the ways humans—and nonhuman animals, too—represent time, value and respond to expected punishment. in summary, one might worry that compared to a rl approach pp lacks genuinely explanatory unifying power and provides us with only a gappy bridge between computational function, neural mechanism and phenomenology. rl modelling foregrounds temporal experience, threat response and valuation as general mental functions that serotonin would modulate and that are impaired in anxiety disorders, depression and addiction. interpreting the effects of serotonergic psychedelics through the lenses of rl highlights that psychedelic therapy would be effective mainly when it changes patients’ maladaptive affective responses to anticipated punishment, temporal representations and value. references adams, r. a., stephan, k. e., brown, h. r., frith, c. d., & friston, k. j. (2013). the computational anatomy of psychosis. frontiers in psychiatry, 4(47). https://doi.org/10.3389/fpsyt.2013.00047 boureau, y. l., & dayan, p. (2011). opponency revisited: competition and cooperation between dopamine and serotonin. neuropsychopharmacology, 36, 74–97. https://doi.org/10.1038/npp.2010.151 carhart-harris, r. l., & friston, k. j. (2019). rebus and the anarchic brain: toward a unified model of the brain action of psychedelics. pharmacological reviews, 71(3), 316–344. https://doi.org/10.1124/pr.118.017160 clark, a. (2013). the many faces of precision. replies to commentaries on “whatever next? neural prediction, situated agents, and the future of cognitive science.” frontiers in psychology, 4, 270. https://doi.org/10.3389/fpsyg.2013.00270 colombo, m. (2014). deep and beautiful. the reward prediction error hypothesis of dopamine. studies in history and philosophy of science, part c. studies in history and philosophy of biological and biomedical sciences, 45, 57–67. https: //doi.org/10.1016/j.shpsc.2013.10.006 colombo, m. (2016). social motivation in computational neuroscience: or, if brains are prediction machines, then the humean theory of motivation is false. in the routledge handbook of philosophy of the social mind (pp. 336–356). routledge. colombo, m., & fabry, r. e. (2021). underlying delusion: predictive processing, looping effects, and the personal/subpersonal distinction. philosophical psychology, 1–27. https://doi.org/10.1080/09515089.2021.1914828 colombo, m., & heinz, a. (2019). explanatory integration, computational phenotypes, and dimensional psychiatry: the case of alcohol use disorder. theory & psychology, 29(5), 697–718. https://doi.org/10.1177/0959354319867392 colombo, m., & wright, c. (2017). explanatory pluralism: an unrewarding prediction error for free energy theorists. brain and cognition, 112, 3–12. https://doi.org/10.1016/j.bandc.2016.02.003 corlett, p. r., frith, c. d., & fletcher, p. c. (2009). from drugs to deprivation: a bayesian framework for understanding models of psychosis. psychopharmacology, 206(4), 515–530. https://doi.org/10.1007/s00213-009-1561-0 crockett, m. j., clark, l., hauser, m. d., & robbins, t. w. (2010). serotonin selectively influences moral judgment and behavior through effects on harm aversion. proceedings of the national academy of sciences, 107(40), 17433–17438. https://doi.org/10.1073/pnas.1009396107 colombo, m. (2022). serotonin, predictive processing, and psychedelics. philosophy and the mind sciences, 3, 7. https://doi.org/10.33735/phimisci.2022.9320 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.3389/fpsyt.2013.00047 https://doi.org/10.1038/npp.2010.151 https://doi.org/10.1124/pr.118.017160 https://doi.org/10.3389/fpsyg.2013.00270 https://doi.org/10.1016/j.shpsc.2013.10.006 https://doi.org/10.1016/j.shpsc.2013.10.006 https://doi.org/10.1080/09515089.2021.1914828 https://doi.org/10.1177/0959354319867392 https://doi.org/10.1016/j.bandc.2016.02.003 https://doi.org/10.1007/s00213-009-1561-0 https://doi.org/10.1073/pnas.1009396107 https://doi.org/10.33735/phimisci.2022.9320 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org matteo colombo 6 crockett, m. j., clark, l., & robbins, t. w. (2009). reconciling the role of serotonin in behavioral inhibition and aversion: acute tryptophan depletion abolishes punishment-induced inhibition in humans. journal of neuroscience, 29(38), 11993–11999. https://doi.org/10.1523/jneurosci.2513-09.2009 dayan, p. (2012). twenty-five lessons from computational neuromodulation. neuron, 76(1), 240–256. https://doi.org/10. 1016/j.neuron.2012.09.027 dayan, p., & huys, q. j. (2009). serotonin in affective control. annual review of neuroscience, 32, 95–126. https://doi.org/ 10.1146/annurev.neuro.051508.135607 doya, k. (2002). metalearning and neuromodulation. neural networks, 15(4-6), 495–506. https://doi.org/10.1016/s08936080(02)00044-8 doya, k., miyazaki, k. w., & miyazaki, k. (2021). serotonergic modulation of cognitive computations. current opinion in behavioral sciences, 38, 116–123. https://doi.org/10.1016/j.cobeha.2021.02.003 fitzgerald, t. h., dolan, r. j., & friston, k. (2015). dopamine, reward learning, and active inference. frontiers in computational neuroscience, 9(136). https://doi.org/10.3389/fncom.2015.00136 friston, k., schwartenbeck, p., fitzgerald, t., moutoussis, m., behrens, t., & dolan, r. j. (2014). the anatomy of choice: dopamine and decision making. philosophical transactions of the royal society b: biological sciences, 369(20130481). https://doi.org/10.1098/rstb.2013.0481 hoyer, d., hannon, j. p., & martin, g. r. (2002). molecular, pharmacological and functional diversity of 5-ht receptors. pharmacology biochemistry and behavior, 71(4), 533–554. https://doi.org/10.1016/s0091-3057(01)00746-8 jacobs, b. l., & fornal, c. a. (1999). activity of serotonergic neurons in behaving animals. neuropsychopharmacology, 21(1), 9–15. https://doi.org/10.1016/s0893-133x(99)00012-3 lawson, r. p., mathys, c., & rees, g. (2017). adults with autism overestimate the volatility of the sensory environment. nature neuroscience, 20(9), 1293–1299. https://doi.org/10.1038/nn.4615 letheby, c. (2021a). naturalistic entheogenics: précis of philosophy of psychedelics. psyarxiv. https://doi.org/10.31234/osf. io/ztewb letheby, c. (2021b). philosophy of psychedelics. oxford university press. letheby, c., & gerrans, p. (2017). self unbound: ego dissolution in psychedelic experience. neuroscience of consciousness, 1. https://doi.org/10.1093/nc/nix016 litwin, p., & miłkowski, m. (2020). unification by fiat: arrested development of predictive processing. cognitive science, 44(7), 12867. https://doi.org/10.1111/cogs.12867 liu, z., lin, r., & luo, m. (2020). reward contributions to serotonergic functions. annual review of neuroscience, 43, 141–162. https://doi.org/10.1146/annurev-neuro-093019-112252 maia, t. v., & frank, m. j. (2011). from reinforcement learning models to psychiatric and neurological disorders. nature neuroscience, 14(2), 154–162. https://doi.org/10.1038/nn.2723 moran, r. j., campo, p., symmonds, m., stephan, k. e., dolan, r. j., & friston, k. j. (2013). free energy, precision and learning: the role of cholinergic neuromodulation. journal of neuroscience, 33(19), 8227–8236. https://doi.org/10.1523/ jneurosci.4255-12.2013 seo, c., guru, a., jin, m., ito, b., sleezer, b. j., ho, y. y., & warden, m. r. (2019). intense threat switches dorsal raphe serotonin neurons to a paradoxical operational mode. science, 363(6426), 538–542. https://doi.org/10.1126/science.aau8722 yon, d., & frith, c. d. (2021). precision and the bayesian brain. current biology, 31(17), 1026–1032. https://doi.org/10.1016/ j.cub.2021.07.044 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. colombo, m. (2022). serotonin, predictive processing, and psychedelics. philosophy and the mind sciences, 3, 7. https://doi.org/10.33735/phimisci.2022.9320 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1523/jneurosci.2513-09.2009 https://doi.org/10.1016/j.neuron.2012.09.027 https://doi.org/10.1016/j.neuron.2012.09.027 https://doi.org/10.1146/annurev.neuro.051508.135607 https://doi.org/10.1146/annurev.neuro.051508.135607 https://doi.org/10.1016/s0893-6080(02)00044-8 https://doi.org/10.1016/s0893-6080(02)00044-8 https://doi.org/10.1016/j.cobeha.2021.02.003 https://doi.org/10.3389/fncom.2015.00136 https://doi.org/10.1098/rstb.2013.0481 https://doi.org/10.1016/s0091-3057(01)00746-8 https://doi.org/10.1016/s0893-133x(99)00012-3 https://doi.org/10.1038/nn.4615 https://doi.org/10.31234/osf.io/ztewb https://doi.org/10.31234/osf.io/ztewb https://doi.org/10.1093/nc/nix016 https://doi.org/10.1111/cogs.12867 https://doi.org/10.1146/annurev-neuro-093019-112252 https://doi.org/10.1038/nn.2723 https://doi.org/10.1523/jneurosci.4255-12.2013 https://doi.org/10.1523/jneurosci.4255-12.2013 https://doi.org/10.1126/science.aau8722%20 https://doi.org/10.1016/j.cub.2021.07.044 https://doi.org/10.1016/j.cub.2021.07.044 https://doi.org/10.33735/phimisci.2022.9320 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org précis of thinking and perceiving. précis of thinking and perceiving dustin stokesa(dustin.stokes@utah.edu) keywords cognitive architecture ∙ epistemology of perception ∙ malleability of perception ∙ perceptual learning ∙ perceptual expertise ∙ modularity∙ theory-ladenness∙ perceptual content ∙ virtue and understanding this article is part of a symposium on dustin stokes’s book “thinking and perceiving” (routledge 2021), edited by regina fabry and sascha benjamin fink. 1 introduction consider three types of examples of possible cognitive-perceptual phenomena. convinced of a particular chemical theory, georg might claim to see the phlogiston being released from the wood burning in the fireplace. my expectation that you will be in a foul mood at the party might influence me to “read” your body language in a way that seems to confirm my expectations. “i knew it” i might think “they really are in a foul mood.” in both examples of the first type of case, it appears that one’s background beliefs or expectations influence one’s visual experience. the first example is familiar to philosophers of science as a case of the theory-ladenness of scientific observation. the second, everyday example is familiar to philosophers of mind as a possible case of the cognitive penetration of perception. one unifying feature of these anecdotal examples (despite differences in context) is that they appear to have a circular structure, where a background cognitive state c influences a sensory perceptual state p and, in turn, that cognitively influenced perceptual state reinforces or somehow further justifies the antecedent cognitive state c. this, many have worried, isn’t just circular but viciously so. for many epistemologies, given the epistemic importance of sensory perception, this is cognitive bias of a most worrisome kind. a university of utah, salt lake city. stokes, d. (2023). précis of thinking and perceiving. philosophy and the mind sciences, 4, 10. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/ 0000-0001-5341-3157 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dustin stokes 2 in the second type of case, subjects are studied in experimentally controlled laboratory circumstances. subjects are given some perceptual task—identify the colour of an object, adjust a report item to match a target stimulus in size—and for the relevant stimuli the subjects’ beliefs or concepts appear to influence their task performance. in a bit more detail, here is one example. subjects are asked to perform a fairly rapid visual search task. in one trial structure, subjects are instructed of the target item of their search (e.g. motorbike), then briefly view a search screen containing four items, and then report whether the target was present or absent. in some trials, a semantically related, but non-visually resembling item appeared in the search screen (e.g. a helmet). in such trials, semantically related items are better recalled and identified, they sometimes hinder target identification, and reduce recall of unrelated distractor images. further, eye-tracking results suggest that semantically related items modulate saccadic eye movement. here it appears that rapid visual attention is being grabbed in a way that is independent of the search task, but that depends on cognitive information—relations of meaning, stored in long term memory—and this affects the perception of the subjects (moores et al., 2003). similar results were found when items with names homophonous to the target image appeared in the search screen (meyer et al., 2007). in the third type of case, subjects are studied both in experimentally controlled lab circumstances, and in more ecologically valid circumstances. and in these studies, performance of experts from specialized domains is contrasted with performance of novice or naïve subjects relative to that domain. a wide range of domains have been studied—radiology, surgery, forensic examination, bird watching, elite athletics, to name just a few—using a range of behavioral, neurophysiological, and computational measures. these experts, as one would predict, perform markedly better than control subjects. what’s interesting is that this performance is importantly perceptual (the radiologist has to spot the anomaly in the x-ray, the footballer has to see a shift in the opposing team’s attack), but it also appears to be importantly cognitive. that is, the convergence of evidence across the measures mentioned above suggests that these perceptual experts enjoy perceptual advantages that depend upon their domain-specific cognitive learning. “that’s what you think!” is the thought that some readers will have upon receipt of the above descriptions. that is, the action here is in how these cases are interpreted: are these genuine cases of top-down influence on perception, or something else? some points of these debates will surface below. but first i want to identify some similarities and differences in the three case types, since they help to illustrate some broad themes of thinking and perceiving. cases of the first type are anecdotal, in the sense that they are either imagined or based on some ordinary, observed phenomenon. in the most recent literature, they are employed and indeed sometimes assumed so questions can be asked about normative implications, say about epistemology or value theory. cases of the second type are sometimes similarly motivated, but they involve empirical study and, for some, an argument for a particular style of explanastokes, d. (2023). précis of thinking and perceiving. philosophy and the mind sciences, 4, 10. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org truth, success, and epistemology 3 tion, namely, one where thought has some top-down influence on perception. defenders of those arguments sometimes then go on to consider the same kinds of implications as those for the first type. i prefer the second type of case, since i want to ground treatments of those normative questions in sound empirical study of the human mind. the feature common to the first and second type of case is that they both appear to involve some error on the part of the hypothesized or studied subject; they appear to be cases involving perceptual misrepresentation or some other epistemic flaw. this of course varies from case to case. for instance, studies on the so-called ‘memory color effect’ (hansen et al., 2006) standardly involve error—subjects mis-match colour-diagnostic items such as a banana or a smurf, by contrast to colour-neutral control items. by contrast, the visual search studies involving linguistic memory appear to involve some error and some enhancement. cases of the third type are distinct in this way: they are all cases of apparent perceptual improvement, indeed of perceptual expertise. these cases are the least discussed in recent philosophical literature (by contrast to cases of types 1 and 2, which are now familiar from literature on mental architecture, modularity, cognitive penetrability, the epistemology of perception, etc.). and cases of this third type are ultimately the centerpiece of thinking and perceiving. these are empirically grounded cases that, i argue, support the claim that thought improves perception. the book thus offers some groundwork on the importance of perception and on possible dimensions for distinguishing perception from cognition. it then turns a critical eye to the most obvious opponent, namely, the modularity of mind. finally, it offers a positive defence of an alternative architecture, first by engaging and hopefully enlightening the modularity/cognitive penetrability debate, and then by partly shifting away from that debate to cases of perceptual improvement, learning, and expertise. that’s the general arc of the book. and the agenda is to shift scientific and philosophical theories of perception from the orthodoxy that is modularity to the alternative i favor: malleability. now for a bit more detail. 2 modularity and beyond the relevant controversy about thought and perception concerns modularity of the strong fodorian variety, which claims that perception is informationally encapsulated and thus cognitively impenetrable (fodor, 1983). according to such a view, cognitive states like belief or desire or intention cannot (or, at least, rarely) influence perception in some important way. challenges to modularity take the form of empirically grounded counterexample. a theorist identifies a case or set of studies where it appears that cognition is affecting perception and argues that the case is best explained as an instance of cognitive penetration, rather than a mere intra-perceptual effect, an effect on pre-perceptual attention, or an stokes, d. (2023). précis of thinking and perceiving. philosophy and the mind sciences, 4, 10. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dustin stokes 4 effect on post-perceptual cognitive states such as judgment or belief. what this state of play reveals, i suggest, is a default position assumption. modularity (and, thereby, cognitive impenetrability) remains the default theory for the architecture of perception; it is a kind of litmus test for an interesting cognitive influence on perception. and this has been true for both proponents and opponents of modularity. to be a default view, a theory must be supported either by strong arguments or by superior explanatory power. modularity enjoys neither. i defend this evaluation by, first, criticizing the clearest and most formidable arguments for informationally encapsulated perception. those arguments center around the stability and reliability of perception. there are multiple variations on each type of argument. here is a brief mention of two examples. sometimes modularity is motivated by counterexample to an opponent view. famously, the müller-lyer illusion is held up as a counterexample to a view that universally quantifies cognitive influence on perception. that view claims that all of one’s beliefs and other cognitive “utilities” influence (or can influence) perception. since one’s knowledge of the equal length of the lines in the müller-lyer does not undermine the persistent perceptual illusion, the modularist concludes, that view is false and modularity true. problem: that’s a fictional opponent view: nobody claims that the influence of thought on perception is universal. the actual opponent view—for example in the form of new look psychology—made an existentially quantified claim about cognitive influence on perception. and a (culturally sensitive) persistent illusion is no counterexample to that claim. and further, there are plausible non-modularist explanations of such illusions. more direct arguments for modularity concern the apparent reliability of perception. there are multiple versions of this argument too. cognitive effects would make perceptual information redundant; cognitive effects would bias perception in ways that undermine its accuracy; cognitive effects would make perception maladaptive. since perceptual information is non-redundant, typically accurate, and adaptive, perception is cognitively impenetrable. ergo, modular. each version of the argument requires distinctive treatment and i offer that in the book. one feature common to those treatments, though, is to show how each argument relies to some degree on the pernicious cognitive effects assumption. short of question-begging, that assumption cannot be taken for granted. and there are plausible examples of cognitively enhancing effects on perception. indeed, this comes into very clear view once consideration of perceptual learning and expertise are seriously considered. 3 towards malleability the rest of the book then makes the case that a malleable architecture better explains a large range of recent empirical studies and data. the first prong of this approach is more in keeping with the existing modularity/cognitive penetrability stokes, d. (2023). précis of thinking and perceiving. philosophy and the mind sciences, 4, 10. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org truth, success, and epistemology 5 debate. but there is still an attempt to shed new light and hopefully re-frame some key features of that debate. i suggest that extant attempts to define cognitive penetrability as such have largely failed and result in unhelpful theoretical cross-talk. in place of a “real definition”, we should characterize the phenomenon in terms of its consequences. thus, ψ is cognitive penetration if and only if ψ is a cognitive-perceptual relation, and ψ implies consequences for theory-ladenness or the epistemic role of perception or the behavioural role of perception or mental architecture. this returns attention to scientific and epistemic concerns that have animated parties on both sides of relevant debates and, accordingly, may mitigate cross-talk in a debate-neutral way. i also argue that the modularists’ standard story on attention-mediated instances of cognitive influence on perception is misguided. there are plausible cases where cognition influences covert selective attention—such as feature based and object based attention—and whereby this in turn influences conscious perceptual experience. the studies by moores et al. (2003) and meyer et al. (2007) mentioned above are two related examples. but for a simpler and more intuitive example, think of where’s waldo? puzzles. if you know what waldo looks like, the red and white stripey features of the puzzle pop out in your visual experience. this perceptual organization (phenomenally distinct from that of the waldo-ignorant perceiver) is automatic but dependent on background knowledge of waldo. there are two ways to argue that cases with this structure are instances of cognitive penetration. first, selective attention of this kind is part of perception rather than a gatekeeper between cognition and perception. therefore, a cognitive influence on attention of this kind is a direct cognitive influence on perception. second, one can argue that such influences amount to cognitive penetration by virtue of implying important consequences for our epistemological and/or architectural theories of perception. the second prong in the argument for malleability moves largely beyond talk about modularity and cognitive penetration. here is where i shift to empirical studies of apparent cases of perceptual improvement. these are cases of the third type discussed above, studies on perceptual expertise. perceptual experts perform in a specific domain of training, their performance success is above a threshold set by the standards of that domain, and their performance non-trivially involves sensory perception. such experts have been studied across a wide range of domains. researchers use a variety of measures and methods to study both these “real-world” and lab-trained experts. and in training subjects to become experts regarding “real-world” objects (e.g. cars or birds) and lab-created objects (e.g. “greebles”), they effectively study the acquisition and development of expertise. the best explanation of many of these phenomena is that the expertise is partly resident in the perceptual experiences of the expert, and those perceptual differences (by contrast to novices or the naïve) depend upon the richly cognitive training of the expert. stokes, d. (2023). précis of thinking and perceiving. philosophy and the mind sciences, 4, 10. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dustin stokes 6 it is in this way that i support the claim that thinking affects perceiving (the tap thesis), and in many cases thinking improves perceiving (the tip thesis). the tap thesis divides into two architectural claims. claim 1 says that some cases of perceptual expertise are genuinely perceptual, insofar as they involve differences in perceptual experience. studies on expertise show robust behavioral and neural markers of, and similarities and interactions with, facial recognition—an undeniably perceptual phenomenon. experts enjoy rapid and often “automatic”, successful performance and display significant differences in eye movement patterns. they enjoy advantages in visual short-term memory. this convergence of data is best explained perceptually. (this is a sizeable empirical literature; for two helpful reviews, see (bukach et al., 2006) and (scott, 2011)). claim 2 says that those perceptual differences are sensitive to the cognitive learning specific to the domain of expertise; they depend upon the cognitive etiology of the expert. experts’ performance success, and persistence of that success, varies with fine grained learning of concepts, and those changes are corroborated by lasting neural changes. accordingly, mere “practice” or exposure to relevant stimuli is often insufficient for expert performance. and these skills tend not to “transfer” to similarly complex tasks in domains outside of the expert’s field. perceptual experts are, genuinely, perceptual experts. the epistemology follows from this mental architecture, and one that emphasizes cognitive improvements to perception (by contrast to cases of types 1 and 2 above, which emphasize apparent cases of perceptual error). within a domain, perceptual experts perform reliably, rapidly, and with less distraction. they approach optimality. i argue that this successful performance, qua performance of the agent, is best understood in virtue-theoretic terms. this requires that perception can genuinely improve, and not merely as a matter of normal development or exposure to stimuli. some experts acquire, through concept-rich cognitive training, through deliberate activity, a skill. the expert radiologist or goalkeeper or forensics officer performs better visually because of what she has done, because of her actions, as a responsible epistemic agent. as a consequence of this training, her perceptual systems perform in exceptional ways within that domain. and those levels of performance near maximally satisfy the natural norms for perception, thus fulfilling the representational function of perception in optimal or near-optimal ways (in that domain). the important epistemic difference between this case and the cases of mere development or exposure is that the agent herself is responsible for the relevant etiology and, accordingly, for the perceptual improvement. the epistemic virtue is therefore attributable to the agent herself. in cases of expertise, thinking thus improves perceiving (the tip thesis). 4 some consequences the tap and tip theses are the descriptive and normative components of a theory of how thought may affect perception, of how the mind is richly malleable. imporstokes, d. (2023). précis of thinking and perceiving. philosophy and the mind sciences, 4, 10. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org truth, success, and epistemology 7 tant consequences follow. here are a few such consequences, tailored to readers with interests in specialized topics. 4.1 perceptual content and accuracy accuracy is but one facet of perceptual success and thus just one possible determinant of perceptual content. perceptual success can also involve increased sensitivity to gestalts, patterns, and feature types, achieved rapidly and efficiently, integrated with action, and with less distraction. success along these measures can vary from domain to domain and so it follows that perceptual content is not determined in a purely mind-independent objective* way, but instead in an inter-subjective, objective way, to include facts about the perceiver’s environment, but also facts about the perceiver’s epistemic community. 4.2 admissible contents (and aesthetics) the analysis in the book motivates an argument for rich perceptual content but without admission of natural kinds into that content. enhanced perceptual sensitivity as enjoyed by experts– to patterns, gestalts, and organizational features – engenders rich perceptual content. importantly, this lesson is partly learned by considering cases of perceiving aesthetic properties: the ballet instructor sees not only the colours, edges, shapes, and motion of their students but also how those features are organized in ways that are balanced or serene or graceful. some aesthetic experts are perceptual experts. and indeed, perhaps some of what the aesthetician calls ‘taste’, sometimes elusively, is just some degree of perceptual expertise about aesthetic properties qua aesthetic gestalts. 4.3 theory-ladenness as should be clear, all of this is to accept the theory-ladenness of perceptual observation, both in science and in the ordinary course. with this comes potential virtues, but also important risks and possible vices. virtuous cases are the many cases of perceptual expertise, the expert surgeon or radiologist or arborist. threats of individual bias can be mitigated by the same inter-subjective, communal mechanisms that philosophers of science have offered to mitigate the general threats of theory-ladenness in scientific investigation. an example of a more worrisome vice is implicit bias and the “cross-race effect” in facial recognition. the latter is a well-studied visual phenomenon (young et al., 2011), and the somewhat speculative suggestion is that some cases of implicit bias might be grounded in this kind of visual effect; some racial bias might be, or at least importantly depend upon, culturally influenced perceptual bias (rather than being entirely at the level of judgments or beliefs or values). although worrisome, the malleability of this effect opens space for optimism: subjects can be trained to recognize “cross-race” faces. stokes, d. (2023). précis of thinking and perceiving. philosophy and the mind sciences, 4, 10. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dustin stokes 8 4.4 epistemology and virtue the view in the book is that some cases of perceptual expertise are epistemically virtuous or contribute to epistemic virtue. this aligns with recent views that defend the epistemic evaluability of perceptual experience itself; if perception itself can improve then it can also go badly. at the same time, the virtue-theoretic treatment highlights the agent and her actions. perception on this view is an activity and the agent can be lauded or not for how she has affected that activity through her own cognitive actions. this puts a lot of the perceiver and her experiences into the space of epistemic evaluation. 4.5 understanding the perceptual expert is especially perceptually sensitive to the patterns, organization, and gestalts of whole objects from their domain of expertise. understanding, as an epistemic value, involves a grasp of patterns and organization, of the relation of parts to whole, of how things cohere or hang together. perceptual experts thus enjoy perceptual understanding. a malleable architecture that makes space for this perceptual phenomenon provides better explanation of the important epistemic achievement, of what it is to understand in skillful ways. 4.6 self and self-making from candrakīrti to sartre, hume to reid, locke to parfit, whatever their differences, a notion of self involves a notion of our own minds, as autonomous agents who bear some responsibility for the mental events that populate the mind, all through the continuing process of waking life. we live through our lives, making contact with the world. if thinking affects, sometimes improves perceiving, and we play an active role in these effects, then we play an active role in making up our selves qua minds. and this goes all the way down to how we perceive the world. accordingly, we gain better self-understanding if we conceive of our minds as richly malleable in these ways. 5 conclusion to conclude, let me answer a question that may have come to the minds of some readers. “isn’t an emphasis on perceptual expertise an emphasis on highly special cases? can we generalize to human perception from such specialized, domainspecific improvements to perception?” yes, we can. and this is because the phenomenon is general. although the experts described here are remarkably accomplished, and often studied in elaborately controlled ways, they are not that special. they are not, after all, super-humans. stokes, d. (2023). précis of thinking and perceiving. philosophy and the mind sciences, 4, 10. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org truth, success, and epistemology 9 all humans are habit forming and many of those habits involve perception in nontrivial ways. we populate our work and non-work life with activities, hobbies, and skills that crucially involve seeing, hearing, tasting, and so on. and if i am right about the nature of perceptual expertise, then it is in these daily contexts that our cognitive activities can improve our perception of the world. therefore, genuine perceptual expertise is, i suggest, a pervasive phenomenon. we are all of us potentially perceptual experts and in a variety of contexts. indeed i suspect that most if not all of us are perceptual experts, and probably each of us in multiple domains. to accept this kind of malleability is, therefore, to better understand human perception. it is to better understand ourselves and our place in the world. references bukach, c. m., gauthier, i., & tarr, m. j. (2006). beyond faces and modularity: the power of an expertise framework. trends in cognitive sciences, 10(4), 159–166. https://doi.org/10.1016/j.tics.2006.02.004 fodor, j. a. (1983). the modularity of mind: an essay on faculty psychology. mit press. hansen, t., olkkonen, m., walter, s., & gegenfurtner, k. r. (2006). memory modulates color appearance. nature neuroscience, 9(11), 1367–1368. https://doi.org/10.1038/nn1794 meyer, a. s., belke, e., telling, a. l., & humphreys, g. w. (2007). early activation of object names in visual search. psychonomic bulletin & review, 14(4), 710–716. https://doi.org/10.3758/bf03196826 moores, e., laiti, l., & chelazzi, l. (2003). associative knowledge controls deployment of visual selective attention. nature neuroscience, 6(2), 182–189. https://doi.org/10.1038/nn996 scott, l. s. (2011). face perception and perceptual expertise in adult and developmental populations. in oxford handbook of face perception (pp. 255–286). oxford university press. https://doi.org/10.1093/oxfordhb/9780199559053.013.0011 young, s. g., hugenberg, k., bernstein, m. j., & sacco, d. f. (2011). perception and motivation in face recognition: a critical review of theories of the cross-race effect. personality and social psychology review, 16(2), 116–142. https://doi.org/10.1177/1088868311418987 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. stokes, d. (2023). précis of thinking and perceiving. philosophy and the mind sciences, 4, 10. https://doi.org/10.33735/phimisci.2023.10250 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.tics.2006.02.004 https://doi.org/10.1038/nn1794 https://doi.org/10.3758/bf03196826 https://doi.org/10.1038/nn996 https://doi.org/10.1093/oxfordhb/9780199559053.013.0011 https://doi.org/10.1177/1088868311418987 https://doi.org/10.33735/phimisci.2023.10250 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction modularity and beyond towards malleability some consequences perceptual content and accuracy admissible contents (and aesthetics) theory-ladenness epistemology and virtue understanding self and self-making conclusion psychedelics favour understanding rather than knowledge psychedelics favour understanding rather than knowledge sascha benjamin finka,b,c (sfink@ovgu.de) abstract chris letheby argues in philosophy of psychedelics that psychedelics and knowledge are compatible. psychedelics may cause new mental states, some of which can be states of knowledge. but the influence of psychedelics is largely psychological, and not all psychological processes are epistemic. building on the distinction between processes of discovery and processes of justification, i aim to criticise some aspects of letheby’s epistemology of psychedelics. unarguably, psychedelics can trigger processes of discovery. yet, i hold, they can hardly contribute either to the epistemic success (i.e., truth, veridicality, aptness, skillfulness, etc.) of a mental state or to processes of justification. as these are central for a mental state to be a state of knowledge and are largely uninfluenced by psychedelics, the contributions of psychedelics to knowledge are rather indirect than direct: the heavy epistemic lifting—what turns a mental state into a state of knowledge—is, in its epistemic aspects, largely independent of the influences of psychedelics on our psyche. positively, while the mechanisms that letheby points to need not be associated with knowledge, they do provide crucial epistemic benefits if they are associated with understanding. reading them as facilitating understanding provides a broader picture because it covers also cases of comforting delusions or placebo insights where truth and justification are lacking. keywords psychedelics ∙ philosophy ∙ epistemology ∙ understanding ∙ knowledge ∙ rebus model ∙ process of discovery ∙ process of justification ∙ epistemic benefits ∙ epistemic standards this article is part of a symposium on chris letheby’s book “philosophy of psychedelics” (oup 2021), edited by chiara caporuscio and sascha benjamin fink. chris letheby’s book philosophy of psychedelics (letheby, 2021) delivers an empirically informed, philosophically subtle, and very clear and illuminating explanation for the efficacy of psychedelic-assisted therapy. importantly, letheby rejects the aphilosophy departement university of magdeburg. bcenter for behavioral brain sciences (cbbs), university of magdeburg. chumboldt-universität zu berlin, research training group 2386 “extrospection”. fink, s. b. (2022). psychedelics favour understanding rather than knowledge. philosophy and the mind sciences, 3, 8. https://doi.org/10.33735/phimisci.2022.9264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/ 0000-0001-7535-1882-6251 https://doi.org/10.33735/phimisci.2022.9264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 2 idea that the success of psychedelic-assisted therapy rests on the acquisition of comforting delusions, i.e. persistent false beliefs that make us feel good. one thesis defended by letheby in chapter 8 is that the use of psychedelics can be epistemically beneficial: “some claims to psychedelic-assisted epistemic benefit are accurate” (letheby, 2021, p. 160). so there is knowledge to be gained beyond the doors of perception (huxley, 1954). here, i will critique one general aspect of this chapter before expanding on letheby’s theme of possible epistemic benefits from psychedelic use: i suggest that they further understanding rather than knowing. for reasons of space, i will focus on only a few of the forms of knowledge discussed, mainly knowing-that and knowing-how. still, i suspect that most of the points i raise generalise. as a starting point, it is important to emphasise that even if knowledge is possible under the influence of psychedelics, it is not guaranteed. that is, one does not necessarily gain knowledge by using psychedelics. in some cases, we may end up with comforting delusions or placebo insights, i.e. beliefs that are comforting but probably false (letheby, 2021, p. 28; see also jopling, 2001). what we primarily gain by using psychedelics are mental states, most of them new, original, and unusual if compared to our everyday mind. the influences of psychedelics on the neural and psychological side of these mental processes are becoming more and more fully understood. the outstanding epistemic question is whether and why some of these new mental states amount to knowledge. letheby claims that, under the right circumstances, psychedelics can contribute to our epistemic state. he thereby rejects an epistemic impossibility claim, which he characterises as follows: “it is impossible to gain epistemic benefits by taking psychedelics” (letheby, 2021, p. 163). i agree that this epistemic impossibility claim is false. as letheby (2021, p. 165) points out, we can distinguish between the processes of discovery of a fact and the processes of justification for knowing that fact (hoyningen-huene, 1986).1 via some processes (for example, during a psychedelic episode), we may step into new cognitive territory. but access to these new mental states need not be lost after the initial eliciting process (e.g. the trip). indeed, once we have gotten to know these states, we may be able to re-elicit them, even without the tools we used the first time. now, if these mental states are epistemic, then the psychedelic episode can count as a process of discovery. although the underlying neural mechanism in the case of psychedelic episodes are special, they are not epistemically different from other unusual contexts of discovery, such as kekulé’s “ouroboros dream” of a snake eating its own tail that led him to discover the chemical structure of the benzene ring. yes, psychedelics offer the benefit of getting us into new cognitive territory. but given the vast range of routes to discovery, this benefit cannot be considered unique to psychedelics. 1letheby calls these processes “contexts”, but hoyningen-huene (1986, p. 503f) convincingly argues that they are better called “processes” because talking of “contexts” is too vague—a point he attributes to herbert feigl. fink, s. b. (2022). psychedelics favour understanding rather than knowledge. philosophy and the mind sciences, 3, 8. https://doi.org/10.33735/phimisci.2022.9264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org psychedelics favour understanding rather than knowledge 3 furthermore, because some of the processes altered by psychedelic compounds that lead us to novel mental states may also lead us to comfortable illusions or placebo insights, i would not consider this contribution necessarily epistemic. the contribution of psychedelics to knowledge, if there is any, is coincidental.2 i argue that the heavy epistemic lifting—what makes the difference between these new mental states constituting states of knowledge or not—is done by processes of justification. letheby (2021, p. 165) suggests that justification (as well as truth) can come from outside the psychedelic episode. for example, we may gain new ideas during a trip, but the justification for them being knowledge is not dependent on psychedelics; it can be added after the trip. if psychedelics need not contribute to these justificatory aspects of knowledge in order to be epistemically beneficial, then it is hard to see why psychedelics and knowledge need be incompatible. letheby is right to attest that the impossibility claim must be rejected. but could a more modest version of the impossibility claim survive letheby’s criticism? i wonder whether the current version needs rejecting simply because the formulation is too strong. the distinction between processes of discovery and justification may hint at a more defendable version of impossibilism. consider a more subtle impossibility claim: not that psychedelics preclude the formation of mental states that have the status of knowledge, but that psychedelics cannot contribute to their status as knowledge. that is, psychedelics may have psychological benefits, indirect epistemic benefits, or be beneficial for processes of discovery (letheby, 2021, pp. 191–194), but they lack direct epistemic benefits for processes of justification. call this psychedelic justification impossibilism. an impossibilist about psychedelic justification may point out that, generally, the mechanisms underlying the formation of mental states, especially those altered under psychedelics, are primarily a matter of psychology, not necessarily a matter of epistemology. even if psychedelically formed mental states are largely correct, apt, true, veridical, etc., it does not follow that these states are states of knowledge. for this, we need processes of justification. and a psychedelic justification impossibilist holds that psychedelics will not contribute there. if we maintain the distinction between processes of discovery and justification, this impossibilist will deny that the psychological mechanisms that are altered by psychedelic compounds can be classified as processes of justification. this type of impossibilist may successfully deal with some of letheby’s examples, intended to illustrate the epistemic benefits of psychedelics, e.g. that psychedelics allow one to experience one’s potential or reveal the mutability of the self-model. here, those impossibilists grant indirect epistemic benefits in that psychedelic compounds, experiences, or episodes may contribute to forming states that, as a matter of fact, are knowledge – but deny direct epistemic benefits in that they contribute to such states’ status as knowledge, i.e. to processes of justification or to them being correct, apt, successful, true, veridical, etc. i sympathise with this type of justification impossibilism. 2in the sense of accidental, i.e. neither necessarily related nor necessarily incompatible. fink, s. b. (2022). psychedelics favour understanding rather than knowledge. philosophy and the mind sciences, 3, 8. https://doi.org/10.33735/phimisci.2022.9264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 4 psychedelic justification impossibilism will re-interpret most of letheby’s examples of epistemic benefits as being indirect, but there are some caveats and comments to be added. first, any indirect epistemic benefits not contributing to processes of justification (e.g. being more open-minded, see letheby, 2021, ch. 8.7) are untouched by this type of impossibilism. second, there might be some instances where what is known concerns psychedelics directly, e.g. if a person believes that they have taken psychedelics. here, psychedelics may contribute not only to a belief but also to its truth.3 that psychedelics contribute to truth also extends to cases where psychedelics change psychic processes such that they fit our beliefs. if the self-model, for example, is altered during psychedelics and then coincidentally fits our beliefs about ourselves, then psychedelics contributed directly to the truth of a belief. there are also borderline cases, e.g. where psychedelic ingestion caused a belief 𝑏𝜓 about an element of one’s psyche and 𝑏𝜓 subsequently affects, via processes of cognitive penetration, its subject matter such that 𝑏𝜓 comes out true. for example, if psychedelic ingestion caused me to believe that i am a more kind person and that belief causes me to be a more kind person, i then end up knowing that i am a kind person. in this case, psychedelics affected the truth of a belief, but indirectly. some of these examples closely mirror cases that letheby discusses. but i will ignore these because here psychedelics directly affect the subject matter rather than the status of a mental state as knowledge. instead, i focus on justification. third, some forms of knowledge do not allow for the distinction between processes of discovery and justification. one example is phenomenal knowledge, knowing what it feels like. simply by virtue of having an experience, we immediately get the requisite justification for the phenomenal knowledge gained. for these cases, the justification impossibilist must grant psychedelics a direct contribution to knowledge. fourth, this type of impossibilism crucially depends on the distinction between processes of discovery and processes of justification. one may, as e.g. quine (1969) did, simply reject the distinction by pointing out that epistemology is, as a discipline, dependent on our theories of psychological mechanisms. however, this claim seems too strong, since justification has a social and normative dimension that does not reduce to psychology. a more defensible position would be that psychology is relevant to epistemology even though none reduces to the other. under these circumstances, what amounts to a good or adequate justification will still be largely independent of any given individual, because the standards for justi3one may even construe psychedelic gettier cases. consider, for example, that 𝐴 meets 𝐵 for tea and cookies. 𝐴 sees 𝐵 enter a room. subsequently, 𝐴 believes that 𝐵 is in the other room because 𝐴 saw 𝐵 enter it. 𝐴 may then infer the following: 𝐵 is in the other room or 𝐴 is high. this proposition is made true by either disjunct, and it is justified by the observation of the first disjunct. now, unbeknownst to 𝐴, 𝐵 has left the other room to get something from the shop. so 𝐵 is not in the room—and the first disjunct is false. however, 𝐴 did not realise that the cookies 𝐵 served were laced with a psychedelic compound. so the inferred proposition 𝐵 is in the other room or 𝐴 is high is true and it is true due to psychedelics. fink, s. b. (2022). psychedelics favour understanding rather than knowledge. philosophy and the mind sciences, 3, 8. https://doi.org/10.33735/phimisci.2022.9264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org psychedelics favour understanding rather than knowledge 5 fication are established socially or extra-individually. and because psychedelics affect only an individual’s state, leaving the standards of justification untouched, they do not affect which processes can be counted as processes of justification— even if processes of justification are in the end psychological processes and are indeed affected by psychedelics. the epistemic lifting is done not by psychedelically influenced processes, but by the extra-individual circumstances that turn these processes into processes of justification.4 4the rebus model (carhart-harris et al., 2015) is an interesting case to discuss here. in some sections of chapter 8, this model is presented as playing an epistemic role and in some sense, it can be read as a psychological processes that is a process of justification by default. however, i have doubts. why might it be read epistemically? as a predictive coding model, it builds on the idea that hypotheses are updated based on their fit to incoming signals in a largely bayesian manner. as such, it seems to follow the principles of bayesian epistemology. rebus, specifically, models the psychological influences of psychedelic compounds in the following manner: bottom-up processing (i.e. information flow from levels closer to sensory input to those more distant to it) gains greater influence either by (i) a relaxation of the precision of higher-level priors or (ii) by an increase in the precision of the ascending signals. the idea is that, if we are in such states, our preconceptions have less influence on how we experience the world. thereby, we are more in touch with the ways the world beyond our markov blanket affects us. but can rebus, a descriptive psychological model, truly provide such heavy epistemic lifting? i have doubts. after all, rebus is a prominent and highly influential model on the algorithmic level of description. in marr’s account (marr, 1982), such algorithmic models are constrained by observations on the implementational and descriptive levels. on both these accounts, rebus appears to have some difficulties, i suggest. first, concerning the implementation: the influences on precision (both of the priors or of the ascending signal) is supposedly facilitated by the 5ht2ar receptor (carhart-harris et al., 2015, p. 318). these receptors, to which most serotonergic psychedelics bind, are widespread in the cortex—but not only in the higher areas of the neural hierarchy but also in lower sensory areas. it is “richly expressed in the visual cortex” (carhartharris & friston, 2019, p. 321). if so, it is unclear why a compound like lsd, which binds all over the brain, should affect only one part of the hierarchy or one direction of processing: the precision of higher-level priors or ascending signals. in principle, if the mechanisms are the same elsewhere, serotonergic compounds should increase precision everywhere, not only in the ascending signal or higher levels of the hierarchy. then, depending on circumstances, bottom-up or top-down processing can be the driver of one’s phenomenal experiences in a psychedelic state. entropy is increased in the brain, not only locally but globally. (letheby is not averse to the idea that psychedelics may affect higher cognitive processes as well: he stresses that we are more willing to entertain other possibilities, contemplating them as true.) but, first, if the uncertainty is increased all over, this does not allow us to presume that we are more in tune with perception per se. and second, the contemplation of new possibilities is an indirect not a direct epistemic benefit: i might end up believing fewer truths than before. i also have reservations that rebus accounts well for phenomenology. why? psychedelic states are not always characterised by greater influence of bottom-up processes. indeed, some reports suggest that psychedelic experiences can be driven by previously held beliefs about the metaphysics of the universe rather than by incoming sensory evidence. how about behavioral evidence? even experimental results suggesting that, under psilocybin, participants are more inclined to see a hollow face as hollow need not be attributed to a relaxation of higher priors. these data might instead be explained by the fact that individuals know inferentially that the face must be hollow and are able, thanks to their cognitive insight, to finally see it as hollow. fink, s. b. (2022). psychedelics favour understanding rather than knowledge. philosophy and the mind sciences, 3, 8. https://doi.org/10.33735/phimisci.2022.9264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 6 fifth, the alterations in our psychological processes triggered by psychedelics could be processes of justification if they were reliable (goldman, 1979). in reliabilism, what justifies the results of some processes as being knowledge is the amount of past successes of these processes for tracking truth—even if they have involved some failures. however, it is the history of the processes in the individual, i.e. what caused and causally sustains the mental state in a specific individual (goldman & beddor, 2021), that determines the reliability of that individual’s processes. in practice, the few doses that we are concerned with in psychedelic-assisted therapy do not suffice to establish reliability of psychedelically influenced processes in an individual.5 additionally, the high prevalence of placebo insights and comforting delusions as well as the increased suggestibility that is characteristic of these states (carhart-harris et al., 2015) makes it implausible that psychedelically altered mechanisms are reliable in an epistemic sense. beyond these caveats, psychedelic justification impossibilists grant that ingesting psychedelics may be a way to form mental states that amount to knowledge— maybe even the easiest or most reliable way to attain some states of knowledge. but they will also hold that (a) any process of justification for a psychedelically attained state of knowledge is independent of psychedelics and (b) what determines which processes are processes of justification is also independent of psychedelics. except for some special cases, whatever we got to know by psychedelics, it is not knowledge due to psychedelics. in my view, this is a more sensible impossibility claim. because the distinction between processes of discovery and processes of justification has been applied primarily to propositional knowledge, one may read this as a very restricted form of impossibilism—a form that leaves most of letheby’s epistemic points untouched. however, i am convinced that some of it carries over to other forms of knowledge. there, justification is needed as well, e.g. when we need to justify that the success of an act—e.g. hitting bullseye—is not just lucky but due to know-how or skill. unfortunately, i lack the space to expand on this line of argument here. there is something to be said, though, for letheby’s proposal that thanks to psychedelic use, one gains new knowledge of old facts, i.e. sees already known although rebus is stimulating, its attachment to neural mechanisms (the implementational level) or to phenomenology (the descriptive level) is still too loose to be read realistically rather than instrumentally (see also zednik & jäkel, 2016). so i am not denying that there is something we can describe as a weakening of priors in higher parts of the neural hierarchy. however, in an instrumental reading, it is unclear how much is gained for the epistemological story letheby tells. i am therefore sceptical that rebus model of altered belief formation is one that establishes psychedelics’ direct epistemic benefits. 5letheby circumvents this by citing studies using data from many individuals, akin to how we experimentally establish, for example, that some processes of perception like seeing are reliable. but i remain skeptical that these studies are capable of showing this because reliability cannot be established with only a few trials in individuals. using data from many individuals does not solve this problem, it seems to me. fink, s. b. (2022). psychedelics favour understanding rather than knowledge. philosophy and the mind sciences, 3, 8. https://doi.org/10.33735/phimisci.2022.9264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org psychedelics favour understanding rather than knowledge 7 facts in a new way. reading buddhist texts may lead me to a belief that the self is not a real thing but rather an idea or model, and that this model is mutable. but this differs from experiencing the self as a mutable model. i find this the most compelling case for the epistemic benefits of psychedelics. but i propose that we reconceptualise this process as understanding rather than knowing. let me elaborate. despite my doubts about the direct benefits of psychedelics to processes of justification, i agree with letheby that psychedelics can contribute indirectly to our epistemic state in a meaningful way. letheby offers two examples: that psychedelics increase psychosocial functioning, which in turn influences epistemic functioning; and that they increase our capacity for modal knowledge or reduce biases or narrow mindedness (letheby, 2021, ch. 8.7). i would add that, thirdly, psychedelics may contribute to our understanding of a proposition. in what way may the usage of psychedelics favour understanding? following davidson (1973), understanding can be conceived of as distinct from knowledge: while a known proposition p has to be true in order to be known, we can understand p even if it is false. but we understand p if and only if we know what needs to be the case in order for p to be true. that is, proper understanding involves homing in on what makes a proposition true, even if that proposition is de facto false. such an “understanding of what is already believed” seems to be captured in some trip reports. for example, we see evidence of it in two reports contributed to the erowid experience vaults: “i can’t articulate now the depth of what i learned (or what i discovered i already knew intrinsically to be true)” (erowid.org/exp/53057) or “the dmt death and rebirth experience had just confirmed what i already knew deep down” (erowid.org/exp/96614). the mechanisms behind this type of new understanding of a pre-existing belief are still unclear. but i suggest that the increased brain connectivity of sensory-somatomotor areas, mediated by the 5-ht2a receptor (preller et al., 2018), may lead to an endogenously facilitated, nonconceptual “me-in-a-world-where-p-is-the-case” representation of p (or to a simulation of oneself in a world where 𝑝 is true) rather than to a conceptual representation of p. i may already have abstractly known thanks to evolutionary theory that i and a bird are distantly related, but now i can feel related to the bird. i may already have known by studying buddhist texts that the self is an illusion, but now i can feel it as an illusion. that we gain new understanding of old beliefs mirrors and builds on letheby’s suggestion that we may gain new knowledge of old facts. but it should not be confused with it: knowing old facts in a new light presupposes that what comes in a new disguise is indeed a fact. but the same mechanisms that letheby (2021, ch. 8.6) points to for facilitating new knowledge of old facts also extend to falsities. in contrast, understanding does not share the presupposition that what is understood is a fact: i can also understand counterfactuals and falsities. this reconceptualisation has the benefit of capturing not only the good but also the bad cases, fink, s. b. (2022). psychedelics favour understanding rather than knowledge. philosophy and the mind sciences, 3, 8. https://doi.org/10.33735/phimisci.2022.9264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 8 i.e. comforting delusions and placebo insights. i therefore suggest that some of the mechanisms contributing to epistemic benefits derived from psychedelic use should be associated with understanding rather than with knowing. the knowing/understanding distinction also gives us purchase on the flipsides of epistemic growth. because understanding is not reliant on factuality, it may to some degree also be detrimental to our epistemic state. “old delusions in new disguises” may be misunderstood as new evidence and thereby even strengthen some false beliefs. this, i believe, provides a good explanation for the persistence of some comforting delusions, independent of the comfort they may offer. it also raises the question of why processes facilitating understanding should be seen as epistemically beneficial. let me propose an argument in favour of this. first, and importantly, understanding is not needed for knowledge. i can believe that some statement s expresses a true proposition because some expert told me. if that person truly is an expert, i may even be justified in believing that i express a justified truth when uttering s. in that sense, i know that p if s expresses p. but i may not quite understand what s means. in that sense, i don’t know which p i now know when muttering s to myself. that is, i may know that what s expresses is true in this world, but i may not be able to track its truth into other possible worlds. i may fail to exclude some worlds where p is false and fail to track it in others where p is true—simply because i don’t understand which p this statement s expresses. second, understanding is not needed for believing either. in mathematics, this is an old hat (williams, 1982). archimedes, cusanus, and snellius believed that they could find a way to square the circle by constructive means. that is, they believed that for any given circle one can construe a square with the corresponding area only with a straightedge and a compass. in 1882, however, lindemann proved that there cannot be such a correspondence because of the transcendence (and, thereby, non-constructability) of π. the idea of such constructive means to find the square/circle correspondence, then, is not only false, it is self-contradictory if we grasp the transcendental nature of π. mathematicians of old believed that they could square the circle, but they did not quite understand what they were believing. if they had, they would have seen the futility of it—and would thereupon have stopped believing it. similar examples are found in all areas of science, e.g. in the belief that heavy objects fall faster than lighter ones (galilei, 1638, p. 334), as well as in religion, e.g. the widespread christian belief that humans ascend in their own bodies at the day of reckoning (aquinas, 1259-65, pp. 1259–1265, iv.80). these beliefs are self-contradictory: accepting them immediately allows us to construe paradoxes. for example, if we attach a lighter body attached to a heavier one by a chain, this compound object must fall both faster and slower than than the heavier one—on the one hand, because the lighter ball slows the heavier ball down and, on the other hand, because the lighter and the heavier ball joined together must fall faster than the heavier one by itself. but seeing this self-contradiction means understanding the proposition and tracking its implications for such unusual cases. fink, s. b. (2022). psychedelics favour understanding rather than knowledge. philosophy and the mind sciences, 3, 8. https://doi.org/10.33735/phimisci.2022.9264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org psychedelics favour understanding rather than knowledge 9 we are now in a position to see the epistemic benefits of understanding. consider this: if an expert is discredited for whatever reason (e.g. plagiarism, political affiliation, association with psychedelics), a person may start to disbelieve what the expert says. say the discredited expert asserted the sentence s. if we don’t understand s, we may start to disbelieve that s expresses a truth, even if p (which is expressed by s) remains true. however, if we understand s, we are in a better position to understand whether these allegations should lead us to stop thinking that s expresses a truth. knowledge that is also understood is therefore more stable by virtue of being more attuned to those facts that determine whether it is still justified or not: i know when i should reject a statement s if i understand which proposition s expresses. conversely, understanding a falsity q may lead to easier rejection if understanding a proposition entails being more attuned to its possible truth-makers: if none of them fits the actual world, we are more likely to reject q. additionally, understanding also means seeing connections to other propositions: semantic inferences from p can be drawn only if we understand p. for example, lindemann’s better understanding of π allowed for the proof that one cannot square the circle. understanding therefore has two direct epistemic benefits: greater incorporation into the web of belief and better attunement to truth-makers. ultimately, understanding p means knowing what a truth-maker for p might be. and this may be close to how letheby conceives of new knowledge of old facts: i may have known before that p, but now i have seen what it would mean for p to be true. however, the notion of understanding also covers all the cases where p is false, i.e. where i have only believed that p (but not known). in those cases, i may still have an experience of what it would mean for p to be true. the epistemic benefit of psychedelics therefore accrues, i argue, not by gaining new knowledge of any kind, but simply by increasing our understanding of what we believe. and if what we believed was true and justified, we gain an epistemic understanding, a deeper appreciation of what it is that we knew before or have come to know after the trip. in that sense, psychedelics “contribute to the context of justification, in which ideas are scrutinised for plausibility” (letheby, 2021, p. 172) because understanding allows for a deeper appreciation of plausibility. the contribution of psychedelics is, however, still indirect because false beliefs can also be plausible and understood. that psychological mechanisms influenced by psychedelics favour understanding but not knowledge intersects with letheby’s suggestion that we gain understanding of how to perform (e.g. certain meditative practices, see letheby, 2021, p. 178) as well as his suggestion that psychedelics make us more open to alternatives: if psychedelic states favour our understanding of these alternatives, we may more easily come to reject some false ones. but understanding does not encourage the rejection of most false alternatives, because mechanisms for rejection are not always truth-sensitive: comforting delusions may not be rejected because of their comfort. instead, they may even become more stable, simply because they fink, s. b. (2022). psychedelics favour understanding rather than knowledge. philosophy and the mind sciences, 3, 8. https://doi.org/10.33735/phimisci.2022.9264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 10 are now understood and felt as well as believed. again, that psychedelics favour understanding rather than knowledge provides a broader picture, giving us tools to explain both the cases where our horizon broadens or our delusions harden. in this sense, even some impossibilists might accept that psychedelics have epistemic benefits—but that these are decidedly indirect. references aquinas, t. (1259-65). summa contra gentiles. vat. lat. 9850. carhart-harris, r. l, & friston, k. j. (2019). rebus and the anarchic brain: toward a unified model of the brain action of psychedelics. pharmacological reviews, 71(3), 316–344. https://doi.org/https://doi.org/10.1124/pr.118.017160 carhart-harris, r. l., kaelen, m., whalley, m., bolstridge, m., feilding, a., & nutt, d. j. (2015). lsd enhances suggestibility in healthy volunteers. psychopharmacology, 232(4), 785–794. https://doi.org/https://doi.org/10.1007/s00213-014-3714-z davidson, d. (1973). radical interpretation. dialectica, 27(1), 314–328. https://doi.org/10.1111/j.1746-8361.1973.tb00623.x galilei, g. (1638). discorsi. leida, apresso gli elsevirii. goldman, a. (1979). what is justified belief? in g. pappas (ed.), justification and knowledge (pp. 1–25). boston: d. reidel. goldman, a., & beddor, b. (2021). reliabilist epistemology. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (summer 2021). https://plato.stanford.edu/archives/sum2021/entries/reliabilism/; metaphysics research lab, stanford university. hoyningen-huene, p. (1986). context of discovery and context of justification. studies in history and philosophy of science part a, 18(4), 501–515. https://doi.org/10.1016/0039-3681(87)90005-7 huxley, a. (1954). the doors of perception. chatto & windus. jopling, d. a. (2001). placebo insight: the rationality of insight-oriented psychotherapy. journal of clinical psychology, 57(1), 19–36. https://doi.org/https://doi.org/10.1002/1097-4679(200101)57:1%3c19::aid-jclp4%3e3.0.co;2-z letheby, c. (2021). philosophy of psychedelics. oxford, uk: oxford university press. marr, d. (1982). vision. w. h. freeman. preller, k. h., burt, j. b., ji, j. l., schleifer, c. h., adkinson, b. d., stämpfli, p., et al. (2018). changes in global and thalamic brain connectivity in lsd-induced altered states of consciousness are attributable to the 5-ht2a receptor. elife, 7, e35082. https://doi.org/10.7554/elife.35082 quine, w. v. (1969). epistemology naturalized. in j. kim & e. sosa (eds.) ontological relativity and other essays. new york: columbia university press. williams, j. n. (1982). believing the self-contradictory. american philosophical quarterly, 19(3), 279–285. zednik, c., & jäkel, f. (2016). bayesian reverse-engineering considered as a research strategy for cognitive science. synthese, 193(12), 3951–3985. https://doi.org/10.1007/s11229-016-1180-3 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. fink, s. b. (2022). psychedelics favour understanding rather than knowledge. philosophy and the mind sciences, 3, 8. https://doi.org/10.33735/phimisci.2022.9264 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1124/pr.118.017160 https://doi.org/10.1007/s00213-014-3714-z https://doi.org/10.1111/j.1746-8361.1973.tb00623.x https://plato.stanford.edu/archives/sum2021/entries/reliabilism/ https://doi.org/10.1016/0039-3681(87)90005-7 https://doi.org/10.1002/1097-4679(200101)57:1%3c19::aid-jclp4%3e3.0.co;2-z https://doi.org/10.7554/elife.35082 https://doi.org/10.1007/s11229-016-1180-3 https://doi.org/10.33735/phimisci.2022.9264 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation aidan lyona (a.s.lyon@uva.nl) anya farennikovaa (a.farennikova@uva.nl) abstract what makes psychedelic psychotherapy work? is it the induction of psychedelic experience, with its distinct patterns of hallucinations and insights, or is it the neural ‘shakeup’ that moves the brain out of its regular mode of functioning and into a more disordered state? we consider the role that attention-related phenomenological changes play in psychedelic transformation and psychotherapy. we review letheby’s account of psychedelic psychotherapy, which appeals to increases in phenomenal opacity as the central mechanism of psychotherapeutic transformation. we argue that there is an alternative vehicle of psychedelic transformation that this account overlooks, involving radically transparent experiences. we outline the common kinds of phenomenal transparency shifts typical of psychedelic experiences, and argue that in many cases, such shifts are responsible for the psychotherapeutic benefits. this argument motivates an alternative approach to possible mechanisms of psychedelic self-transformation, and opens up a new venue of empirical research into the role of attention and phenomenology in psychedelic psychotherapy. keywords psychedelics ∙ philosophy ∙ psychotherapy ∙ phenomenology ∙ attention this article is part of a symposium on chris letheby’s book “philosophy of psychedelics” (oup 2021), edited by chiara caporuscio and sascha benjamin fink. what we need is a new language … that is a new kind of poetry, that’s the poetry of the dancing bee that tells us where the honey is … in order to create that language, you’re going to have to learn how you can go through a looking glass, into another kind of perception, where you have that sense of being united to all things, and suddenly you understand everything. — andre, my dinner with andre auniversity of amsterdam lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-4319-3554 https://orcid.org/0000-0002-0314-3934 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org aidan lyon & anya farennikova 2 1 introduction in philosophy of psychedelics, letheby seeks to explain the benefits of psychedelic therapy in terms that are acceptable to a strictly naturalistic worldview. in his pursuit, letheby does a wonderful job of surveying the latest scientific evidence and marshals compelling arguments for his view. while we agree with many of letheby’s arguments, our main contention is that the general view of psychedelics that he offers is rather incomplete. in this commentary, we will focus on one respect in which this is the case: letheby’s claim that psychedelics increase phenomenal opacity and that this is inextricably tied to their therapeutic and transformative effects. we’ll argue that psychedelics also regularly increase phenomenal transparency and that this is at least—if not more—distinctive of the effects of psychedelics and integral to their therapeutic and transformative effects. let’s begin, then, with defining phenomenal opacity and transparency. 2 opacity and transparency phenomenal opacity and transparency are somewhat subtle notions, so let’s start with an example to help fix our ideas. imagine you are enjoying a comedy sitcom, laughing along with the jokes and relating to the characters. now suppose that your attention, for whatever reason, is suddenly drawn to the show’s laugh track. this shift in your attention dramatically affects your experience of the show. suddenly, the jokes don’t seem as funny, and you recognise that the laugh track had been shaping your experience of the jokes, making them seem funnier than they are by themselves. in fact, the whole show may acquire a feeling of artifice—you may even feel that the characters have been replaced by the actors that play them. this change in your experience of the show corresponds to an increase in phenomenal opacity. before the change, there was a sense in which you were “seeing through” the actors to the characters, or “hearing through” the laugh track to the intended humour of the jokes. in this sense, your experience of the show was transparent. after the change, when your attention is drawn to the laugh track, there is a sense in which your experience “hits” the laugh track and stops there. in this sense, your experience has become opaque: you can’t see the show beyond the laugh track and the actors. it can be difficult to shake off this phenomenological change, but most people eventually forget about the laugh track and soon return to the world of the show with its funny jokes and relatable characters. when this happens, the initial increase in opacity is reversed, and one’s experience regains its transparency. this example helps fix our ideas, but we should also have some definitions of these notions. an immediate problem, however, is that various authors have given somewhat different definitions and embedded them within different theoretical lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org through the psychedelic looking-glass 3 frameworks. thus, letheby (2021) takes his inspiration from metzinger (2003) and frames his definitions within the predictive processing framework. before metzinger, the notions appear in the work of apter (1982, and 1989), and the idea (but not the terminology) appears to trace back to polanyi (1966). vervaeke & ferraro (2016) have also recently explicated these notions as part of a reformulation of the concept of mindfulness and analysis of mystical experience. the concept of transparency has also been regularly deployed in the philosophy of perception, particularly in connection to the debate over naive realism (martin, 2002), going back to at least the work of moore (1903). as with any substantial idea in philosophy, there are many nuances that need to be acknowledged, but this would require a longer discussion than is presently possible. so, for now, we will make do with a characterisation of transparency and opacity that’s coarse grained enough to capture their common usage. we first need to say what kind of an entity can be transparent or opaque. in short, anything mental: perceptions, imaginings, thoughts, beliefs, desires, emotions, and the underlying processes constructing them. for lack of a better term, we’ll call these items representations. while some of these items may not be representational, this choice fits well with how the notions are used in the literature. using the terminology from our sitcom example, one definition goes as follows: a representation is opaque if attention “stops” at that representation, and it is transparent if attention can “pass through” that representation to others in one’s representational matrix. although somewhat metaphorical, this definition captures the essence of the notions in a relatively neutral way. it also has intuitive appeal, as it fits well with the popular flashlight and camera-lens models of attention: attention is like light that can illuminate representations. if a representation is opaque, it prevents the light of attention from shining through it, thus keeping other representations in the dark (figure 1). we should note that this isn’t the only way to define these notions. for example, letheby says that a representation is transparent if it appears to one as reality and not as a representation, and opaque otherwise (p. 140). while slightly different, the definition relies on the same idea of “seeing through.”1 thus, we see through transparent representational structures to experiences they construct. because the underlying representations are invisible, their product appears as reality to us. vervaeke & ferraro (2016) also give a similar definition in terms of attention scaling back (opacity) and forth (transparency). we think it is reasonably clear that all authors in question have the same basic understanding in mind. we can now give some broad characterisations of what increases in opacity and transparency involve. one hallmark feature of opacity is that it tends to involve a 1the difference between this definition and ours is that the latter allows for a representation to become opaque without the subject being aware of it as a representation. arguably, this is what happens with the sitcom example: for instance, the characters appear as actors, but the actors are not experienced as mental representations—they are simply experienced as actors. modulo this minor difference, the definitions are in close alignment. lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org aidan lyon & anya farennikova 4 figure 1: attentional resources pass through transparent representations (dotted lines) and stop at opaque ones (solid lines). sense of unreality and optionality of that which is being represented. its character is somewhat akin to that of moral relativism: instead of seeing an evil act, one sees an act that can be evaluated as evil, but doesn’t have to be. in contrast, a hallmark feature of transparency is that it tends to involve a sense of reality and necessity. many of our moral judgements are like this: the wrongness of an act— such as kicking a baby—can be so immediate and so viscerally real and undeniable, that the judgement is essentially a perception and we are compelled to intervene without any thought. a second, and closely related, feature of transparency is that it tends to involve a kind of unity, absorption, or connection with one’s representations. correspondingly, opacity tends to involve duality, separateness, or disconnection. another example will help highlight this. recall the time when you first learned to ride a bike. when you begin learning, your attention is often focused on your immediate points of contact with the bike and their dynamics: the handle bars, the pedals, the bike seat, etc.2 in this early stage, you experience yourself and the bike as being quite separate (perhaps even combative!) things. however, as you learn to ride, your sensations of the bike parts dissolve, and you enter into a kind of unity or oneness with the bike. you can even begin to experience other things through the bike: bumps on the ground, the surface’s traction and inclination, etc. you become so connected with the bike, that riding becomes a kind of flow state, which is another common feature of highly transparent experiences (vervaeke & ferraro, 2016). as always, much more could be said, but we have enough detail to proceed with our discussion. 2for convenience, we have switched to a realist way of speaking, but nothing hangs on this. the example can be translated into purely representational terms. lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org through the psychedelic looking-glass 5 3 the benefits of psychedelic-induced opacity the gist of letheby’s argument is as follows. first, our default mental state is largely stuck in what we may call transparency mode: most of our mental representations are transparent. second, although a representation may be transparent, it nevertheless colours our experience—in much the same way that looking through orange-tinted glasses makes the world appear a particular way. third, some of our transparent representations colour our experience in ways that are unnecessarily unpleasant and unhealthy. fourth, by increasing the opacity of these representations, psychedelics allow us to see them as just representations, which involves seeing them as optional, which then facilitates changing them. using our glasses analogy, it’s as though the world appears “of” to us, but we don’t know why, and then we realise that this is due to the glasses, and so we remove them (and perhaps put different ones on). we agree that psychedelics have this sort of effect and that it explains many of their therapeutic benefits. psychedelic experiences frequently involve features that draw one’s attentional resources up out of one’s experiential rabbit holes and to the surface of one’s representational structures. one obvious example is the tendency for psychedelics to induce hallucinations and other forms of mental imagery. since these alterations to perceptual experience are often recognised as not being veridical, they immediately highlight the representational nature of experience (metzinger, 2003). more generally, by increasing the repertoire of conscious experience, psychedelics call attention to the fact that many of our thoughts, emotions, and perspectives, are optional and therefore can be altered. similarly, by relaxing the grip that certain beliefs have on us (carhart-harris & friston, 2019), psychedelics allow us to see that those beliefs are unwarranted and unhealthy, and that they can be revised. as letheby points out, when these beliefs are selforiented, seeing that they are optional can pave the way for an especially therapeutic transformation of the individual and their sense of self. although we agree with all of these points, we don’t think they do full justice to phenomenology of psychedelic trips, or to the causal processes involving their transformation of the individual. psychedelics do cause increases in phenomenal opacity, but this is only a fraction of a deeper and more general effect that is doing the real work—which is that psychedelics disrupt one’s attentional system. this disruption often triggers increases in opacity, but it also regularly has the opposite effect of inducing increases in transparency. in our view, the attentional system is somewhat like a coin balanced on its edge: if you tip it over, it may land heads (increased opacity), but it may equally land tails (increased transparency). indeed, if there is any bias at all, it may lean towards increased transparency. whereas certain meditative practices seem to be distinctly (but not exclusively) oriented toward increases in opacity (albahari, 2006), psychedelic trips appear to be more distinctly oriented towards increases in transparency. lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org aidan lyon & anya farennikova 6 4 psychedelic transparency here we will present five classes of examples that demonstrate that psychedelics regularly cause substantial increases in transparency and that these are integral to the transformative and therapeutic process. the first involves a radical increase in transparency of one’s representation of nature. consider the following reports: it was like being inside of nature, and i could’ve just stayed there forever—it was wonderful. all kinds of other things were coming, too, like feelings of being connected to everything, i mean, everything in nature. everything—even like pebbles, drops of water in the sea … it was like magic. it was wonderful, and it wasn’t like talking about it, which makes it an idea, it was, like, experiential. it was like being inside a drop of water, being inside of … a butterfly’s wing. and being inside of a cheetah’s eyes. (belser et al., 2017, p. 371). before i enjoyed nature, now i feel part of it. before i was looking at it as a thing, like tv or a painting. you’re part of it, there’s no separation or distinction, you are it. (watts et al., 2017, p. 534). [it] just opens you up and it connects you … it’s not just people, it’s animals, it’s trees—everything is interwoven, and that’s a big relief … i think it does help you accept death because you don’t feel alone […] that’s the number one thing—you’re just not alone. (swift et al., 2017, p. 499). such reports are extremely common, quite distinctive of psychedelic trips, and recognisable as increases in transparency. they are also often intimately bound up in the therapeutic transformation. the second class of examples is those that involve a radical increase in transparency of one’s representation of music. this is a particularly important class of examples, because music is often integral to the therapeutic process. for example, in summarising their results, belser et al. (2017) write: all 13 participants described how the music that was played during their psilocybin session took on a central role in their phenomenal experience. participants described the music as a conduit, vehicle, or carrier of their experience. (belser et al., 2017, p. 373). when we look at the reported experiences, it is clear they involve increases in transparency. for example: i felt that i was part of the music and i was in the music, it was me. it was it just me listening to music. i was the music. i was the drum, or the flute, or the violin, and i was really part of that. (belser et al., 2017, p. 374). lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org through the psychedelic looking-glass 7 the music was fantastic, actually—the music that they had. i remember at one point, i thought, ‘it’s not the drug doing this—it’s the music.’ the music felt like it was what was making things happen. (belser et al., 2017, p. 374). music was really how everything was conveyed to me, it all came through the music … like everything that i experienced did not really happen in the english language, it kind of happened through the music, like the music was the conduit for this experience to happen. (belser et al., 2017, p. 373). to summarise, psychedelics appear to plunge the individual deeply into the world of the music, absorbing them to such an extent that they become the music. simultaneously, the music becomes a process that functions as a vehicle for transformation and therapeutic change. at one point, letheby briefly recognises the tension between this observation and his view (p. 44), but he dismisses it by pointing out that while such experiences do occur, they may bookend the kind of “introvertive” opaque experience that he believes does the real therapeutic work. we agree that this may happen (sometimes), but this characterisation doesn’t do justice to the powerfully transformative process carried out by the music, and which often culminates in a highly transparent, “extrovertive” mystical experience.3 the third class of examples is those that involve increased transparency of mental imagery. while we agree that hallucinations can draw attention to the representational nature of experience (as per the previous section), it is clear that they often involve substantial increases in transparency. indeed, much like music, the visual aspect of the experience is often the vehicle for transformation. for example, in summarising their results, belser et al. (2017) write: these complex visualizations are not merely interesting scenes painted on the proscenium curtain of the theater of the mind. rather, among study participants, these visions served as principle organizing motifs of subjective experience with multifold vectors (e.g., audiovisual, relational, autobiographical, spiritual, epistemological, ontological). such visions were often deeply coded with layers of meaning by participants. (belser et al., 2017, p. 379). indeed, people become so absorbed in their visual experience that it takes a reality of its own, and becomes a world that is rich in meaning and provides a medium for personal and spiritual exploration. the fourth class of examples involves increases in transparency that give rise to an amplified sense of reality. people often report psychedelic trips as involving 3letheby doesn’t explain how he uses the term ‘introvertive,’ but we assume this is taken from stace’s distinction between introvertive and extrovertive mystical experience (stace, 1960). in contrast to letheby, stace argued that drug-induced mystical experiences are mostly extrovertive in nature. in our opinion, they can easily be either—depending on the context and individual. lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org aidan lyon & anya farennikova 8 experiences that are “realer than real” or encountering an ultimate reality (griffiths et al., 2019; yaden et al., 2017). for example: normally, human beings have a certain feel of the real. this feature is so fundamental to our perception of the world and our being in it that in general we are totally oblivious of it. there are situations, however, when our sense of reality is reduced—for instance, extreme fatigue and being in the dark. this also happens with ayahuasca. the ayahuasca inebriation, however, also presents cases in which the feel of reality is enhanced. many drinkers report that what they see during the course of the intoxication seems to be ‘more real than real.’ […] the feeling is so strong and compelling that it may be coupled by the assessment, very common with ayahuasca, that what is seen and thought during the course of the intoxication defines the real, whereas the world that is normally perceived is actually an illusion. […] indeed, the feeling is that the knowledge one gains is ultimate and veridical. it is achieved not by means of analysis and reflection, but rather by means of a direct contact, or even identification, with the objects to be known. (shanon, 2002, p. 205). clearly, this direct contact and identification with something “realer than real” is a form of increased transparency. as we noted in section 2, a hallmark feature of transparency is this feeling of being in direct contact with a reality, an immersion in or connection with that reality, and a strong sense of necessity or undeniability to what is being experienced. this relates to the noetic quality of mystical experience that william james famously identified (james, 1902, p. 293). the fifth and final class of examples builds on the previous one by flipping the light of attention around and shining it inward. that is, the sense of reality that people often encounter isn’t always oriented towards the world. it can also be directed inwards, to reveal a truer, more authentic self.4 for example, in summarising their results, noorani et al. (2018) write: session experiences were described as revealing a deeper, better, or more essential self that either led to a decreased desire to smoke, or to smoking not making sense anymore. (noorani et al., 2018, p. 759). the individual reports often demonstrate an enhanced transparency of one’s selfrepresentation, marked by a sense of self that is revealed and feeling more real than the usual one: [it was] me revealing myself, like actually showing myself to the world. this is who i am, this is who i really am. (noorani et al., 2018, p. 759). 4actually, we think that the light of attention can be diffused in both directions simultaneously— inwards and outwards—but we lack the space to develop this point here. lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org through the psychedelic looking-glass 9 i feel more in touch with who i really am—my real self, myself that’s connected to everyone and everything. (belser et al., 2017, p. 37). i just feel that definitely it [ayahuasca] got me more in touch with my spirit … it brought it out more because it was there. but now it is there even more. (argento et al., 2019, p. 784). on this point, it is worth noting that another idea that is central to letheby’s view is that psychedelics cause an unbinding of one’s self-model. indeed, for letheby, increased opacity and unbinding of the self-model go hand-in-hand (p. 183). however, experiences like those described above do not fit his interpretation. instead, psychedelics appear to cause a substantial increase in transparency of one’s selfrepresentation, so that they (seemingly) bring the individual into direct contact with some ultimate reality or truth concerning one’s self. this isn’t to say that opacity shifts related to self-unbinding are not important for self-transformation. the point, rather, is that there is another powerful mechanism for self-transformation: transparency shifts corresponding to manifestations of the self. we’ve presented five examples of how psychedelics radically increase transparency: connection with nature, unity with music, absorption in mental imagery, encountering an ultimate reality, and the revelation of a true, authentic self. these examples are highly representative of psychedelic trips, demonstrating the ubiquity of increased transparency. they also help demonstrate how increased transparency is intimately bound up with, and integral to, the therapeutic process. indeed, it is becoming increasingly recognised that disconnection, in its varying forms, is often at the core of much mental suffering (e.g., hari, 2018), and that psychedelics are effective in part because of their tendency to promote experiences of connectedness and unity (carhart-harris et al., 2018). to put our central point another way, the nomenclature of a ‘psychedelic trip’ captures something fundamental to the experience: one is taken on a journey that is itself transformative and therapeutic. along the way, one has all sorts of novel encounters that are revelatory and which transfigure the individual—much like alice’s adventures in wonderland. while we agree that increases in opacity do occur, they alone cannot explain the effects of psychedelics, nor are they especially distinctive of them. therefore, increases in transparency need to be included as an integral part of the the explanation of how psychedelics achieve their effects. 5 objections and replies let’s now consider some objections against our claim that increased transparency is integral to psychedelic transformation. objection 1: one may grant our point about transparency, but object that opacity increases are nevertheless more integral to the therapeutic process. in particular, one may argue that transparency shifts causally depend on opacity shifts. lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org aidan lyon & anya farennikova 10 we think the current qualitative data lacks the resolution that would allow us to discern such a causal dependence, but one might nevertheless try to make the argument based on evidence concerning neural processes. for example, letheby is committed to the self-unbinding mechanism and its correlation with experiences of increased opacity (p. 183). one might argue that this mechanism kickstarts the subsequent therapeutic transformation (which may involve transparency shifts). for the sake of brevity, let’s grant this point. nevertheless, it wouldn’t undermine the importance of transparency increases. consider the following analogy. in order to fly to another country for a relaxing vacation, one must first book a plane ticket. therefore, the relaxation that results from one’s vacation causally depends on having first booked a ticket. however, in explaining why people tend to feel more relaxed after such vacations, it would be odd to focus on the fact that they all booked plane tickets. clearly, the transformation lies in the experiences one has during the vacation. and so it goes, we argue, with psychedelic therapy—the transformation often lies in the journey afforded by the increased transparency. objection 2: another way to object that opacity increases are more integral to the therapeutic process is to flip the previous objection around and argue that experiences involving transparency increases provide the material for subsequent increases in opacity, which ultimately do the therapeutic work. for letheby, these increases in opacity take the form of a new awareness of the mutable nature of one’s mind or the constructed nature of one’s self. again, we agree that this often happens, but it isn’t the only way in which psychedelics transform the individual, and nor is it particularly distinctive of the psychedelic process. whereas increased opacity was like the purchasing of a plane ticket in the previous objection, here, it is more like writing a travel diary after the vacation is over. indeed, any transformative change—whether it be a vacation, psychedelic therapy, or even traditional therapy—is likely to cause a moment in which one takes a more opaque stance towards their mindset, behaviour, or beliefs. such an outcome is practically inevitable, and is itself therapeutically important, for it may help maintain the transformation. however, the crux of the change is the experience itself. what, then, explains the emphasis on opacity in letheby’s account? after all, letheby clearly recognises the importance of increased connectedness and acceptance, which are paradigmatic increases in transparency. this choice may stem from viewing transparency as our psychological default. hence the solution— stepping away from experience and disidentifying from it. this approach, however, neglects another aspect of our experience: the fact that it is often too opaque. for this reason, increasing transparency can also be beneficial and may be an important therapeutic alternative to opacity shifts. objection 3: there is another possible reason for this emphasis. increases in transparency often don’t fit well with a strictly naturalist position—e.g., when they involve revelations of spirits or unions with cosmic consciousness, etc. one may object, then, that any explanation of psychedelic therapy mustn’t involve these lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org through the psychedelic looking-glass 11 experiences as explanans. they must, therefore, be viewed as epiphenomenal, or somehow explained away. we suspect this may be part of letheby’s focus on opacity increases—they seem better suited to the naturalist program. we agree that increases in opacity may be more naturalistically acceptable— at least, at first glance. however, this doesn’t entail that transparency increases should be disregarded or downplayed. more work needs to be done before we can assess their scientific status—see, for example, taves (2020), sanders & zijlmans (2021), and breeksema & elk (2021)—and we are optimistic that many such experiences will turn out to be scientifically acceptable. with that said, one reason to be excited about psychedelic experiences is that they have the potential to expand the domain of scientific inquiry. to paraphrase stanislav grof, just as the telescope and microscope expanded astronomy and biology, the psychedelic looking-glass has the potential to expand psychology in ways that we don’t yet fathom. 6 conclusion letheby is right to emphasise that our cognitive models tend to become ossified in unhealthy forms of transparency and that psychedelics can help us see them opaquely—as optional. however, people can equally get stuck in opaque modes of representation: hiding behind screens, removed from nature, superficially sampling choices (much like the sampling encouraged by some dating apps), and shying away from an authentic engagement with life.5 increases in transparency involve deep dives into commitment, meaning, and connection. they can go wrong and are therefore risky, but human life is severely diminished without them. much of the transformative power of psychedelics lies in their ability to help us step through their looking glass into a deeper and more meaningful engagement with the world, others, and ourselves. acknowledgments we’d like to thank chris letheby, michiel van elk, logan fletcher, and ruben laukkonen for helpful discussions and feedback. references albahari, m. (2006). analytical buddhism: the two-tiered illusion of the self. palgrave macmillan. https://doi.org/10.1057/ 9780230800540 apter, m. j. (1982). metaphor as synergy. in d. s. miall (ed.), metaphor: problems and perspectives (pp. 55–70). harvester press. apter, m. j. (1989). reversal theory: a new approach to motivation, emotion and personality. anuario de psicologia/the ub journal of psychology, 42(3), 19–29. argento, e., capler, r., thomas, g., lucas, p., & tupper, k. w. (2019). exploring ayahuasca‐assisted therapy for addiction: a qualitative analysis of preliminary findings among an indigenous community in canada. drug and alcohol review, 38(7), 781–789. https://doi.org/10.1111/dar.12985 5vervaeke et al. (2017) have called this state of affairs the meaning crisis. lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1057/9780230800540 https://doi.org/10.1057/9780230800540 https://doi.org/10.1111/dar.12985 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org aidan lyon & anya farennikova 12 belser, a. b., agin-liebes, g., swift, t. c., terrana, s., devenot, n., friedman, h. l., guss, j., bossis, a., & ross, s. (2017). patient experiences of psilocybin-assisted psychotherapy: an interpretative phenomenological analysis. journal of humanistic psychology, 57(4), 354–388. https://doi.org/10.1177/0022167817706884 breeksema, j. j., & elk, m. van. (2021). working with weirdness: a response to “moving past mysticism in psychedelic science”. acs pharmacology & translational science, 4(4), 1471–1474. https://doi.org/10.1021/acsptsci.1c00149 carhart-harris, r. l., erritzoe, d., haijen, e., kaelen, m., & watts, r. (2018). psychedelics and connectedness. psychopharmacology, 235(2), 547–550. https://doi.org/10.1007/s00213-017-4701-y carhart-harris, r. l., & friston, k. j. (2019). rebus and the anarchic brain: toward a unified model of the brain action of psychedelics. pharmacological reviews, 71(3), 316–344. https://doi.org/10.1124/pr.118.017160 griffiths, r. r., hurwitz, e. s., davis, a. k., johnson, m. w., & jesse, r. (2019). survey of subjective ‘god encounter experiences’: comparisons among naturally occurring experiences and those occasioned by the classic psychedelics psilocybin, lsd, ayahuasca, or dmt. plos one, 14(4), e0214377. https://doi.org/10.1371/journal.pone.0214377 hari, j. (2018). lost connections: why you’re depressed and how to find hope. bloomsbury publishing. james, w. (1902). the varieties of religious experience: a study in human nature. longmans, green; co. https://doi.org/10 .1037/10004-000 letheby, c. (2021). philosophy of psychedelics. oxford university press. martin, m. g. f. (2002). the transparency of experience. mind & language, 17(4), 376–425. https://doi.org/10.1111/14680017.00205 metzinger, t. (2003). being no one: the self-model theory of subjectivity. mit press. https://doi.org/10.7551/mitpress/1551. 001.0001 moore, g. e. (1903). the refutation of idealism. mind, 12(4), 433–453. https://doi.org/10.1093/mind/xii.4.433 noorani, t., garcia-romeu, a., swift, t. c., griffiths, r. r., & johnson, m. w. (2018). psychedelic therapy for smoking cessation: qualitative analysis of participant accounts. journal of psychopharmacology (oxford, england), 32(7), 756– 769. https://doi.org/10.1177/0269881118780612 polanyi, m. (1966). the tacit dimension. doubleday. sanders, j. w., & zijlmans, j. (2021). moving past mysticism in psychedelic science. acs pharmacology & translational science, 4(3), 1253–1255. https://doi.org/10.1021/acsptsci.1c00097 shanon, b. (2002). the antipodes of the mind: charting the phenomenology of the ayahuasca experience. oxford university press on demand. stace, w. t. (1960). mysticism and philosophy. palgrave macmillan. swift, t. c., belser, a. b., agin-liebes, g., devenot, n., terrana, s., friedman, h. l., guss, j., bossis, a. p., & ross, s. (2017). cancer at the dinner table: experiences of psilocybin-assisted psychotherapy for the treatment of cancer-related distress. journal of humanistic psychology, 57(5), 488–519. https://doi.org/10.1177/0022167817715966 taves, a. (2020). mystical and other alterations in sense of self: an expanded framework for studying nonordinary experiences. perspectives on psychological science, 15(3), 669–690. https://doi.org/10.1177/1745691619895047 vervaeke, j., & ferraro, l. (2016). reformulating the mindfulness construct: the cognitive processes at work in mindfulness, hypnosis, and mystical states. in a. raz & m. lifshitz (eds.), hypnosis and meditation: towards an integrative science of conscious planes (pp. 241–268). oxford university press. vervaeke, j., mastropietro, c., & miscevic, f. (2017). zombies in western culture: a twenty-first century crisis. open book publishers. https://doi.org/10.11647/obp.0113 watts, r., day, c., krzanowski, j., nutt, d., & carhart-harris, r. (2017). patients’ accounts of increased “connectedness” and “acceptance” after psilocybin for treatment-resistant depression. journal of humanistic psychology, 57(5), 520–564. https://doi.org/10.1177/0022167817709585 yaden, d. b., nguyen, k. d. l., kern, m. l., belser, a. b., eichstaedt, j. c., iwry, j., smith, m. e., wintering, n. a., hood, r. w., & newberg, a. b. (2017). of roots and fruits: a comparison of psychedelic and nonpsychedelic mystical experiences. journal of humanistic psychology, 57(4), 338–353. https://doi.org/10.1177/0022167816674625 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3, 10. https://doi.org/10.33735/phimisci.2022.9323 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1177/0022167817706884 https://doi.org/10.1021/acsptsci.1c00149 https://doi.org/10.1007/s00213-017-4701-y https://doi.org/10.1124/pr.118.017160 https://doi.org/10.1371/journal.pone.0214377 https://doi.org/10.1037/10004-000 https://doi.org/10.1037/10004-000 https://doi.org/10.1111/1468-0017.00205 https://doi.org/10.1111/1468-0017.00205 https://doi.org/10.7551/mitpress/1551.001.0001 https://doi.org/10.7551/mitpress/1551.001.0001 https://doi.org/10.1093/mind/xii.4.433 https://doi.org/10.1177/0269881118780612 https://doi.org/10.1021/acsptsci.1c00097 https://doi.org/10.1177/0022167817715966 https://doi.org/10.1177/1745691619895047 https://doi.org/10.11647/obp.0113 https://doi.org/10.1177/0022167817709585 https://doi.org/10.1177/0022167816674625 https://doi.org/10.33735/phimisci.2022.9323 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction opacity and transparency the benefits of psychedelic-induced opacity psychedelic transparency objections and replies conclusion positive affect and letheby's naturalization of psychedelic therapy positive affect and letheby’s naturalization of psychedelic therapy sarah hoffmana (sarah.hoffman@usask.ca) abstract letheby’s naturalistic theory of psychedelic therapy argues that the therapeutic power of psychedelics lies in their ability to allow individuals “to discover the contingency, mutability and simulatory nature of their own sense of identity and habitual modes of attention.” the general shape of this project is persuasive; it is hard to see how the claim that successful therapy must involve changes to the self could be objected to, and letheby sketches a consistent, if speculative, picture of psychedelic experience. but the role of affect in psychedelic therapy is insufficiently explored by letheby as a comparison with mdma-assisted therapy indicates. this comparison further suggests that letheby’s reliance on a particular conceptualization of ego dissolution experiences is in need of further explanation and justification. keywords mdma therapy ∙ positive affect ∙ psychedelic ∙ psychedelic therapy ∙ self-model this article is part of a symposium on chris letheby’s book “philosophy of psychedelics” (oup 2021), edited by chiara caporuscio and sascha benjamin fink. given the resurgence of interest in psychedelic drugs as therapeutic adjuncts and in personal transformation, letheby’s naturalizing account of psychedelic experience and therapy philosophy of psychedelics is timely and welcome. the theory he presents claims that psychedelic drugs facilitate positive change via opening a window through which individuals can “learn directly that there are other ways of being, and other ways of seeing, because their ordinary ways of being and seeing result from a malleable modeling process” (letheby, 2021, p. 125). alterations in a subject’s self-model during the acute phase of their use and the capacity of these drugs to facilitate the consolidation of positive changes over the longer term are thus the mechanisms through which he asserts psychedelics can effect positive change. this approach strikes me as on the right track. for one thing, the auniversity of saskatchewan hoffman, s. (2022). positive affect and letheby’s naturalization of psychedelic therapy. philosophy and the mind sciences, 3, 9. https://doi.org/10.33735/phimisci.2022.9285 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://doi.org/10.33735/phimisci.2022.9285 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sarah hoffman 2 therapeutic effects of psychedelics are psychological in nature and thus must in some sense involve the self, which then, if we follow letheby’s view, implicates changes to the self-model. while such changes could be brought about without the causal mechanism itself being psychological or experiential, as is perhaps the case for treatments with selective serotonin reuptake inhibitors such as prozac, this appears not to be the case here. as letheby also points out, psychedelics seem to be working in a way that involves increasing the capacity for exercising mindfulnessrelated skills, which indicates the mechanism involves at least some alterations to the way someone is paying attention to their mental contents. moreover, as long as evidence continues to support the induction of a mystical-type experience as the best predictor of therapeutic effects, it makes sense that the profound changes to the sense of self that occur in psychedelic experience are part of the story of how this happens.1 understanding these experiences, uncovering which aspects of them are responsible for their benefits, and explaining just how they are responsible requires handling both neuroscientific evidence and some fairly speculative hypotheses about normal and psychedelic consciousness. the model letheby develops thus draws on a wide range of often very recent and tentative data about the effects of psychedelics on the brain, with both first-person accounts of psychedelics experiences in therapy and ideas from contemporary philosophy of mind and cognitive sciences playing an essential role. there are of necessity a lot of moving parts and speculative aspects to the theory. i won’t be challenging most of this. as already indicated i think the general shape of his project is persuasive. it is hard to see how the claim that successful therapy must involve changes to the self could be objected to, and the clinical and imaging data marshaled by letheby sketch a consistent, if speculative, picture of psychedelic experience involving changes to the default mode network that facilitate a healthier resetting of the narrative self. the claim that psychedelic experience results in increased psychological insight and increased mindfulness-related capacities is also supported by his self-representation change theory. but once the mechanism he proposes to account for this — his self-unbinding — is examined more closely, some elements of the account seem more persuasive than others. in particular, and to its detriment, the role of affect in psychedelic therapy is insufficiently explored by letheby. general considerations about the psychometrics of psychedelic-induced mystical-type experiences suggest that the positive valence of these experiences play a central role in successful psychedelic therapy. and a comparison with mdma-facilitated therapy further supports this. affect can of course be accommodated by letheby’s account given its intimate connection to the self and self-modeling. but the role it plays in psychedelic experience and the psychological processes that lead to positive change indicate to me that this is an 1i use letheby’s phrase “mystical-type experience” for simplicity and to indicate agreement that experiences scoring high enough on psychometric measures like the mystical experiences questionnaire to count as complete mystical experiences can be naturalized and need not “involve the apparent apprehension of a non-natural or supernatural reality” (letheby, 2021, p. 73). hoffman, s. (2022). positive affect and letheby’s naturalization of psychedelic therapy. philosophy and the mind sciences, 3, 9. https://doi.org/10.33735/phimisci.2022.9285 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9285 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org positive affect and letheby’s naturalization of psychedelic therapy 3 important dimension in need of more investigation. further, the relative absence of focus on feeling connects to the way letheby conceptualizes psychedelic ego dissolution as the limit case of, and so the model for, the unbinding of the self. this is absolutely central to letheby’s account given that self-unbinding is the mechanism he claims to play the major role in the positive outcomes of psychedelic experience and therapy. what i argue is thus not intended to challenge the framework through which letheby explains psychedelic experience and therapy nor his confidence that these can be naturalized and the comforting delusion objection countered. but it may be that the soundness of his approach taken at a general level is independent of some of his commitments in philosophy of mind, and it is worth teasing out elements that warrant more discussion and defense. on the one hand, why think that the role of positive valence and affect deserves more attention than letheby has given it? one reason is found right in psychedelicinduced mystical-type experiences. given that evidence currently suggests the best predictor of therapeutically positive outcomes is the induction of these experiences, it is reasonable to think that some aspects of mystical-type experiences are where we will locate the responsible psychological mechanisms, and so understanding their nature and structure is crucial. letheby’s discussion tries to come to grips with the complicated current state of the science of measuring and understanding the factors at work, but fails to feature positive valence as a demonstrable dimension of mystical-type experiences. positive mood is a measurably independent factor in the psychometrics of the mystical experiences questionnaire (meq), independent, that is, from the other three factors—transcendence of space and time, ineffability and mystical quality (barrett et al., 2015). while there is some dispute about the correct way to understand the current data, no one to my knowledge has questioned that positive mood is an independent factor and nothing letheby says challenges this. on the other hand, if positive affect is the main mechanism of psychedelic therapy one might expect the positive mood factor of the meq to be the best predictor of good therapeutic outcomes. while to my knowledge this strong claim has not been supported, there is evidence indicating that affect is nevertheless central to mystical experiences, warranting further exploration. the mystical and positive mood factors of the meq are independent, but they are also very highly correlated (barrett et al., 2015). this is unsurprising given that many items in the mystical factor seem to involve positive affect (for example, feelings of the sacred and holiness, freedom from limitations, experience of pure being, a sense of being at a spiritual height, and a sense of reverence). additionally, roseman et al. (2017) shows many of the items from the altered states of consciousness (asc) most highly correlated with clinical outcomes to be those with positively valenced feeling—notably, “i felt particularly profound” and “i experienced a profound inner peace” come in first and third. that positive valence is playing a role makes sense if we think in the abstract about positive therapeutic outcomes. for one, many of the mental health problems helped by psychedelic therapy involve dysfunctional patterns of engagement with hoffman, s. (2022). positive affect and letheby’s naturalization of psychedelic therapy. philosophy and the mind sciences, 3, 9. https://doi.org/10.33735/phimisci.2022.9285 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9285 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sarah hoffman 4 difficult material or its outright suppression and avoidance. inducing a blissful state to facilitate productive engagement with painful material might be as good a description of what happens in the psychedelic state and therapy as facilitating changes to self-representation. there is research investigating this line of thinking, some of which might be taken to conflict with my focus on positive valence because it associates negative feeling states with therapeutic outcomes. the data, however, is not univocal as to their relationship. carbonaro et al. (2016) indicates that the occurrence of negatively valanced affect as measured by the challenging experiences questionnaire is associated with good therapeutic outcomes, whereas haijen et al. (2018) shows a negative correlation between well-being and higher scores on the ceq. carhart-harris et al. (2018) indicates that peak experiences are less predictive of psychological well-being when challenging experiences also occur. roseman et al. (2017) shows that higher scores on the oceanic boundlessness factor of the asc, combined with low scores on the dread of ego dissolution factor (which includes negative affect), predict positive therapeutic outcomes. roseman et al. (2019) shows challenging experiences are beneficial if they resolve with an emotional breakthrough and sense of relief. so positive affect seems also to be playing an important role in conjunction with challenging experiences here which highlights the importance of affect to the therapeutic process, and thus the importance of expanding letheby’s theory to more explicitly consider it. ultimately, engaging with difficult material through the lens of positive valence does have to do with a person’s self-experience and thus helping alleviate problems is, as letheby has asserted, a matter of revising self-models. so what i am pointing out is consistent with his general theory. letheby say that he “is not claiming that changes to self-representation are the only psychological mechanism in psychedelic therapy” (letheby, 2021, p. 87). the same observation holds for my suggestion, which is not that positive affect independently accounts for good therapeutic outcomes, but that including it is crucial. the ultimate aim of psychedelic research is confidence we have identified all the relevant factors and understood their interaction. certainly it would be unfair to demand of letheby that he produces such an account, as the state of the field is such that no one could currently do this. my suggestion is not that, but rather that more attention to the role of affect in the process of positive change to self-representation is an essential and potentially fruitful line of inquiry. mdma-assisted therapy provides a reason to focus on positive affect in psychedelic therapy and, as i will argue below, also points to a possible weakness in letheby’s account. mdma is not a classic psychedelic, so strictly falls outside the scope of letheby’s theory. he focuses specifically on those substances which are thought to exert their psychoactive effects “primarily by mimicking the action of the neurotransmitter serotonin (5-ht) at a specific receptor subtype: the serotonin-2a (5-ht2a) receptor” (letheby, 2021, p. 9). but as with classic psychedelics, a growing data set supports the therapeutic use of mdma. mdma is pharmacologically complex, having known effects on many subtypes of serotonin receptors (including sharing with classic psychedelics action at 5-ht2a hoffman, s. (2022). positive affect and letheby’s naturalization of psychedelic therapy. philosophy and the mind sciences, 3, 9. https://doi.org/10.33735/phimisci.2022.9285 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9285 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org positive affect and letheby’s naturalization of psychedelic therapy 5 receptors); in addition to increasing levels of serotonin mdma also increases levels of dopamine and noradrenaline, and it may cause the release of oxytocin (sessa, 2018). as an adjunct to therapy it shows promise for not only the treatment of ptsd, but also anxiety and anxiety disorders (danforth et al., 2018; wolfson et al., 2020), substance use disorder (sessa et al., 2019, 2021) and, more generally, trauma-related psychosocial difficulties (brewerton et al., 2021; mithoefer et al., 2019). while it is not currently understood exactly how mdma facilitates the reduction of ptsd symptoms, suggested mechanisms involve the role of serotonin and oxytocin in social cognition, feelings of connection, and feelings of compassion, particularly self-compassion that it induces. the enhancement of feelings of sociability, trust and communicative motivation may accelerate productive interpersonal and intrapersonal therapeutic work. in particular it has also been proposed that the positive feeling states that mdma induces allow productive revisiting of fearful and negative memories without becoming overwhelmed by them. moreover, though most of the current clinical evidence for mdma-assisted therapy relates to ptsd, recent small scale and exploratory studies of mdma-assisted therapy for alcohol use disorders demonstrated not only reduction in problematic drinking but, importantly for the comparison i am making, also “improvements in quality of life, mindfulness, self-compassion, anxiety and depression scores” (sessa et al., 2019, p. 3; see also sessa et al., 2021). there are thus important parallels between classic psychedelic and mdma-assisted therapeutic experiences with positive outcomes. the underlying neurochemical mechanisms are of course very unlikely to be identical given their differing psychopharmacology, but the important point for my argument is that at the experiential level, both involve experiences that can be understood as unbinding the self in letheby’s terms, they share the induction of emotional breakthroughs and psychological insights, and, crucially, the aspect of positive valence of both seems to be an important element of the mechanism. but there is a revealing contrast between mdma and classic psychedelics in their relative capacity to cause mystical-type experiences. a recent prospective study comparing lsd and mdma in both compassionate use patients and healthy subjects indicates that while acute effects of mdma significantly increase scores on the mystical experiences questionnaire, unlike with lsd, these effects are insufficient to count as the induction of a total mystical experience (schmid et al., 2021). this comports with anecdotal accounts of mdma in which we do not find the sorts of descriptions paralleling those in classic psychedelic experience that letheby cites to support his conjecture of psychedelics “inducing global phenomenal opacity” (letheby, 2021, p. 144). i think this contrast then reveals something else in need of further explanation in letheby: his conceptualization of psychedelic ego dissolution as the limit case of the therapeutic process of self-unbinding. his model posits self-unbinding as a process in which self-related mental contents become less phenomenally transparent. this also seems to be why “simulatory nature” appears in the claim that the power of psychedelics lies instead in their ability to allow individuals “to discover the contingency, mutability and simulatory hoffman, s. (2022). positive affect and letheby’s naturalization of psychedelic therapy. philosophy and the mind sciences, 3, 9. https://doi.org/10.33735/phimisci.2022.9285 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9285 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sarah hoffman 6 nature of their own sense of identity and habitual modes of attention.” (letheby, 2021, p. 125) given the predictive processing framework to which letheby is committed, the self is a posit the neurocognitive system makes in its attempts to minimize prediction errors and in order to organize the ongoing flow of information. that is, the self performs a binding function. this framework additionally has the consequence that the self is a simulation, albeit an indirectly constrained one. according to letheby’s account, then, in the psychedelic experience the subject becomes aware of the simulatory nature of this process because these drugs disrupt the normal neurocognitive functioning that constitutes the self as a binder. the self-model is normally transparent but under psychedelics becomes more opaque and is thereby experienced as a simulatory cognitive or modeling process rather than as a self actually in the world. that the self is a mental simulation, in other words, is directly experienced. fair enough, if we are willing to endorse the predictive processing framework and the way letheby uses it to understand the nature of the self and self-modeling, something he has argued for in previous work (see letheby & gerrans, 2017). but this way of describing things requires a commitment to the predictive processing account, which opens the theory to objections coming from philosophy of mind and cognitive science more generally. it also hides an ambiguity between cognitive and phenomenal opacity. and this ambiguity complicates the idea that a direct experience of the simulatory nature of the self is doing the therapeutic work in psychedelic therapy. that psychedelics induce phenomenal opacity is a plausible way to understand certain aspects of psychedelic experience, particularly perceptual. for example, understanding phenomenal opacity in terms of the attentional availability of earlier processing stages of some process of mental representation explains what is happening in the typical morphing geometric pseudohallucinations that classic psychedelics induce (metzinger, 2003, p. 243). the phenomenology of these pseudohallucinations is then one in which our visual experience, something that is normally transparent in that we see ‘through’ it to what it represents, has become phenomenally more opaque as we have phenomenal experience of aspects of the modelling process normally unavailable to us. that psychedelics cause such experiences lends some support to the idea that they have the capacity to opacify aspects of consciousness in this way, and this is what letheby extends in his understanding of the case of full psychedelic ego dissolution as kind of selfless phenomenal consciousness. the first-person testimonial evidence can also be explained in this way, as he asserts, and this may be of significant philosophical interest in what it implies about the possibility of selfless consciousness. these experiences could be understood as the limit case of opacification in which the neurocognitive system implemented in the brain “operates under a phenomenally opaque system-model, but not under a self-model.” (metzinger, 2003, p. 566). but it isn’t clear to me that this is necessarily what is going on in self-unbinding at the experiential level in the less-than-full ego dissolution but still therapeutically effective psychedelic experiences. after all, metacognitions about the self will always make those parts of the self-model that are their objects hoffman, s. (2022). positive affect and letheby’s naturalization of psychedelic therapy. philosophy and the mind sciences, 3, 9. https://doi.org/10.33735/phimisci.2022.9285 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9285 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org positive affect and letheby’s naturalization of psychedelic therapy 7 more opaque but not typically phenomenally opaque. and if this is right, then the phenomenal opacity of the self-model is not necessary for positive therapeutic outcomes. full ego dissolution, psychedelically-induced or otherwise, is philosophically interesting and might shed some light on questions about the self, consciousness and self-consciousness, but may have less importance to psychedelic therapy than letheby seems to give it. in light of this, and given the data i discussed above implicating the importance of the affective dimensions of therapeutic psychedelic experience, a shift to the ways they contribute to changes in the self-model looks to me like a promising direction in which to develop letheby’s account. references barrett, f. s., johnson, m. w., & griffiths, r. r. (2015). validation of the revised mystical experience questionnaire in experimental sessions with psilocybin. journal of psychopharmacology, 29(11), 1182–1190. https://doi.org/10.1177/02 69881115609019 brewerton, t. d., lafrance, a., & mithoefer, m. c. (2021). the potential use of n-methyl-3,4-methylenedioxyamphetamine (mdma) assisted psychotherapy in the treatment of eating disorders comorbid with ptsd. medical hypotheses, 146, 110367. https://doi.org/10.1016/j.mehy.2020.110367 carbonaro, t. m., bradstreet, m. p., barrett, f. s., maclean, k. a., jesse, r., johnson, m. w., & griffiths, r. r. (2016). survey study of challenging experiences after ingesting psilocybin mushrooms: acute and enduring positive and negative consequences. journal of psychopharmacology, 30(12), 1268–1278. https://doi.org/10.1177/0269881116662634 carhart-harris, r. l., roseman, l., haijen, e., erritzoe, d., watts, r., branchi, i., & kaelen, m. (2018). psychedelics and the essential importance of context. journal of psychopharmacology, 32(7), 725–731. https://doi.org/10.1177/0269881118 754710 danforth, a. l., grob, c. s., struble, c., feduccia, a. a., walker, n., jerome, l., yazar-klosinski, b., & emerson, a. (2018). reduction in social anxiety after mdma-assisted psychotherapy with autistic adults: a randomized, double-blind, placebo-controlled pilot study. psychopharmacology, 235(11), 3137–3148. https://doi.org/10.1007/s00213-018-5010-9 haijen, e. c. h. m., kaelen, m., roseman, l., timmermann, c., kettner, h., russ, s., nutt, d., daws, r. e., hampshire, a. d. g., lorenz, r., & carhart-harris, r. l. (2018). predicting responses to psychedelics: a prospective study. frontiers in pharmacology, 9, 897. https://doi.org/10.3389/fphar.2018.00897 letheby, c. (2021). philosophy of psychedelics. oxford university press. letheby, c., & gerrans, p. (2017). self unbound: ego dissolution in psychedelic experience. neuroscience of consciousness, 2017(1), nix016. https://doi.org/10.1093/nc/nix016 metzinger, t. (2003). being no one: the self-model theory of subjectivity. mit press. mithoefer, m. c., feduccia, a. a., jerome, l., mithoefer, a., wagner, m., walsh, z., hamilton, s., yazar-klosinski, b., emerson, a., & doblin, r. (2019). mdma-assisted psychotherapy for treatment of ptsd: study design and rationale for phase 3 trials based on pooled analysis of six phase 2 randomized controlled trials. psychopharmacology, 236(9), 2735–2745. https://doi.org/10.1007/s00213-019-05249-5 roseman, l., haijen, e., idialu-ikato, k., kaelen, m., watts, r., & carhart-harris, r. (2019). emotional breakthrough and psychedelics: validation of the emotional breakthrough inventory. journal of psychopharmacology, 33(9), 1076–1087. https://doi.org/10.1177/0269881119855974 roseman, l., nutt, d. j., & carhart-harris, r. l. (2017). quality of acute psychedelic experience predicts therapeutic efficacy of psilocybin for treatment-resistant depression. frontiers in pharmacology, 8, 974–974. https://doi.org/10.3389/fphar. 2017.00974 schmid, y., gasser, p., oehen, p., & liechti, m. e. (2021). acute subjective effects in lsdand mdma-assisted psychotherapy. journal of psychopharmacology, 35(4), 362–374. https://doi.org/10.1177/0269881120959604 sessa, b. (2018). why mdma therapy for alcohol use disorder? and why now? neuropharmacology, 142, 83–88. https://doi.org/10.1016/j.neuropharm.2017.11.004 sessa, b., higbed, l., o’brien, s., durant, c., sakal, c., titheradge, d., williams, t. m., rose-morris, a., brew-girard, e., burrows, s., wiseman, c., wilson, s., rickard, j., & nutt, d. j. (2021). first study of safety and tolerability of 3,4-methylenedioxymethamphetamine-assisted psychotherapy in patients with alcohol use disorder. journal of psychopharmacology, 35(4), 375–383. https://doi.org/10.1177/0269881121991792 hoffman, s. (2022). positive affect and letheby’s naturalization of psychedelic therapy. philosophy and the mind sciences, 3, 9. https://doi.org/10.33735/phimisci.2022.9285 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1177/0269881115609019 https://doi.org/10.1177/0269881115609019 https://doi.org/10.1016/j.mehy.2020.110367 https://doi.org/10.1177/0269881116662634 https://doi.org/10.1177/0269881118754710 https://doi.org/10.1177/0269881118754710 https://doi.org/10.1007/s00213-018-5010-9 https://doi.org/10.3389/fphar.2018.00897 https://doi.org/10.1093/nc/nix016 https://doi.org/10.1007/s00213-019-05249-5 https://doi.org/10.1177/0269881119855974 https://doi.org/10.3389/fphar.2017.00974 https://doi.org/10.3389/fphar.2017.00974 https://doi.org/10.1177/0269881120959604 https://doi.org/10.1016/j.neuropharm.2017.11.004 https://doi.org/10.1177/0269881121991792 https://doi.org/10.33735/phimisci.2022.9285 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sarah hoffman 8 sessa, b., sakal, c., o’brien, s., & nutt, d. (2019). first study of safety and tolerability of 3,4-methylenedioxymethamphetamine (mdma)-assisted psychotherapy in patients with alcohol use disorder: preliminary data on the first four participants. bmj case reports, 12(7), e230109. https://doi.org/10.1136/bcr-2019-230109 wolfson, p. e., andries, j., feduccia, a. a., jerome, l., wang, j. b., williams, e., carlin, s. c., sola, e., hamilton, s., yazarklosinski, b., emerson, a., mithoefer, m. c., & doblin, r. (2020). mdma-assisted psychotherapy for treatment of anxiety and other psychological distress related to life-threatening illnesses: a randomized pilot study. scientific reports, 10(1), 20442–20442. https://doi.org/10.1038/s41598-020-75706-1 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. hoffman, s. (2022). positive affect and letheby’s naturalization of psychedelic therapy. philosophy and the mind sciences, 3, 9. https://doi.org/10.33735/phimisci.2022.9285 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1136/bcr-2019-230109 https://doi.org/10.1038/s41598-020-75706-1 https://doi.org/10.33735/phimisci.2022.9285 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org distinguishing absence of awareness from awareness of absence. distinguishing absence of awareness from awareness of absence matan mazora (mtnmzor@gmail.com) stephen m. fleminga,b (stephen.fleming@ucl.ac.uk) abstract contrasting brain states when subjects are aware compared to unaware of a presented stimulus has allowed researchers to isolate candidate neural correlates of consciousness. here we propose that an important next step in this research program is to investigate, perhaps paradoxically, brain states that covary with reports of absences of awareness. specifically, we propose that in order to distinguish content-specific and content-invariant neural correlates of consciousness, a distinction needs to be made between the neural correlates of awareness of stimulus absence, and the neural correlates of absence of awareness (of either stimulus presence or absence). we ground this distinction in higher-order computational models of consciousness, where the state of higher-order nodes is invariant to the specific contents of awareness. to map the different levels of these models to neurophysiological correlates, we suggest two empirical approaches – inverted designs and two-dimensional awareness reports – in which reports about awareness and stimulus presence can be dissociated. keywords absence ∙ awareness ∙ consciousness ∙ ncc this article is part of a special issue on “the neural correlates of consciousness”, edited by sascha benjamin fink. 1 introduction contrasting brain states in trials where subjects are aware or unaware of a physical stimulus (the aware-unaware contrast; a-u; also known as contrastive analysis; baars, 1993) has been a crucial tool for the study of neural correlates of consciousness (ncc). in a typical experiment, participants’ brain activity is monitored while awellcome centre for human neuroimaging, ucl bmax planck ucl centre for computational psychiatry and ageing research; department of experimental psychology, ucl mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-3601-0644 https://orcid.org/0000-0003-0233-4891 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org matan mazor and stephen m. fleming 2 they observe a series of stimuli presented at or near perceptual threshold. contrasting average neural activation during trials in which subjects did or did not have conscious awareness of the stimulus is then taken to capture the neural processes that contribute to perceptual awareness above and beyond what is necessary for subliminal processing. in the past two decades the introduction of clever experimental manipulations such as meta-contrast masking (lau & passingham, 2006) and no-report paradigms (pitts et al., 2012) has afforded further isolation of the effects of differences in conscious experience from the neural correlates of task relevance, task performance, and explicit report. inspired by the a-u contrast, computational models of visual awareness have in turn provided theoretical frameworks for the interpretation of empirical findings, focusing on what makes some stimuli available to consciousness while others remain unconscious. baars (1993) proposed that conscious awareness is the result of a broadcasting of information to a global neural workspace. drawing an analogy between consciousness and a working theatre, at any given moment only some of the actors (percepts, thoughts, memories, etc.) on a stage (working memory) are under the spotlight (attention) and visible to the audience (other cognitive faculties). this descriptive cognitive model was later grounded in candidate neural implementations, constrained by neurophysiological data (aru et al., 2019; dehaene et al., 2003). over the years, these models have been refined and challenged by findings of increased activity in a widespread frontoparietal network in studies of the a-u contrast on the one hand (consistent with implementation of long-range neural connections supporting the global workspace; del cul et al., 2007; dehaene & changeux, 2011), and through debates over the extent of changes in consciousness in patients with prefrontal lesions on the other hand (see koch et al., 2016; boly et al., 2017; odegaard et al., 2017, for recent reviews). in parallel to these empirical and theoretical advances, researchers and philosophers have identified distinct subdivisions within the overarching terminology of the ncc. marvan and polak, in this volume, provide a comprehensive overview of recent ncc classification schemes, highlighting the important distinction between content-specific and general nccs. in this article, we focus on a related distinction between content-specific and content-invariant nccs (also resembling the distinction made between differentiating and non-differentiating nccs; bayne & hohwy, 2013). we use these terms to refer to nccs that are specific or invariant to the intentional contents of consciousness. examples of content-specific nccs are the neural correlates of being aware of a face, or the neural correlates of being aware of a stimulus on a computer screen. they are content-specific because they tell us something about the contents of experience (the presence of a face in the first case, or of a stimulus in the second). on the other hand, content-invariant nccs include neural markers that distinguish different global or local states of awareness, independent of the intentional content of experience. these contentinvariant states of awareness may fluctuate even within experimental paradigms that are designed to measure the content of consciousness, and with fully awake mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org distinguishing absence of awareness from awareness of absence 3 and sober healthy participants. for instance, changes in neural excitability, attention, or beliefs about attention can affect reports of stimulus awareness, even if such factors are independent of how a stimulus is represented or encoded. as we will see later, content-invariant nccs can in some cases carry content (such as subpersonal beliefs about one’s attentional state), as long as this content is not the content of consciousness. as pointed out by others (aru et al., 2012), the a-u contrast used in experimental research does not exclusively identify either content-specific or contentinvariant aspects of the ncc. the fact that brain region x is more activated in trials where participants were aware of the stimulus may reflect a difference in conscious perceptual contents (stimulus presence or absence), and/or a difference in other content-invariant properties (for example, level of receptiveness to incoming sensory information, and/or meta-level beliefs about sensory processing, which may fluctuate and change even when awake and performing a task). the distinction between these two interpretations is especially clear when zeroing in on trials in which subjects report being unaware. let’s take a closer look at what the participant might be communicating when they report being unaware of the target. we will put aside cases in which participants’ responses do not reflect their actual state of awareness, for example due to motor errors in response. in the remaining veridical ‘unaware’ trials, participants may be expressing something along the lines of “even though i was aware of other objects and events (for example the background), the target was not in the content of my awareness” (content-specific interpretation; awareness of absence), or alternatively something more like “my awareness level was low” (content-invariant interpretation; absence of awareness). this content-invariant interpretation can be further subdivided into lapses of awareness that are global (“my global awareness level was low”) and local (“my awareness of visual stimuli in the fovea of my visual field was low”).1 borrowing from baar’s theatre analogy, the participant may be reporting that there was no actor under the spotlight (content-specific), that the spotlight was dimmed or turned off (content-invariant, global), or alternatively that the spotlight was on but directed away from where the actor was supposed to stand (content-invariant, local). in the remainder of this article, we will argue that in order to better understand the commonalities, differences, and interactions between content-specific and content-invariant nccs, more research is needed into the different causes of ‘unaware’ responses. in what follows we will unpack what this may mean both theoretically and practically. theoretically, in section 2 we describe firstand higherorder approaches to the modeling of awareness reports, and explain how these approaches handle the difference between awareness of absence and absence of awareness. practically, in section 3 we propose two experimental approaches to correlate neural activity with distinct components of higher-order computational models: inverted designs and two-dimensional reports. 1we thank an anonymous reviewer for pointing out this important distinction. mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org matan mazor and stephen m. fleming 4 2 awareness of absence is not absence of awareness a powerful approach to the modeling of awareness reports is grounded in decision theory, and specifically in signal detection theory (sdt; green & swets, 1966). in its simplest form, sdt treats percepts as scalar values on a one-dimensional ‘evidence’ axis. evidence in sdt can be thought of as mean-field activation in some sensory brain region or as the firing rate of single neurons, although no commitments are made to specific implementations. evidence tends to be higher when a stimulus is present and lower when a stimulus is absent, but only probabilistically so: sometimes a stimulus leads to weak evidence, or the absence of a stimulus to strong evidence. subjects then generate awareness reports by comparing the sensory evidence sample to an internal criterion. only samples that fall above the criterion are reported as reflecting awareness to stimulus presence (see fig. 1, panel a). already in its basic form, sdt successfully accounts for several core observations. first, not all percepts reach awareness, and sometimes illusory awareness of a stimulus occurs without a stimulus being present. these two error types are known as ‘misses’ and ‘false alarms’, respectively. second, the ability to differentiate between the presence and absence of a stimulus and the overall tendency to report being aware of a stimulus may vary independently. the former is modeled as the shape of and distance between the distributions of strength of evidence when a stimulus is present or absent (sensitivity, or 𝑑′), and the second as the position of the criterion with respect to these distributions (𝑐), both of which can be straightforwardly estimated from empirical data if equal-variance gaussians are assumed. this simple sdt model has been further extended to account for the relation between subjective reports and objective discrimination. for example, multidimensional sdt models treat percepts as points on a multi-dimensional space, with different dimensions representing different stimulus properties such as color or size (see fig. 1, panel b). king and dehaene (2014) have shown that such a model accounts for central findings on visual awareness, such as above-chance discrimination of stimuli that are reported as unseen, and the nonlinear relation between subjective visibility reports and sensory strength. importantly, the multidimensional sdt model accounted for these findings without postulating additional processes supporting conscious awareness. in the tradition of ‘perception as inference’ (helmholtz, 1948), the application of sdt models in consciousness research implicitly equates perceptual awareness reports with subpersonal beliefs about the presence of signal. similarly, reports of being unaware of a stimulus are modeled as reflecting a high posterior probability of a stimulus being absent. importantly, these models do not differentiate between having no experience of a signal, and having an experience of no signal. both are modeled as the result of a probabilistic inference where the winning hypothesis is mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org distinguishing absence of awareness from awareness of absence 5 figure 1: sdt models. of awareness reports. (a) in its traditional one-dimensional form, sdt describes awareness reports as a comparison of sensory samples against a decision criterion. sensory samples probabilistically vary as a function of signal presence. (b) in multi-dimensional sdt, samples vary along more than one dimension, and different decisions can be made by applying different decision criteria. (c) in higher-order models, awareness reports are affected by the sensory samples and by beliefs about the underlying distributions. for example, the decision criterion shifts as a function of one’s beliefs about their current attention state. ‘stimulus absence’. for a computational model to make the important distinction between absence of awareness and awareness of absence that we identify above, it must allow for content-invariant and content-specific aspects of awareness to independently vary. only then can a distinction be made between trials in which the participant was not aware of the target nor of its absence (low content-invariant consciousness), and trials in which they were clearly aware of the absence of the stimulus (high content-invariant consciousness in the absence of content-specific consciousness of the stimulus). in what follows, we provide examples of multidimensional or multi-layered computational architectures, and explain how they handle the distinction between absence of awareness and awareness of absence. one example of such a multi-dimensional computational architecture is provided by the attention schema theory (ast) of consciousness (graziano, 2013; mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org matan mazor and stephen m. fleming 6 graziano & webb, 2015). in ast, the proposition “i am aware of x” can be broken down into three parts: knowledge of myself (denoted s), knowledge of x, and finally, a schematic knowledge of the relation of ‘being aware of’ (denoted a). while states of awareness are critically dependent on a (and more specifically, on the schematic description of one’s own mental attributes, including attention), in the ast framework they are not dependent on any specific x (graziano, 2013, ch. 9). in other words, awareness of absence may be formalized as the presence of a in the absence of x, whereas absence of awareness is formalized as the absence of a (with x either present or absent)2. the utility of this distinction between absence of awareness and awareness of absence is not limited to states of pure consciousness without intentional content (which may or may not exist, or be theoretically possible in principle; metzinger, 2020), but also extends to modeling awareness of specific absences, such as being aware of the absence of a stimulus in a display. similar higher-order accounts of consciousness have posited that consciousness emerges from a higher-order representation of one’s belief state and its relation with the external world (lau, 2019), or a content-invariant representation in a higher-order state-space (fleming, 2020). like the attention schema theory, these proposals dissociate between the specific contents of consciousness and being conscious of these contents (lau & rosenthal, 2011), and as a result allow for a representation of absence that is more than the absence of a representation. some higher-order models have identified an abstract representation of presence vs. absence as the apex of a representational hierarchy (fleming, 2020). here we make a finer observation regarding the autonoetic nature of absence representations more specifically. the capacity of higher-order accounts of consciousness to model the distinction between awareness of absence and absence of awareness speaks to the idea that a representation of absence, more so than representations of presence, is in essence higher-order. the critical difference between the belief state x is absent and the absence of the belief state x is present is my counterfactual belief that i would have detected x had it been presented. this depends on my ability to represent myself and my mental states under varying conditions, a capacity that is not available to first-order models. this distinction between higher-order and perceptual aspects of consciousness can be further formalized using more detailed computational models. for example, lau (2007) proposed an extended sdt framework to explain how the phenomenon of blindsight can result from an improper placement of the decision criterion. blindsight is the loss of subjective awareness following damage to the primary visual cortex despite relatively preserved discrimination performance. according to lau’s account, damage to the striate cortex results in a global decrease in the strength of evoked responses to visual stimuli. if subjects know about this 2note that in ast, the critical content-invariant aspect of awareness a is not attention, but higherorder beliefs about one’s own attentional state. therefore, both a and x represent specific contents, but x corresponds to the content of consciousness, and a to meta-level content about one’s own state that is itself not necessarily accessible to consciousness. mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org distinguishing absence of awareness from awareness of absence 7 change, they can adjust their criterion to be more liberal and be more likely to report experiencing a stimulus even for relatively weak signals. but if subjects don’t know about this change to the signal distribution, they will not update the position of their criterion and will report not being aware of stimuli, even when these stimuli are objectively registered by the visual system as measured with 2-alternative forced choice tasks (ko & lau, 2012). lau’s original model was designed to explain a persistent change following brain damage, but a similar logic can be applied to transient fluctuations in the signal distributions and in subjects’ meta-representation of this. for example, mazor et al. (2020) modeled subjective confidence in detection decisions using an extended sdt model, where beliefs about fluctuations in overall attention states affect the expected signal strength, which in turn affects the placement of a decision criterion in a perceptual detection task (see fig. 1, panel c). in more elaborate bayesian models, meta-inference about current attentional states not only affects beliefs about world states, but also drives the active deployment of endogeneous attention (sandved smith et al., 2020). common to these higher-order models of awareness is a hierarchical structure of beliefs, with higher-level beliefs about content-invariant aspects of awareness (for example, beliefs about attentional state or expected firing rate following a lesion) informing lower-level inference about external world states (for example, the presence or absence of a stimulus). the expected precision or strength of sensory signals can be further conditioned on the interaction between attention and signal properties such as spatial location and modality. for example, when listening to a concert, we may represent our sensory precision as being high for auditory signals, but low for visual signals. similarly, when focusing on the center of a computer screen, we may represent our sensory precision as high near the center of the screen, but low at the periphery. as a result, and perhaps counterintuitively, detection threshold may be lowered for visual stimuli in the first case, or for visual stimuli in the periphery in the second case. the rationale behind this adjustment of the detection threshold as a function of higher order beliefs about sensory precision is in essence the same, regardless of whether the threshold is adjusted globally or in a more constrained manner. in both cases, the positioning of the criterion contributes to the conscious percept in a way that is independent of the actual content of the percept itself. as such, we refer to this higher level of the hierarchy as content-invariant. 3 using awareness of absence to disentangle content-specific and content-invariant aspects of consciousness in the previous section we outlined two approaches to the modeling of awareness reports within the framework of signal detection theory. both firstand higherorder models treat awareness reports as the result of subpersonal inference about mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org matan mazor and stephen m. fleming 8 the most likely state of the world given noisy sensory evidence, but only in higherorder models is this inference informed by (and potentially informs) beliefs about content-invariant aspects of awareness. we showed that this hierarchical structure endows higher-order models with a capacity to express awareness of absence. we now turn to examine how judgments of awareness of absence can be used to differentiate between content-specific and content-invariant nccs. to illustrate our proposal, we will start with an example, based on the experiment described in pitts et al. (2012). imagine that you are a participant in a psychological experiment. you just finished the first phase of the experiment, where you were asked to identify an occasional dim disc target on a red ring while your brain activity was monitored using eeg. you found this part pretty demanding but your general feeling is that you did well on the task. you are now asked whether you noticed any patterns within an array of little white lines displayed within the ring (see fig. 2). you haven’t noticed any patterns, so you tick ‘no’, but this question makes you think that you might have missed something. in the second phase you are asked to perform the same task again, but this time you are more attentive to the little white lines and you notice a few squares and diamonds appear. when analyzing the data, the researchers will contrast neural responses to squares and diamonds in the first and the second phases and attribute any differences they find to your awareness of the white shapes. in similar experimental designs an occipital-parietal eeg signal at around 200 milliseconds after shape appearance is typically more negative in those trials where participants notice the appearance of the shapes (known as visual awareness negativity or van; förster et al., 2020, see fig. 2, lower panel). in our example, the researchers could confidently classify this neural signature, if found, as the neural signature of visual awareness. this is because, due to the careful task design, other explanations of a difference in activity can be eliminated: it can’t be differences in task relevance (the shapes are irrelevant in both phases of the experiment), it can’t be the need to report (in both phases no report is needed of a single shape), and it can’t be the physical properties of the visual display (as they are matched). the only remaining explanation is that the activation is specific to awareness of the patterns emerging in the second phase while being absent in the first phase. but there are still two possible interpretations of this effect. one is that the contrast reveals brain activity that is correlated with the perceptual contents of conscious experience: geometric shapes compared to random line configurations, or more abstractly, presence of a stimulus object versus absence of a stimulus object. a different interpretation is that the contrast reveals brain activity that is correlated with content-invariant aspects of awareness that are not directly linked to any specific aspect of the display, but to the state of being aware of something. such changes in content-invariant aspects of awareness may comprise shifts in attentional state and/or metacognitive beliefs about attentional state, regardless of specific perceptual contents. mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org distinguishing absence of awareness from awareness of absence 9 figure 2: upper panel:. stimuli used in pitts et al. (2012). in the original experiment, neural responses were recorded following the occasional appearance of a square or a diamond in the central display, as a function of awareness and task relevance. in our hypothetical inverted design, neural responses are recorded following the occasional disappearance of an existing square. reproduced based on figure 1, pitts et al. (2012). lower panel: typical erp over occipital electrodes in aware-unaware contrasts (illustration). based on figure 1, förster, koivisto, and revonsuo (2020). 3.1 the inverted design to make the distinction between these two alternatives clear, imagine a novel variant of pitts’ experiment. it begins just like the original one, except that the first display you see includes a diamond shape composed of a subset of the little white lines. after the first phase, you are asked whether you noticed the disappearance (rather than appearance) of the diamond at any point during the experiment. you haven’t (you assumed the diamond was there all along), so you tick ‘no’, but this question makes you think that you might have missed something. in the second phase you are asked to perform the same task again, but this time you are more attentive to the little white lines and you notice that the diamond disappears occasionally for a brief time. what should we expect for the van component of the eeg in our alternative variant of the experiment? if the van correlates with content-invariant aspects of awareness – i.e. the state of being aware of the presence or absence of specific contents – this contrast should yield similar results to mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org matan mazor and stephen m. fleming 10 the ones in the original experiment, with a more negative erp for aware trials. however, if the van corresponds to content-specific aspects of awareness (for example, the appearance of a distinct stimulus in the visual array), we should expect a different pattern in the second experiment: the occipital-parietal signal at this time window should in fact be more similar to responses evoked by stimulus offset – stimulus disappearance – in the original experiment. this hypothetical experiment is an example of what we refer to as an inverted design, in which participants actively report the disappearance of a previously visible target stimulus. inverted designs resolve the content-specific/contentinvariant ambiguity inherent in interpreting the aware-unaware contrast by incorporating awareness of stimulus absence as an experimental condition of interest. for the awareabsence–unawareabsence contrast, a content-invariant ncc is predicted to show a similar activation profile to the more typical awarepresence– unawarepresence contrast, whereas a content-specific ncc is predicted to show divergent results (because a content-specific ncc should track the contentspecific awareness of stimulus presence). importantly, this differential prediction for content-specific and content-invariant ncc is achieved without pharmacological or invasive interventions (in contrast with aru et al., 2012; klein & barron, 2020), circumventing ethical concerns and practical challenges. furthermore, using an inverted design allows detecting content-invariance in a different, potentially stronger sense than what is afforded by using a broad set of stimuli that spans multiple semantic categories (rutiku et al., 2016). a neural marker of awareness that survives design inversion is invariant not only to various types of stimulus presence, but also to the more abstract notions of both stimulus presence and absence. inverting the typical awareness-of-presence design has the potential to advance our ability to tell apart these two classes of nccs. however one concern is that reports of stimulus disappearance may conflate detecting stimulus absence and detecting the presence of the event of stimulus disappearance (for example, by relying on sensory transients). such alternative accounts can be ruled out by introducing control conditions (for example, changes in the display that do not correspond with the appearance or disappearance of a stimulus), or by adopting gradual and illusory disappearances, where noticing the absence of a stimulus is not directly linked with any physical change to the display. recently, davidson and colleagues (2020) combined a variant of this latter approach with a clever experimental manipulation to show that a particular neurophysiological signal is coupled to content-invariant rather than content-specific aspects of awareness. ssvep (steady-state visually evoked potentials) reflect the entrainment of neural oscillations to rhythmic stimuli that can be recorded using eeg. in their design, participants were asked to report the illusory disappearance of peripheral targets, driven either by perceptual filling-in or by phenomenally matched physical disappearance. by allowing the background and targets to flicker at different frequencies, davidson and colleagues could separately quanmazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org distinguishing absence of awareness from awareness of absence 11 tify ssveps induced by the target and background both before and during the disappearance event. consistent with a content-invariant interpretation of the ssvep, they found that ssveps for both target and background increased before the illusory disappearance of targets. if these ssveps were tracking the perceptual content of visual awareness, entrainment to the target frequency should have decreased before reports of disappearance, given that representations of specific content should presumably have become weaker when such content is absent. instead, an increase in ssveps prior to target disappearance is in line with them tracking the focus of attention, but also more broadly with tracking content-invariant aspects of visual awareness. increased alpha suppression (an independent marker of attentional effort) prior to events of illusory disappearance provided further support for this content-invariant interpretation. in summary, by inverting the a-u contrast and asking participants to report the disappearance of visual targets, davidson and colleagues were able to decide between content-specific and contentinvariant interpretations of a candidate ncc. 3.2 two-dimensional report scheme in typical experimental designs, awareness reports are given on a one-dimensional scale, from reports of minimal awareness of a stimulus, to full awareness of a stimulus. for instance, the commonly used “perceptual awareness scale” (pas) proposes four categories of experience going from “no experience”, to “brief glimpse” to “almost clear experience” to “clear experience” (ramsøy & overgaard, 2004; sandberg & overgaard, 2015). as we have seen, the problem with such a scale is that it conflates both content-specific and content-invariant aspects of awareness: a report of “no experience” can be consistent with both absence of awareness and awareness of absence. an alternative to such one-dimensional report schemes tackles this twodimensional nature of awareness reports head-on, and allows participants to report both first-order perceptual contents, and their second-order assessment of awareness of these contents. this approach allows participants to report being fully aware of the absence of a stimulus, for example in the context of a visual detection task (see fig. 3). we are not familiar with previous studies that have explicitly incorporated such a two-dimensional awareness report scheme. however, a related approach is to elicit visual detection judgments in conjunction with subjective confidence ratings (kanai et al., 2010; kellij et al., 2020; mazor et al., 2020; meuwese et al., 2014). such a design allows subjects to report both high or low confidence in stimulus presence, and also high or low confidence in stimulus absence. these studies have consistently found generally lower confidence ratings for judgments about stimulus absence, even when these judgments are correct. participants also exhibit poorer metacognitive ability to discriminate between missing an existing stimulus and correctly identifying the absence of that stimulus. this is consistent with an interpretation of ‘no’ mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org matan mazor and stephen m. fleming 12 figure 3: (a) in a typical aware-unaware contrast, trials in which participants perceived the stimulus are contrasted with trials in which they failed to perceive it. in this design both higher-order and perceptual states of awareness are interwined. (b) a twodimensional report scheme allows a contrast of trials in which participants were aware of stimulus absence with trials in which they were unaware of stimulus presence (or absence). responses as mostly reflecting absence of awareness of a stimulus, rather than (accurate) awareness of absence. there was, however, one intriguing exception to this pattern. when stimulus visibility was manipulated using attentional manipulations such as the attentional blink or spatial uncertainty, rather than manipulations of the stimulus itself, kanai et al. (2010) found that participants were more confident in correct-rejection trials compared to misses (and as a result had better metacognitive sensitivity for reporting absence). one interpretation of this effect is that with their attention fluctuating, participants were better able to know when they were likely to have missed the presence of a target, and therefore also whether they were likely to have correctly perceived stimulus absence. in other words, these attentional manipulations enhance the dependence of stimulus visibility on content-invariant aspects of consciousness, such that they begin to influence behavioural reports. this focus on awareness of absence under disturbed attention neatly reveals a potential contribution of beliefs about one’s attention to the subjective content of awareness. the use of similar paradigms in conjunction with neuroimaging may open up new avenues for dissociating content-specific and content-invariant nccs. 4 a neurofunctional separation between contentinvariance and content-specificity which nccs are sensitive to the presence or absence of an actor, and which to the presence or absence of a spotlight? based on currently available data, we can mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org distinguishing absence of awareness from awareness of absence 13 make some crude guesses. sensory regions are likely to represent local contents of awareness, such as subpersonal beliefs about stimulus presence or absence (boly et al., 2017; koch et al., 2016). in contrast, lateral and anterior aspects of prefrontal cortex may track higher-order, global aspects of conscious states, and have been associated both with reality monitoring and criterion-setting (del cul et al., 2009; lau & rosenthal, 2011; simons et al., 2017; vallesi, 2012). more specifically, associations between activation in the lateral prefrontal cortex and subjective awareness may reflect beliefs about global states of attention and the translation of these beliefs into policy changes (such as setting the criterion for awareness reports), rather than beliefs about the presence of a specific target stimulus. consistent with a role for these regions in attention monitoring and criterion setting, activation in parietal and prefrontal cortical areas have been shown to decrease not only in trials where participants reported a stimulus to be highly visible, but also in trials where participants reported the stimulus not to be visible at all, compared to intermediate subjective visibility ratings (binder et al., 2017; christensen et al., 2006). this nonmonotonic effect may plausibly reflect a projection of a monotonic effect of a higher-order construct (such as beliefs about sensory precision or subjective confidence) onto a content-specific ‘visibility’ dimension. for example, trials in which participants report extremely low or high visibility ratings may be similar in that in both cases participants believe they were attentive and expected high fidelity sensory signals (high precision, in the parlance of predictive coding models) as a result. indeed, when given the option to report stimulus presence and absence independently from subjective confidence, modulation of prefrontal activation by confidence was highly similar for ‘yes’ and ‘no’ responses in a perceptual detection task (mazor et al., 2020). a role for the prefrontal cortex in higher-order monitoring of internal states is also consistent with relative increases in prefrontal activation in paradigms that require explicit report, compared to no-report paradigms (frassle et al., 2014). the hypothesis that the lateral prefrontal cortex is representing these higher-order aspects of awareness rather than content-specific percepts can be directly tested using an inverted design or a two-dimensional report scheme similar to the ones we discuss above. 5 conclusion a popular approach in the quest for the ncc has been to contrast aware and unaware trials during near-threshold perception. however, such an aware-unaware contrast conflates neural correlates of transient fluctuations in content-invariant aspects of consciousness with neural correlates of the specific perceptual contents of consciousness. this problem is most pertinent in trials where participants report being unaware of a stimulus, as such trials can reflect either awareness of absence or absence of awareness. here we advocate for the adoption of theoretical frameworks that make this distinction explicit, such as higher-order or multi-level models. to map the different levels of these models to neurophysiological markers, mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org matan mazor and stephen m. fleming 14 we suggest the use of tasks in which beliefs about stimulus presence and absence are manipulated independently of reports of awareness. acknowledgments we thank the two anonymous reviewers for their illuminating comments and suggestions. this work was supported by a philip leverhulme prize from the leverhulme trust. s.m.f. is supported by a sir henry dale fellowship jointly funded by the wellcome trust and the royal society (206648/z/17/z). the wellcome centre for human neuroimaging is supported by core funding from the wellcome trust (203147/z/16/z). references aru, j., bachmann, t., singer, w., & melloni, l. (2012). distilling the neural correlates of consciousness. neuroscience & biobehavioral reviews, 36(2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.12.003 aru, j., suzuki, m., rutiku, r., larkum, m. e., & bachmann, t. (2019). coupling the state and contents of consciousness. frontiers in systems neuroscience, 13, 43. https://doi.org/10.3389/fnsys.2019.00043 baars, b. j. (1993). a cognitive theory of consciousness. cambridge university press. bayne, t., & hohwy, j. (2013). consciousness: theoretical approaches. in a. e. cavanna, a. nani, h. blumenfeld, & s. laureys (eds.), neuroimaging of consciousness (pp. 23–35). springer berlin heidelberg. binder, m., gociewicz, k., windey, b., koculak, m., finc, k., nikadon, j., derda, m., & cleeremans, a. (2017). the levels of perceptual processing and the neural correlates of increasing subjective visibility. consciousness and cognition, 55, 106–125. https://doi.org/10.1016/j.concog.2017.07.010 boly, m., massimini, m., tsuchiya, n., postle, b. r., koch, c., & tononi, g. (2017). are the neural correlates of consciousness in the front or in the back of the cerebral cortex? clinical and neuroimaging evidence. the journal of neuroscience, 37(40), 9603–9613. https://doi.org/10.1523/jneurosci.3218-16.2017 christensen, m. s., ramsøy, t. z., lund, t. e., madsen, k. h., & rowe, j. b. (2006). an fmri study of the neural correlates of graded visual perception. neuroimage, 31(4), 1711–1725. https://doi.org/10.1016/j.neuroimage.2006.02.023 davidson, m. j., mithen, w., hogendoorn, h., van boxtel, j. j., & tsuchiya, n. (2020). the ssvep tracks attention, not consciousness, during perceptual filling-in. elife, 9, e60031. https://doi.org/10.7554/elife.60031 dehaene, s., & changeux, j.-p. (2011). experimental and theoretical approaches to conscious processing. neuron, 70(2), 200–227. https://doi.org/10.1016/j.neuron.2011.03.018 dehaene, s., sergent, c., & changeux, j.-p. (2003). a neuronal network model linking subjective reports and objective physiological data during conscious perception. proceedings of the national academy of sciences, 100(14), 8520–8525. https://doi.org/10.1073/pnas.1332574100 del cul, a., baillet, s., & dehaene, s. (2007). brain dynamics underlying the nonlinear threshold for access to consciousness. plos biology, 5(10), e260. https://doi.org/10.1371/journal.pbio.0050260 del cul, a., dehaene, s., reyes, p., bravo, e., & slachevsky, a. (2009). causal role of prefrontal cortex in the threshold for access to consciousness. brain, 132(9), 2531–2540. https://doi.org/10.1093/brain/awp111 fleming, s. m. (2020). awareness as inference in a higher-order state space. neuroscience of consciousness, 2020(1), niz020. https://doi.org/10.1093/nc/niz020 förster, j., koivisto, m., & revonsuo, a. (2020). erp and meg correlates of visual consciousness: the second decade. consciousness and cognition, 80, 102917. https://doi.org/10.1016/j.concog.2020.102917 frassle, s., sommer, j., jansen, a., naber, m., & einhauser, w. (2014). binocular rivalry: frontal activity relates to introspection and action but not to perception. journal of neuroscience, 34(5), 1738–1747. https://doi.org/10.1523/jneuro sci.4403-13.2014 graziano, m. s. a. (2013). consciousness and the social brain. oxford university press. graziano, m. s. a., & webb, t. w. (2015). the attention schema theory: a mechanistic account of subjective awareness. frontiers in psychology, 06. https://doi.org/10.3389/fpsyg.2015.00500 green, d. m., & swets, j. a. (1966). signal detection theory and psychophysics. john wiley. helmholtz, h. von. (1948). concerning the perceptions in general, 1867. in w. dennis (ed.), readings in the history of psychology (pp. 214–230). appleton-century-crofts. mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neubiorev.2011.12.003 https://doi.org/10.3389/fnsys.2019.00043 https://doi.org/10.1016/j.concog.2017.07.010 https://doi.org/10.1523/jneurosci.3218-16.2017 https://doi.org/10.1016/j.neuroimage.2006.02.023 https://doi.org/10.7554/elife.60031 https://doi.org/10.1016/j.neuron.2011.03.018 https://doi.org/10.1073/pnas.1332574100 https://doi.org/10.1371/journal.pbio.0050260 https://doi.org/10.1093/brain/awp111 https://doi.org/10.1093/nc/niz020 https://doi.org/10.1016/j.concog.2020.102917 https://doi.org/10.1523/jneurosci.4403-13.2014 https://doi.org/10.1523/jneurosci.4403-13.2014 https://doi.org/10.3389/fpsyg.2015.00500 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org distinguishing absence of awareness from awareness of absence 15 kanai, r., walsh, v., & tseng, c.-h. (2010). subjective discriminability of invisibility: a framework for distinguishing perceptual and attentional failures of awareness. consciousness and cognition, 19(4), 1045–1057. https://doi.org/10.1 016/j.concog.2010.06.003 kellij, s., fahrenfort, j., lau, h., peters, m. a. k., & odegaard, b. (2020). an investigation of how relative precision of target encoding influences metacognitive performance. attention, perception, & psychophysics. https://doi.org/10.3758/s134 14-020-02190-0 king, j.-r., & dehaene, s. (2014). a model of subjective report and objective discrimination as categorical decisions in a vast representational space. philosophical transactions of the royal society b: biological sciences, 369(1641), 20130204. https://doi.org/10.1098/rstb.2013.0204 klein, c., & barron, a. b. (2020). how experimental neuroscientists can fix the hard problem of consciousness. neuroscience of consciousness, 2020(1), niaa009. https://doi.org/10.1093/nc/niaa009 ko, y., & lau, h. (2012). a detection theoretic explanation of blindsight suggests a link between conscious perception and metacognition. philosophical transactions of the royal society b: biological sciences, 367(1594), 1401–1411. https: //doi.org/10.1098/rstb.2011.0380 koch, c., massimini, m., boly, m., & tononi, g. (2016). neural correlates of consciousness: progress and problems. nature reviews neuroscience, 17(5), 307–321. https://doi.org/10.1038/nrn.2016.22 lau, h. (2019). consciousness, metacognition, & perceptual reality monitoring. psyarxiv. https://doi.org/10.31234/osf.io/ ckbyf lau, h. c. (2007). a higher order bayesian decision theory of consciousness. in progress in brain research (vol. 168, pp. 35–48). elsevier. https://doi.org/10.1016/s0079-6123(07)68004-2 lau, h. c., & passingham, r. e. (2006). relative blindsight in normal observers and the neural correlate of visual consciousness. proceedings of the national academy of sciences, 103(49), 18763–18768. https://doi.org/10.1073/pnas.0607716103 lau, h., & rosenthal, d. (2011). empirical support for higher-order theories of conscious awareness. trends in cognitive sciences, 15(8), 365–373. https://doi.org/10.1016/j.tics.2011.05.009 mazor, m., friston, k. j., & fleming, s. m. (2020). distinct neural contributions to metacognition for detecting, but not discriminating visual stimuli. elife, 9, e53900. https://doi.org/10.7554/elife.53900 metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 1–44. https://doi.org/10.33735/phimisci.2020.i.46 meuwese, j. d. i., loon, a. m. van, lamme, v. a. f., & fahrenfort, j. j. (2014). the subjective experience of object recognition: comparing metacognition for object detection and object categorization. attention, perception, & psychophysics. https: //doi.org/10.3758/s13414-014-0643-1 odegaard, b., knight, r. t., & lau, h. (2017). should a few null findings falsify prefrontal theories of conscious perception? the journal of neuroscience, 37(40), 9593–9602. https://doi.org/10.1523/jneurosci.3217-16.2017 pitts, m. a., martínez, a., & hillyard, s. a. (2012). visual processing of contour patterns under conditions of inattentional blindness. journal of cognitive neuroscience, 24(2), 287–303. https://doi.org/10.1162/jocn_a_00111 ramsøy, t. z., & overgaard, m. (2004). introspection and subliminal perception. phenomenology and the cognitive sciences, 3(1), 1–23. https://doi.org/10.1023/b:phen.0000041900.30172.e8 rutiku, r., aru, j., & bachmann, t. (2016). general markers of conscious visual perception and their timing. frontiers in human neuroscience, 10. https://doi.org/10.3389/fnhum.2016.00023 sandberg, k., & overgaard, m. (2015). using the perceptual awareness scale (pas). in m. overgaard (ed.), behavioral methods in consciousness research (pp. 181–196). oxford university press. sandved smith, l., hesp, c., lutz, a., mattout, j., friston, k., & ramstead, m. (2020). towards a formal neurophenomenology of metacognition: modelling meta-awareness, mental action, and attentional control with deep active inference. psyarxiv. https://doi.org/10.31234/osf.io/5jh3c simons, j. s., garrison, j. r., & johnson, m. k. (2017). brain mechanisms of reality monitoring. trends in cognitive sciences, 21(6), 462–473. https://doi.org/10.1016/j.tics.2017.03.012 vallesi, a. (2012). organisation of executive functions: hemispheric asymmetries. journal of cognitive psychology, 24(4), 367–386. https://doi.org/10.1080/20445911.2012.678992 mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.concog.2010.06.003 https://doi.org/10.1016/j.concog.2010.06.003 https://doi.org/10.3758/s13414-020-02190-0 https://doi.org/10.3758/s13414-020-02190-0 https://doi.org/10.1098/rstb.2013.0204 https://doi.org/10.1093/nc/niaa009 https://doi.org/10.1098/rstb.2011.0380 https://doi.org/10.1098/rstb.2011.0380 https://doi.org/10.1038/nrn.2016.22 https://doi.org/10.31234/osf.io/ckbyf https://doi.org/10.31234/osf.io/ckbyf https://doi.org/10.1016/s0079-6123(07)68004-2 https://doi.org/10.1073/pnas.0607716103 https://doi.org/10.1016/j.tics.2011.05.009 https://doi.org/10.7554/elife.53900 https://doi.org/10.33735/phimisci.2020.i.46 https://doi.org/10.3758/s13414-014-0643-1 https://doi.org/10.3758/s13414-014-0643-1 https://doi.org/10.1523/jneurosci.3217-16.2017 https://doi.org/10.1162/jocn_a_00111 https://doi.org/10.1023/b:phen.0000041900.30172.e8 https://doi.org/10.3389/fnhum.2016.00023 https://doi.org/10.31234/osf.io/5jh3c https://doi.org/10.1016/j.tics.2017.03.012 https://doi.org/10.1080/20445911.2012.678992 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org matan mazor and stephen m. fleming 16 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. mazor, m., & fleming, s. m. (2020). distinguishing absence of awareness from awareness of absence. philosophy and the mind sciences, 1(ii), 6. https://doi.org/10.33735/phimisci.2020.ii.69 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.69 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction awareness of absence is not absence of awareness using awareness of absence to disentangle content-specific and content-invariant aspects of consciousness the inverted design two-dimensional report scheme a neurofunctional separation between content-invariance and content-specificity conclusion cotard syndrome, self-awareness, and i-concepts cotard syndrome, self-awareness, and i-concepts rocco j. gennaroa (rjgennaro@usi.edu) abstract various psychopathologies of self-awareness, such as somatoparaphrenia and thought insertion in schizophrenia, might seem to threaten the viability of the higher-order thought (hot) theory of consciousness since it requires a hot about one’s own mental state to accompany every conscious state. the hot theory of consciousness says that what makes a mental state a conscious mental state is that there is a hot to the effect that “i am in mental state m.” i have argued in previous work that a hot theorist can adequately respond to this concern with respect to somatoparaphrenia and thought insertion. there is also cotard syndrome which is a rare neuropsychiatric disorder in which people hold the delusional belief that they are dead, do not exist, or have lost their blood or internal organs. in this paper, i argue that hot theory has nothing to fear from it either and can consistently account for what happens in such unusual cases. i analyze cotard syndrome in light of my previous discussion of somatoparaphrenia and thought insertion, and argue that hot theory can provide a somewhat analogous account without the worry of inconsistency. it is crucial to recognize that there are multiple “self-concepts” and levels of hots which can help to provide a more nuanced explanation. with regard to the connection between consciousness and self-consciousness, it is proposed that cotard patients are indeed capable of having some “i-thoughts” about their bodies and mental states. keywords consciousness ∙ cotard syndrome ∙ higher-order thoughts ∙ i-concepts ∙ self-awareness ∙ somatoparaphrenia ∙ thought insertion this article is part of a special issue on “radical disruptions of self-consciousness”, edited by thomas metzinger and raphaël millière. 1 introduction various psychopathologies of self-awareness, such as somatoparaphrenia and thought insertion in schizophrenia, might seem to threaten the viability of auniversity of southern indiana gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-8053-0279 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org rocco j. gennaro 2 the higher-order thought (hot) theory of consciousness since it requires a hot about one’s own mental state to accompany every conscious state (which might also be viewed as some form of self-consciousness). the hot theory of consciousness says that what makes a mental state a conscious mental state is that there is a hot to the effect that “i am in mental state m” (rosenthal, 2005). i have argued in previous work that a hot theorist can adequately respond to this concern with respect to somatoparaphrenia and thought insertion (gennaro, 2015, 2020). somatoparaphrenia is a monothematic delusion and a disorder of self-awareness characterized by the sense of alienation from parts of one’s body (vallar & ronchi, 2009). it is a bizarre body delusion where one denies ownership of a limb or an entire side of one’s body. schizophrenia is a mental disorder which most commonly manifests itself through auditory hallucinations, paranoid or bizarre delusions, or disorganized speech and thinking. thought insertion is the delusion that some thoughts are not “one’s own” in some sense and are inserted into one’s mind by someone else (frith, 1992). thus, it may seem that a hot with a self-referential “i” in its content could not be present. there is also cotard syndrome which is a rare neuropsychiatric disorder in which people hold the delusional belief that they are dead (either figuratively or literally), do not exist, are putrefying, or have lost their blood or internal organs (billon, 2016). once again, there seems to be a stunning abnormality of self-awareness in these patients which may also shed light on the very connection between conscious states and self-consciousness. although it will be useful to compare and contrast cotard syndrome with somatoparaphrenia and thought insertion, i will argue that hot theory also has nothing to fear from it and can consistently account for what happens in such unusual cases. more specifically, in section 2, i explain hot theory emphasizing the features which play a prominent role in later sections. in section 3, i present a relatively brief discussion of somatoparaphrenia and thought insertion in order to set the stage for addressing cotard syndrome. in section 4, i analyze cotard syndrome in light of the previous section and argue that hot theory can provide a somewhat analogous account without fear of inconsistency. it will also be crucial to recognize that there are multiple “self-concepts” (that is, “i-concepts”) and levels of hots which can help to provide a more nuanced explanation of the relationship between conscious mental states and self-awareness. in section 5, i shift to a related and important ambiguity in self-awareness with respect to the problem of personal identity. in section 6, i offer some concluding thoughts. 2 hot theory one question that should be answered by any theory of consciousness is: what makes a mental state a conscious mental state? there is a long tradition that has attempted to understand consciousness in terms of some kind of higher-order awareness. for example, john locke (1689) once said that “consciousness is the percepgennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org cotard syndrome, self-awareness, and i-concepts 3 tion of what passes in a man’s own mind.” this intuition has been revived by a number of contemporary philosophers (lycan, 1996; rosenthal, 1997, 2005). the general idea is that what makes a mental state conscious is that it is the object of some kind of higher-order representation (hor). a mental state m becomes conscious when there is a hor of m. so, for example, my desire to drink some water becomes conscious when i am (non-inferentially) “aware” of the desire. intuitively, it seems that conscious states, as opposed to unconscious ones, are mental states that i am “aware of” in some sense. this is sometimes referred to as the transitivity principle (tp): (tp) a conscious state is a state whose subject is, in some way, aware of being in. conversely, the idea that i could be having a conscious state while totally unaware of being in that state seems odd or perhaps even contradictory. a mental state of which the subject is completely unaware is clearly an unconscious state. for example, i would not be aware of having a subliminal perception and thus it is an unconscious perception. any theory which attempts to explain consciousness in terms of higher-order states is known as a higher-order representational theory of consciousness. the most common division of hor theories is between higherorder thought (hot) theories and higher-order perception (hop) theories. hot theorists, such as david rosenthal, think it is better to understand the hor as thoughts involving some kind of conceptual component. hop theorists urge that the hor is a perceptual or experiential state of some kind which does not require the kind of conceptual content invoked by hot theorists (lycan, 1996). although hot and hop theorists agree on the need for a hor theory of consciousness, they disagree on the superiority of their respective positions (such as in rosenthal, 2004). i focus on hot theory here because i have defended the theory in previous publications, but also because it is the hot concepts that generate the apparent objections and hot theory lends itself to a far more interesting discussion regarding varieties of self-awareness. it should be noted that higher-order thought theorists agree that one must become aware of the lower-order (lo) mental state noninferentially. we might suppose with rosenthal, for example, that the hot must be caused noninferentially by the lo state in order to make it conscious. the subject is directly aware of the lo state. the point of this condition is mainly to rule out some alleged counterexamples to hot theory, such as cases where i become aware of my unconscious desire to kill my boss because i have consciously inferred it from a session with a psychiatrist, or where my envy becomes conscious after making inferences based on my own behavior. the characteristic feel of such a conscious desire or envy may be absent in these cases, but since awareness of them arose via conscious inference, the hot theorist accounts for them by adding this noninferential condition. it might seem that hot theory results in circularity by defining consciousness in terms of hots. it also might seem that an infinite regress results because a conscious mental state must be accompanied by a hot, which, in turn, must be gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org rocco j. gennaro 4 accompanied by another hot ad infinitum. however, the widely accepted reply is that when a conscious mental state is a first-order world-directed state the higherorder thought (hot) is not itself conscious. when the hot is itself conscious, there is a yet higher-order (or third-order) thought directed at the second-order state. in this case, we have introspection which involves a conscious hot directed at an inner mental state. when one introspects, one’s attention is directed back into one’s mind. as we shall see, this distinction is crucial to the discussion which follows. it is worth mentioning that most versions of representationalism are reductionistic regarding consciousness which, for many, is part of the attraction. one motivation for representationalism is that a naturalistic account of intentionality or representation can arguably be more easily attained, such as via causal theories whereby mental states are understood as representing outer objects by virtue of some reliable causal connection. the idea, then, is that if consciousness can be explained in representational terms and representation can be understood in purely physical terms, then there is the promise of a naturalistic theory of consciousness. most representationalists think that there is then room for a “second-step” reduction to be filled in later by neuroscience. hot theorists also tend to endorse this line of argument (gennaro, 2012, chapter two). 3 somatoparaphrenia and thought insertion somatoparaphrenia is a bizarre delusion characterized by the sense of alienation from parts of one’s body (vallar & ronchi, 2009). patients with it often deny ownership of a limb or an entire side of their body. somatoparaphrenia might seem to threaten the viability of the higher-order thought (hot) theory of consciousness since it requires a hot about one’s own mental state to accompany every conscious state. liang and lane (2009), for example, initially argued that at least one specific case of somatoparaphrenia is problematic for hot theory because it contradicts the notion that, when i am in a conscious state, i must have the accompanying hot that “i am in mental state m.” the “i” is not only importantly self-referential but essential in tying the conscious state to oneself and thus to one’s ownership of m. rosenthal (2010) responds that one can be aware of bodily sensations in two ways which, normally at least, go together: (a) awareness of a bodily sensation as one’s own, and (b) awareness of a bodily sensation as having some bodily location, like a hand or foot. patients with somatoparaphrenia still experience the sensation as their own but also as having a mistaken bodily location. these patients still do have the awareness in (a), which is the main issue at hand, but they have a very strange awareness in sense (b) which leads them to misidentify the bodily location of the sensation in someone else. lane and liang (2010) protest that rosenthal has still not explained why the identification of the bearer of the sensation cannot also go astray in somatoparaphenia, especially since rosenthal clearly holds that gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org cotard syndrome, self-awareness, and i-concepts 5 misrepresentation can otherwise occur between a hot and its target mental state. a hot theorist might also argue that there are really two conscious states that seem to be at odds (gennaro, 2015). there is a conscious feeling in a limb but also the (conscious) attribution of the limb to someone else. it is crucial to emphasize that somatoparaphrenia is often characterized as a delusion of belief under the broader category of anosognosia (prigatano, 2010). a delusion is often defined as a false belief that is held based on an incorrect (and probably unconscious) inference about external reality or one self that is firmly sustained despite what almost everyone else believes and despite what constitutes incontrovertible and obvious proof or evidence to the contrary (bortolotti, 2009). beliefs are themselves often taken to be intentional states integrated with other beliefs and mental states. they are typically understood to be caused by perceptions or experiences which, in turn, lead to action or behavior. thus, somatoparaphrenia seems closer to self-deception and involves frequent confabulation. if this is a reasonable interpretation, then a hot theorist can argue that the patient actually has the following two conscious states: s1: a conscious feeling (i.e., a tactile sensation) in the limb in question, and s2: a conscious belief that the limb (and thus sensation) belongs to someone else. having both s1 and s2, especially if both conscious at any given time, is indeed strange and perhaps even contradictory in some sense but the puzzlement has nothing to do with hot theory itself (but see also lane, 2015, for much more on this debate). the patient would simply have a different unconscious hot directed at s1 and s2. schizophrenia is a mental disorder which most commonly manifests itself through auditory hallucinations, paranoid or bizarre delusions, or disorganized speech and thinking. thought insertion is the delusion that some thoughts are not “one’s own” in some sense or are being inserted into one’s mind by someone else (frith, 1992). the puzzle of thought insertion is thus perhaps somewhat analogous to somatoparaphrenia. for example, just as we might distinguish between experiencing a sensation as one’s own as opposed to its bodily location in patients with somatoparaphrenia, so we might distinguish between experiencing a thought as one’s own as opposed to its causal origin in patients with schizophrenia. those with somatoparaphrenia attribute a limb or a sensation to someone else whereas some schizophrenics attribute thoughts (or at least their causal origin) to someone else. if this rough similarity is correct, then a hot theorist might argue that the schizophrenic patient has the following two conscious states: s1: a conscious thought in one’s own mind, and s2: a conscious belief that the thought in s1 has been inserted by someone else. gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org rocco j. gennaro 6 having both s1 and s2, especially if both are conscious at any given time, is again very strange and perhaps even irrational, but the puzzlement is not a problem for hot theory. in s1, we can suppose that the patient has a hot about her own firstorder conscious thought, i.e. she is aware of a mental state occurring in her mind. in s2, the patient also has a conscious belief that s1 has been inserted into her mind by someone else. once again, and analogous to somatoparaphrenia, there is an essentially delusional element involved in schizophrenia and inserted thoughts. why or how s2 is generated is of course an interesting and difficult question in its own right. for example, as stephens and graham (2000) suppose, perhaps the best explanation is that s2 is consciously or unconsciously inferred from a lack of sense of agency (or “passivity” experience) or perhaps it is inferred from a perceived lack of causal origin. but none of the alternative explanations cause trouble specifically for hot theory’s ability to explain s1. just as hot theory says about all conscious states, when a patient is aware of having a (first-order) conscious thought, there is a hot about that thought (gennaro, 2020). in both of the above disorders, there is a conscious thought (s1) which can be explained by a hot theorist in the usual way, but there is a second belief or thought which apparently contradicts the first one (s2). recall also that hot theory comes with a well-known noninferentiality condition, such that a hot must involve awareness of its target mental state noninferentially, that is, in an unmediated way. this is important because if there is any kind of conscious inference to a belief, such as in s2, then the hot would not arise in the requisite manner for its target mental state to be conscious. more importantly, however, using conscious inference and reasoning to become aware of a mental state is much more likely to occur via introspection, that is, by consciously thinking about one’s own mental states. notice that s2 is more of an introspective state with a belief directed at a mental state. this, in turn, increases the likelihood that delusional thinking will generate further false beliefs. notice that this also fits nicely with hot the ory which can explain why there would seem to be a phenomenological sense of “myness” or “ownership” when one introspects, namely, that the hot is itself conscious. by contrast, it is doubtful that such a sense is present when one has an unconscious hot. the concept “i” is part of a conscious hot in the introspective case but it is part of an unconscious thought in the first-order case. nonetheless, it is certainly true that when there is a disturbance or abnormality in one’s i-concept, such as one’s bodily representation, one’s consciousness will be altered and result in some odd feelings of body or thought disownership. with regard to any potential misrepresentation between a hot and its target, recall the worry that there could equally be the possibility of a mismatch between the “i” in the hot and the “i” in the first-order mental state. but it is unclear how this could be so. consider wittgenstein’s (1958) distinction between the “ias-subject” (e.g., “i have a pain”) and the “i-as-object” (“i have a broken arm”). note that what he called the “i-as-subject” seems to involve some kind of minimal regennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org cotard syndrome, self-awareness, and i-concepts 7 maining sense of “ownership” even in the most extreme cases of thought insertion. this minimal sense can be the extent to which a hot theorist allows for a hot that “i am in mental state m” with regard to s1 even if the patient also has the other conscious state as in s2. crucially, however, there is never an i-as-object in the content of the first-order conscious state while there is an implicit (and unconscious) i-as-subject at the second-order level as well as an i-as-object in a typical hot. according to hot theory, there would only be an i-as-subject concept in a first-order state whereas the content of the state refers to the outer world. after all, if we assume that any mental state must have a bearer, then even first-order states should involve some primitive concept of i. the same is true for the unconscious hot that accompanies a first-order conscious state, but here there is also an i-as-object referenced in the content of the hot (i.e., “i think that i am in m”). still, these i-concepts are usually parts of unconscious thoughts and so there is little reason to suppose that there is any phenomenological sense of “myness” or “ownership” in these first-order cases. however, when one introspects and has a conscious hot directed at a mental state, there is not only a conscious i-as-subject concept but also a conscious i-as-object concept in the content of the hot which may account for any subjective sense of myness. so there could be no mis match between an i-as-object in the content of a mental state m and its hot because there isn’t an i-as-object concept at all in the content of m itself. 4 cotard syndrome cotard syndrome is a rare neuropsychiatric disorder in which people hold the delusional belief that they are dead, do not exist, are putrefying, or have lost their blood or internal organs (billon, 2016). importantly, most patients also have severe depression and anxiety problems, not to mention the way that they describe their self-related and very odd experiences. some bizarre patient accounts are as follows: she had the constant experience of having no identity or “self” and being only a body without content….she was convinced that her brain had vanished, her intestines had disappeared, and her whole body was translucent. (debruyne, portzky, van den eynde, & audenaert, 2009, p. 197) he said “i speak, breathe and eat, but i am dead”. (nejad & toofani, 2005, p. 250) it was as if it was not me walking, it was not me talking, as if it was not me living […] i can look at me, i am somehow bothered by my body, as if it wasn’t me, as if i lived on the side of my body, on the side of myself if you like. i don’t know how to explain. (janet & raymond, 1898, p. 70) gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org rocco j. gennaro 8 one general difficulty is just how literally to take the reports of cotard patients. should we take them at face value or treat them with some skepticism given the delusional aspect in question? young and leafhead (1996, p. 154), for example, note that there is no symptom or symptom complex present in all the patients described by cotard and that this “gives grounds for arguing that the term ‘cotard’s syndrome’ should be used cautiously; at best it represents an idealized pattern which in practice is not found even in pure cases.” further, cotard patients sometimes speak in more qualified terms, such as it is “as if” i do not exist or have a body (as above in the third quotation). like somatoparaphrenia and thought insertion, one might again suppose that cotard syndrome threatens the hot theory of consciousness. recall that hot theory requires that a hot to the effect that “i am in mental state m” accompanies each conscious state. but if i believe that i don’t exist, for example, then it would seem that i do not really have a self-referencing hot. depending upon the specific cotard patient, the “i-concept” would not seem to refer to her “self” or her “body” or her “mind.” so any analogy to thought insertion or somatoparaphrenia might depend to some extent on which quotation we focus on. when some aspect of one’s body is denied or feels alienated, perhaps it is more like somatoparaphrenia. when a mental state is denied or feels alienated, it is somewhat closer to thought insertion. it is difficult to grasp what it must be like to experience these abnormal conditions but we must also keep in mind the delusional element found in all three cases. i will mainly focus here on cases where a patient apparently denies her own existence or “self,” sometimes called the “nihilistic delusion” (radovic, 2017) which goes back originally to jules cotard (1882); see also (metzinger, 2003). this is one difference between someone with cotard syndrome and those with somatoparaphrenia and thought insertion. billon (2016) argues that cotard syndrome is the delusional counterpart of depersonalization disorders which are dissociative disorders marked by periods of feeling disconnected or detached from one’s body and thoughts. a cotard patient also has an impaired self-awareness and the feeling that there is no “i” in some sense. billon says that “…depersonalised patients are not delusional: they refuse to endorse the belief that they do not think and exist” (2016, p. 370) but he also points out that the “border between mere depersonalisation and the cotard syndrome might not be always clear-cut” (2016, fn. 22). he then goes on to explain that “[g]iven the similarity between the experience of depersonalised patients and that of cotard patients, what differentiates [them] is plausibly some kind of rationality deficit […]” (billon, 2016, p. 380). in any case, consider the following statements: at times she would say, “i am no longer alive”; “i am dead”. (enoch & ball, 2001, p. 159) he often said, “i am a ghost” or “i am dead”, and sometimes denied the existence of certain parts of his body saying, for example, “i have no blood”. (enoch & ball, 2001, p. 165) gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org cotard syndrome, self-awareness, and i-concepts 9 [it is like] seeing life as if it were played like a film in a movie. but in that case where am i? who is watching the film? (simeon & abugel, 2006, p. 15) but notice first that even the statement that “i am dead” still contains an initial reference to oneself (at least as a grammatical fact). it seems to me that there will always be some reference to oneself when expressing any kind of thought or belief (even if delusional, irrational, or self-contradictory in some other way). of course this is precisely what leads to an apparent contradiction. imagine: “who thinks that you don’t exist?” answer: “i do”? “who is dead?” answer: “i am”? again, perhaps this is merely a misleading way of speaking but the same might be true of other patient reports. still, we can see how the pattern of reply above to somatoparaphrenia and thought insertion can be applied here as well. that is, we may suppose that at least some cotard patients have the following two conscious states: s1: a conscious feeling or experience of unfamiliarity with myself (or body or mental states); and s2: a conscious belief or thought that i am dead or i do not exist. here, once again, having both s1 and s2 is indeed strange and perhaps even selfcontradictory in some sense, but it still wouldn’t tell against hot theory. in s1, we can suppose that the patient has a hot about her own feeling or experience, that is, she is aware of her feeling or mental state. in s2, the patient also has a conscious belief or thought that she does not exist. how or why s2 is generated is of course interesting in its own right. for example, perhaps the best explanation is that s2 is simply generated by the unfamiliar feeling or experience described in s1. this would be a so-called “endorsement” account. but there are at least two possibilities here, namely, the endorsement and the explanationist theories of delusion formation. the endorse ment account holds that delusions are formed by endorsing as veridical the content of the unusual experience (bayne & pacherie, 2004) whereas the explanationist account holds that delusions are formed as a way to explain (or rationalize) an unusual experience (maher, 1999). one way to frame these two theories of delusion formation is by asking the following question: is the delusional belief formed sometime prior to the person having the abnormal conscious experi ence or sometime afterward? alternatively, we might ask: are delusions bizarre convictions that alter one’s way of seeing the world, or are they hypotheses formulated to account for some unusual experiences and then endorsed as beliefs? in the previous discussion i have perhaps shown more sympathy with the explanationist account but, as langdon and bayne (2010) point out, most delusions are likely hybrids of both views especially when we consider them over periods of time. they propose a continuum from “received” to “reflective” delu sion, whereby the former largely come to pass initially gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org rocco j. gennaro 10 via the endorsement process and the latter are formed mainly by an explanationist processes. langdon and bayne (2010) also importantly distinguish between “spontaneous” and “provoked” confabulation where the latter arises only in response to direct questioning. after all, what makes these patient dialogues so compelling and bizarre are their rather stunning confabulations in response to questions. thus, in some ways, i am rather sympathetic with young’s two-stage “interactionist account” of the cotard delusion (young, 2012). he tells us that the interactionist explanation incorporates, for example, ratcliffe’s (2008) emphasis on existential change and disturbed phenomenology, but at the same time acknowledges the necessary role played by some form of second-stage cognitive disruption […] furthermore, the interactionist approach holds that the formation of the delusional belief alters the nature of the phenomenal experience, such that what is perceived […] is actually oneself dead or immortal (in the case of the cotard delusion). (young, 2012, p. 136) additionally, we might even suppose that there are unconscious inferences which we, by definition, wouldn’t be aware of making. it may even be that s2 is initially generated, at least in part, via unconscious reasoning designed to “make sense of” or rationalize the experience involved in s1. surely delusional rationalizing need not always be a conscious process especially during spontaneous confabulations. of course, once s2 is generated, then it may be used thereafter to affect or interpret first-order feelings or experiences. notice also that the noninferential condition on hot theory does not rule out unconscious inferences playing some role in generating an introspective state. rather, it only rules out that s1 itself is or feels conscious in the same way when it arises via conscious inference. in any case, regardless of the alternative explanations for the genesis of s2, cotard syndrome as such does not seem to cause trouble specifically for hot theory’s ability to explain s1 or s2. just as hot theory says about all conscious states, when a patient is aware of having a (first-order) conscious feeling or experience, there is a hot about that state. this is what we have in the case of s1. but we should also recognize the delusional element involved in the formation of s2. there are of course situations where those with cotard syndrome might even realize that they have come to think or believe contradictory things based on their abnormal experiences. to be clear: i do not mean to suggest that the discussion above on endorsement and explanationist theories of delusion results either in some novel view of mine in relation to cotard syndrome or that my view depends essentially on arguments contained in one of those accounts. the main point is rather that no matter how or when s2 is generated with respect to s1, hot theory is not threatened and is indeed consistent with both s1 and s2. still, hot theory does have a natural way to account for s2 as more of an introspective state. again, if there is an additional gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org cotard syndrome, self-awareness, and i-concepts 11 and particularly unique issue posed by cotard, it is with respect to cases of selfdenial (“nihilistic delusions”), not with respect to hot theory’s ability to account for s1 and s2. with regard to any potential misrepresentation between a hot and its own target, it is again important to note that some kind of minimal “i-as-subject” seems to remain even in the most extreme cases of cotard syndrome. this minimal iconcept can merely be the extent to which a hot theorist allows for a hot that “i am in mental state m” with regard to s1. but there is never an i-as-object in the content of the first-order conscious state; rather, there is an implicit (and unconscious) i-as-subject at the second-order level as well as an i-as-object in a typical hot. according to hot theory, there would only be an i-as-subject concept in the first-order state. however, when one introspects and has a conscious hot directed at a mental state, there is not only a conscious i-as-subject concept but also a conscious i-as-object concept in the content of the conscious hot that may account for any further subjective sense of myness or ownership. again, there would be no potentially problematic mis match between an i-as-object in the content of a mental state m and its hot because there isn’t an i-as-object concept at all in the content of m itself. in terms of the general topic of this special issue, “can there be states of consciousness that lack any kind of self-consciousness?”, then, the answer partly depends upon how one defines “self-consciousness”, and there are clearly various notions of self-consciousness or self-awareness ranging from very minimal to very sophisticated. i have previously argued at length that conscious states and conscious creatures (including most animals) are self-conscious in at least one sense; namely, that each conscious state is accompanied by a hot and so any conscious creature has hots (gennaro, 1996). for various reasons i won’t repeat here, i also urged that a plausible minimal form of self-consciousness is simply having any kind of higher-order or meta-psychological thought. this can include any hot, conscious or not, and even those containing such minimal i-concepts as “bodily self-awareness”; that is, distinguishing one’s own body from other things. so we are continuously “aware of” our mental states but this awareness is generally not a conscious awareness unless we are also attending to our mental states, in which case we have an unconscious third-order awareness of the second-order introspective awareness of the mental state. the answer to the above question would thus be “no” according to this way of understanding “self-consciousness” and according to hot theory. however, my rather rudimentary notion of self-consciousness will likely not be robust enough to satisfy some philosophers. this is understandable and it is also partly why we must distinguish various levels of self-consciousness including the more sophisticated notions discussed in this paper, such as “introspection” as found in hot theory. in this sense, of course, each conscious state does not involve self-consciousness according to hot theory because we can have conscious states without introspecting them (because conscious states are most often outer-directed). so if one identifies self-consciousness with introspection, then gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org rocco j. gennaro 12 the answer to the above question would be “yes.” still, cotard patients do seem capable of some level of self-awareness and even having some introspective states, albeit often including some wildly bizarre beliefs or thoughts about themselves (as in s2). (actually, it is sometimes also useful to distinguish further between two kinds of introspection: momentary focused and deliberate (gennaro, 1996, p. 19). the former is a brief conscious focus on a mental state such as a pain or desire, whereas the latter involves consciously reasoning about our first-order mental states for a period of time.) it is worth mentioning here that hot theory, like many contemporary theories of consciousness in analytic philosophy of mind, tends to focus on explaining “state” consciousness, that is, “what makes a mental state conscious?” we sometimes speak of an individual mental state, such as a pain or perception, as conscious. this is in contrast to “creature” consciousness since we also often speak of organisms or creatures as conscious. creature consciousness is simply meant to refer to the fact that an organism is awake or aware of its surroundings, as opposed to sleeping or in a coma. however, state consciousness is often implied by creature consciousness and vice versa. for example, a conscious creature will typically have conscious states and having conscious states are likely only to occur in generally conscious creatures. one might wonder how the above discussion relates to what some authors characterize as “consciousness of oneself” or an “i-concept” specifically involving the capacity to form i*-sentences or i*-thoughts. the asterisk used here was introduced by castañeda (1966) and can be found in the literature on de se propositional attitudes where an intentional state is directed at oneself (perry, 1979). it is not just that i can have thoughts or beliefs about myself but that i myself (or i*) realize that the thought or belief is about me. i may come to realize that the messy shopper accidentally dropping food in the aisles is me or suddenly realize that the person in the store mirror is me. so this is clearly a more sophisticated notion of selfconsciousness where one is thinking about oneself as an object. as baker (2013) explains, “i* is typically embedded in the”that-clause” of a complex first-person sentence with a psychological or linguistic main verb” (p. 33), such as “i believe that i am short.” “i am not only the thinker of the thought but i am also part of the object of my thought” (pp. 32-33). with regard to cotard’s syndrome, on my view, i*-concepts enter the picture at the point i have called the “i-as-object” at the level of introspection. however, as we have seen, the delusional element in cotard patients leads them to have radically false beliefs about themselves. further, such beliefs and thoughts are sometimes directed at their bodies (or body parts) and sometimes directed at their mental states. as we have seen, this only complicates the matter (more on this in section 5). but, for example, in extreme forms of cotard syndrome, as metzinger (2003, p. 456) puts it, “we are faced with a delusional belief that can be expressed as follows: [*i am certain that i* do not exist].” we might similarly suppose that there is the thought “i think that i* do not exist” which echoes the kind of intentional state reflected in s2. gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org cotard syndrome, self-awareness, and i-concepts 13 my main problem with baker and others, however, is that she holds that the “capacity to use i*-language (i.e. a robust first-person perspective) is constitutive of self-consciousness” (baker, 2013, p. 39) where “the robust first-person perspective is to conceive of oneself as oneself* in the first person” (baker, 2013, p. 35). it is unclear to me why she doesn’t think that it’s preferable to suppose that there are degrees of i-concepts and self-consciousness ranging from more basic to rather sophisticated. although baker does allow for a “rudimentary first-person perspective,” she doesn’t consider anything less than having the more robust complete mastery of an i-concept as being a genuine self-concept at all. why suppose that complete mastery is required for self-consciousness or for having an i-concept? and why should all self-concept possession be so closely tied to a linguistic ability? surely there is room for degrees of concept possession not only for self-concepts but for most concepts (see gennaro, 2012, chapters six through eight). this would also, in turn, open the door to the view that many animals and human infants are not only conscious but also possess concepts and self-consciousness to some degree. it is worth mentioning in this context the way that guillot explores our sense of self-awareness, that is, by making a threefold distinction within the more general notion of “subjective character” as follows: 1. “for-me-ness” is when the “subjective character is a subject’s characteristic awareness of her experience” (guillot, 2017, p. 31); that is, the object of selfawareness is the experience itself. 2. “me-ness” is when “subjective character is a subject’s awareness of herself as part of having the experience” (guillot, 2017, pp. 31–32); that is, what makes an experience special for its subject is the fact that the subject is somehow aware of herself. 3. “mineness” is when subjective character is a subject’s “awareness of herself as having the experience” (guillot, 2017, p. 32); that is, subjective character presents it as a phenomenal awareness that my experiences are mine. guillot holds that “the experience of prototypical cotard patients exhibits for-me-ness, but neither me-ness nor mineness. she also suggests”that schizophrenic inserted thoughts might exhibit for-me-ness and me-ness, but not mineness” (2017, p. 44). this analysis seems plausible to some extent. for example, cotard patients do seem to be aware of a (rather unusual) experience (as reflected in s1) and thus retain some very minimal sense of self-awareness, but they do not seem to relate that experience to themselves in any coherent or significant way. as a matter of fact, they even often explicitly deny that there is a self to which that experience can be ascribed. gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org rocco j. gennaro 14 however, in relation to hot theory, i have argued elsewhere for independent reasons that there isn’t ubiquitous self-awareness (i.e. for-me-ness) that accompanies each normal conscious first-order state. for example, i argue specifically against kriegel’s self-representationalist view at length (gennaro, 2008, 2012, chapter five), that is, against the notion that “for-me-ness” is a form of peripheral (i.e. non-attentive but conscious) inner awareness. in doing so, i address the differences between us with respect to reductionist motivations, as was briefly mentioned early in this paper. i also take issue with the phenomenological argument he adduces in favor of his view. it seems to me that our conscious attention is often so focused at the world and its objects that it seems unlikely that we are continuously consciously self-aware. it is also clear that such a view lacks plausibility especially as compared to the widely accepted and fairly obvious fact of peripheral outer-directed conscious awareness, such as the awareness in one’s peripheral visual field while watching a concert or working on one’s computer. much the same goes for me-ness and mineness. again, the hot is typically unconscious when directed at a first-order conscious state. me-ness and mineness, on my view, enter the picture only upon introspection, that is, when a hot is consciously directed at one’s own mental state. a similar line of response might also be raised against ford and smith’s (2006) argument in favor of the self-representationalist theory of consciousness. the claim here again is that there is an element of self-awareness – guillot’s “forme-ness” seemingly – internal to every conscious state. ford and smith contend that, for example, cases of depersonalization show that self-representationalism is correct. but just because the removal of something – for example, normal proprioception – causes deficits in one’s conscious mental states, it surely does not follow that the awareness of that thing is part of normal conscious experience. the relation could be causal instead of constitutive. that is, the typical abilities and awareness in question might merely, in the normal case, causally contribute to the phenomenology of one’s conscious mental states without being part of the conscious state itself (see also howell & thompson, 2017). with regard to “bodily self-awareness,” bermúdez (2011, pp. 161–166), for example, distinguishes between a “sense” and a “judgment” of ownership. he argues, in the end, that the so-called “sense” of ownership is best viewed as a “judgment” rather than as a feeling itself. in contrast to de vignemont (2007), bermúdez rejects the “inflationary” conception of the sense of ownership according to which there is a distinctive positive phenomenology of bodily ownership: “there are facts about the phenomenology of bodily awareness […] and there are judgments of ownership, but there is no additional feeling of ownership” (bermúdez, 2011, p. 166). there are still of course conscious bodily sensations and proprioceptive states. what one actually feels is the first-order conscious state accompanied by the unconscious thought about the mental state, as opposed to any phenomenology of so-called “myness”. recall that hot theory can explain why there is a phenomenological sense of myness when one introspects, namely, that the hot is gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org cotard syndrome, self-awareness, and i-concepts 15 itself conscious whereas no such sense is present when one has an unconscious hot. the “i-concept” is part of a conscious thought in the introspective case but part of an unconscious thought in the outer-directed first-order case. nonetheless, it is again certainly true that when there is a disturbance in one’s i-concept (including bodily self-awareness), one’s consciousness will likely be altered and result in some very odd beliefs and feelings of disownership. but, like bermúdez, i do not find it compelling to argue that if a deficit of bodily awareness is manifested in consciousness, then that aspect of bodily awareness is normally part of our conscious states, albeit even in some peripheral way. 5 self-awareness and personal identity: an ambiguity with respect to cotard patients specifically, radovic (2017) hints at an interesting and related aspect of self-awareness which might well explain at least some of their odd statements: in order to grasp in what sense cotard patients believe themselves to be dead, it might be helpful to take a closer look at different conceptions of death and, in particular, the received notion of biological death. biological death may be viewed from two different perspectives: a first-person point of view and a third-person point of view. from a naturalistic first-person point of view, death is commonly understood as the permanent absence of consciousness and the disintegration of personhood. death is the end of our selves as conscious persons. by contrast, from a naturalistic third-person point of view, animals such as humans are dead when the heart does not beat, when there is no respiration, when the body is permanently still, or when the brain exhibits no neural activity. (radovic, 2017, p. 688) this line of analysis brings to mind a related distinction between two conceptions of “personal identity,” namely, psychological continuity and bodily continuity. the question raised is often framed as “what makes a person the same person over time?” any such discussion in western philosophy tends to begin with the work of john locke whose account of personal identity famously appealed to psychology and, more specifically, to consciousness and memory (locke, 1689). on his view, a later person (p2) is identical to an earlier person (p1) just in case p2’s consciousness “can be extended backwards” to p1. this is taken to mean that p2 remembers p1’s thoughts and experiences, “from the inside” so to speak. my personhood goes with my first-person perspective and especially with my consciousness and memory, not necessarily with my body. the basic idea is familiar enough: i have memories from my first campus visit to the university of southern indiana, to experiences back in graduate school at syracuse university, and so on. you do not gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org rocco j. gennaro 16 have those same memories and so my having them makes me the same person as those previous persons (as opposed to you). they are episodic memories, that is, i remember them as experienced from the first-person point of view. episodic memory is different than what is sometimes called “procedural memory,” that is, memory for skills and abilities like playing the piano or riding a bike. of course, locke is not saying that a fifty-year-old person must (episodically) remember everything from one’s earlier life since there is some normal forgetting of one’s own past. although locke’s view has been dominant in western philosophy, other philosophers disagree and have put forth what has come to be known as “animalism” which holds that being the same “biological organism” or having the “same body” is essential to one’s personal identity (olson, 2007). on this view, one continues to exist “just in case one’s purely animal functions – metabolism, the capacity to breathe and circulate one’s blood, and the like – continue” (olson, 1997, p. 16). to elaborate, olson explains that “an animal’s life is understood as a self-organizing biological event that maintains the organism’s complex internal structure. […] an organism must constantly take in new particles, reconfigure and assimilate them into its living fabric, and expel those that are no longer useful to it. an organism’s life enables it to persist and retain its characteristic structure despite constant material turnover” (olson, 2007, p. 28). interestingly, locke himself recognized the ambiguity and thus distinguished between “same person” and “same body” or “same biological organism.” for one thing, locke was clearly open to the possibility of something like what is now called dissociative identity disorder (did) where “two or more persons” can be associated with “one body.” locke used the term “same man” to denote to latter. radovic (2017) correctly points out that cotard patients often do seem to indicate that they are dead in the biological sense as opposed to the psychological sense. consider: i find myself regarding existence as though from beyond the tomb, from another world; […] i am as it were, outside my own body and individuality; i am depersonalized. (simeon & abugel, 2006, p. 133) i felt that my brain was somewhere else and from there was just watching me […] i was completely unable to tell whether i was still present or whether i was the part that was gone. in short, there were two different beings, the one watching the other. (roberts, 1960, p. 481) much of the above, therefore, highlights the ambiguity of self-reference talk and the use of “i”. “i am dead” might be taken to mean that “my body is no longer alive or functioning” or “i am no longer in my body” as opposed to “i am no longer conscious” or “my conscious existence has ceased.” this could also partly explain the apparent contradiction in some statements made by cotard patients. that is, if such a patient is saying “i am dead” and means it in the biological sense, then there could still be a sense in which “i am alive” and detached from my body in gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org cotard syndrome, self-awareness, and i-concepts 17 some way, such as immortal or more like a “ghost.” thus, it seems that something like “bodily self-awareness” still lingers in the background (perhaps more like somatoparaphrenia): at one point he had stated that he was a ghost and that no one could see him. he explained that his physical body had been transformed into the immaterial body of a ghost. nevertheless, he indicated that he visualized his body without any difficulty. about 2 months prior to admission he had concluded that eating was unnecessary since he was already dead. (silva, leong, weinstock, & gonzales, 2000, p. 188) he often said, “i am a ghost” or “i am dead”, and sometimes denied the existence of certain parts of his body saying, for example, “i have no blood.” (enoch & ball, 2001, p. 165) this suggests that cotard patients often refer to themselves in third-person terms rather than in first-person terms. recall our two earlier statements: s1: a conscious feeling or experience of unfamiliarity with myself (or body parts or mental states); and s2: a conscious belief or thought that i am dead or i do not exist. in light of this discussion, it may even be the case that s1 and s2 are not really in direct conflict at all. that is, cotard patients may sometimes be referring to their bodies (or body parts) in s2 whereas s1 reflects a more first-person perspective on their own mental states. this is somewhat similar to wittgenstein’s distinction between the “i-as-subject” and the “i-as-object” whereby the former may indicate a first-person point of view and the latter a third-person point of view. thus, with respect to hot theory, the i-concept referent in a hot might vary depending upon several factors, including which disorder or delusion we are addressing. it might refer to one’s own body as a sort of bodily self-awareness, perhaps more like somatoparaphrenia. it might refer to one’s own mental state, perhaps closer to thought insertion. but it might even refer to one’s own existence. it seems rather uncontroversial to say that are different self-concepts and some are more relevant than others under certain conditions. billon makes a related point when he suggests that at least sometimes, cotard patients “feel more entitled to refer to themselves in the third-person than in the first-person” (2016, p. 376). along somewhat similar lines, it is worth mentioning that some cotard patient reports seem to resemble descriptions of out-of-body experiences (obes). an obe occurs when one experiences one’s own body (or even the world) from above or outside one’s body. you feel like you left your body and are “floating” above it. this seems to indicate that your “mind” or “soul” is separating out from your body. a materialist may of course explain how very similar experiences can be induced gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org rocco j. gennaro 18 by stimulating neurons in the temporo-parietal junction (blanke & arzy, 2005). although cotard patients can have a wide variety of neurological and psychiatric pathologies, at least some have been known to have abnormalities in the temporoparietal area (kudlur, george, & jaimon, 2007). 6 conclusion in closing, then, we have seen that various psychopathologies of self-awareness might seem to threaten the viability of the higher-order thought (hot) theory of consciousness since it requires a hot about one’s own mental state to accompany every conscious state. the same might be said of cotard syndrome but, by analogy with somatoparaphrenia and thought insertion, i have argued that hot theory can answer this objection and can consistently account for what happens in this admittedly unusual case. there are possibly irrational combinations of mental states but hot theory has the resources to explain why they do not cause trouble for the theory. there are also multiple “i-concepts” and levels of hots which can help to provide a more nuanced explanation. for example, there would not really be any “self-concept” mismatch between a conscious state and its hot. in addition, wittgenstein’s distinction between “i-as-subject” and “i-as-object” as well as different approaches to the problem of personal identity can aid us in understanding how apparently contradictory claims can be explained away. finally, with regard to the connection between consciousness and self-consciousness, it seems clear from the above discussion that cotard patients are capable of having some “i-thoughts” about their bodies and minds which seem to indicate various degrees of self-awareness. as we have seen, they do think about themselves in various ways. so i do not suppose that cotard syndrome is characterized by a total lack of self-consciousness. however, perhaps not surprisingly given their delusional nature, many such thoughts about themselves are often rather bizarre and clearly false. acknowledgments i would like to thank tom metzinger, raphaël millière, two anonymous reviewers, and the editors-in-chief of this journal for numerous comments on previous drafts of this paper.  i would also like to thank andrea wright and laura bernhardt for their assistance in preparing the manuscript for publication in accordance with the specific submission format guidelines. references baker, l. r. (2013). naturalism and the first-person perspective. new york: oxford university press. bayne, t., & pacherie, e. (2004). bottom up or top down? philosophy, psychiatry, and psychology, 11(1), 1–11. https://doi.org/10.1353/ppp.2004.0033 bermúdez, j. l. (2011). bodily awareness and self-consciousness. in s. gallagher (ed.), the oxford handbook of the self (pp. 157–179). new york: oxford university press. gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1353/ppp.2004.0033 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org cotard syndrome, self-awareness, and i-concepts 19 billon, a. (2016). making sense of the cotard syndrome: insights from the study of depersonalization. mind and language, 31(3), 356–391. https://doi.org/10.1111/mila.12110 blanke, o., & arzy, s. (2005). the out-of-body experience: disturbed self-processing at the temporo-parietal junction. the neuroscientist, 11(1), 16–24. https://doi.org/10.1177/1073858404270885 bortolotti, l. (2009). delusions and other irrational beliefs. oxford: oxford university press. castañeda, h.-n. (1966). “he”: a study in the logic of self-consciousness. ratio, 8, 130–157. cotard, j. (1882). du délire des négations. archives de neurologie, 4, 152–170. debruyne, h., portzky, m., van den eynde, f., & audenaert, k. (2009). cotard’s syndrome: a review. current psychiatric reports, 11(3), 197–202. https://doi.org/10.1007/s11920-009-0031-z enoch, m. d., & ball, h. n. (2001). uncommon psychiatric syndromes (4th ed.). london: arnold, cop. ford, j., & smith, d. w. (2006). consciousness, self, and attention. in u. kriegel & k. williford (eds.), self-representational approaches to consciousness (pp. 353–377). cambridge, ma: the mit press. frith, c. (1992). the cognitive neuropsychology of schizophrenia. east sussex, uk: psychology press. gennaro, r. (1996). consciousness and self-consciousness: a defense of the higher-order thought theory of consciousness. philadelphia; amsterdam: john benjamins publishers. gennaro, r. (2008). representationalism, peripheral awareness, and the transparency of experience. philosophical studies, 139, 39–56. https://doi.org/10.1007/s11098-007-9101-4 gennaro, r. (2012). the consciousness paradox: consciousness, concepts, and higher-order thoughts. cambridge, ma: the mit press. gennaro, r. (2015). somatoparaphrenia, anosognosia, and higher-order thoughts. in r. gennaro (ed.), disturbed consciousness: new essays on psychopathology and theories of consciousness (pp. 55–74). cambridge, ma: mit press. gennaro, r. (2020). inserted thoughts and the higher-order thought theory of consciousness. in p. angel gargiulo & h. mesones-arroyo (eds.), psychiatry and neurosciences update: bridging the divide vol 4. cham, switzerland: springer. guillot, m. (2017). i me mine: on a confusion concerning the subjective character of experience. review of philosophy and psychology, 8(1), 23–53. https://doi.org/10.1007/s13164-016-0313-4 howell, r., & thompson, b. (2017). phenomenally mine: in search of the subjective character of consciousness. review of philosophy and psychology, 8(1), 103–127. https://doi.org/10.1007/s13164-016-0309-0 janet, p., & raymond, f. (1898). névroses et idées fixes, vol. 2. paris: felix alcan. kudlur, s. n. c., george, s., & jaimon, m. (2007). an overview of the neurological correlates of cotard syndrome. european journal of psychiatry, 21(2), 99–116. retrieved from http://scielo.isciii.es/pdf/ejpen/v21n2/original2.pdf lane, t. (2015). self, belonging, and conscious experience: a critique of subjectivity theories of consciousness. in r. gennaro (ed.), disturbed consciousness: new essays on psychopathology and theories of consciousness (pp. 103–140). cambridge, ma: the mit press. lane, t., & liang, c. (2010). mental ownership and higher-order thought: response to rosenthal. analysis, 70(3), 496–501. https://doi.org/10.1093/analys/anq022 langdon, r., & bayne, t. (2010). delusion and confabulation: mistakes of perceiving, remembering and believing. cognitive neuropsychology, 15(1-3), 319–345. https://doi.org/10.1080/13546800903000229 liang, c., & lane, t. (2009). higher-order thought and pathological self: the case of somatoparaphrenia. analysis, 69(4), 661–668. https://doi.org/10.1093/analys/anp094 locke, j. (1689). an essay concerning human understanding (p. h. nidditch, ed.). oxford: clarendon. lycan, w. (1996). consciousness and experience. cambridge, ma: the mit press. maher, b. a. (1999). anomalous experience in everyday life: its significance for psychopathology. the monist, 82(4), 547–570. https://doi.org/10.5840/monist199982428 metzinger, t. (2003). being no one. cambridge, ma: mit press. nejad, a., & toofani, k. (2005). co-existence of lychanthropy and cotard’s syndrome in a single case. acta psychiatrica scandinavica, 111(3), 250–252. https://doi.org/10.1111/j.1600-0447.2004.00438.x olson, e. (1997). the human animal: personal identity without psychology. new york: oxford university press. olson, e. (2007). what are we? a study in personal ontology, new york: oxford university press. perry, j. (1979). the problem of the essential indexical. nous, 13(1), 3–21. https://doi.org/10.2307/2214792 prigatano, g. (2010). the study of anosognosia. new york: oxford university press. radovic, f. (2017). the sense of death and non-existence in nihilistic delusions. phenomonology and the cognitive sciences, 16(4), 679–699. https://doi.org/10.1007/s11097-016-9467-x gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1111/mila.12110 https://doi.org/10.1177/1073858404270885 https://doi.org/10.1007/s11920-009-0031-z https://doi.org/10.1007/s11098-007-9101-4 https://doi.org/10.1007/s13164-016-0313-4 https://doi.org/10.1007/s13164-016-0309-0 http://scielo.isciii.es/pdf/ejpen/v21n2/original2.pdf https://doi.org/10.1093/analys/anq022 https://doi.org/10.1080/13546800903000229 https://doi.org/10.1093/analys/anp094 https://doi.org/10.5840/monist199982428 https://doi.org/10.1111/j.1600-0447.2004.00438.x https://doi.org/10.2307/2214792 https://doi.org/10.1007/s11097-016-9467-x https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org rocco j. gennaro 20 ratcliffe, m. (2008). feelings of being: phenomenology, psychiatry, and the sense of reality. oxford: oxford university press. roberts, w. (1960). normal and abnormal depersonalization. the british journal of psychiatry, 106(443), 478–493. https://doi.org/10.1192/bjp.106.443.478 rosenthal, d. (1997). a theory of consciousness. in n. block, o. flanagan, & g. guzeldere (eds.), the nature of consciousness (pp. 729–753). cambridge, ma: the mit press. rosenthal, d. (2004). varieties of higher-order theory. in r. gennaro (ed.), higher-order theories of consciousness: an anthology (pp. 17–44). amsterdam: john benjamins. rosenthal, d. (2005). consciousness and mind. new york: oxford university press. rosenthal, d. (2010). consciousness, the self and bodily location. analysis, 70(2), 270–276. https://doi.org/10.1093/analys/anp167 silva, j. a., leong, g., weinstock, r., & gonzales, c. (2000). a case of cotard’s syndrome associated with self-starvation. journal of forensic sciences, 45(1), 188–190. https://doi.org/10.1520/jfs14658j simeon, d., & abugel, j. (2006). feeling unreal: depersonalization disorder and the loss of the self. oxford: oxford university press. stephens, g. l., & graham, g. (2000). when self-consciousness breaks: alien voices and inserted thoughts. cambridge, ma: the mit press. vallar, g., & ronchi, r. (2009). somatoparaphrenia: a body delusion. a review of the neuropsychological literature. experimental brain research, 192(3), 533–551. https://doi.org/10.1007/s00221-008-1562-y vignemont, f. de. (2007). habeas corpus: the sense of ownership of one’s own body. mind and language, 22(4), 427–449. https://doi.org/10.1111/j.1468-0017.2007.00315.x wittgenstein, l. (1958). the blue and brown books (2nd ed.). oxford: basil blackwell. young, a., & leafhead, k. (1996). betwixt life and death: case studies of the cotard delusion. in p. halligan & j. marshall (eds.), method in madness: case studies in cognitive neuropsychiatry (pp. 147–171). hove: psychology press. young, g. (2012). delusions of death and immortality: a consequence of misplaced being in cotard patients. philosophy, psychiatry, and psychology, 19(2), 127–140. retrieved from https://muse.jhu.edu/article/484349 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1192/bjp.106.443.478 https://doi.org/10.1093/analys/anp167 https://doi.org/10.1520/jfs14658j https://doi.org/10.1007/s00221-008-1562-y https://doi.org/10.1111/j.1468-0017.2007.00315.x https://muse.jhu.edu/article/484349 https://doi.org/10.33735/phimisci.2020.i.41 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction hot theory somatoparaphrenia and thought insertion cotard syndrome self-awareness and personal identity: an ambiguity conclusion how radical is perceptual malleability? a reply to commentators. how radical is perceptual malleability? a reply to commentators dustin stokesa(dustin.stokes@utah.edu) abstract the unifying theme across all four commentaries is the question: just how radical are the ideas contained in, and implied by, thinking and perceiving? does the abandonment of the modularity of mind, and an embrace of the malleability of mind, have wide reaching consequences for empirical studies of sensory perception, for cognitive architecture, for the metaphysics of mind and the epistemology of perception? and which of those consequences are welcomed, and which of those are problematic? these general questions anchor in (at least) the following topics: learning and the role of cognition, object recognition, perceptual expertise, early versus late vision, perception and truth, virtue epistemology, and computational architectures of mind. this reply attempts to address the overarching question from this variety of angles taken in the four commentaries. keywords cognitive architecture ∙ epistemology of perception ∙ malleability of perception ∙ perceptual learning ∙ perceptual expertise ∙ modularity this article is part of a symposium on dustin stokes’s book “thinking and perceiving” (routledge 2021), edited by regina fabry and sascha benjamin fink. 1 introduction ‘radical’ simply means grasping things at the root. angela davis i first want to offer my sincere gratitude to zed adams, zoe drayson, chris mole, and aleksandra mroczko-wąsowicz for spending the time with my book, and for offering such a wide array of insightful commentary and criticism. it is truly appreciated. auniversity of utah, salt lake city. stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/ 0000-0001-5341-3157 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dustin stokes 2 if there is a unifying theme across all commentaries, it is to ask the question: just how radical are the ideas contained in, and implied by, thinking and perceiving? does the abandonment of the modularity of mind, and an embrace of the malleability of mind, have wide-reaching consequences for empirical studies of sensory perception, for computational models of thought, for the metaphysics of mind and the epistemology of perception? and which of those consequences are welcomed, and which of those are problematic? in what follows, i’d like to address this overarching question from the variety of angles taken in the four commentaries. my hope is that i address the most central, even if not all, of the criticisms and suggestions. 2 beyond the default a key pivot point in the book is to move away from central emphasis on modularity, abandoning the assumption that modularity should continue to be the marker relative to which interesting relations between cognition and perception are judged. this, hopefully, is an obvious consequence of the book. it is also, at least for this philosopher, one of the hardest talks to walk. most of relevant research in the last two decades betrays this assumption: attempts to show theoretically important instances of cognitive influence on perception have been attempts to show the cognitive penetration of perception, ergo, to show counterexamples to the informational encapsulation of perception. in previous work i have made the assumption myself, arguing for empirically grounded cases of cognitive penetration. and indeed some of the discussion in thinking and perceiving continues to engage relevant debates in those terms. in particular, i argue that cognitive penetration should be characterized in terms of its important (possible) consequences and that we should re-think how perceptual attention plays a role in the (possible) cognitive penetration of perception. so it would be natural to hold up modularity (and its defeat) as a test for the success of the book. natural, but off the mark. this surfaces in some of the choice of terminology. as adams characterizes things, the debate is one between “fodorian modularists (fms)” and “stokean anti-modularist (sams)”. the latter term gives me pause since it characterizes the architecture of mind negatively, as simply a denial of the orthodox view. (this terminology is echoed in mroczko-wąsowicz’s commentary.) that’s the very assumption to be resisted. instead, the many cases of apparent cognitive influence on perception analyzed in the book, with emphasis on cases of perceptual expertise, are given a novel architectural and epistemological story. the epistemology will center in the discussion below. for now it’s important to reemphasize that in terms of mental architecture, these cases are plausibly explained in a plurality of ways. only some of those ways may clearly be anti-modularist in the sense that they “count as” cognitive penetration; but all of them will plausibly motivate a view whereby visual and other sensory perceptual modalities are richly malleable, and improvable in contextand content-sensitive ways. stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how radical is perceptual malleability? a reply to commentators 3 on mole’s analysis, malleability is the “rival picture” to modularity. and if we are to abandon the default, he suggests, there are important choice points regarding the strength of the replacement rival view. the choice that mole emphasizes concerns whether malleability just revises possible relations between broad mental categories (cognition and perception) or goes deeper to revise the metaphysics fundamental to the orthodox mental architecture. the deeper or more radical reading will be discussed below. the weaker, less radical reading comes to “rejecting only the idea that these processes [cognition and perception] are encapsulated from one another” (mole p. 3). under one description, this view would not be surprisingly radical since it would just be a victory for the opponents of modularity, a victory that was the central aim for those who have made the case for the cognitive penetration of perception. but that description is, again, loaded with the problematic default position assumption. in various places, thinking and perceiving does make the rejection in question, offering empirical cases that modularity inadequately explains and giving critical analysis of the central arguments for informational encapsulation. but those same analyses go one step further. they support not just a rejection of the view but ousting the view (or the view’s defeat) from holding court over the theoretical space. the success conditions for malleability (or for any other novel alternative) need not include counterexamples to that putative orthodoxy. it may start with showing how the modular view can be rejected, but once that ladder is climbed, the more radical aim is to kick the ladder away. 3 expertise, learning, and recognition adams writes that “distinctions are the lifeblood of philosophy” (p. 1). this seems right, while by the same token we should be cautious about the proliferation of distinctions in a theoretical space. sometimes distinctions are made to mark genuine and important differences in the world. and this is a good way to sharpen a theoretical problem and to challenge theories. while in other instances distinctions are made, and seem to multiply, in ways that lack principled grounds and may create theoretical cross-talk. and of course the devil is always in the details, in where the precise lines of the distinction are drawn. mroczko-wąsowicz identifies object recognition as an important explanandum for a theory of mental architecture. she then either identifies or intimates distinctive stages of perceptual processing that may impact how expert cases of recognition are explained. the relevant big question here is what we mean by ‘recognition’ and where it should be placed in the architecture. as mroczkowąsowicz suggests, the term can denote object categorization or identifying an object as an individual. recognition is thus a capacity that is not (or not always) strictly perceptual nor strictly cognitive. and this comports with what one finds in the cognitive neuropsychology on the topic (humphreys et al., 1999). one way this surfaces is that object recognition can fail in cases of agnosia that are apperceptive (apparently involving a visual deficit) and those that are associative stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dustin stokes 4 (apparently involving a memory or semantic deficit) (riddoch & humphreys, 2001). accordingly, we shouldn’t expect recognition to fit discretely and exclusively into one mental stage or process. the question then becomes: which broad architecture—modularity or malleability—better explains expert recognition in its various guises? a familiar appeal for modularity here is to “go early”: identify recognitional capacities with some early stage of visual processing. since early vision is arguably informationally encapsulated, the modularist can then locate expert recognition as occurring there, and this would comport well with some of the automatic and effortless perceptual skills that we find in experts. the modularist can make this appeal, the trouble is that it is not going to explain much with respect to expert performance, nor explain it in modular terms. to see this, consider what “early vision” is supposed to be and what informationally encapsulated early vision should be capable of. non-exclusively, early vision has been characterized computationally (it computes, say, basic features such as color, edges, depth) (marr, 1982) or temporally (say, less than 100ms post stimulus onset) (raftopoulos, 2009). it is plausible that some such stage/s of visual processing are strongly modular, that is, entirely informationally encapsulated. it is furthermore plausible that feature detecting components—groups of simple and complex cells in primary visual cortex—as well as many other neural circuits and low-level components are informationally encapsulated. this, however, is not tantamount to a defense of the modularity of perception, and is compatible with malleability, as we will see shortly. even circumscribing ourselves to this early stage of visual perception, though, how well can it explain cases of expert recognition? mroczko-wąsowicz writes “for modularists perceptual expertise does not make object recognition more sensitive to cognitive influences than it is already the case for regular (non-expert) recognition. on their view, perceptual expertise makes the lowlevel component of object recognition – early visual sensory processing – more fluent and facilitates perceptual discriminations of fine grained subordinate categories” (p. #; emphasis added). the second claim is incompatible with the defense captured in the first claim. if early visual processing, in some expert cases, involves sensitivity to, discrimination of, subordinate categories (e.g. ‘kingbird’) or even sub-subordinate categories (e.g. ‘eastern kingbird’), the better explanation of this enhanced sensitivity will appeal to cognition. and this for the simple reason that the empirical evidence shows, at least for some domains of expertise, that these are not “pure” visual categories, and not learned just on the basis of mere exposure. instead, relevant behavioral and neural changes do not occur without learning those specific categorical concepts (gauthier & tarr, 1997; scott et al., 2008; tanaka et al., 2005). so if the modularist wants to explain these kinds of expert recognitional capacities for all experts, by placing them in “early vision”, it looks like they will have to allow cognition into the posited visual stage after all. now, mroczko-wąsowicz rightly notes that some modularists will admit that “such early visual representations do not alone determine late vision phenomena stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how radical is perceptual malleability? a reply to commentators 5 like object recognition” (p. #). this is why a defense of the informational encapsulation of some sub-components of visual processing—“early vision”—does not amount to a defense of the informational encapsulation of visual representation or experience. and since the modularist often motivates their view by appeal to epistemic concerns, they too are concerned with those personal-level visual phenomena. but now the modularist is in a pinch, since the wanted early visual explanations of expert recognition violate informational encapsulation (as argued just above) and so-called “late” vision allows for cognitive influence. it is here that the advantages of malleability come clearly into view, since it will better explain object recognition at any “later” stage. suppose that some instances of expert recognition occur at the (late) perceptual level, that is, at the level of perceptual experience. malleability has the advantage of allowing for rich perceptual content. thinking and perceiving defends the claim that the expert sometimes enjoys enriched perceptual content specific to her domain of expertise, a visual sensitivity to patterns, gestalts, and organizational features, be they of birds, mammograms, or patterns of play in sport. and this is partly a result of the cognitive-rich etiology of their learning the relevant perceptual skills. this well explains the efficiency and accuracy with which experts act and judge in category-sensitive ways. and it is an explanatory advantage not open to modularity. suppose that some instances of expert recognition occur “post-perceptually”, as they surely do. given malleability, experts are especially attentionally tuned to features, patterns, and gestalts that aid in this recognition. and again this explains why (even if sometimes “post-perceptually”) the expert is exceptional at making relevant identifications and diagnoses in her domain. those visual attentional advantages are, or so i’ve argued, cognitively sensitive as well. the modularist can always add some additional stages of perceptual processing, or locate the performance advantages in something even “later”, say in visual judgment or perceptual seemings, but it’s here that the proliferation of distinctions really does look unprincipled. adams focuses on varieties of perceptual improvement, and asks “are they exclusively the result of developmental changes (i.e., acquisition) or can they result from”agent-driven, accuracy enhancing training” (stokes 2021, 182) (i.e., learning)?” (p. 2). he then uses, as i understand it, the distinction to put modularity on the first side and malleability on the second. although the distinction is clear enough, 1 i would resist that it will neatly divide modularity from malleability as suggested. my view allows for both kinds of perceptual improvement: cases of what adams calls “acquisition”, and cases of what he calls “learning”. if there is a relevant demarcation here, it is that modularity seems not to allow for perceptual improvement that involves adams’s form of learning when it is agent-driven and 1it is worth noting that although this distinction is attributed to me, it’s not one that i make. in the book, i talk about acquisition as concerning a learning period. so i don’t oppose the terms in this way. stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dustin stokes 6 cognitive. while, again, malleability does allow for it, and such cases are central to the kind of perceptual improvement that populate the book. there are a few other places where i would resist adams’s characterization of my view. i highlight these not to quibble, but because doing so will hopefully bring to light some lessons that can be drawn from the suggested acquisition/learning distinction. adams asks “what, exactly, is involved in ‘cognitive or conceptual learning’ and why is it required for learning?”(p. 2) in answer to the first question, in brief, cognitive learning involves semantic concepts, categories specific to a domain, typically some kind of linguistic or auditory feedback. it is thoughtful in some intuitive way and involves personal level effort or engagement. in answer to the second question, cognitive learning is not required for learning (even circumscribing learning in the way that adams does). referring to a pair of quotations, adams writes, “as these quotations make clear, stokes thinks that in order for”conceptual learning” (i.e., learning) to occur, “expert training” must be involved. this immediately raises the question: what exactly is “expert training” and why is it required for learning?” (p. 2). what i would hope these quotations instead make clear is that learning can occur without uptake of conceptually rich categories. there are genuine cases of perceptual learning, ones that occur as a result of repeated exposure to instances of a kind (this is the point of the sommelier taster and color chip examples). accordingly, expert training is not required for learning (in the general sense or in adams’s circumscribed sense). but it is required for some instances of perceptual expertise. and it’s those instances (of the kind discussed in the précis and in this reply) that require more than mere exposure, that require the cognitive learning described just above. so, malleability allows both for perceptual improvement that occurs just as a result of a natural developmental trajectory (adams’s “acquisition”), and for perceptual improvement that involves “pure” perceptual learning (i.e., the kind that involves mere exposure). adams may then ask: why the emphasis on perceptual improvement that involves cognitive learning in its etiology? the answer to this question brings us to central themes in the book. first, these are cases where the “pure” perceptual explanation favored by the modularist falls short. second, the central claim—that thinking improves perceiving—is best defended by focusing on agent-driven, concept-rich cases of perceptual improvement. here perception becomes more skillful, and as a result of cognitive activity that clearly involves the agency of the perceiver. here, the individual achieves epistemic virtue through her cognitively sensitive perceptual improvement. 4 agency and epistemic virtue virtue theory centers on achievement, and this is no less true of a virtue epistemology than a virtue ethics. traditionally, virtues are tied to individual flourishing. one can trace this all the way back to aristotle. an agent flourishes epistemically as they improve their skills in acquiring information from the environment. stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how radical is perceptual malleability? a reply to commentators 7 the information that enables flourishing will vary with the individual and the context. that is, an agent’s intellectual achievement will be partly tied to their goals, tasks, and behaviorally relevant needs. everybody is different, as the saying goes, so how could it be otherwise? perception plays a crucial role in such epistemic success, and that success is broader than (but includes) truth and accuracy. as will become clear below, contrary to drayson’s reading of the theoretical space, i don’t think this puts the view at odds with traditional or contemporary virtue theory. and it may depart from some of the points of emphasis of “most traditional epistemologists” (by which i assume drayson means non-virtue epistemologies in the anglo-analytic tradition), but that departure is intended. adams is right to underscore the varieties of perceptual improvement. sometimes these improvements are ones that occur without the agent’s doing. increased sensitivity to distinct categories of stimuli, can simply occur as a consequence of natural sensory development and/or repeated exposure. but many cases of expertise are different. here the agent has, deliberately and laboriously, undergone training in a specific domain: in radiology or fingerprint examination, in tennis or football, in ornithology or visual depiction. as a consequence of this training, her perceptual systems perform in exceptional ways within that domain. and those levels of performance near-maximally satisfy the norms for perception, thus fulfilling the representational function of perception in optimal or near-optimal ways (in that domain). the important epistemic difference between this case and the cases of mere development or exposure (“acquisition”) is that the agent is herself responsible for the relevant etiology and, accordingly, for the perceptual improvement. i highlight these cases of perceptual improvement because they highlight the individual, and the achievements that clearly extend from the individual’s agency. it is because of the expert’s efforts, including cognitive efforts, that she makes the perceptual improvements she does. some of those improvements concern accurate perceptual representations, but the improvements are not limited to accuracy or truth since those are not the limits of epistemic interest. if they were, then epistemology would have it that true belief is the final epistemic value. but of course it’s not: there is supposed to be some additional ingredient that elevates mere true belief to what we call knowledge. the virtue theorist’s solution to this “value problem” is a natural one. because neither getting to the truth, nor doing so reliably, exhausts intellectual flourishing, the additional “internal” ingredient is the agent and her credit-worthy cognitive achievements (see zagzebski, 2003). perceptual experts understood as virtuous agents better find truth, and accurately perceive the world, but which truths (which objective features of the world) they find may be partly determined by their aims and tasks within an epistemic context. hopefully this recapitulates the motivation to characterize perceptual expertise not in terms of so-called traditional epistemology but instead in virtuetheoretic terms. cases of perceptual expertise (those that involve the cognitive learning that adams hones in) are richly active. they are cases where one plays stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dustin stokes 8 an active role not just in how one perceives in the present, but in how one will perceive in the future, and by virtue of the (cognitive) efforts that one makes now. this makes perceptual success richly agential. it makes us agents in how we make contact with the world. 5 (non-)traditional epistemology: beyond truth and knowledge as should be clear, virtue epistemologies are agent-centered, not truth-focused as drayson argues. this does not imply that truth is not important. it only means that it’s not the central focus of importance. the focal point of epistemic assessment is the agent rather than some state of the agent. to miss this is to miss the motivation for a virtue theory, and the value that comes with an achievement-based epistemology. the agent matters centrally because, on a virtue theory, her flourishing is the central value. forming true beliefs on the basis of accurate perceptual representations is an important way to flourish. but it is not the only way. sometimes it’s which accurate perceptions that matter most, of an array of possibilities. it’s which features of the world that one picks up on, how efficiently, and with less distraction. it’s patterns and diagnostic features that suit one’s behavioral needs and goals. and it’s not always more-is-better: some patterns or diagnostic features are more important, to an individual, than others. accordingly, different individuals will enjoy distinctive conditions for intellectual flourishing. given their distinctive goals, the expert goalkeeper and the expert sports doctor will be sensitive to distinctive features of the striker making penalty kicks. their respective perceptual success depends on this, and when optimally successful, they will each perceive (accurately or verdically) different features of the striker’s actions. this, again, is why the account given in the book encourages a notion of perceptual success that subsumes, but broadens beyond, mere truth or accuracy. to be clear, it is thus not an abandonment of those epistemic values, but it is a push beyond them. i don’t have the sense that this renders the view radically pragmatist, even if it doesn’t comport perfectly with all other epistemologies. to what degree, then, does this put my view at odds with (non-virtue theoretic) traditional epistemologies, as drayson suggests? she writes, “if we switch to a success-based notion of epistemology, then claims about the architecture seem less relevant. the more we focus on success rather than truth as our epistemic goal, the less reason we have to suppose that cognitive architecture will provide the answers to our epistemological questions” (p. 5). there is a lot to say here. first, i would be cautious about a descriptive cognitive architecture providing answers, by itself anyway, to our normative epistemological questions. second, if we characterize the success conditions for perception in the broadened ways i’ve suggested, why does architecture seem less relevant? one obvious point of relevance is this: if such stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how radical is perceptual malleability? a reply to commentators 9 success is partly agentand context-sensitive, then the kinds of influences that the agent, her context, and her cognition can have on her perceptual processes are of central importance. this is a question (or set of questions) of cognitive architecture. finally, the tip thesis (that thinking improves perceiving) is the normative claim defended in the book. but one can adopt the descriptive tap claim—that thinking affects perceiving—without adopting the particular tip thesis defended, perhaps pursuing a different version of that normative claim. this, indeed, is what the first portion of the epistemology chapter (chapter 7) offers: a sketch of how epistemic internalists, reliabilists, and rationality of perception theorists might treat cases of perceptual expertise understood in terms of malleability. those sketches, however, are not adopted or defended; the virtue theory is defended. this last point is worth a bit more emphasis. drayson worries that my “analysis cannot be understood merely as a supplement to existing epistemological models” (p. 2) and, in particular that if “stokes were to take this pragmatist approach to rationality, he could retain his claim that ‘experiences of perceptual experts contribute in a positive way to their rational standing as epistemic agents’ (178), but he would no longer be engaging with siegel’s concept of theoretical rationality” (p. 4-5). siegel’s innovation in this space is to offer an avowed proof of concept for the rationality of perception (siegel, 2017 ). what the malleable mental architecture of expertise points to is some empirical support for that model. it is in this sense that it is a possible supplement to and, accordingly, an engagement with the view. admittedly, it is not an endorsement of the view. a couple paragraphs below the quoted passage, to transition to the endorsed virtue-theoretic account, i write “what each of the approaches lack, however, is a sufficiently robust emphasis on the role of the expert, qua agent, in the epistemic successes of her expertise. the most natural and illuminating way to highlight the agent is to turn to theories of epistemic virtue” (p. 179). 6 how radical is malleability? openness to theoretical radicalness is not unlike risk tolerance. as we’ve learned in these recent pandemic years, such things can vary quite dramatically from individual to individual. most of the allegedly radical features of thinking and perceiving are ones that i didn’t find to be so radical when i wrote them, and most of them were explicit in those pages. but my critics are certainly correct to point to some of the revisions to orthodoxy and to question just how far they go. to close, i’d like to consider two of the most substantive of those (possible) revisions: one epistemic and one metaphysical. these ideas are not fully developed in the book, but they are there in short form, and drayson and mole, respectively, have drawn me out on them. there is a continued myopic emphasis in contemporary, orthodox epistemology of perception (by which i do mean non-virtue epistemologies in the angloanalytic tradition), on propositional knowledge. this is, or is close to, goldman’s stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dustin stokes 10 “familiar assumption”, as i understand it. accordingly, such epistemologies hyperfocus on truth and accuracy, as attributed to discrete states of the perceiver, often described in computational terms, as static outputs of perceptual processing. what this misses is the wider range of uses to which we put perception, and the ongoing and active process that is perception. we perceive to know, to get to the truth. but, and probably more often, we perceive to navigate, act on, and understand the world. and we do this through our agency. a distinctive but non-exclusive epistemic value is understanding. understanding involves a mental grasp of order, patterns, cohesiveness, of how the parts of the world hang together. it is not, on most analyses, reducible to some set of propositional knowledge. put another way, by contrast to knowledge, it does not target single propositions (kvanvig, 2003; zagzebski, 2001). perceptual expertise comports with, indeed encourages emphasis on this epistemic value. the expert, through ongoing and active cognitive and perceptual engagement, recognizes and grasps patterns, coherence, structure, and relations between the elements that comprise their domain of expertise. some of this understanding, on the malleable architecture, is perceptual. the high performing expert is able to visually recognize and grasp the patterns, organization, and coherence of those domain-specific elements. since this kind of understanding is an ongoing process, and one involving an achievement attributable to the expert qua agent, it aligns well with the virtue-theoretic analysis defended in the book. it encourages us to think about the perceiver and her cognitive and perceptual processes, rather than about discrete perceptual states and their epistemic standing. accordingly, it deemphasizes “getting it right” in the form of just getting accurate perceptual experiences and true beliefs. and this does depart from contemporary orthodoxy in “traditional” epistemology, both in terms of the central target for epistemic assessment and the central epistemic value. thinking about perception as an active process may also encourage revisions to the underlying metaphysics of perception, and perhaps some of them the ones to which mole points. indeed, mole suggests that the more radical reading of thinking and perceiving and the malleability it endorses is to take “perception or cognition to be something other than a computational process that generates representations” (p. 3). drawing on remarks made in the epilogue of the book, he writes “when arguing for the rejection of encapsulation, [stokes] is hoping to point us towards a picture in which the relationship between facts about beliefs and facts about representations in the brain is metaphysically different in kind from the relationship between facts about perceptual experiences and such facts about the brain” (p. 4). and then finally, mole offers both a safe route out of a familiar problem in this territory—the “qua problem”—plus further suggestion on a more radical metaphysics: the effects of expertise that stokes cites can therefore be taken as support for his claim that thinking has a beneficial effect on perception, but they should not be taken as establishing that the processes of stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how radical is perceptual malleability? a reply to commentators 11 perception are influenced by one’s conscious beliefs as such. the way is therefore open for him to adopt a position in which conscious attitudes stand in a metaphysical relationship to the brain that is different in kind from the relationship that is seen in the case of experiences (p. 7). the safe route is one that i would take. cases of perceptual expertise as characterized do not require cognitive influence on perception in the form of cognitive penetration by conscious beliefs (as such). the plurality of explanation of cases of expertise should make this clear. and the most plausible kinds of cognitive influence are going to be diachronic, where cognitive learning figures in the etiology of perceptual improvement. but this isn’t mole’s most provocative suggestion, as i understand it. that provocation concerns whether we should abandon a classically computational metaphysics of the mind. the safe answer is that the contents of thinking and perceiving do not commit me to that abandonment. one can accept the architecture and epistemology contained therein and keep some kind of ctm. “but, wink wink, nudge nudge,” mole will press, “what do you really think? should we get beyond (out of?) the boxes?” well, yes. i think we should. (if we are getting confessional here: i’ve never liked the boxes, not least for the stares of incredulity that come from my students when i teach boxology.) but, have i given an explicit argument or analysis for doing that? no, i have not. at best, then, i can offer some sketchy thoughts in reply to mole’s invitation. what i think the book encourages is abandoning thinking of perceptual experiences as passive, static snapshots. 2 as i suggest in that same discussion in the epilogue, the static snapshot characterization is ecologically and phenomenologically unsound: we do not perceive the world in isolable moments, through discrete experiences. we instead make mental contact with the world in a dynamic, ongoing process. accordingly, this observation sits oddly with the view that treats experiences as discrete states outputted by a computational process. that is not to suggest that perceptual experiences are not representations; i maintain that they are. but it does suggest that they do not fit naturally in boxes, in the classical sense. if perception is a process, active and temporally extended, phenomenologically non-discrete, and the world is a non-static constantly variable “stimulus”, then we might expect the neural correlates or grounds to take the same form. such representations might accordingly be distributed across neural activity, rather than being discrete, atomic outputs of linear computations. there are familiar and formidable options here, taking some connectionist rather than classical computational form. indeed, some have gone as far as denying that such psychological processes are computational at all (at least in the classical 2interestingly, some of mole’s suggestions center on analogous changes to the metaphysics of conscious propositional attitudes such as belief. but i think this probably does stretch well beyond the claims and analyses in the book. stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dustin stokes 12 sense), but instead dynamic, evolving features of the overall biological organism or system (van gelder, 1995). again, nothing in the book commits me to such a view. but working through some of the ideas in the book have made me better recognize its appeal. is this a radical lesson? perhaps. it will depend upon the context in which it is received. if steeped in two orthodox theories—modularity and classical computationalism—then it is a clear departure from the basic claim that perceptual representations (like all mental representations) are atomic outputs of rule-bound computational processes. from that perspective, then, it is radical. but when not entrenched in such theories, the following observations seem almost unavoidable. life is a process. mental life is a process. and perceiving with it. malleability embraces these observations, and further embraces the notion that we are active agents in the process, including the perceptual process. it puts more of us, as individuals, in what we perceive. it says that our contact with the world is very much our own. to some ears this may be radical in some pejorative sense. i would like to think, however, that it is radical in the sense that it is getting down to grasping reality at the roots. references gauthier, i., & tarr, m. j. (1997). becoming a “greeble” expert: exploring mechanisms for face recognition. vision research, 37(12), 1673–1682. https://doi.org/10.1016/s0042-6989(96)00286-6 humphreys, g. w., price, c. j., & riddoch, m. j. (1999). from objects to names: a cognitive neuroscience approach. psychological research, 62(2-3), 118–130. https://doi.org/10.1007/s004260050046 kvanvig, j. l. (2003). the value of knowledge and the pursuit of understanding. cambridge university press. marr, d. (1982). vision: a computational investigation into the human representation and processing of visual information. w.h. freeman. raftopoulos, a. (2009). cognition and perception: how do psychology and neural science inform philosophy. mit press. riddoch, m., & humphreys, g. w. (2001). object recognition. in b. rapp (ed.), handbook of cognitive neuropsychology. (pp. 45–74). psychology press. scott, l. s., tanaka, j. w., sheinberg, d. l., & curran, t. (2008). the role of category learning in the acquisition and retention of perceptual expertise: a behavioral and neurophysiological study. brain research, 1210, 204–215. https: //doi.org/10.1016/j.brainres.2008.02.054 siegel, s. (2017). the rationality of perception. oxford university press. tanaka, j. w., curran, t., & sheinberg, d. l. (2005). the training and transfer of real-world perceptual expertise. psychological science, 16(2), 145–151. https://doi.org/10.1111/j.0956-7976.2005.00795.x van gelder, t. (1995). what might cognition be, if not computation. journal of philosophy, 92(7), 345–381. https://doi.or g/10.2307/2941061 zagzebski, l. (2001). recovering understanding. in m. steup (ed.), knowledge, truth, and duty: essays on epistemic justification, responsibility, and virtue. oxford university press. zagzebski, l. (2003). the search for the source of epistemic good. metaphilosophy, 34(1‐2), 12–28. https://www.jstor.org/ stable/24439221 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. stokes, d. (2023). how radical is perceptual malleability? a reply to commentators. philosophy and the mind sciences, 4, 15. https://doi.org/10.33735/phimisci.2023.10954 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/s0042-6989(96)00286-6 https://doi.org/10.1007/s004260050046 https://doi.org/10.1016/j.brainres.2008.02.054 https://doi.org/10.1016/j.brainres.2008.02.054 https://doi.org/10.1111/j.0956-7976.2005.00795.x https://doi.org/10.2307/2941061 https://doi.org/10.2307/2941061 https://www.jstor.org/stable/24439221 https://www.jstor.org/stable/24439221 https://doi.org/10.33735/phimisci.2023.10954 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction beyond the default expertise, learning, and recognition agency and epistemic virtue (non-)traditional epistemology: beyond truth and knowledge how radical is malleability? in search of lost time: integrated information theory needs constraints from temporal phenomenology. in search of lost time: integrated information theory needs constraints from temporal phenomenology ishan singhala,∗ (ishan20@iitk.ac.in) ramya mudumbaa,∗(mudumba20@iitk.ac.in) narayanan srinivasana (nsrini@iitk.ac.in) abstract integrated information theory (iit) of consciousness proposes an identity between its causal structure and phenomenology. through this assertion, iit aims to explain consciousness by prioritizing first-person experience. however, despite its phenomenology-first stance, developments in iit have overlooked temporality. as such, we argue that at present iit’s phenomenological analysis is incomplete. in this critique, we show how iit takes a non-identical illusionist stance towards the experiences of continuity, flow, and extent of our experiences. moreover, in isolating temporal grains of experience to a single timescale, iit misses out on capturing the multi-scale nested nature of temporal phenomena. hence, we contend that iit needs an axiom for time, through which its causal structure can be refined to account for temporal experiences. here, we propose an axiom to address these concerns. we conclude by discussing how iit may need to be revised if our concerns hold true. keywords consciousness ∙ integrated information theory ∙ phenomenology ∙ time 1 introduction integrated information theory (iit) of consciousness has become an established contender for explaining consciousness based on three key unique proposals in its manifesto (oizumi et al., 2014). they are: (1) iit asserts that a measure (phi, φ) that represents a system’s ability to integrate and differentiate information can be used to measure the degree of consciousness of that system (the so-called consciousnessmeter). (2) iit identifies its causal conceptual structure with subjective first-person experience, essentially conjecturing a psycho-neural identity for consciousness. (3) finally, iit also claims to be a more parsimonious explanation of consciousness adepartment of cognitive science, indian institute of technology, kanpur, india ∗equal contributions. singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-5342-0381 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ishan singhal, ramya mudumba, and narayanan srinivasan 2 because it prioritizes phenomenology and hence constrains (via axioms) its causal structure by keeping phenomenology first. iit has been criticized on all three of these grounds in the past. there have been arguments that iit’s measure(s) can be arbitrarily manipulated by changing a system’s design from a recurrent feedback loop to a feedforward structure (aaronson, 2014; doerig et al., 2019), or that iit measures only network efficiency in the brain and not consciousness itself (merker et al., 2021), or that iit’s consciousness meter entails or is somewhat consistent with panpsychism (merker et al., 2021; tononi & koch, 2015). furthermore, iit’s axioms are criticized for failing to account for phenomenology, being underdeveloped, heuristic-like, and being tautological in saying the same thing in different ways (bayne, 2018; merker et al., 2021; pokropski, 2018, 2019). oddly still, the phenomenology-first nature of iit is used to defend against the theory being circular, with the circularity being broken apparently by grounding the theory in phenomenology (negro, 2020; tsuchiya et al., 2020). it has been previously argued, however, that iit has no phenomenological grounding or framework around which it builds its theory (pokropski, 2019). perhaps nothing makes this more apparent than the absence of a cornerstone of phenomenological analysis, namely temporality (husserl, 1964). at the very heart of phenomenological analysis is the temporal horizon, which is the necessary and inherent breadth of our experience over time. for not only husserl who called it the most important and difficult problem in understanding the phenomenology of conscious experience (husserl, 1964), time has been at the center of phenomenology also for kant, heidegger, hegel, merleau-ponty, sartre, and james, and most recently varela and gallagher (andersen & grush, 2009; dainton, 2010). in fact, it is hard to find a phenomenologist who does not think time is a fundamental aspect of conscious experience. perplexingly, however, temporal phenomenology literature and frameworks developed over three centuries remain absent from iit. not only has there been no place for temporal phenomenology in iit (kent & wittmann, 2021), but also the work done by those mentioned above has not been incorporated into iit. this is further reflected in the absence of any reference to these phenomenologists or their work on time in any of the main original papers of iit (oizumi et al., 2014; tononi, 2004, 2008). this paper is thus aimed at investigating and critiquing the phenomenologyfirst and identity claims of iit, specifically from the perspective of temporal phenomenology. in this paper, we critique the lack of temporality in iit and what this absence entails. we argue that iit does not adequately account for phenomenology and consequently cannot justify itself as giving priority to or being identical with phenomenological experience. moreover, in ignoring temporality both in and of experience (i.e. both time-consciousness and timing of consciousness), iit misses out on constraining the purported temporal evolution of conscious content and its structure. the theory fails to make adequate use of the existing literature on the timing of consciousness as well as time-consciousness. previously, iit has singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org in search of lost time 3 been criticized on these grounds by kent & wittmann (2021), where they point out the demerits of relying on a single time-point value of phi measured over narrow time scales ranging a few hundred milliseconds and thus falling short of explaining the gamut of temporal phenomenology of experience. we extend and add to their criticisms comprehensively in this paper. before moving on to making these arguments, we briefly discuss why temporality has an important status in understanding phenomenological experience. in the sections after this, we offer our specific criticisms of iit. we point out several temporal properties lacking in iit, for example, continuity, extent, nestedness, multiscale time, and direction. we argue that an absence of a phenomenological account of these phenomena in iit renders it incomplete and in need of revision at best or falsified at worst. finally, we conclude by proposing an axiom of time that may help the theory in accommodating key aspects of phenomenology and reworking its conceptual structure. 2 why is temporality important for a phenomenological theory of consciousness? why should iit or any theory of consciousness care about temporality? if there is any property of consciousness that can be argued to be common between a causal structure of experience and consciously experienced content, it is time (cohen, 2011; james, 1890; kent & wittmann, 2021; kiverstein, 2010; kiverstein & arstila, 2013; phillips, 2010, 2014; varela, 1999). to experience the color red, nothing in the causal structure underlying the experience of “red” itself has to be colored. this is not only true for color but other kinds of experiences as well; the causal structure of experiential states of emotion or hunger themselves are not emotional or hungry. however, it has been posited that the temporal evolution of content is isomorphic1 to the temporal evolution of the structure that generates it (dainton, 2008; kiverstein, 2010; phillips, 2010, 2014). that is, not only does our experience represent temporal properties, but the structure responsible for our experience has temporal properties of its own. by virtue of this, if one has to understand the relationship between a causal structure of experience and experience itself, one has to do it via temporality (singhal & srinivasan, 2021). thus, if there is indeed something common between the vehicle and content of experience (as also claimed by iit), time or temporality is the ideal starting point (or perhaps the only starting point). another reason why time has a special status in phenomenology is because of the experience of a “now” (james, 1890). a purported organizing principle of our experiences is that they occur in a manner that is both extended and present-like (droege, 2009; kelly, 2005; pockett, 2003). literature in phenomenology is resplendent with illustrations of experiences such as watching a bird glide across the sky, 1for an opposing view, see dennett & kinsbourne (1992). singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ishan singhal, ramya mudumba, and narayanan srinivasan 4 listening to an opera singer holding a note, tracking the feather’s fall to the ground, pursuing a high-speed projectile in sports, etc. these experiences, at least, seem to occur over a continuous and extended time interval. more formally, an extended present-like organization of our experiences is accompanied by the ability to perceive order, duration, persistence, simultaneity, flow, succession, and continuity both within and between them (kelly, 2005). these have been argued to be not only fundamental properties of consciousness but also necessary aspects of a phenomenological experience. recent proposals (singhal & srinivasan, 2021; varela, 1999; windt, 2015) have argued that these temporal properties can be deployed to describe the necessary properties of consciousness and unify existing findings in the field under a ‘minimal unifying model’ approach (wiese, 2020). to do justice and make use of the rich history of phenomenological analysis, iit needs to account for these minimal properties of consciousness. this requirement is critically important for any theory that claims to be a phenomenology-first and structureexperience identity theory. in its current formalization (oizumi et al., 2014), temporal phenomenology remains significantly ignored. in the following two sections, we show where and how time is missing in iit and what are the implications of this for the theory. 3 continuity 3.1 phenomenology a property of conscious experience that fails to be accounted for by iit is the continuity or flow of experience. in a sole mention of temporal frameworks, it is claimed that iit is parsimonious with a dynamic snapshot theory of time-consciousness (see tononi & koch, 2015, p. 16, also quoted below). according to this view, experience is made up of frame-like snapshots. however, unlike the traditional cinematic theory, where the snapshots are static, it allows for temporal information to be ‘represented’ in these snapshots (phillips, 2017). both these views nevertheless are against the naive view that our experience seems continuous because it is so. for instance, to represent motion, one could have a representation that itself changes or represents motion over an instant (like the speedometer in a car). iit does not discuss the implications and motivations of choosing the latter. to a skeptic, it appears that the choice of a snapshot was made of convenience since iit’s causal structure is a discrete state system and would require a discretetime framework (oizumi et al., 2014). the convenience only ends here. quickly, iit runs into a major philosophical complication, which is further exacerbated by the combination of two claims one that of identity between structure and phenomenal experience (oizumi et al., 2014); and the other of disallowing overlapping time windows (to avoid multiple causations) (tononi & koch, 2015). being a phenomenology-first theory, iit tries to incorporate the property of continuity by positing that experience is “seemingly” continuous even though it is a succession singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org in search of lost time 5 of discrete snapshots, thus giving continuity the status of a mere illusion (see the quote below). the exclusion postulate requires that the set of mechanisms that specify one particular experience do so over the time window at which φ reaches a maximum. if the next experience involves an over-lapping set of mechanisms, it would seem that, to avoid multiple causation, it should be specified over a non-overlapping time window. accordingly, the seemingly continuous ‘stream’ of consciousness would actually be constituted by a discrete succession of ‘snapshots’, in line with some psychophysical evidence. note that each snapshot has motion and other dynamic percepts associated with it. (tononi & koch, 2015, p. 16, footnote 11) in taking an illusionist stance, iit inadvertently hurts its own identity claim because now there is an illusion of continuity (still a phenomenological explanandum), which is no longer identical to iit’s causal structure. further, the exclusion postulate necessitates that our experiences are bound in isolation and encapsulated in their own ‘time windows’ (tononi & koch, 2015). to unpack the nature of the underlying temporal structure that iit advocates, we consider three frameworks from timeconsciousness literature: the cinematic, retentional, and extensional theories (for a detailed review, see dainton, 2010). from a naive phenomenological viewpoint, if the structure of consciousness is claimed to be identical to experiential content (as in iit), prima facie, one would opt for an extensional framework. since the phenomenology of continuity is itself not contested, and if the identity must hold, it is essential that the underlying causal structure unfolds continuously as well. however, iit surprisingly champions a cinematic framework. however dynamic the snapshots are, continuity still has no other way but to end up on a shelf of illusions. furthermore, it is unclear what the phrase ‘time-windows‘ means within a cinematic framework. suppose it is the time-window at the end of which everything within the window is rendered conscious (herzog et al., 2020). in that case, it is not clear whether the time window (or scale) selected is the one ‘over which’ or ‘at the end of which’ φ reaches maximum. one could argue that iit entails a retentional framework where acts of consciousness are still momentary. however, it has to be noted that according to the husserlian framework (husserl, 1964), the content is still extended and has a nonuniform, tripartite temporal structure within a ‘now’. this temporal structure is constituted by a primal impression of the momentary now together with retention of the just-past and protention of the immediate-future (dainton, 2010). for example, husserl (1964, p. 202) says: “but each such momentary perception is the nuclear phase of a continuity, a continuity of momentary gradated retentions on the one side, and a horizon of what is coming on the other side: a horizon of ‘protention,’ which is disclosed to be characterized as a constantly gradated coming.” such an account would mean that even though the acts are momentary, the consinghal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ishan singhal, ramya mudumba, and narayanan srinivasan 6 tent of the ‘now’ is not punctate but overlapping. a husserlian conception of time poses a problem for iit in two ways. one, iit does not have an inherent temporal structure within a ‘now’. second, it does not allow for overlapping time windows to occur (tononi & koch, 2015). the latter has more serious implications in a cinematic framework because this makes it hard to explain the diachronic unity of consciousness, i.e., the unity of experience across time (dainton, 2010). even the ‘integration axiom’ talks about a kind of integration that is synchronic in nature (integration of different contents at a discrete moment of experience (tononi, 2015) but remains silent about the integration of content over time (kent & wittmann, 2021). we do not experience choppy snapshots but a continuous flow of experience. we experience change, succession, and persistence of a musical note held in time. suppose the snapshots are dynamic and changes over the time window are still “represented”. but if there is no overlap between two consecutive windows (or a mechanism connecting them), how would one explain the continuity of experience? hence, the combination of claims made by iit is not coherent enough to explain the phenomenology of temporal passage and continuity. 3.2 transitions between contents of conscious experience any theory of consciousness not only needs to explain how one becomes conscious of something but also how within our consciousness things come to be, change, and disappear (aru & bachmann, 2017; fekete et al., 2018). how do we go from perceiving one thing to perceiving something else? surprisingly, proponents of iit pay little attention to answering this question from phenomenology. as mentioned earlier, state transitions of experience for iit are discrete and nonoverlapping. it is important that iit provides the rationale behind employing a discrete snapshot theory based on phenomenological evidence because the evidence for a discrete cinematic framework is not straightforward or self-evident at all. the phenomenological evidence should take precedence over computational tractability. it has been argued that the experimental data that apparently shows that perception is discrete could also be explained by assuming continuous models (fekete et al., 2018; piper, 2019). the existence of finite-length windows of integration does not necessarily imply that conscious experience is discrete in nature (piper, 2019; vanrullen & koch, 2003). in fact, theories that posit an isomorphic relationship between experience and the substrate and take phenomenology seriously are mostly extensional theories. for example, the temporo-spatial theory of consciousness (ttc) claims that the common currency between the brain’s neural activity and dynamics of experience is the shared temporo-spatial properties between both (northoff et al., 2020; northoff & huang, 2017), that is there is an isomorphic correspondence between temporo-spatial expansion of neural activity and that of phenomenal experience beyond the duration of the stimuli (see section 4.3, northoff & huang, 2017). the reentrant oscillatory multiplexing (rom) theory also proposes an extensional framework to explain the temporal properties of singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org in search of lost time 7 experience like continuity via the sustainability of the phase coherence between multiple neuro-oscillatory frequency bands (piper, 2019). consider this quote that highlights the discrete account of experience in iit: according to iit, an experience is identical with a maximally irreducible conceptual structure specified over itself by a complex of elements in a state, and it exists at a discrete interval of time at which cause-effect power reaches a maximum. it is important to establish how this relates to the dynamics of the elements of the complex at a faster time scale and to the temporal evolution of the states of the complex (barrett and seth 2011). it should be noted, however, that a complex in a state may specify a conceptual structure even when it is dynamically at a fixed point. (tononi, 2015, footnote 22) the often-made mistake in time-consciousness literature by cinematic models is to mix a succession of experiences with the experience of succession while they are regarded as phenomenologically different (dainton, 2010; hoerl, 2013). a cinematic model inescapably treats those two experiences as one, given that the snapshots are discrete and extension-less. assuming that iit posits a dynamic snapshot theory (tononi & koch, 2015), the experience of succession can only be salvaged within a time-window without mixing it with succession of experiences. however, it raises the question of how iit can explain the experience of succession across windows without making it synonymous with the succession of experiences. there is no other option left but to fall back on the succession of two snapshots to explain the experience of succession between consecutive windows. consider a case where there is a beat1-noisebeat2noise-beat3. we do integrate the beats across two non-consecutive windows (if one may say) and over time (bregman, 1994, chapter 3; koenderink et al., 2012). such cases prove much harder for a cinematic theory to explain the experience of succession without employing a separate mechanism (probably operating at a different time scale than the windows) that stitches content over windows. integration of content over non-consecutive windows poses a problem to the identity between the dynamics of content and structure that iit celebrates. what we hear is a succession of beat1→beat2→beat3, but the succession of windows would imply beat1→noise→beat2→noise→beat3. this creates an inconsistency between how the content unfolds in experience and how the structure does. 3.3 implications the current workings of iit entail a flexible timescale over which conscious experiences stabilize, in the range of a few hundred milliseconds (tononi & koch, 2015; for a criticism, see barrett barrett, 2016). however, this flexibility does not explain how two moments of experience are connected over time. this is problematic given that iit’s conceptual structure postulates isolated moments of experience in singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ishan singhal, ramya mudumba, and narayanan srinivasan 8 non-overlapping windows. such a postulation is neither derived from nor explains the phenomenological experience of continuity. more importantly, iit does not develop or adhere to existing doctrines in temporal phenomenology. for instance, previous ‘phenomenology-first’ attempts at explaining consciousness have adopted either a jamesian or husserlian view of time (grush, 2006). in its current proposal, iit does not strictly fit into existing proposals of timeconsciousness or propose one of its own. this leaves the phenomenological foundations of the theory and its conceptual structure incomplete and perhaps incompatible with phenomenological analyses of time. ideally, the theory needs to explicate its phenomenological grounds with respect to temporal unfolding (fekete et al., 2018; kent & wittmann, 2021); otherwise, it remains incomplete and does not allow us to understand consciousness. let’s assume iit would want to adhere to a dynamic snapshot conception of temporality (tononi & koch, 2015). then a more nuanced and detailed analysis of iit’s compatibility with the dynamic snapshot idea is needed (for instance, see phillips, 2017). another similar proposal is herzog, drissi-daoudi & doerig (2020), wherein they propose a discrete retentional framework for visual experiences. though much ink has been spilled between the proponents of iit and herzog et al., being at odds on the spatial organization of a system (aaronson, 2014; doerig et al., 2019; negro, 2020; tsuchiya et al., 2020), their respective temporal underpinnings seem to be entirely in agreement. that is, the unfolding argument posits that any recurrent feedback system can be ‘unfolded’ into a purely feedforward system such that their input-output functions match entirely. several proposed measures of φ in iit would assign different values for these systems despite them having the same input-output functions. on the other hand, it is argued that iit’s causal structure is not arbitrary. rather by virtue of being phenomenologically identical, it is grounded by a phenomenological structure of experience (negro, 2020; tsuchiya et al., 2020). moreover, iit is about ‘being’ and not ‘doing’, i.e., it is phenomenologically grounded and not a functional theory of consciousness (albantakis & tononi, 2021). despite these proclamations, iit and herzog et al. share the same illusionist (phenomo-temporal-antirealist) stance when it comes to temporality. 4 temporal hierarchies and nestedness 4.1 multiple timescales of experience multiple timescales have historically been postulated to understand not only different temporal phenomena but also different properties of temporality (andersen & grush, 2009; dorato & wittmann, 2020; varela, 1999). an intuitive and oft-cited example shows a difference in timescales for which we are directly aware of changes (in our perceptual experiences) and timescales in which we must make inferences to be aware of changes. consider, for instance, being aware of the second hand of singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org in search of lost time 9 a clock moving vs. being aware of the hour hand of the clock moving. presumably, in both cases, we are aware that a change has occurred; however, in the former, it seems as though we can see the change occurring, while in the latter, the change is inferred. over three millennia, different estimates have been postulated for the breadth of time for which we can notice a change directly in our experience. in contemporary phenomenology, these estimates range from a couple of seconds to ~12 seconds. neuroscientific methods, however, have appropriated this breadth of the ‘present moment’ to mean either large-scale integration cycles in the brain or the time it takes for us to become conscious of a change (white, 2018). studies here usually postulate values between 100-300 milliseconds (hogendoorn, 2022). this range is also precisely where iit claims consciousness operates (i.e., at ~100 msec scale of windows). however, as one can perhaps more strikingly appreciate when a perceiver becomes conscious of something, it is not the same as the extent of the moment they are perceiving. to give one last example, before moving ahead, consider listening to a song that has polyrhythms. it could be mozart’s don giovanni being played in two different meters, or the constant meter switching in brahms’s violin sonata, all of the indian carnatic music, or perhaps even the fast double-time swings of a drummer. one of the magical experiences of hearing these music pieces is the nested experience of change in auditory content at two different rhythms, arguably being experienced under the nesting of an extended present moment (grush, 2006). it is as if what we were listening to in the finest temporal frame was embedded inside a hierarchical temporal content structure. thus, the notes being played at any instance seem inseparable from their temporal context. a more intuitive way to appreciate the above example is to think of it as a figureground separation also in time and not just in space. therefore, when iit postulates that conscious content updates at a single timescale of 100ms in non-overlapping windows, it does not seem like this postulate is based on self-evident truths from experience (as posited by multiple phenomenologists like husserl, merleau-ponty, varela, etc.). 4.2 nestedness within experience in psychology and cognitive science, the mechanisms responsible for and those that work around consciousness (attention, perception, memory, actions, predictions) have different but consistent time scales. the mechanisms by which we recognize and identify changes, estimate duration intervals, perceive motion, infer causes, and integrate scenes or information from multiple modalities seem to all operate in different but interrelated hierarchical and nested timescales (montemayor & wittmann, 2014; pöppel, 1997; singhal & srinivasan, 2021; varela, 1999). moreover, the environment in which we act also consists of events happening at multiple nested timescales. singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ishan singhal, ramya mudumba, and narayanan srinivasan 10 to recognize speech, for instance, an individual simultaneously needs to parse out phonemes, syllables, words, prosody, tone, and pauses. of course, this also involves localizing sound differentiating pitch and timbre. if we look at the timescales over which these processes stabilize, we would have a range that goes from 1-2 ms (localizing sound) to a couple of seconds (prosody and tone information). while iit does allow for a flexible time-window of experience, it does so exclusively. that is, they tie experience to one timescale at any given moment of conscious experience. an iit-derived dummy construction of this experience would perhaps postulate that we go from localizing sounds progressively to understanding the prosody of speech. essentially, φ reaches a maximum at different time scales for different properties of speech, serially. moreover, by the axiom of exclusion, only one timescale is tied to experience at any given moment. hence, iit would not be able to account for multiple timescales and multiple facets of auditory experience (at any moment). however, a theory of consciousness that postulates that each moment is a temporal whole that underlies a nested hierarchy of multiple timescales would not face this problem (see northoff & zilio (northoff & zilio, 2022) specifically for a comparison with iit). note that an objection to the above argument might be that micro-level temporal properties at a finer grain are subsumed by a macro-level timescale in iit and thus maintained. to address this objection, let us revisit the example cited in grush (2006) about the temporal context of a melody. this melody repeats the same five notes in three consecutive bars. however, the notes in the third bar sound distinctively different, even though they are the same notes. grush (2006) explains this via the ‘temporal context’ set by the first two bars. the first two bars here persist as retentions during the present moment experience of the third bar. the example here shows how the perceptual content experienced at the moment is phenomenologically constitutive of both the temporal context over a longer time scale (1-2 seconds) and also by the notes themselves being played at a quicker time scale (tens of milliseconds). at the time of hearing the third bar, we hear not only the succession of the individual notes but also the persisting melody over the three bars. it is unclear how the framework of iit would subsume or embed phenomenological content that evolves over multiple timescales at one temporal grain. since the snapshots in iit represent content in isolated windows (dynamic snapshots), embedding multi-timescale phenomenology within this framework would lead to asynchrony or lags in perceiving this melody (hogendoorn, 2022; white, 2018; white, 2021). this would be remarkably unlike how we hear the melody. if our argument from this example is coherent, it will point to an instance of differences in how iit’s causal structure changes vis-à-vis our experience. this non-identical dynamical structure in iit and our experience has previously been proposed as a falsification criterion (negro, 2020). this is yet another way in which iit’s casual structure is not the same as our phenomenological experience. singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org in search of lost time 11 4.3 implications in its current formulation, iit’s temporal structure (or lack thereof) leaves out an enormous literature of temporal regularities in and of experience. evidence of mentality unfolding over multiple timescales is not accommodated in iit, either from psychology (pöppel, 1997; white, 2018; wittmann, 2011), neuroscience (northoff & huang, 2017), or even phenomenology (kent, 2019; montemayor & wittmann, 2014; varela, 1999). moreover, a single index measure (φ) is used to capture not only the vividness of the quality of an experience but also its inherent dynamic structure. iit represents time the same way it is represented on a postal letter via a date. this is a tag that refers to another tag, in this case, a calendar. this is exactly how dynamic snapshot theories argue mental time is represented in a static, non-punctate manner. a similar debate exists in neuroscience too (johnston & nishida, 2001), where it is contested whether neural representations of time are like event tags (punctuate) or dynamic (nonpunctuate and extended). here too, iit currently falls in the event-time group, where temporal properties of experience are represented separately from the experience themselves (which is itself inconsistent with how iit conceptualizes structure and content of experience). these implications of ignoring multiscale phenomena in our experience and their mechanisms require serious consideration and fleshing out by proponents of iit. from the integration axiom2, it follows that despite the phenomenal distinctions, and heterogeneous types of content, we still experience a unified whole, irreducible to these distinctions. however, it has to be noted here that the literature points out that different kinds of content unfold at different timescales, possibly nested within each other. given this, it makes one wonder how isolating a temporal grain at which conscious content updates in time (as in the exclusion axiom) would help achieve this feat of having such a complex texture of experience. we speculate that the exclusion axiom3 as it stands does not seem to handle multiple time scales. 5 past, present, & future of time in iit 5.1 status quo whether an axiomatic approach to study consciousness is an appropriate endeavor is not contested here (for such arguments, refer to bayne, 2018). we proceed henceforth, assuming that it is, limiting our discussion within the scope of iit’s axioms. 2“consciousness is unified: each experience is irreducible to non-interdependent, disjoint subsets of phenomenal distinctions.” (tononi, 2015). 3“consciousness is definite, in content and spatio-temporal grain: each experience has the set of phenomenal distinctions it has, neither less (a subset) nor more (a superset), and it flows at the speed it flows, neither faster nor slower.” (tononi, 2015). singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ishan singhal, ramya mudumba, and narayanan srinivasan 12 proponents of iit argue that time needs no axiom of its own based on the criteria of essentiality, according to which an axiom must apply to all experiences (tononi, 2015). since there exists an experience of timelessness (costines et al., 2021; josipovic & miskovic, 2020; metzinger, 2020) in some states of meditation, temporality is not an essential property of consciousness; hence there is no need to consider an axiom for time (see below). for example, one can argue that assuming the existence of a subject of the experience, in addition to the experience itself, is unnecessary; that an experience may stay the same without vanishing; that some experiences may seem timeless; that there may be experiences lacking spatial dimensions, as in some dreams, or situatedness (disoriented patients), or figure-ground distinctions (ganzfeld); that some experiences, such as boredom, may not refer to something in the world; that causality, change, and time may be derived from the existence axiom, and so on. (tononi, 2015, footnote 3) another notable property of consciousness is that experience changes all the time. however, since “timeless” experiences can occur, at least “for a short time”, it is arguable whether change should constitute an axiom/postulate. (tononi, 2015, footnote 12) timelessness and experience of other such states, however, are not unambiguously posited in the current times. their description and thus conceptualization are still underway (costines et al., 2021; metzinger, 2020; srinivasan, 2020; windt, 2015; woods et al., 2021). with our scientific understanding of such minimally phenomenal experiences (mpe) still evolving and rudimentary, it may be too early to entirely rule out an axiom for time based on an exceptional case. even if such a state has to be considered seriously by iit in its current formulation, we argue that it would imply complications for its existing axioms as well. the phenomenological descriptions that accompany such states are not only that such an experience is timeless (josipovic & miskovic, 2020; metzinger, 2020; srinivasan, 2020) but also perspective-less, composition-less, non-flowing, non-specific, contentless (where there is nothing to be integrated or differentiated), etc. these descriptions would thus entail invalidation of iit’s existing axioms. the properties of experience captured by iit’s axioms would also fail to be universal and fail the essentiality criterion test. mpes, in general, pose a problem for many theories of consciousness (srinivasan, 2020). 5.2 all axioms have “now” temporality and moments of ‘now’ are so inexorably linked to consciousness that an attempt to detail the properties of experiences necessitates an anchor of ‘now’. this is a feature either explicitly or implicitly in all of iit’s five axioms. one can singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org in search of lost time 13 appreciate this by looking closely at the axioms themselves and how the word ‘now’ is used in defining those axioms themselves. the existence axiom explicitly details that ‘consciousness exists here and now’. the second axiom of composition states that ‘within one experience i may distinguish’ several different components. here again, since an experience cannot exist without having a temporal extent, detailing something happening within an experience is again impossible without carving out a ‘now’. the third axiom of iit individuates each experience from another; by virtue of this, an experience is said to be informative. however, the specificity and differentiated-ness of one’s experience can only be via and inside a ‘now’. the fourth axiom (integration) illustrates how multiple phenomenal distinctions co-exist as a unified whole, and experience is irreducible to these individual distinctions. this is true not only in space (inside a visual scene) but also in time (inside a moment of now). the final axiom captures the idea that our experiences seem to have a specific spatio-temporal grain and flow at a particular speed, not faster or slower. interpreted from the view of temporal phenomenology, this axiom could be read as saying that our experiences have (1) a certain temporal breadth and (2) speed at which they change. this is just another way to describe a ‘now’ moment, albeit the breadth of this now speculated to be at ~100 milliseconds is not derived from any feature of experience, which is self-evident. a more intuitive way to understand what we mean by arguing that all of iit’s axioms necessitate an anchor of a ‘now’, is to ask: when?. iit posits that experiences are real, informative, compositional, integrated, and exclusive, but when?. ‘now’ may seem like a trivial answer to this question. our point of contention is, however, that the now is a temporal carving of experience via phenomenology. it is a present-like moment with an extent (james, 1890; kelly, 2005; kent & wittmann, 2021). a dialectic contradiction where aspects of just-past and about-tooccur futures co-exist within a temporal whole (dainton, 2010). it is one possible phenomenon that differentiates consciousness from unconsciousness (blackmore, 2012). without understanding the temporal properties of this phenomenological and psychological “now” and the flow of conscious experience, we cannot understand the structure of experience. this conclusion has been reached multiple times in the last few hundred years (andersen & grush, 2009; dainton, 2010; whitrow, 1980). 5.3 proposed axiom of time we concur with many previous proposals which argue that the study of consciousness can be constrained and meaningfully improved through phenomenology (van gelder, 1999; varela, 1999). as such, we appreciate the phenomenologically driven stance of iit. our contention here has been only that iits phenomenological analysis is incomplete. we propose a way to address at least one shortcoming in the singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ishan singhal, ramya mudumba, and narayanan srinivasan 14 phenomenological foundations of iit. we think that there is a need to propose a temporality axiom. previously, an axiom for time was proposed by hunt (2016) addressing the shortcomings of iit’s axioms in accommodating temporal passage of experience: “along with the undeniable reality of consciousness itself, the passage of time from moment to moment of our consciousness is known to each of us as certainly as anything else is known.” the corresponding postulate supposes a strong processtime philosophy, which assumes that there is temporal passage in the external world and our experience is a mapping onto it. further in this section, we propose our own axiom and contrast it with hunt’s. our proposed axiom comes from a first-person phenomenology approach. it would enable iit, and future theories that would be of similar nature, to fill the gaps pointed out by our critique. it takes temporality in our experience as primary and is based on a naive account, wherein we “experience” only in terms of a temporal whole. moreover, it makes explicit that our experience has a temporal extent and a direction. time axiom: consciousness is such that experience occurs to us as a temporal whole; i.e., experience always has an extension, is continuous, and has an inherent direction that is asymmetric. if this axiom is tenable, the conceptual framework in iit would need to be altered to accommodate temporal phenomenology. it also might be necessary to alter other axioms of iit to ensure the set is complete and independent, which we do not attempt to do in this paper. we are presently neutral to whether an axiomatic approach to consciousness is the way to develop theories of consciousness. our aim here is only to add an important but missing piece given the basic approach of iit. an objection to this axiom might be exceptional experiences (mpe, for instance), which are reported to be timeless (tononi, 2015). we have mentioned earlier in the paper how these experiences are problematic for other existing axioms of iit as well. perhaps such experiences may need to be understood better before they are used as justifications for excluding axioms. perhaps that may even require different ontological commitments. moreover, it is possible that mpes may lack many temporal properties but not all temporal properties4. for instance, these experiences may still have the temporal property of persistence, endurance, or extent5. these claims, of course, need to be formally investigated and remain only speculative for now. our larger 4vogel et al. (2020) proposed a temporal framework consisting of two time-layers (micro and macro) and two time-formats (structure and flow) to distinguish how the implicit temporal properties of experience form a basis for the explicit ones. within this theory, it is possible for only selective temporal properties to be distorted. 5for an example of this, see windt (2015) for a discussion on how mpes may involve losing a reference point in time where succession of events is lost. but the temporal extent of a phenomenological now is maintained. singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org in search of lost time 15 aim here is to point out that a phenomenology-first theory like iit cannot ignore temporality in and of experience. our proposal differs from hunt’s (2016) proposal in significant ways. one, we do not prescribe to a process-time philosophy or make ontological claims about time passage in proposing our axiom. our approach to time consciousness is via a first-person phenomenological perspective. two, we detail how passage of time is not the only temporal property of importance and we argue in this section how absence of passage (in the case of timelessness) does not have to entail an absence of temporality. since hunt’s proposal only includes one property of time (i.e.) passage, it does not capture other temporal properties. third, his proposal does not consider the nested hierarchical nature of temporal consciousness. fourth, the axiom as proposed by hunt (2016) is explicitly formulated in epistemological terms (“known to each of us as certainly as anything else is known”) potentially raising further questions especially given the cinematic conception of iit. the aim of hunt’s approach was to rewrite iit’s exclusion axiom while allowing it to account for temporal passage. our objective in this paper is not to rewrite a specific axiom of iit. however, the current formulation of iit is based on a cinematic conception of temporality where our experiences only ‘seem’ continuous (see section 3.). thus, the axioms need to be modified if they have to come from the truth of the first-person experience. how the axioms need to be rewritten or modified based on our proposal is a matter of future debate and is contingent on accepting that currently iit is unconstrained by temporal phenomena. 6 implications & conclusion in its current formulation of iit, we have comprehensively described how despite claiming to be a phenomenology-first theory, iit takes neither phenomenology nor its literature seriously. as a theory that champions the study of the intrinsic structure of consciousness, not just its extrinsic functions (ellia et al., 2021), proponents of iit ignore and misconstrue the one property that phenomenologists, psychologists, and neuroscientists have been spilling ink on for centuriestemporality. it is important in the context of iit, given it is also the only property conscious experience that is common between its structure and content and is argued to be one of the three candidates for a minimal unifying model of consciousness (mum) (wiese, 2020). upon close inspection, as we did, one can find the glaring inconsistencies that arise when one sits down to chart out temporal properties of experience within iit’s constraints and axioms. with the inconsistencies we point out and the axiom we have proposed to address them, we believe that at least one of four possible actions must be chosen by the proponents of iit. two of these entail iit either giving up its phenomenologyfirst and/or identity-based groundings, at least when it comes to experienced temporality. a third option is for iit to make accommodations for a time axiom and singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ishan singhal, ramya mudumba, and narayanan srinivasan 16 revisit the workings of its cause-effect repertoire in this light. of course, our criticisms could be entirely misplaced. showing we are wrong and time/temporality is not needed for explaining consciousness could be a fourth possible action taken by proponents of iit. tononi, massimini, boly, & koch (2016) end their paper by saying, “time will tell whether this account (iit) is anywhere near the mark”. they perhaps meant that the correctness of iit will only be established in the future as more work is done on it. however, in this paper, we have argued that (phenomenological) ‘time’ is already telling us that iit is not anywhere near the mark. in this critique, we have argued that iit, as proposed, does not address temporality explicitly to constrain its conceptual structure. specific temporal properties of experience are ignored or missing presently in the theory. further, the implicit suggestion of a discrete framework undermines several commitments made within the theory and falls short of explaining various temporal phenomena. thus, in the present form, there is neither a phenomenology-first stance nor a phenomenology-identity for time in the framework of iit. to partly rectify this, we have proposed here a new axiom for time. to accommodate this, we feel that iit needs significant revision. 7 acknowledgements ishan singhal and ramya mudumba were supported by individual prime minister’s research fellowship grants (government of india). we thank michael madary for his comments. references aaronson, s. (2014). why i am not an integrated information theorist (or, the unconscious expander). https://www.scottaaronson.com/blog/?p=1799 albantakis, l., & tononi, g. (2021). what we are is more than what we do. arxiv. https://doi.org/10.48550/arxiv.2102.04219 andersen, h. k., & grush, r. (2009). a brief history of time-consciousness: historical precursors to james and husserl. journal of the history of philosophy, 47(2), 277–307. https://doi.org/10.1353/hph.0.0118 aru, j., & bachmann, t. (2017). in and out of consciousness: how does conscious processing (d)evolve over time? frontiers in psychology, 8, 128. https://doi.org/10.3389/fpsyg.2017.00128 barrett, a. b. (2016). a comment on tononi & koch (2015)“consciousness: here, there and everywhere?” philosophical transactions of the royal society b: biological sciences, 371(1687), 20140198. https://doi.org/10.1098/rstb.2014.0198 bayne, t. (2018). on the axiomatic foundations of the integrated information theory of consciousness. neuroscience of consciousness, 2018, 1. https://doi.org/10.1093/nc/niy007 blackmore, s. (2012). turning on the light to see how the darkness looks. in s. kreitler & o. maimon (eds.), consciousness: its nature and functions (pp. 109–124). nova. bregman, a. s. (1994). auditory scene analysis: the perceptual organization of sound. mit press. https://doi.org/10.1121/1.408434 cohen, m. x. (2011). it’s about time. frontiers in human neuroscience, 5, 2. https://doi.org/10.3389/fnhum.2011.00002 costines, c., borghardt, t. l., & wittmann, m. (2021). the phenomenology of “pure” consciousness as reported by an experienced meditator of the tibetan buddhist karma kagyu tradition. analysis of interview content concerning different meditative states. philosophies, 6, 2. https://doi.org/10.3390/philosophies6020050 dainton, b. (2008). the experience of time and change. philosophy compass, 3(4), 619–638. https://doi.org/10.1111/j.1747-9991.2008.00153.x singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.scottaaronson.com/blog/?p=1799 https://doi.org/10.48550/arxiv.2102.04219 https://doi.org/10.1353/hph.0.0118 https://doi.org/10.3389/fpsyg.2017.00128 https://doi.org/10.1098/rstb.2014.0198 https://doi.org/10.1093/nc/niy007 https://doi.org/10.1121/1.408434 https://doi.org/10.3389/fnhum.2011.00002 https://doi.org/10.3390/philosophies6020050 https://doi.org/10.1111/j.1747-9991.2008.00153.x https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org in search of lost time 17 dainton, b. (2010). temporal consciousness. in e. n. zalta (ed.), the stanford encyclopedia of philosophy. metaphysics research lab, stanford university. http://plato.stanford.edu/archives/fall2010/entries/consciousness-temporal dennett, d., & kinsbourne, m. (1992). time and the observer: the where and when of consciousness in the brain. behavioral and brain sciences, 15, 183–247. https://doi.org/doi:10.1017/s0140525x00068229 doerig, a., schurger, a., hess, k., & herzog, m. h. (2019). the unfolding argument: why iit and other causal structure theories cannot explain consciousness. consciousness and cognition, 72, 49–59. https://doi.org/10.1016/j.concog.2019.04.002 dorato, m., & wittmann, m. (2020). the phenomenology and cognitive neuroscience of experienced temporality. phenomenology and the cognitive sciences, 19(4), 747–771. https://doi.org/10.1007/s11097-019-09651-4 droege, p. (2009). now or never: how consciousness represents time. consciousness and cognition, 18(1), 78–90. https://doi.org/10.1016/j.concog.2008.10.006 ellia, f., hendren, j., grasso, m., kozma, c., mindt, g., p. lang, j., m. haun, a., albantakis, l., boly, m., & tononi, g. (2021). consciousness and the fallacy of misplaced objectivity. neuroscience of consciousness, 2021(2). https://doi.org/10.1093/nc/niab032 fekete, t., van de cruys, s., ekroll, v., & leeuwen, c. van. (2018). in the interest of saving time: a critique of discrete perception. neuroscience of consciousness, 2018, 1. https://doi.org/10.1093/nc/niy003 grush, r. (2006). how to, and how not to, bridge computational cognitive neuroscience and husserlian phenomenology of time consciousness. synthese, 153(3), 417–450. https://doi.org/10.1007/s11229-006-9100-6 herzog, m. h., drissi-daoudi, l., & doerig, a. (2020). all in good time: long-lasting postdictive effects reveal discrete perception. trends in cognitive sciences, 24(10), 826–837. https://doi.org/10.1016/j.tics.2020.07.001 hoerl, c. (2013). a succession of feelings, in and of itself, is not a feeling of succession. mind, 122(486), 373–417. https://doi.org/10.1093/mind/fzt070 hogendoorn, h. (2022). perception in real-time: predicting the present, reconstructing the past. trends in cognitive sciences, 26(4), 128–141. https://doi.org/10.1016/j.tics.2021.11.003 hunt, t. (2016). taking time seriously in tononi’s integrated information theory. journal of consciousness studies, 23(9-10), 88–110. https://www.ingentaconnect.com/content/imp/jcs/2016/00000023/f0020009/art00005 husserl, e. (1964). the phenomenology of inner time consciousness. indiana university press. james, w. (1890). the principles of psychology. macmillan. johnston, a., & nishida, s. y. (2001). time perception: brain time or event time? current biology, 11(11), 427–430. https://doi.org/10.1016/s0960-9822(01)00252-4 josipovic, z., & miskovic, v. (2020). nondual awareness and minimal phenomenal experience. frontiers in psychology, 11, 2087. https://doi.org/10.3389/fpsyg.2020.02087 kelly, s. (2005). the puzzle of temporal experience. in a. brook & k. akins (eds.), cognition and the brain: the philosophy and neuroscience movement (pp. 208–238). cambridge university press. kent, l. (2019). duration perception versus perception duration: a proposed model for the consciously experienced moment. timing & time perception, 7(1), 1–14. https://doi.org/10.1163/22134468-20181135 kent, l., & wittmann, m. (2021). time consciousness: the missing link in theories of consciousness. neuroscience of consciousness, 2021, 2. https://doi.org/10.1093/nc/niab011 kiverstein, j. (2010). making sense of phenomenal unity: an intentionalist account of temporal experience. royal institute of philosophy supplements, 67, 155–181. https://doi.org/10.1017/s1358246110000081 kiverstein, j., & arstila, v. (2013). time in mind. in h. dyke & a. bardon (eds.), a companion to the philosophy of time (pp. 444–469). john wiley & sons. koenderink, j., richards, w., & doorn, a. j. van. (2012). space-time disarray and visual awareness. i-perception, 3(3), 159–165. https://doi.org/10.1068/i0490sas merker, b., williford, k., & rudrauf, d. (2021). the integrated information theory of consciousness: a case of mistaken identity. behavioral and brain sciences, 1–72. https://doi.org/10.1017/s0140525x21000881 metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1, 1–44. https://doi.org/10.33735/phimisci.2020.i.46 montemayor, c., & wittmann, m. (2014). the varieties of presence: hierarchical levels of temporal integration. timing & time perception, 2(3), 325–338. https://doi.org/10.1163/22134468-00002030 negro, n. (2020). phenomenology-first versus third-person approaches in the science of consciousness: the case of the integrated information theory and the unfolding argument. phenomenology and the cognitive sciences, 19(5), 979–996. https://doi.org/10.1007/s11097-020-09681-3 singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 http://plato.stanford.edu/archives/fall2010/entries/consciousness-temporal https://doi.org/doi:10.1017/s0140525x00068229 https://doi.org/10.1016/j.concog.2019.04.002 https://doi.org/10.1007/s11097-019-09651-4 https://doi.org/10.1016/j.concog.2008.10.006 https://doi.org/10.1093/nc/niab032 https://doi.org/10.1093/nc/niy003 https://doi.org/10.1007/s11229-006-9100-6 https://doi.org/10.1016/j.tics.2020.07.001 https://doi.org/10.1093/mind/fzt070 https://doi.org/10.1016/j.tics.2021.11.003 https://www.ingentaconnect.com/content/imp/jcs/2016/00000023/f0020009/art00005 https://doi.org/10.1016/s0960-9822(01)00252-4 https://doi.org/10.3389/fpsyg.2020.02087 https://doi.org/10.1163/22134468-20181135 https://doi.org/10.1093/nc/niab011 https://doi.org/10.1017/s1358246110000081 https://doi.org/10.1068/i0490sas https://doi.org/10.1017/s0140525x21000881 https://doi.org/10.33735/phimisci.2020.i.46 https://doi.org/10.1163/22134468-00002030 https://doi.org/10.1007/s11097-020-09681-3 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ishan singhal, ramya mudumba, and narayanan srinivasan 18 northoff, g., & huang, z. (2017). how do the brain’s time and space mediate consciousness and its different dimensions? temporo-spatial theory of consciousness (ttc). neuroscience & biobehavioral reviews, 80, 630–645. https://doi.org/10.1016/j.neubiorev.2017.07.013 northoff, g., wainio-theberge, s., & evers, k. (2020). is temporo-spatial dynamics the “common currency” of brain and mind? in quest of “spatiotemporal neuroscience.” physics of life reviews, 33, 34–54. https://doi.org/10.1016/j.plrev.2019.05.002 northoff, g., & zilio, f. (2022). from shorter to longer timescales: converging integrated information theory (iit) with the temporo-spatial theory of consciousness (ttc). entropy, 24(2), 270. https://doi.org/10.3390/e24020270 oizumi, m., albantakis, l., & tononi, g. (2014). from the phenomenology to the mechanisms of consciousness: integrated information theory 3.0. plos computational biology, 10, 5. https://doi.org/10.1371/journal.pcbi.1003588 phillips, i. (2010). perceiving temporal properties. european journal of philosophy, 18, 2. https://doi.org/10.1111/j.14680378.2008.00299.x phillips, i. (2014). experience of and in time. philosophy compass, 9(2), 131–144. https://doi.org/10.1111/phc3.12107 phillips, i. (ed.). (2017). rethinking the specious present. in the routledge handbook of philosophy of temporal experience (pp. 146–156). routledge. piper, m. s. (2019). neurodynamics of time consciousness: an extensionalist explanation of apparent motion and the specious present via reentrant oscillatory multiplexing. consciousness and cognition, 73(10275), 1. https://doi.org/10.1016/j.concog.2019.04.006 pockett, s. (2003). how long is “now”? phenomenology and the specious present. phenomenology and the cognitive sciences, 2(1), 55–68. https://doi.org/doi.org/10.1023/a:1022960122740 pokropski, m. (2018). commentary: from the phenomenology to the mechanisms of consciousness: integrated information theory 3.0. frontiers in psychology, 9, 101. https://doi.org/10.3389/fpsyg.2018.00101 pokropski, m. (2019). phenomenology and mechanisms of consciousness: considering the theoretical integration of phenomenology with a mechanistic framework. theory & psychology, 29(5), 601–619. https://doi.org/10.1177 pöppel, e. (1997). a hierarchical model of temporal perception. trends in cognitive sciences, 1, 56. https://doi.org/10.1016/s1364-6613(97)01008-5 singhal, i., & srinivasan, n. (2021). time and time again: a multi-scale hierarchical framework for time-consciousness and timing of cognition. neuroscience of consciousness, 2021, 2. https://doi.org/10.1093/nc/niab020 srinivasan, n. (2020). consciousness without content: a look at evidence and prospects. frontiers in psychology, 11, 1992. https://doi.org/10.3389/fpsyg.2020.01992 tononi, g. (2004). an information integration theory of consciousness. bmc neuroscience, 5(1), 1–22. https://doi.org/10.1186/1471-2202-5-42 tononi, g. (2008). consciousness as integrated information: a provisional manifesto. the biological bulletin, 215(3), 216– 242. https://doi.org/10.2307/25470707 tononi, g. (2015). integrated information theory. scholarpedia, 10(1), 41–64. https://doi.org/10.4249/scholarpedia.4164 tononi, g., boly, m., massimini, m., & koch, c. (2016). integrated information theory: from consciousness to its physical substrate. nature reviews neuroscience, 17(7), 450–461. https://doi.org/10.1038/nrn.2016.44 tononi, g., & koch, c. (2015). consciousness: here, there and everywhere? philosophical transactions of the royal society b: biological sciences, 370(1668), 20140167. https://doi.org/10.1098/rstb.2014.0167 tsuchiya, n., andrillon, t., & haun, a. (2020). a reply to “the unfolding argument”: beyond functionalism/behaviorism and towards a science of causal structure theories of consciousness. consciousness and cognition, 79(107), 2877–10287. https://doi.org/10.1016/j.concog.2020.102877 van gelder, t. (1999). wooden iron? husserlian phenomenology meets cognitive science. in f. j. v. j. petitot, b. pachoud, & j. m. roy (eds.), naturalizing phenomenology (pp. 245–265). stanford university press. vanrullen, r., & koch, c. (2003). is perception discrete or continuous? trends in cognitive sciences, 7(5), 207–213. https://doi.org/10.1016/s1364-6613(03)00095-0 varela, f. j. (1999). the specious present: a neurophenomenology of time consciousness. in f. j. v. j. petitot, b. pachoud, & j. m. roy (eds.), naturalizing phenomenology (pp. 266–329). stanford university press. vogel, d. h., falter-wagner, c. m., schoofs, t., krämer, k., kupke, c., & vogeley, k. (2020). flow and structure of time experience–concept, empirical validation and implications for psychopathology. phenomenology and the cognitive sciences, 19(2), 235–258. https://doi.org/10.1007/s11097-018-9573-z white, p. a. (2018). is conscious perception a series of discrete temporal frames? consciousness and cognition, 60, 98–126. https://doi.org/10.1016/j.concog.2018.02.012 white, p. a. (2021). perception of happening: how the brain deals with the no‐history problem. cognitive science, 45, 12. https://doi.org/10.1111/cogs.13068 singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neubiorev.2017.07.013 https://doi.org/10.1016/j.plrev.2019.05.002 https://doi.org/10.3390/e24020270 https://doi.org/10.1371/journal.pcbi.1003588 https://doi.org/10.1111/j.1468-0378.2008.00299.x https://doi.org/10.1111/j.1468-0378.2008.00299.x https://doi.org/10.1111/phc3.12107 https://doi.org/10.1016/j.concog.2019.04.006 https://doi.org/doi.org/10.1023/a:1022960122740 https://doi.org/10.3389/fpsyg.2018.00101 https://doi.org/10.1177 https://doi.org/10.1016/s1364-6613(97)01008-5 https://doi.org/10.1093/nc/niab020 https://doi.org/10.3389/fpsyg.2020.01992 https://doi.org/10.1186/1471-2202-5-42 https://doi.org/10.2307/25470707 https://doi.org/10.4249/scholarpedia.4164 https://doi.org/10.1038/nrn.2016.44 https://doi.org/10.1098/rstb.2014.0167 https://doi.org/10.1016/j.concog.2020.102877 https://doi.org/10.1016/s1364-6613(03)00095-0 https://doi.org/10.1007/s11097-018-9573-z https://doi.org/10.1016/j.concog.2018.02.012 https://doi.org/10.1111/cogs.13068 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org in search of lost time 19 whitrow, g. j. (1980). the natural philosophy of time. clarendon. wiese, w. (2020). the science of consciousness does not need another theory, it needs a minimal unifying model. neuroscience of consciousness, 2020. https://doi.org/10.1093/nc/niaa013 windt, j. m. (2015). just in time—dreamless sleep experience as pure subjective temporality. in t. metzinger & j. m. windt (eds.), open mind. mind group. doi.org/10.15502/9783958571174 wittmann, m. (2011). moments in time. frontiers in integrative neuroscience, 5, 66. https://doi.org/10.3389/fnint.2011.00066 woods, t., windt, j., & carter, o. (2021). the path to contentless experience in meditation: an evidence synthesis based on expert texts. mindrxiv. https://doi.org/10.31231/osf.io/6j928 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. singhal, i., mudumba, r., & srinivasan, n. (2022). in search of lost time: integrated information theory needs constraints from temporal phenomenology. philosophy and the mind sciences, 3, 13. https://doi.org/10.33735/phimisci.2022.9438 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1093/nc/niaa013 https://doi.org/10.15502/9783958571174 https://doi.org/10.3389/fnint.2011.00066 https://doi.org/10.31231/osf.io/6j928 https://doi.org/10.33735/phimisci.2022.9438 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction why is temporality important for a phenomenological theory of consciousness? continuity phenomenology transitions between contents of conscious experience implications temporal hierarchies and nestedness multiple timescales of experience nestedness within experience implications past, present, & future of time in iit status quo all axioms have "now" proposed axiom of time implications & conclusion acknowledgements radical disruptions of self-consciousness: editorial introduction radical disruptions of self-consciousness editorial introduction raphaël millièrea (raphael.milliere@philosophy.ox.ac.uk) thomas metzingerb (metzinger@uni-mainz.de) this article is part of a special issue on “radical disruptions of self-consciousness”, edited by thomas metzinger and raphaël millière. 1 consciousness and self-consciousness human beings are not only conscious creatures, but also self -conscious creatures: they have the capacity to be conscious of themselves, and indeed, to be conscious of themselves as themselves. one obvious way in which one can be conscious of oneself (as oneself) is to consciously think about oneself (as oneself), by making use of a concept of self. however, many philosophers share the intuition that selfconsciousness is more pervasive in our conscious mental lives than sophisticated cognitive states that involve conceptually representing oneself as oneself. indeed, some go so far as to suggest that a more basic form of self-consciousness or sense of self is ubiquitous in all conscious experiences. while it is easy to feel the pull of this intuition, it might be more difficult than it seems to provide compelling arguments and empirical evidence to support it. let us call the general claim that some basic form of self-consciousness or sense of self is ubiquitous is all conscious experiences the ubiquity thesis. this claim has a long history. in the west, early psychologists took it to be relatively obvious; thus, william james argued that “whatever i may be thinking of, i am always at the same time more or less aware of myself ” (james, 1892/1961, p. 42). a few auniversity of oxford bjohannes gutenberg university, mainz millière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-6965-6073 https://orcid.org/0000-0003-2514-7191 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière and thomas metzinger 2 years later, mary whiton calkins confidently stated that “all consciousness is selfconsciousness, that is, one never is conscious at all without an awareness, however vague, confused, unanalysed, and unexpressed, of oneself-being-conscious” (calkins, 1908, p. 68). in addition, versions of the ubiquity thesis can be found throughout the phenomenological tradition, for example in husserl’s work: “to be a subject is to be in the mode of being aware of oneself” (husserl, 1973, p. 151). clearly, it would be historically inaccurate to say that the ubiquity thesis is a marginal position. a number of contemporary scientists and philosophers also explicitly endorse the ubiquity thesis. here are just a few characteristic examples among many others: if “self-consciousness” is taken to mean “consciousness with a sense of self”, then all human consciousness is necessarily covered by the term – there is just no other kind of consciousness as far as i can see. (damasio, 1999, p. 19) i can recognize […] a sense in which a special form of selfconsciousness is built into the character of experience. [for example] some sort of peripheral consciousness of oneself as a situated, active perceiver […] is essentially involved in the ordinary experience of looking. (siewert, 2013, p. 256) [s]elf-consciousness is an integral and constitutive feature of phenomenal consciousness […]. (zahavi, 2014, p. 62) if the ubiquity thesis was true, it should be impossible to be in a conscious mental state without thereby being self-conscious at all. the aim of this special issue is to scrutinize this hypothesis: could one be conscious without being self-conscious in any way? is there any empirical evidence for the existence of conscious states that lack self-consciousness altogether? if one could undergo such “selfless” states, would one even be able to reliably remember and report them? these are some of the main questions that run through this collection of articles. 2 philosopher’s syndrome and the refrigerator light fallacy this special issue is about something most of us might find very hard to conceive: states of consciousness in which self-consciousness is radically disrupted or altogether missing. there clearly seems to be something to the ubiquity thesis; it has a kind of raw appeal, even though it is difficult to give a positive account of the basic form of self-consciousness or sense of self that supposedly lurks in the background of every conscious experience. why is the ubiquity thesis intuitively appealing? one possible explanation is that it is simply very difficult – if not outright impossible – to imagine what it would be like to be in a state of consciousness millière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org radical disruptions of self-consciousness: editorial introduction 3 lacking any kind of self-consciousness. indeed, the appeal of the ubiquity thesis might be a particularly interesting instance of what dennett calls philosopher’s syndrome: “mistaking a failure of imagination for an insight into necessity” (dennett, 1991, p. 401). one explanation of this failure of imagination might go as follows: every deliberate attempt at imagining a state of consciousness that genuinely lacks self-consciousness is a form of mental action; therefore, as an attempt to actively control your own state of mind, it automatically creates the phenomenology of mental agency, goal-directedness, and a sense of effort – thereby sustaining a sense of self precisely when one is trying to imagine its absence. many of us have a deep-seated philosophical intuition that consciousness without self-consciousness is simply inconceivable, and this intuition might be rooted in the functional architecture underlying human consciousness, which in turn is a product of natural evolution and sociocultural priors. but as dennett tells us one should resist the temptation to mistake a failure of imagination for an insight into necessity: from the fact that one cannot imagine being in a state of consciousness entirely lacking self-consciousness, it presumably does not follow that it is nomologically (or even metaphysically) impossible to be in such a state. there is a sense in which the difficulty of thinking about states of consciousness entirely lacking self-consciousness is deeper than a mere failure of imagination. consider the light inside your refrigerator: is it always on? certainly, you can imagine it being off whenever the refrigerator door is closed. but every time you try to peek into the refrigerator – as if to take it by surprise – the light is invariably on. as we all know, the very act of opening the refrigerator automatically turns on the light. but suppose you didn’t know how refrigerators work: how would you know, then, that the refrigerator light even turns off? this toy example illustrates the infamous “refrigerator light fallacy” (see schear, 2009, p. 101, as well as jaynes, 1976, p. 23), for an early version of the metaphor). just because one cannot actually observe the refrigerator light when it is off, because of the functional architecture of refrigerators, one should not conclude that the light is always on. to paraphrase dennett, one should not mistake a failure of observation for an insight into necessity. just like the light that turns off whenever one looks in the refrigerator, it might be the case that deliberately attending to one’s occurrent conscious mental state is sufficient to sustain a form of self-consciousness. if so, it might be impossible to assess and validate the existence of states of consciousness entirely lacking selfconsciousness through introspection. the very act of introspecting is a form of mental action that involves controlling the focus of inner attention. as such, it might always involve a phenomenal quality of “attentional agency” (metzinger, 2017), together with a subtle but consciously experienced sense of effort. this sense of effort might be enough to sustain a consciousness of oneself as the subject attending to one’s experience. consider hume’s infamous denial that one can find anything like a sense of self when introspecting one’s occurrent experience: millière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière and thomas metzinger 4 for my part, when i enter most intimately into what i call myself, i always stumble on some particular perception or other, of heat or cold, light or shade, love or hatred, pain or pleasure. i never can catch myself at any time without a perception, and never can observe any thing but the perception. (hume, 1978, p. 252) it could be argued that in this passage hume misses an important aspect of introspection. perhaps the very act of “entering” into oneself, “stumbling” upon conscious contents, and trying to “catch” or “observe” oneself sustains the very sense of self that hume denies finding in his experience. but if this is the case, then it should be difficult if not impossible to assess whether one is undergoing a selfless state of consciousness, for the very act of attending to one’s experience might be sufficient to endow it with a form of self-consciousness. perhaps we are constitutively unable to introspectively assess the existence of states of consciousness entirely lacking self-consciousness. thus, one can see why the ubiquity thesis is prima facie attractive: most of us have probably never experienced, nor even successfully imagined, what it would be like to be conscious without being self-conscious in any way. furthermore, it might impossible to introspect a genuinely selfless state of consciousness as such because of the very nature of introspective mechanisms. but this should not deter us from asking whether there are counterexamples to the ubiquity thesis. indeed, there is an increasing amount of work on empirical conditions in which self-consciousness is disrupted. analyzing such cases is crucial for assesssing the empirical plausibility of the ubiquity thesis, but it also raises their own set of concerns: how credible are reports of states of consciousness lacking selfconsciousness? can we take them at face value? and even if some form of selfconsciousness can be disrupted in certain cases, are there conditions in which all forms of self-consciousness are entirely missing? 3 the need for a new research program over the past two decades, philosophers and scientists have paid close attention to a number of empirical cases portrayed as disruptions of self-consciousness, including schizophrenic thought insertion (e.g. metzinger, 2003; billon, 2013; parrott, 2017), alienation symptoms (e.g. lane, 2014, 2015), depersonalisation disorder (e.g. metzinger, 2003; gerrans, 2014; simeon & abugel, 2006), somatoparaphrenia (e.g. metzinger, 2003; liang & lane, 2009; vignemont, 2013), autoscopic phenomena (e.g. blanke, landis, spinelli, & seeck, 2004; metzinger, 2013), and full-body illusions (e.g. blanke & metzinger, 2009; alsmith, 2010). however, most if not all of these conditions cannot adequately be described as involving conscious states in which self-consciousness is radically disrupted, let alone entirely missing. take thought insertion, for example, which refers to schizophrenic patients’ reported feeling that some of their thoughts are not remillière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org radical disruptions of self-consciousness: editorial introduction 5 ally theirs and have been inserted or implanted in their heads. how should we characterize the experience described in such reports? it has been argued that the relevant reports might be prompted, at least in part, by the patients’ lack of a sense of agency over the relevant thoughts (gallagher, 2004; o’brien & opie, 2003; stephens & graham, 1994). according to this explanation, ordinary thoughts in healthy individuals come with the sense that one is the author of these thoughts; by contrast, inserted thoughts in schizophrenic patients are not accompanied by this sense of agency. the relevant feature that is allegedly disrupted in this case – the sense of being in control of one’s mental activity and of causing one’s thoughts – is arguably a form of self-consciousness: it refers to the experience of one’s thoughts as originating in one’s own mental activity. however, a flood of recent data on mind wandering and spontaneous, task-independent thought (fox & christoff, 2018) shows that the phenomenal qualities of “mental ownership” and “mental agency” can be dissociated even in the healthy population: when lost in discursive thought or immersed in a manifest daydream we experience ownership of our thoughts, but without a sense of cognitive control over them. empirical research shows that this may actually be the case for two thirds of our conscious mental life (metzinger, 2015). accordingly, schizophrenic thought insertion might be better explained by the presence of an additional experience of “alienation” with respect to one’s thoughts, or a feeling that one’s thoughts are controled by an external agent, rather than the mere loss of the sense of agency over one’s thoughts (as in mind wandering). in any case, there are presumably other forms of self-consciousness than the sense of agency, some of which might not be missing during instances of schizophrenic thought insertion. for example, it might be the case that we often or always have a sense of bodily ownership – an experience of our bodies as our own (vignemont, 2018). one might fail to experience one’s thoughts as one own, and still experience one’s body as one’s own. conversely, one might lack the experience of one’s body as one’s own, as in somatoparaphrenia and perhaps depersonalization disorder, and yet still be self-conscious in some other way(s). more generally, there might be a number of different ways of being selfconscious that may individually go missing in specific conditions; but to cast doubt on the empirical plausibility of the ubiquity thesis, one would have to show that some states of consciousness lack any form of self-consciousness. none of the cases described above seem to fit that description. while it is plausible that some forms of self-consciousness are individually missing in these conditions, none of them appear to lack any forms of self-consciousness at the same time. at best, they offer examples of partially selfless states of consciousness. looking only at the conditions listed above, one might doubt that totally selfless states of consciousness ever actually occur. however, other conditions that have received less attention in recent philosophy of mind might provide more compelling examples of totally selfless states of consciousness. one of the goals of this special issue is to shed light on some of these cases. for example, the millière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière and thomas metzinger 6 subjective effects of certain psychoactive drugs, and particularly those of classic psychedelic drugs such as lsd, psilocybin, or 5-meo-dmt, present a special interest for the assessment of the ubiquity claim. indeed, these drugs are known to have dramatic effects on self-consciousness, and some reports even suggest that they might temporarily suppress any form of self-consciousness – a phenomenon known as ‘drug-induced ego dissolution’ in the scientific literature (letheby & gerrans, 2017; millière, 2017; nour & carhart-harris, 2017). likewise, several meditation techniques explicitly aim at reaching a state in which self-consciousness is entirely missing; there is increasing empirical evidence that advanced meditators might actually reach such states (millière, carhart-harris, roseman, trautwein, & berkovich-ohana, 2018; winter et al., 2020). beyond drug-induced and meditationinduced states, it has recently been suggested that conscious states might occur during deep sleep outside of dreams, and might lack ordinary content, including any form of self-representation (thompson, 2015; windt, 2015; windt, nielsen, & thompson, 2016). anecdotal evidence also suggests that totally selfless states of consciousness might occur during acute psychotic episodes (saks, 2007) and partial epileptic seizures (johanson, valli, revonsuo, & wedlund, 2008). finally, it is intriguing to consider whether patients in the minimally conscious state, as well as patients with cotard syndrome who believe that they do not exist and often fail to refer to themselves with the first-person pronoun, might lack any form of self-consciousness (billon, 2016; metzinger, 2003, pp. 454–461). these are just a few examples of actual conditions that might be plausible candidates of states of consciousness entirely lacking self-consciousness; more research is needed to determine whether any of them actually satisfy this description. this special issue is a first step in establishing this research program. 4 summary of contributions in his article entitled “dissolving the self: active inference, psychedelics, and ego-dissolution”, george deane (2020) agrees with many authors in the field that the capacity of classic psychedelic drugs to disrupt and even “dissolve” self-consciousness is of great philosophical and scientific interest. while there is a growing corpus of empirical research on the neural correlates and therapeutic benefit of psychedelic experiences, we still stand in need of a conceptually finegrained theoretical account of their underlying mechanisms. in deane’s paper, psychedelic-induced ego-dissolution is accounted for within the framework of active inference: his central thesis is that the experience of drug-induced ego dissolution is mediated by the collapsing of the “temporal thickness” of the agent’s deep temporal model, which is equivalent to lowered precision on high-level priors. deane’s argument here is composed of three steps: first, a view of the self-model is proposed as arising within a temporally deep generative model of an embodied organism navigating an affordance landscape in the service of allostasis. second, a view of the action of psychedelics as lowering the precision millière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org radical disruptions of self-consciousness: editorial introduction 7 of high-level priors within the generative model is unpacked in terms of a high bayesian learning rate. finally, deane argues that the relaxation of high-level priors causes a “collapse” in the temporal thickness of the generative model, resulting in a collapse in the self-model and a loss of the ordinary sense of being a self. this account has major implications for our understanding of ordinary self-consciousness and disruptions in self-consciousness present in psychosis, autism, depression, and dissociative disorders. accordingly, george deane also discusses the wider context of this novel conceptual approach, which has a rich spectrum of philosophical, theoretical and therapeutic implications. retrospective reports of conscious episodes entirely lacking self-consciousness might seem to have an air of paradox: how can subjects confidently report that they underwent such an episode, if they lacked any consciousness of themselves at the time? in his contribution, “look who’s talking! varieties of ego-dissolution without paradox”, sascha fink (2020) considers different ways to tackle this question and to interpret the relevant reports. fink begins by pointing out how the main source of evidence for the existence of selfless states of consciousness are first-person reports. one might question whether these reports can be taken at face value. for example, metzinger (2003, p. 566), gennaro (2008), and foster (2016, p. 6) have hinted at the self-defeating nature of such statements if we take them to be genuine reports: indeed, people reporting the relevant experiences (a) explicitly deny having been self-conscious during the relevant experience, but (b) suggest that they witnessed selfless experience in so far as they can report it (but see millière (2020), this issue, for a defense of the claim that there is no inconsistency in retrospective first-person reports of experiences lacking self-consciousness). as fink puts it, the content of such reports seems to conflict with the pragmatics of reporting. he argues that self-ascriptions of selfless states of consciousness can be explained in a number of ways, some of which do not entail that these selfascriptions are self-defeating. only some explanations for such utterances rely on an actual change in the content of phenomenal experience, and of those that do rely on a change in consciousness, only one (total ego-dissolution) is incoherent. but its alternatives do not lead to contradictions. fink ends by arguing that when the relevant reports genuinely describe a past experience, it is plausible that this was an experience in which the sense of self expands rather than disappears. in “cotard syndrome, self-awareness, and i-concepts”, rocco gennaro (2020) defends the higher-order thought (hot) theory of consciousness against philosophical criticism relying on empirical data from various psychopathologies of self-awareness, such as somatoparaphrenia and thought insertion in schizophrenia. according to the hot theory of consciousness, what makes a mental state m a conscious mental state is the occurrence of a higher-order thought with the content <i am in mental state m>. having already argued in previous work that a hot theorist can adequately respond to the conceptual challenges posed by somatoparaphrenia and thought insertion, gennaro now turns to cotard syndrome which is a rare neuropsychiatric disorder in which people hold the delusional bemillière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière and thomas metzinger 8 lief that they are dead, do not exist, or have lost their blood or internal organs. in his contribution he analyzes cotard syndrome in light of his previous discussion of somatoparaphrenia and thought insertion, arguing that hot theory can provide a somewhat analogous account. his central point is that there are multiple concepts of self and corresponding levels of hots that provide a more nuanced explanation of psychopathologies such as the cotard syndrome. thus, on his view, cotard patients are still capable of having some kinds of “i-thoughts” about their bodies and mental states. in his article “being for no-one: psychedelic experience and minimal subjectivity”, chris letheby (2020) asks whether reports of drug-induced ego dissolution provide us with solid evidence against so-called “subjectivity theories of consciousness”, according to which phenomenal consciousness constitutively involves a minimal form of self-awareness or “subjectivity”. billon & kriegel (2015) have previously denied that thought insertion and depersonalization disorder constitute counterexamples to such theories. in particular, they suggested that putatively selfless states of consciousness associated with depersonalization might not in fact be phenomenally conscious at all, in which case they would not threaten subjectivity theories. letheby argues that this line of defense does not work equally well for reports of selfless states of consciousness induced by classic psychedelic drugs, particularly potent and fast-acting psychedelics such as dmt and 5-meo-dmt, because there is little doubt that the relevant states are phenomenally conscious. letheby also considers an objection according to which drug-induced states may lack “me-ness” (an awareness of oneself as oneself ) and “mineness” (an awareness of one’s conscious mental state as one’s own) but cannot lack “for-me-ness” (an awareness of one’s mental state that presupposes a minimal form of experiential selfhood). letheby addresses this objection with a dilemma: if “for-me-ness” is a component of phenomenology, then reports of drug-induced ego dissolution should suggest that this component is present in the relevant experiences. however, these reports suggest that this is not the case. if, by contrast, “for-me-ness” is not a component of phenomenology, then the claim that states of consciousness can lack any form of self-consciousness is vindicated. jakub limanowski and karl friston (2020) analyze reports of selfless states of consciousness within the framework of active inference. in “attenuating oneself: an active inference perspective on ‘selfless’ experiences”, they address reports of “selfless” experiences from the perspective of active inference and predictive processing. they argue that phenomenal self-modelling is functionally grounded in active inference as action planning and precision control within deep generative models. this establishes a link between computational mechanisms and phenomenal selfhood. limanowski and friston propose that putatively “selfless” states of consciousness constitute rare cases in which normally congruent processes of computational and phenomenal self-modeling diverge in an otherwise conscious system. they think that two potential candidate mechanisms might lead to such a divergence by attenuating the conscious experience of selfhood: “self-flattening” millière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org radical disruptions of self-consciousness: editorial introduction 9 via a reduction in the depth of active inference, and “self-attenuation” via a reduction of the expected precision of self-evidence. interestingly, limanowski and friston – like only one other author in this collection, namely miguel ángel sebastián – conclude in favour of the ubiquity thesis. thomas metzinger (2020) proposes a new methodological alternative to classical reductionist research strategies. these strategies target conscious experience by isolating minimally sufficient neural correlates and developing a functional analysis, ultimately leading to computational descriptions and mechanistic explanations. by contrast, he aims at a “minimal model explanation” for conscious experience, taking the phenomenal character of “pure consciousness” or “pure awareness” in meditation as its entry point. in “minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of ‘pure’ consciousness”, he tries to take the phenomenology of “awareness of awareness” or “consciousness as such” as seriously as possible – although for many centuries exactly this phenomenology has been claimed to be either non-existent or strictly ineffable. metzinger develops the theoretical concept of “minimal phenomenal experience” (mpe) as a candidate for the simplest form of consciousness, substantiating it by extracting six semantic constraints from the existing literature and using sixteen phenomenological case-studies to incrementally flesh out the new working concept. one empirical hypothesis is that the phenomenological prototype of “pure awareness”, to which all such reports refer, really is the content of a specific predictive model, namely, a bayesian representation of tonic alertness. on a more abstract conceptual level, the phenomenal character of the relevant – and selfless – states can be described as referring to a model of an unpartitioned epistemic space. in “the varieties of selflessness”, raphaël millière (2020) argues against the ubiquity thesis through a divide-and-conquer strategy. after distinguishing six notions of self-consciousness commonly discussed in the literature, he argues on the basis of empirical evidence that each of them can individually fail to be instantiated at least in some states of consciousness. at a minimum, such states of consciousness can be said to be “partially selfless” insofar as they lack at least one of the ways in which one can be self-conscious. millière subsequently argues that there is preliminary evidence that some states of consciousness lack all six forms of self-consciousness. these include states induced by certain psychoactive drugs and meditation practices, as well as – more speculatively – conscious states that might occur during dreamless sleep. such states of consciousness might be said to be “totally selfless”, insofar as they might lack all of the ways in which one can be self-conscious. millière also considers four objections to the possibility and reportability of totally selfless states of consciousness. he concludes that there is no inconsistency in the idea that such states might exist and be retrospectively reported, and that it is in fact empirically plausible that such states actually occur – which undermines the ubiquity thesis.  taking the opposite view, miguel ángel sebastián (2020) defends a version of the ubiquity thesis against empirical objections in “perspectival millière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière and thomas metzinger 10 self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness”. he begins by pointing out how it is often claimed that a minimal form of self-awareness is constitutive of our conscious experience, and how some critics have replied that such a claim is only plausible for ordinary experience but seems to be contradicted by empirical evidence on altered states of consciousness. sebastián argues that given a proper understanding of a minimal form of self-awareness that he labels “perspectival first-person awareness” (or pfp-awareness), it becomes apparent that even putatively selfless states of consciousness do not entirely lack self-awareness. pfp-awareness is anchored in a non-conceptual, identification-free self-attribution that defines the ultimate origin of the first-person perspective of conscious experience. after offering a semantic account of pfp-awareness, sebastián reviews the empirical literature on altered states of consciousness that have been claimed to lack any form of selfconsciousness (including some discussed in this special issue). he concentrates on the phenomenology of states induced by psychedelic drugs, meditation and dreams, as they have been claimed to present the biggest threat to the ubiquity thesis. first, he argues that although there are good reasons to think that some forms of self-awareness that typically accompany our ordinary experience can be compromised in altered states of consciousness, this does not mean pfp-awareness is absent in these states. secondly, he argues that most reports that seem to describe states of consciousness entirely lacking self-consciousness (including pfp-awareness) are probably due to confirmation bias grounded in expectations and metaphysical beliefs – and hence that we should not take these reports at face value. as pointed out at the outset of this introduction, it seems difficult to imagine undergoing a state of consciousness lacking any form of self-consciousness. there is a sense in which such states are genuinely inconceivable, although we should resist philosopher’s syndrome and conclude from this alone that they cannot exist. but even if one can provide a convincing conceptual account of such states, they might simply never occur in neurotypical humans, as a matter of brute empirical fact. in “breaking the self: radical disruptions of self-consciousness and impossible conscious experiences”, wanja wiese (2020) asks whether there are logically possible types of conscious experience that are nomologically impossible, given independently justified assumptions about the neural underpinnings of consciousness in human beings. wiese points out that in a sense, the answer is trivial: just consider the fact that the types of perceptual experiences we can have are limited by our sensory organs. but there may also be non-trivial types of conscious experience that are impossible. for instance, if a basic form of self-consciousness is nomologically necessary, then experiences lacking this phenomenal property are (nomologically) impossible. more generally, there might be deep causal dependencies between the neural mechanisms that are required to instantiate distinct phenomenal properties (in human beings). accordingly, instantiating one of these phenomenal properties without certain others may be nomologically impossible. millière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org radical disruptions of self-consciousness: editorial introduction 11 in either case, there would be non-trivial cases of nomologically impossible types of conscious experience. wiese’s article clarifies this hypothesis and outlines a general methodology to assess it. finally, he relates this discussion to the question of whether there can be states of consciousness lacking self-consciousness. references alsmith, a. (2010). minimal conditions for the simplest form of self-consciousness. in t. fuchs, p. henningsen, & h. sattel (eds.), the embodied self (pp. 35–41). stuttgart: schattauer. billon, a. (2013). does consciousness entail subjectivity? the puzzle of thought insertion. philosophical psychology, 26(2), 291–314. https://doi.org/10.1080/09515089.2011.625117 billon, a. (2016). making sense of the cotard syndrome: insights from the study of depersonalisation. mind & language, 31(3), 356–391. https://doi.org/10.1111/mila.12110 billon, a., & kriegel, u. (2015). jaspers’ dilemma: the psychopathological challenge to subjectivity theories of consciousness. in r. j. gennaro (ed.), disturbed consciousness: new essays on psychopathology and theories of consciousness (pp. 29–54). cambridge, ma: mit press. blanke, o., landis, t., spinelli, l., & seeck, m. (2004). out-of-body experience and autoscopy of neurological origin. brain, 127(2), 243–258. https://doi.org/10.1093/brain/awh040 blanke, o., & metzinger, t. (2009). full-body illusions and minimal phenomenal selfhood. trends in cognitive sciences, 13(1), 7–13. https://doi.org/10.1016/j.tics.2008.10.003 calkins, m. w. (1908). psychology as science of self: ii. the nature of the self. the journal of philosophy, psychology and scientific methods, 5(3), 64–68. https://doi.org/10.2307/2012219 damasio, a. (1999). the feeling of what happens: body and emotion in the making of consciousness. san diego, ca: houghton mifflin harcourt. deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 dennett, d. c. (1991). consciousness explained. london: penguin books. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 foster, c. (2016). being a beast. london: profile books. fox, k. c. r., & christoff, k. (eds.). (2018). the oxford handbook of spontaneous thought: mind-wandering, creativity, and dreaming. new york, ny: oxford university press. gallagher, s. (2004). neurocognitive models of schizophrenia: a neurophenomenological critique. psychopathology, 37(1), 8–19. https://doi.org/10.1159/000077014 gennaro, r. j. (2008). are there pure conscious events? in c. chakrabarti & g. haist (eds.), revisiting mysticism (pp. 100–120). cambridge: cambridge scholars press. gennaro, r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https: //doi.org/10.33735/phimisci.2020.i.41 gerrans, p. (2014). all the self we need. in t. metzinger & j. windt (eds.), open mind. https://doi.org/10.15502/978395857 0078 hume, d. (1978). a treatise of human nature (2nd edition; l. a. selby-bigge & p. h. nidditch, eds.). oxford: oxford university press. husserl, e. (1973). zur phänomenologie der intersubjektivität: texte aus dem nachlass. zweiter teil: 1921-1928 (i. kern, ed.). dordrecht: springer. james, w. (1961). psychology: the briefer course (g. allport, ed.). new york: harper & row. (original work published 1892.) jaynes, j. (1976). the origin of consciousness in the breakdown of the bicameral mind. boston, ma: houghton mifflin. johanson, m., valli, k., revonsuo, a., & wedlund, j.-e. (2008). content analysis of subjective experiences in partial epileptic seizures. epilepsy & behavior, 12(1), 170–182. https://doi.org/10.1016/j.yebeh.2007.10.002 lane, t. (2014). when actions feel alien: an explanatory model. in t.-w. hung (ed.), communicative action (pp. 53–74). singapore: springer science+business. lane, t. (2015). self, belonging, and conscious experience: a critique of subjectivity theories of consciousness. in r. j. gennaro (ed.), disturbed consciousness: new essays on psychopathology and theories of consciousness. mit press. millière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1080/09515089.2011.625117 https://doi.org/10.1111/mila.12110 https://doi.org/10.1093/brain/awh040 https://doi.org/10.1016/j.tics.2008.10.003 https://doi.org/10.2307/2012219 https://doi.org/10.33735/phimisci.2020.i.39 https://doi.org/10.33735/phimisci.2020.i.40 https://doi.org/10.1159/000077014 https://doi.org/10.33735/phimisci.2020.i.41 https://doi.org/10.33735/phimisci.2020.i.41 https://doi.org/10.15502/9783958570078 https://doi.org/10.15502/9783958570078 https://doi.org/10.1016/j.yebeh.2007.10.002 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière and thomas metzinger 12 letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 letheby, c., & gerrans, p. (2017). self unbound: ego dissolution in psychedelic experience. neuroscience of consciousness, 2017(1). https://doi.org/10.1093/nc/nix016 liang, c., & lane, t. (2009). higher-order thought and pathological self: the case of somatoparaphrenia. analysis, 69(4), 661–668. https://doi.org/10.1093/analys/anp094 limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 metzinger, t. (2003). being no one: the self-model theory of subjectivity. cambridge, ma: mit press. metzinger, t. (2013). why are dreams interesting for philosophers? the example of minimal phenomenal selfhood, plus an agenda for future research. frontiers in psychology, 4, 746. https://doi.org/10.3389/fpsyg.2013.00746 metzinger, t. (2015). m-autonomy. journal of consciousness studies, 22(11-12), 270–302. metzinger, t. (2017). the problem of mental action: predictive control without sensory sheets. in t. metzinger & w. wiese (eds.), philosophy and predicitive processing. https://doi.org/10.15502/9783958573208 metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of ”pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 millière, r. (2017). looking for the self: phenomenology, neurophysiology and philosophical significance of drug-induced ego dissolution. frontiers in human neuroscience, 11(245), 1–22. https://doi.org/10.3389/fnhum.2017.00245 millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci. 2020.i.48 millière, r., carhart-harris, r. l., roseman, l., trautwein, f.-m., & berkovich-ohana, a. (2018). psychedelics, meditation and self-consciousness. frontiers in psychology, 9(1475), 1–29. https://doi.org/10.3389/fpsyg.2018.01475 nour, m. m., & carhart-harris, r. l. (2017). psychedelics and the science of self-experience. the british journal of psychiatry, 210(3), 177–179. https://doi.org/10.1192/bjp.bp.116.194738 o’brien, g., & opie, j. (2003). the multiplicity of consciousness and the emergence of the self. in a. s. david & t. t. j. kircher (eds.), the self and schizophrenia: a neuropsychological perspective (pp. 107–120). cambridge: cambridge university press. parrott, m. (2017). subjective misidentification and thought insertion. mind & language, 32(1), 39–64. https://doi.org/10.1 111/mila.12132 saks, e. r. (2007). the center cannot hold: my journey through madness. new york: hachette books. schear, j. k. (2009). experience and self-consciousness. philosophical studies, 144(1), 95–105. https://doi.org/10.1007/s11098009-9381-y sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 siewert, c. (2013). phenomenality and self-consciousness. in u. kriegel (ed.), phenomenal intentionality (pp. 235–257). oxford: oxford university press. simeon, d., & abugel, j. (2006). feeling unreal: depersonalization disorder and the loss of the self. oxford: oxford university press. stephens, g. l., & graham, g. (1994). self-consciousness, mental agency, and the clinical psychopathology of thoughtinsertion. philosophy, psychiatry, and psychology, 1(1), 1–10. thompson, e. (2015). dreamless sleep, the embodied mind, and consciousness. in t. k. metzinger & j. m. windt (eds.), open mind. https://doi.org/10.15502/9783958570351 vignemont, f. de. (2013). the mark of bodily ownership. analysis, 73(4), 643–651. https://doi.org/10.1093/analys/ant080 vignemont, f. de. (2018). mind the body: an exploration of bodily self-awareness. oxford: oxford university press. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 windt, j. m. (2015). just in time – dreamless sleep experience as pure subjective temporality. in t. k. metzinger & j. m. windt (eds.), open mind. https://doi.org/10.15502/9783958571174 windt, j. m., nielsen, t., & thompson, e. (2016). does consciousness disappear in dreamless sleep? trends in cognitive sciences, 20(12), 871–882. https://doi.org/10.1016/j.tics.2016.09.006 winter, u., levan, p., borghardt, t. l., akin, b., wittmann, m., leyens, y., & schmidt, s. (2020). content-free awareness: eeg-fcmri correlates of consciousness as such in an expert meditator. frontiers in psychology, 10, 3064. https://doi.or g/10.3389/fpsyg.2019.03064 zahavi, d. (2014). self and other: exploring subjectivity, empathy, and shame. oxford: oxford university press. millière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://doi.org/10.1093/nc/nix016 https://doi.org/10.1093/analys/anp094 https://doi.org/10.33735/phimisci.2020.i.35 https://doi.org/10.3389/fpsyg.2013.00746 https://doi.org/10.15502/9783958573208 https://doi.org/10.33735/phimisci.2020.i.46 https://doi.org/10.3389/fnhum.2017.00245 https://doi.org/10.33735/phimisci.2020.i.48 https://doi.org/10.33735/phimisci.2020.i.48 https://doi.org/10.3389/fpsyg.2018.01475 https://doi.org/10.1192/bjp.bp.116.194738 https://doi.org/10.1111/mila.12132 https://doi.org/10.1111/mila.12132 https://doi.org/10.1007/s11098-009-9381-y https://doi.org/10.1007/s11098-009-9381-y https://doi.org/10.33735/phimisci.2020.i.44 https://doi.org/10.15502/9783958570351 https://doi.org/10.1093/analys/ant080 https://doi.org/10.33735/phimisci.2020.i.32 https://doi.org/10.15502/9783958571174 https://doi.org/10.1016/j.tics.2016.09.006 https://doi.org/10.3389/fpsyg.2019.03064 https://doi.org/10.3389/fpsyg.2019.03064 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org radical disruptions of self-consciousness: editorial introduction 13 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. millière, r., & metzinger, t. (2020). radical disruptions of self-consciousness: editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.50 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org consciousness and self-consciousness philosopher's syndrome and the refrigerator light fallacy the need for a new research program summary of contributions how level is the `cognitive playing field'? context shapes alterations in self-conception during the psychedelic experience. how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience joshua m. martina,b (joshua.martin2902@gmail.com) philipp sterzerb (philipp.sterzer@charite.de) abstract in philosophy of psychedelics, letheby provides a convincing basis for the idea that psychedelics primarily derive their therapeutic potential through mediating favourable changes to self-related belief systems. in this commentary, we take a closer look at the role that contextual factors (‘set’ and ‘setting’) play in letheby’s two-factor account of psychedelic therapy. while letheby acknowledges that psychedelic effects are highly context-dependent, the exact role that context plays in self-modelling during the acute experience is not entirely clear. we argue that context plays an essential role in shaping the ‘discovery’ of alternate self-conceptions during the acute experience. specifically, users are more likely to experience alterations in self-conception that are consistent with context-dependent features, such as one’s prior mindset (‘set’) and aspects of the external environment (‘setting’). this is consistent with a rebus model of psychedelic effects, where a corollary effect of relaxing high-level prior beliefs is that the system becomes increasingly sensitised to bottom-up information (‘prediction errors’) from lower-level intrinsic sources and the sensory periphery. furthermore, it may explain why a positive well-intentioned state of mind and a supportive therapeutic environment are more likely to result in positive acute experiences and long-term improvements in mental wellbeing. we see this position as being largely compatible with letheby’s proposal, but may shift some aspects of its emphasis and framing. in particular, it may conflict with letheby’s notion that psychedelics lead to a ‘level cognitive playing field’, where self-related hypotheses are thought to be assigned more or less equal probability. keywords context ∙ psychedelics ∙ rebus ∙ self ∙ set and setting this article is part of a symposium on chris letheby’s book “philosophy of psychedelics” (oup 2021), edited by chiara caporuscio and sascha benjamin fink. aberlin school of mind and brain, faculty of philosophy, humboldt-universität zu berlin, berlin, germany. bdepartment of psychiatry and psychotherapy, charité–universitätsmedizin berlin, campus charité mitte, berlin, germany martin, j. m., & sterzer, p. (2022). how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3, 11. https://doi.org/10.33735/phimisci.2022.9326 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-6761-4045 https://orcid.org/0000-0003-4687-2317 https://doi.org/10.33735/phimisci.2022.9326 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org joschua m. martin & philipp sterzer 2 psychedelics are drawing world-wide attention from researchers and clinicians due to a growing appreciation for their therapeutic potential in treating a variety of mental disorders. unlike traditional pharmacological treatments, which can take weeks or months to take effect, psychedelics have the advantage of bringing about rapid positive changes in mental wellbeing, in some cases even after a single therapeutic session (ross et al., 2016). a key area of interest in current psychedelic research concerns their apparent ability to modulate belief-systems (mcgovern et al., 2021; timmermann et al., 2021), which are theorised to play an important role in conferring therapeutic benefits (carhart-harris & friston, 2019). under sufficiently high-doses, psychedelics tend to be accompanied by transformative ‘mystical-type’ experiences, which may include content of a transcendent nature, such as experiencing a cosmic consciousness, or encountering interdimensional beings or godly entities. the induction of mystical-type experiences during a psychedelic session has been found to be an important predictor of improvements in mental health (james et al., 2020; roseman et al., 2018). based on these findings, some researchers have cautioned that while psychedelics may be beneficial for psychological wellbeing, they may only do so through mediating the occurrence of metaphysical beliefs in ‘non-natural’ entities (lavazza, 2017). assuming a naturalistic worldview, one may then object to psychedelic therapy since it’s benefits derive from inducing a ‘comforting delusion’, which may make it epistemically and ethically questionable. in his new book, the philosophy of psychedelics, chris letheby (2021) provides a convincing case for why this ‘comforting delusion” objection is misguided. according to letheby, psychedelics primarily exert their therapeutic benefits, not through the occurrence of non-naturalistic metaphysical beliefs, but rather through favourable changes to self-related belief systems. in terms of a therapeutic mechanism, psychedelics are proposed to loosen over-weighted pathological self-related beliefs underlying symptoms of mental illness, in turn facilitating their beneficial revision. letheby draws on a wide range of theoretical and empirical literature to justify the therapeutic importance of changes to self-representation in psychedelic therapy. he interprets the findings to support a two-factor model of psychedelic therapeutic effect. the first factor is “the induction of plasticity at multiple levels: neural, cognitive, and phenomenological” (p. 147). this “may be sufficient for the shortterm ‘afterglow’ of several weeks, in which subjects report feeling lighter, freer, more energised and unburdened” (p. 147). however, for lasting changes to occur, letheby claims that a second factor is necessary: “the discovery of new forms of self-modelling during the experience, and the consolidation of these during the subsequent period of integration” (pp. 147-148). together, these two factors emphasise that “the cognitive plasticity of the self-model and its phenomenological manifestations are of primary therapeutic importance” (p. 143). we find letheby’s book to be a valuable contribution to the current literature on psychedelic therapy, providing several lines of enquiry to be evaluated by fumartin, j. m., & sterzer, p. (2022). how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3, 11. https://doi.org/10.33735/phimisci.2022.9326 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9326 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how level is the ‘cognitive playing field’? 3 ture theoretical and empirical research. in this commentary, we take a closer look at the role that contextual influences may play in the second factor of letheby’s account. letheby acknowledges that psychedelic effects are “strongly influenced by ‘set and setting: the psychological state of the person taking the drug and the environment in which they take it’ ”. however, the role that contextual factors play in self-modelling during the acute psychedelic experience are not entirely clear, and seem odds with some of his claims, such as the idea that psychedelics lead to a ‘level cognitive playing field’ in terms of one’s felt identity. we aim to show that context plays an essential role in shaping the ‘discovery’ of healthier forms of selfmodelling during the psychedelic experience. to do this, we first outline the role of context (‘set’ and ‘setting’) in psychedelic literature. we then briefly discuss contextual influences in relation to the rebus model (carhart-harris & friston, 2019) – a predictive processing perspective which letheby draws heavily upon in his account. finally, we explore the implications of enhanced context-sensitivity for self-modelling and argue that there are principled differences in the kinds of alterations in self-conception that are likely to be experienced, given the brain’s heightened sensitivity to context under psychedelics. 1 ‘set’ and ‘setting’ in psychedelic research contextual factors are widely acknowledged to exert a powerful influence on psychedelic effects. while the context-dependent nature of psychedelic effects had been appreciated long before (hartogsohn, 2016), the importance of context was popularised through the writings of timothy leary in the 1960s, who introduced the terms ‘set’ and ‘setting’ to describe the non-pharmacological factors that influence the quality of a psychedelic experience (leary et al., 1964). while definitions may slightly differ, ‘set’ generally refers to “the psychological factors of personality, preparation, expectation, and intention; and ‘setting’ refers to”the environmental factors of the physical, social and cultural surroundings in which the experience occurred” (hartogsohn, 2016, p. 1261). together, they are proposed to exert a central role in determining the quality of the psychedelic experience (carhart-harris et al., 2018; hartogsohn, 2016; noorani & alderson-day, 2020), and have been used to explain why psychedelics can promote such diverse and seemingly contradictory subjective reports (hartogsohn, 2016). they carry a long history within psychedelic research (hartogsohn, 2016), and remain an important consideration in modern clinical research designs (gukasyan & nayak, 2021). the important influence of ‘set’ and ‘setting’ is apparent in various descriptions of psychedelics in the literature. for example, psychedelics have been described as “meaning-enhancers” (hartogsohn, 2018), “non-specific catalysts and amplifiers of the psyche” (grof, 1994, p. 11), and “a magnifying glass to consciousness, augmenting whatever already existed in one’s mind” (leary et al., 1963; as cited by hartogsohn, 2016, p. 1261). martin, j. m., & sterzer, p. (2022). how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3, 11. https://doi.org/10.33735/phimisci.2022.9326 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9326 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org joschua m. martin & philipp sterzer 4 several empirical findings support the idea that context influences the quality of the psychedelic experience and subsequent therapeutic outcomes. for example, having clear intentions and expectations is more likely to lead to a ‘peak’ or mystical-type experience (carhart-harris et al., 2018; haijen et al., 2018) which is widely known to relate to improvements in psychological wellbeing (carhartharris et al., 2018; griffiths et al., 2016; roseman et al., 2018; ross, 2018). similarly, having positive expectations is associated with “pleasant and educational experiences” (metzner et al., 1965, p. 19), and has been found to contribute to improvements in mental health in a micro-dosing paradigm (kaertner et al., 2021). furthermore, having a positive set or clear intentions protects against having a challenging experience (carhart-harris et al., 2018). similar findings have been found for the influence of ‘setting’, where taking psychedelics is more likely to lead to a ‘peak experience’ (carhart-harris et al., 2018), and “feeling comfortable in the environment of the psychedelic experience, and with the people who were present during it, had a positive effect on well-being after the experience” (haijen et al., 2018, p. 9). on the other hand, having a negative set (metzner et al., 1965), or taking psychedelics unknowingly or without preparation (masters & houston, 2000; as cited by letheby, 2021), or taking it in an uncomfortable setting (studerus et al., 2012) is more likely to result in adverse reactions. while challenging experiences may sometimes be viewed as therapeutically beneficial (carbonaro et al., 2016; gashi et al., 2021), they may also lead to less favourable outcomes for psychological wellbeing (carhart-harris et al., 2018; haijen et al., 2018). 2 predictive processing, the rebus model, and context in his book, letheby draws heavily upon the predictive processing framework (clark, 2015; hohwy, 2013) and more specifically the rebus model (‘relaxed beliefs under psychedelics,’ carhart-harris & friston, 2019) of psychedelic effects. according to predictive processing, the brain can be conceptualised as a multilevel hierarchical generative model, which formulates predictions at multiple levels of abstraction and spatiotemporal scale. levels at the bottom of this hierarchy capture relatively concrete, domain-specific features that operate over shorter spatiotemporal-scales, while levels at the top capture relatively abstract, multimodal and integrative features that operate over longer spatiotemporal-scales. predictions are thought to be formulated at each hierarchical level in a manner akin to bayesian inference, where incoming sensory evidence and prior beliefs are combined. the overall goal of the brain is to engage in prediction error minimisation — to minimise the difference (‘prediction errors’) between brainbased predictions and incoming sensory input from the world (and the body). to do this, the brain must take into account the estimated reliability or confidence (‘precision’) through ascribing a weighting to the prior and sensory input at each martin, j. m., & sterzer, p. (2022). how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3, 11. https://doi.org/10.33735/phimisci.2022.9326 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9326 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how level is the ‘cognitive playing field’? 5 hierarchical level. this allows the brain to flexibly adapt to context-varying levels of noise derived from changes in the internal and external environment. imbalances in precision-weighting have been theoretically implicated in psychiatric disorders (sterzer et al., 2018), and are thought to occur under the influence of consciousness-altering drugs (carhart-harris & friston, 2019; corlett et al., 2016). the rebus model represents one such model, which aims to characterise psychedelic effects through the lens of altered precision-weighting. on this account, psychedelics exert their effects through weakening or ‘relaxing’ the confidence (i.e. decreasing the precision) in the brain’s highest-level prior beliefs. pharmacologically, this is thought to be brought about by the upregulation of 5-ht2a receptors, which are most densely found in high-level association regions with broad executive influence, such as the default mode network (beliveau et al., 2017). these areas are thought to sit at the very top levels of the generative model, and thus play a strong constraining role on predictions that are made at lower levels. according to letheby (2021, p. 119), these levels “encode our most abstract, fundamental beliefs about self and world—our bedrock, largely unconscious, domain-general assumptions concerning space, time, and causality, the laws of logic, and the existence of the self”. by relaxing the influence of these high-level beliefs, psychedelics may loosen rigid pathological priors under mental illness, which provide an opportunity for their revision. letheby draws upon this central idea to explain a number of psychedelics’ well-known phenomenological effects, such as ego-dissolution, feelings of connectedness and freedom, and the increased opacity of the self-model. however, one upshot of the rebus model which letheby’s account may overlook is its implications for enhanced context-sensitivity. as carhart-harris and friston (2019) describe in their paper, “a corollary of relaxing high-level priors or beliefs under psychedelics is that ascending prediction errors from lower levels of the system (that are ordinarily unable to update beliefs due to the top-down suppressive influence of heavily-weighted priors) can find freer register in conscious experience, by reaching and impressing on higher levels of the hierarchy” (p. 319). this “implies that the system’s (context) sensitivity will be enhanced— as it will be more receptive to bottom-up information, particularly from intrinsic sources” (p. 338). this is consistent with the idea that the effects of 5ht2a receptor agonists induce cortical plasticity and “an associated sensitivity to internal (i.e. endogenous processes and pre-existing mental context) and external influence (i.e. the environment)” (carhart-harris et al., 2018, p. 747; see carhart-harris & nutt, 2017). this enhanced sensitivity under psychedelics may be a plausible mechanism for the amplified influence of ‘set’ and ‘setting’ during the psychedelic experience. specifically, the influence of the physical ‘setting’ may be explained by an increased sensitivity to prediction errors deriving from the immediate surroundings, while the influence of ‘set’ or mental context may reflect an increased sensitisation of martin, j. m., & sterzer, p. (2022). how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3, 11. https://doi.org/10.33735/phimisci.2022.9326 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9326 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org joschua m. martin & philipp sterzer 6 prediction errors from lower-level intrinsic sources1. wider aspects of the social and cultural ‘setting’ may involve some combination of these processes. some authors have argued that the increased influence of ‘set’ and ‘setting’ is consistent with a ‘strong prior’ theory (mcgovern et al., 2021), where precise priors predominate over sensory input (corlett et al., 2019). this may at first seem at odds with the rebus model, which proposes that psychedelics primarily act to weaken high-level prior beliefs. however, this position may be compatible with the rebus model, so long as these strong prior influences emerge at lower levels in response to the relaxed constraining influence of priors at the highest levels of the system (cf. safron, 2020). indeed, imbalances in precision-weighting need not be related across the inferential hierarchy in a linear way (sterzer et al., 2018), and imbalances at certain levels can have compensatory changes in other levels or parts of the system (corlett et al., 2019). from this perspective, the rebus model may involve both ‘strong’ and ‘weak’ prior influences, which are present at different hierarchical levels to different degrees. the relative balance and strength of this relationship may also depend on dose-dependent changes in 5-ht2a receptor activity (safron, 2020). despite these underlying complexities, psychedelics can be generally characterised as weakening the constraining influence of high-level priors encoding rigid highly-abstract beliefs of the world and self, and (relatively) strengthening lower-level prior beliefs associated with context-dependent information sources. these lower-levels act as an internal information source for high-level predictions, and thus their increased prior weighting gives rise to more precise prediction errors at higher hierarchical levels. this may provide a plausible mechanism for the amplified influence of context on high-level prediction formation. 1plausibly, one may question whether the prior beliefs underlying ‘set’ factors, such as intentions and expectations, can be considered to be implemented as a ‘lower-level’ intrinsic source, due to their abstract and self-related nature. however, they may still arguably differ relative to the highest-level ‘bedrock’ beliefs that are theorised to be the primary pharmacological target of psychedelics, which are largely unconscious, domain-general and rigid. by comparison, ‘set’ factors are more flexible, consciously accessible, and tend to operate over shorter timescales. if these properties are to map onto differences in hierarchical level, then it seems plausible to assume that they may be encoded at a lower level relative to the highest-level beliefs that are relaxed under psychedelics. furthermore, ‘set’ factors are likely to contain predictive information that is encoded at multiple hierarchical levels, which may be differentially impacted by the pharmacological action of psychedelics. for example, a user with a positive ‘set’ may carry both high-level abstract self-relevant information, as well as lower-level information, for example reflecting physiological arousal and a general mood. even if psychedelics may act to weaken the influence of certain highlevel beliefs underlying ‘set’ factors, they may still relatively spare (or increase) the influence of information encoded at lower-levels, which may still have a marked effect on the quality of the acute experience. martin, j. m., & sterzer, p. (2022). how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3, 11. https://doi.org/10.33735/phimisci.2022.9326 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9326 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how level is the ‘cognitive playing field’? 7 3 context shapes alterations in self-conception under psychedelics from the perspective of the rebus model, psychedelics are described to flatten the prior landscape (carhart-harris & friston, 2019) or, in self-modelling terms, to level the ‘cognitive playing field’ (letheby, 2021) with respect to one’s identity. as letheby writes, “by reducing the brain’s confidence in its assumptions about who, what, and where ‘i’ am, the cognitive playing field is levelled with respect to one’s own identity, allowing all sorts of hypotheses to be entertained that normally would be ruled out” (p. 143). however, we believe this notion of a ‘level cognitive playing field’ may be potentially misleading, since it may suggest that, under psychedelics, a user may be just as likely to entertain one hypothesis regarding the self as opposed to any other. we agree that the relaxation of higher-order priors will lead to a general flattening of the prior landscape, “thereby expanding the space of phenomenologically possible worlds” (p. 121) and allowing the brain to “slide more readily between world-models” (p. 143). nevertheless, given the system’s enhanced context-sensitivity, we wish to emphasise that the resulting hypothesis-space that is experienced (in self-modelling terms, who ‘i’ am) is not random — there are principled differences in the kinds of hypotheses that are likely to be entertained during the psychedelic experience. specifically, the user will be more likely to experience alterations in world and self-models that are consistent with internal and external contextual influences, due to the increased influence of prediction errors from these information-sources in revising high-level predictions. in his book, letheby develops a two-factor model of psychedelic therapy. for letheby (2021), this two-factor structure is necessary because the benefits from psychedelics do not only derive from the discovery that the self is mutable—“that much of what we think is concrete and can’t be escaped from are actually just ways we think” (p. 148), but also from the discovery of healthy alternate self-conceptions during the experience: “under the influence, patients can gain new perspectives on their lives, see things differently, and access information previously filtered out or ignored. they can thereby discover new and often healthier forms of selfmodelling” (p. 137). however, if psychedelics lead to a ‘level cognitive playing field’ in terms of one’s felt identity, one may question why patients tend to experience healthy alternative self-conceptions during the acute experience as opposed to unhealthy ones. after all, there are a variety of ways in which one may conceive of the self, many of which are likely to be dysfunctional and to carry detrimental outcomes for mental health. these dysfunctional forms of self-representation should also represent viable candidates for self-modelling within a flattened prior landscape of ‘phenomenologically possible worlds’. we believe that the role of context may provide a solution to this, since it suggests that the different alternative self-conceptions that are possible are not equally likely, given the amplified influence of context on self-modelling. unmartin, j. m., & sterzer, p. (2022). how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3, 11. https://doi.org/10.33735/phimisci.2022.9326 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9326 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org joschua m. martin & philipp sterzer 8 der favourable contextual conditions, healthier and positive self-conceptions are preferred to unhealthy ones, because these high-level hypotheses are assigned a greater probability, given their ability to explain away precise prediction errors derived from internal and external context-dependent information sources. from this perspective, context may shape not only what is consolidated during the integration period (although this is certainly important), but also what kinds of alternative self-models are accessed or discovered during the acute experience. this is consistent with the empirical literature cited previously, where positive expectations, clear intentions, and a supportive and safe setting, are significant predictors of positive acute mystical-type experiences, as well as later improvements in psychological wellbeing. furthermore, it may explain why patients with inadequate preparation and support may experience adverse reactions, and reduced benefits for mental health. in these instances, poor contextual conditions reduce therapeutic benefits through hindering the ability of patients to represent themselves in beneficial ways. if this is on track, then context may play an important role in facilitating the kinds of favourable self-related insights that letheby implicates to underlie durable improvements to mental health. 4 implications for self-related insights and epistemic benefits by considering altered self-conception through the lens of ‘set’ and ‘setting’, one may consider alternative interpretations of some of the qualitative reports described in letheby’s book. for example, take the following subjective report in noorani et al. (2018, p. 759) who evaluated the effects of psilocybin on attempts to quit smoking: it felt like i’d died as a smoker and was resurrected as a nonsmoker. because it’s my perception of myself, and that’s how i felt. so i jumped up and i said ‘i’m not a smoker anymore, it’s all done’. according to letheby (2021, p. 89), cases like the above may be explained by “an increase in the felt identification with the minimal, perspective-taking self, and a concomitant decrease in felt identification with specific contents of the narrative self (the ‘patterns and structures and responsibilities’, such as being a ‘smoker’)”. on this account, psychedelics may lead to a levelled cognitive playing field in terms of the felt identification of the narrative self, where the self-related hypotheses of ‘smoker’ and ‘non-smoker’ are now assigned more or less equal prior probability. this allows patients to access or ‘discover’ a healthier altered self-conception under psychedelics (i.e. the felt identification of a ‘non-smoker’), which may then be further consolidated through an integration period after the experience. however, by considering these changes through the heightened sensitivity to context, one may say that this change in self-conception may also be driven by a relative martin, j. m., & sterzer, p. (2022). how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3, 11. https://doi.org/10.33735/phimisci.2022.9326 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9326 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how level is the ‘cognitive playing field’? 9 increase in the felt identification of the narrative self as a ‘non-smoker’, as this hypothesis is congruent with contextual features, such as the therapeutic setting and the patient’s underlying motivations and expectations to quit. the crucial difference from before is that the ‘non-smoker’ hypothesis is here assigned a higher probability relative to the ‘smoker’ one – they are not on a ‘level cognitive playing field’. this may additionally explain why this altered self-conception is felt with relative confidence and why the addict does not spontaneously transition between the different ‘smoker’ and ‘non-smoker’ perspectives, which a flattened cognitive landscape conception may imply. this account also has possible implications for considering the epistemic benefits that psychedelics confer. specifically, it emphasises the role that nonpharmacological factors play in gaining new self-related knowledge or insights. we interpret the influence of ‘set’ to be best captured by letheby’s description of ‘new knowledge of old facts’, as individuals acquire the ability to represent beliefs about themselves “under a new and different mode of presentation” (letheby, 2021, p. 85). in this way, psychedelics may promote epistemic benefits as individuals experience already existing mental contents in a novel and clinically meaningful way. this may explain the findings suggesting that the insights gained under psychedelics are in many cases “not radically ‘new’, but largely seen as a ‘returning’ or ‘remembering’ of something that was in some sense already known to participants” (noorani et al., 2018, p. 759). for their fit to the present account, one may evaluate the extent to which these ‘already known’ insights can be traced back to consciously accessible ‘set’ factors that an individual has leading up to a psychedelic experience. we further speculate that reports of psychedelics may reveal “a deeper, better, or more essential self” (noorani et al., 2018, p. 759) may reflect an increased congruence between one’s sense of self-concept and the expectations and intentions that one carries for self-related change. 5 concluding remarks chris letheby’s new book, the philosophy of psychedelics, provides a valuable contribution to the current literature through highlighting the primary role that changes to self-related belief systems play in promoting therapeutic benefits. here, we have proposed that letheby’s account may be enhanced by considering the role that contextual influences play in shaping the process of discovery of alternate self-conceptions during the psychedelic experience. we see this position as being compatible with letheby’s account, but may shift some aspects of its emphasis and framing. in particular, we critique the notion that psychedelics lead to a ‘level cognitive playing field’, which may obscure the influence of enhanced-context sensitivity on self-modelling. aspects of this criticism may not be specific to letheby’s account, but may also apply to the rebus model, which also emphasises a flattened prior landscape conception of psychedelic effects. in this sense, our commentary may dovetail with other commentators who have critmartin, j. m., & sterzer, p. (2022). how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3, 11. https://doi.org/10.33735/phimisci.2022.9326 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9326 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org joschua m. martin & philipp sterzer 10 icised the rebus model on the grounds that it does not pay sufficient attention to ‘set’ and ‘setting’ in its application (noorani & alderson-day, 2020). we hope this commentary can supplement letheby’s existing account and provide an opening for a further clarification of his position. these and related questions can help further debates concerning the extent to which belief changes under psychedelics (and their psychological benefits) may be driven by context-dependent and context-independent mechanisms (timmermann et al., 2021). references beliveau, v., ganz, m., feng, l., ozenne, b., højgaard, l., fisher, p. m., svarer, c., greve, d. n., & knudsen, g. m. (2017). a high-resolution in vivo atlas of the human brain’s serotonin system. journal of neuroscience, 37(1), 120–128. https: //doi.org/10.1523/jneurosci.2830-16.2016 carbonaro, t. m., bradstreet, m. p., barrett, f. s., maclean, k. a., jesse, r., johnson, m. w., & griffiths, r. r. (2016). survey study of challenging experiences after ingesting psilocybin mushrooms: acute and enduring positive and negative consequences. journal of psychopharmacology, 30(12), 1268–1278. https://doi.org/10.1177/0269881116662634 carhart-harris, r. l., & friston, k. (2019). rebus and the anarchic brain: toward a unified model of the brain action of psychedelics. pharmacological reviews, 71(3), 316–344. https://doi.org/10.1124/pr.118.017160 carhart-harris, r. l., & nutt, d. j. (2017). serotonin and brain function: a tale of two receptors. journal of psychopharmacology, 31(9), 1091–1120. https://doi.org/10.1177/0269881117725915 carhart-harris, r. l., roseman, l., haijen, e., erritzoe, d., watts, r., branchi, i., & kaelen, m. (2018). psychedelics and the essential importance of context. journal of psychopharmacology, 32(7), 725–731. https://doi.org/10.1177/0269881118 754710 clark, a. (2015). surfing uncertainty: prediction, action, and the embodied mind. oxford university press. corlett, p. r., honey, g. d., & fletcher, p. c. (2016). prediction error, ketamine and psychosis: an updated model. journal of psychopharmacology, 30(11), 1145–1155. https://doi.org/10.1177/0269881116650087 corlett, p. r., horga, g., fletcher, p. c., alderson-day, b., schmack, k., & powers iii, a. r. (2019). hallucinations and strong priors. trends in cognitive sciences, 23(2), 114–127. https://doi.org/10.1016/j.tics.2018.12.001 gashi, l., sandberg, s., & pedersen, w. (2021). making “bad trips” good: how users of psychedelics narratively transform challenging trips into valuable experiences. international journal of drug policy, 87, 102997. https://doi.org/10.1016/j. drugpo.2020.102997 griffiths, r. r., johnson, m. w., carducci, m. a., umbricht, a., richards, w. a., richards, b. d., cosimano, m. p., & klinedinst, m. a. (2016). psilocybin produces substantial and sustained decreases in depression and anxiety in patients with life-threatening cancer: a randomized double-blind trial. journal of psychopharmacology, 30(12), 1181–1197. https: //doi.org/10.1177/0269881116675513 grof, s. (1994). crisis intervention in situations related to unsupervised use of psychedelics. in lsd psychotherapy. hunter house publishers. gukasyan, n., & nayak, s. m. (2021). psychedelics, placebo effects, and set and setting: insights from common factors theory of psychotherapy. transcult psychiatry. https://doi:10.1177/1363461520983684 haijen, e. c., kaelen, m., roseman, l., timmermann, c., kettner, h., russ, s., nutt, d., daws, r. e., hampshire, a. d., & lorenz, r. (2018). predicting responses to psychedelics: a prospective study. frontiers in pharmacology, 897. https: //doi.org/10.3389/fphar.2018.00897 hartogsohn, i. (2016). set and setting, psychedelics and the placebo response: an extra-pharmacological perspective on psychopharmacology. journal of psychopharmacology, 30(12), 1259–1267. https://doi.org/10.1177/0269881116677852 hartogsohn, i. (2018). the meaning-enhancing properties of psychedelics and their mediator role in psychedelic therapy, spirituality, and creativity. frontiers in neuroscience, 12, 129. https://doi.org/10.3389/fnins.2018.00129 hohwy, j. (2013). the predictive mind. oxford university press. james, e., robertshaw, t. l., hoskins, m., & sessa, b. (2020). psilocybin occasioned mystical-type experiences. human psychopharmacology: clinical and experimental, 35(5), e2742. https://doi.org/10.1002/hup.2742 kaertner, l. s., steinborn, m. b., kettner, h., spriggs, m. j., roseman, l., buchborn, t., balaet, m., timmermann, c., erritzoe, d., & carhart-harris, r. l. (2021). positive expectations predict improved mental-health outcomes linked to psychedelic microdosing. scientific reports, 11(1), 1–11. https://doi.org/10.1038/s41598-021-81446-7 martin, j. m., & sterzer, p. (2022). how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3, 11. https://doi.org/10.33735/phimisci.2022.9326 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1523/jneurosci.2830-16.2016 https://doi.org/10.1523/jneurosci.2830-16.2016 https://doi.org/10.1177/0269881116662634 https://doi.org/10.1124/pr.118.017160 https://doi.org/10.1177/0269881117725915 https://doi.org/10.1177/0269881118754710 https://doi.org/10.1177/0269881118754710 https://doi.org/10.1177/0269881116650087 https://doi.org/10.1016/j.tics.2018.12.001 https://doi.org/10.1016/j.drugpo.2020.102997 https://doi.org/10.1016/j.drugpo.2020.102997 https://doi.org/10.1177/0269881116675513 https://doi.org/10.1177/0269881116675513 https://doi: 10.1177/1363461520983684 https://doi.org/10.3389/fphar.2018.00897 https://doi.org/10.3389/fphar.2018.00897 https://doi.org/10.1177/0269881116677852 https://doi.org/10.3389/fnins.2018.00129 https://doi.org/10.1002/hup.2742 https://doi.org/10.1038/s41598-021-81446-7 https://doi.org/10.33735/phimisci.2022.9326 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how level is the ‘cognitive playing field’? 11 lavazza, a. (2017). ways of being well: realistic and unrealistic well-being. in l.tadio (ed.), new perspectives on realism, 237–252. mimesis international. leary, t., litwin, g. h., & metzner, r. (1963). reactions to psilocybin administered in a supportive environment. journal of nervous and mental disease, 137(6), 561–573. https://doi.org/10.1097/00005053-196312000-00007 leary, t., metzner, r., & alpert, r. (1964). the psychedelic experience: a manual based on the tibetan book of the dead. new york: university press books. letheby, c. (2021). philosophy of psychedelics. oxford university press. masters, r., & houston, j. (2000). the varieties of psychedelic experience: the classic guide to the effects of lsd on the human psyche. simon and schuster. mcgovern, h. t., leptourgos, p., hutchinson, b., & corlett, p. r. (2021). do psychedelics change beliefs? psyarxiv preprints. https://10.31234/osf.io/3dc2h metzner, r., litwin, g., & weil, g. (1965). the relation of expectation and mood to psilocybin reactions: a questionnaire study. psychedelic review, 5, 3–39. noorani, t., & alderson-day, b. (2020). spotlight commentary: rebus and the anarchic brain. neuroscience of consciousness, 2020(1), niaa007. https://doi.org/10.1093/nc/niaa007 noorani, t., garcia-romeu, a., swift, t. c., griffiths, r. r., & johnson, m. w. (2018). psychedelic therapy for smoking cessation: qualitative analysis of participant accounts. journal of psychopharmacology, 32(7), 756–769. https://doi.or g/10.1177/0269881118780612 roseman, l., nutt, d. j., & carhart-harris, r. l. (2018). quality of acute psychedelic experience predicts therapeutic efficacy of psilocybin for treatment-resistant depression. frontiers in pharmacology, 8, 974. https://doi.org/10.3389/fphar.2017 .00974 ross, s. (2018). therapeutic use of classic psychedelics to treat cancer-related psychiatric distress. international review of psychiatry, 30(4), 317–330. https://doi.org/10.1080/09540261.2018.1482261 ross, s., bossis, a., guss, j., agin-liebes, g., malone, t., cohen, b., mennenga, s. e., belser, a., kalliontzi, k., babb, j., su, z., corby, p., & schmidt, b. l. (2016). rapid and sustained symptom reduction following psilocybin treatment for anxiety and depression in patients with life-threatening cancer: a randomized controlled trial. journal of psychopharmacology, 30(12), 1165–1180. https://doi.org/10.1177/0269881116675512 safron, a. (2020). on the varieties of conscious experiences: altered beliefs under psychedelics (albus). psyarxiv preprints. https://10.31234/osf.io/zqh4b. sterzer, p., adams, r. a., fletcher, p., frith, c., lawrie, s. m., muckli, l., petrovic, p., uhlhaas, p., voss, m., & corlett, p. r. (2018). the predictive coding account of psychosis. biological psychiatry, 84(9), 634–643. https://doi.org/10.1016/j.bi opsych.2018.05.015 studerus, e., gamma, a., kometer, m., & vollenweider, f. x. (2012). prediction of psilocybin response in healthy volunteers. plos one, 7(2), e30800. https://doi.org/10.1371/journal.pone.0030800 timmermann, c., kettner, h., letheby, c., roseman, l., rosas, f. e., & carhart-harris, r. l. (2021). psychedelics alter metaphysical beliefs. scientific reports, 11(1), 1–13. https://doi.org/10.1038/s41598-021-01209-2 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. martin, j. m., & sterzer, p. (2022). how level is the ‘cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3, 11. https://doi.org/10.33735/phimisci.2022.9326 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1097/00005053-196312000-00007 https://10.31234/osf.io/3dc2h https://doi.org/10.1093/nc/niaa007 https://doi.org/10.1177/0269881118780612 https://doi.org/10.1177/0269881118780612 https://doi.org/10.3389/fphar.2017.00974 https://doi.org/10.3389/fphar.2017.00974 https://doi.org/10.1080/09540261.2018.1482261 https://doi.org/10.1177/0269881116675512 https://10.31234/osf.io/zqh4b https://doi.org/10.1016/j.biopsych.2018.05.015 https://doi.org/10.1016/j.biopsych.2018.05.015 https://doi.org/10.1371/journal.pone.0030800 https://doi.org/10.1038/s41598-021-01209-2 https://doi.org/10.33735/phimisci.2022.9326 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org `set' and `setting' in psychedelic research predictive processing, the rebus model, and context context shapes alterations in self-conception under psychedelics implications for self-related insights and epistemic benefits concluding remarks explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs) colin kleina (colin.klein@anu.edu.au) jakob hohwyb (jakob.hohwy@monash.edu) tim bayneb (timothy.bayne@monash.edu) abstract the science of consciousness is currently structured around the search for the neural correlates of consciousness (nccs). one of the alleged advantages of the ncc framework is its metaphysical neutrality—the fact that it begs no contested questions with respect to debates about the fundamental nature of consciousness. here, we argue that even if the ncc framework is metaphysically neutral, it is structurally committed, for it presupposes a certain model—what we call the lite-brite model—of consciousness. this represents a serious liability for the ncc framework, for the plausibility of the lite-brite model is very much an open question, and the science of consciousness would be better served by a framework that does not presuppose it. drawing on interventionist ideas in the philosophy of science, we suggest that the difference-maker framework can provide just such an alternative. instead of searching for the neural correlates of consciousness (nccs), we ought to be searching for the difference makers of consciousness (dmcs). keywords consciousness ∙ difference-making∙ explanation ∙ intervention ∙ neural correlates ∙ nccs this article is part of a special issue on “the neural correlates of consciousness”, edited by sascha benjamin fink. athe australian national university bmonash university klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-7406-4010 https://orcid.org/0000-0003-3906-3060 https://orcid.org/0000-0001-8591-7907 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org colin klein, jakob hohwy, and tim bayne 2 1 introduction since the revival of neuroscientific interest in consciousness in the early 1990s, consciousness science has focused on the search for the neural correlates of consciousness (nccs) (koch et al., 2016; metzinger, 2000). although there is much debate about what and where the nccs are (e.g. boly et al., 2017; odegaard et al., 2017), there has been relatively little debate about whether the science of consciousness ought to be framed in terms of the search for the nccs (although see noë & thompson, 2004; neisser, 2012). this paper challenges that consensus, arguing that the ncc framework presupposes a particular – and, we suggest, implausible – conception of the structure of consciousness. we call the conception in question the lite-brite model of consciousness. the lite-brite model and the ncc framework are mutually reinforcing: adherence to the ncc framework as the only game in town supports the assumption that the lite-brite model must be right, and that in turn underpins the sense that the ncc framework is appropriate. in the first half of this paper we argue that the science of consciousness would be better served by adopting a framework which is neutral between competing conceptions of the structure of consciousness. in the second half of the paper we turn to the task of developing such a framework. drawing on interventionist accounts of explanation (craver & kaplan, 2011; cummins, 2000; pearl, 2000; woodward, 2003), we suggest that the goals of consciousness science would be best-served by focusing on difference-making relations. in a nutshell, the science of consciousness should replace the search for the neural correlates of consciousness with the search for the difference-makers of consciousness (dmcs). 2 the ncc framework at the heart of the ncc framework is the very notion of an ncc. in a representative recent statement, koch et al (2016, p. 308) define an ncc as “… the minimum neuronal mechanisms jointly sufficient for any one specific conscious percept.” this requires some unpacking; we will do our critique in the course of that unpacking. we note from the outset that the notion of a neural correlate of consciousness is a contested one, and there is some variation in how different research groups understand the notion. that said, there is enough of a consensus about what an ncc is for us to intelligibly talk about ‘the ncc framework’ – that is, the idea that consciousness science should be structured around the search for nccs. first, although koch et al. refer here to ‘neural mechanisms’, there are few discussions of mechanisms or mechanistic explanation as such within the ncc literature. indeed, the question of how exactly to characterize the n’s in the nccs has been relatively neglected. early work on nccs tended towards a simple localizationist approach in which ns were categorized in terms of brain regions. for example, v5/mt has often been described as the ncc for motion perception (block, klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org explanation in the science of consciousness 3 2005). more recent work suggests a range of ways of categorizing n-type states. in other areas of neuroscience, for example, authors have emphasized the explanatory importance of structural and dynamic networks (sporns, 2011). second, although koch et al. refer to ‘one specific conscious percept’, the ncc framework has not been restricted to percepts (or even to the contents of consciousness more generally). the nccs that are specific to particular conscious contents are typically called content-nccs. in addition to content-nccs, theorists have also targeted state-nccs – that is, the neural correlates of global, non-specific, conscious states, such as those that are associated with alert wakefulness, remdreaming, light sedation and the minimally conscious state (bayne et al., 2016; laureys, 2005). a third type of conscious state that has been brought within the scope of the ncc framework is the generic state of simply being conscious (chalmers, 2000). an ncc which functions as a minimal neuronal substrate for being conscious is a generic-ncc.1 but it is the third aspect of the notion of an ncc that is perhaps most noteworthy: the idea that a neural state n qualifies as the ncc for a particular conscious state c if and only if n is ‘minimally sufficient’ for c. n is minimally sufficient for c if and only if the existence of n necessitates the existence of c (that’s the sufficiency bit), and no proper part of n necessitates c (that’s the minimal bit). the motivation for characterizing nccs in terms of minimal sufficiency is clear: one wants to isolate only those neural features that are directly implicated in consciousness. from the perspective of the ncc framework, neural states that are causally upstream or downstream of consciousness ought to be screened off as confounds (aru et al., 2012; de graaf et al., 2012; miller, 2007). they might be of interest to the science of consciousness broadly construed, but they aren’t the central quarry of those working within the ncc framework. although nccs are defined in terms of minimal sufficiency, this is a point of tension between the ideology of the ncc framework and its practice, for the methods used to identify nccs don’t allow one to draw inferences about minimal sufficiency (bayne, 2007; fink, 2016; hohwy, 2009; searle, 2000). in searching for content-nccs, theorists contrast a situation in which a participant has one type of experience (say, a face) with a situation in which they have another type of experience (say, a house). the neural state (e.g., activity in the fusiform “face area”, ffa) that is implicated in the face experience (but not the house experience) is then described as ‘the ncc for faces’. however, this paradigm clearly provides 1although most theorists distinguish between these different types of nccs, there is disagreement/confusion about how they are related to each other. for example, koch et al. state that the nccs for generic consciousness (what they call the ‘full nccs’) can be identified with “the union of the sets of content-specific ncc for all possible contents of experience” (koch et al., 2016, p. 308). that suggestion is mistaken, for if there is an a priori connection between generic-nccs and content-specific nccs it is likely to involve the intersection relation rather than the union relation. however, in our view the ncc framework is best developed without appealing to any particular view of the relationship between different types of nccs, for different accounts of the structure of consciousness will entail different relations here. klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org colin klein, jakob hohwy, and tim bayne 4 no reason to think that ffa activity is fully sufficient for face experiences, for it doesn’t allow one to screen off neural activity that is necessary both for experiences of houses and for experiences of faces. further, there are good reasons to doubt whether ffa activity could suffice for an experience of faces. after all, one wouldn’t expect to be able to generate face experiences by placing a section of ffa in a petri dish and running some current through it. all one can conclude from this kind of contrastive experiment is that ffa suffices for face-experiences in an awake conscious participant with an otherwise complete and intact brain. in response to this point, some theorists distinguish between a conscious state’s total ncc (which guarantees its existence) and its core ncc, which is “the part of the total ncc that distinguishes one conscious state from another – the rest of the total ncc being considered as the enabling conditions for the conscious experience” (block, 2005, p. 47; also chalmers, 2000). with this distinction in hand (it is claimed) we can then say that the contrastive method just outlined allows one to draw inferences only about the core component of a content-ncc. although the ncc framework clearly requires something like the distinction between total nccs and core nccs, it is less clear what exactly the distinction amounts to or how it might be drawn. in distinguishing between core and total nccs block and chalmers draw on shoemaker’s (1981) distinction between the core and the total realizer of a functional state. however, neither block nor chalmers espouse functionalism, and it’s unlikely that they intend for the distinction between core and total nccs to be understood in functional terms. block suggests that the non-core component of an ncc can be viewed as an enabling condition for the conscious experience, but that suggestion can be understood in two ways. on one view an enabling condition is merely a causal factor in the generation of the core component. however, that doesn’t seem to be what block has in mind, for that view would imply that activating the core ncc in a petri dish would generate an experience – a state of affairs that he clearly rules out. another possibility is that the non-core component is a background condition: its role isn’t to activate the core component, but to collude with the core factor in generating a conscious state in the way in which the striking of a match colludes with oxygen to generate fire. but if this interpretation is right (and it is arguably what block and chalmers have in mind), then the choice of labels is a somewhat unhappy one, for the contribution of the non-core ncc is no more marginal or peripheral than the activity of the core ncc. the distinction between core nccs and non-core nccs might be important from an explanatory point of view (with core nccs being central to some explanatory projects and non-core nccs central to others), but it is not clear what metaphysical ice it would cut. arguably the fundamental distinction that block and chalmers are driving at here is the distinction between neural activity that is specific to a particular type of content and that which is non-specific – i.e., which occurs across a range of conscious contents. in other words, core nccs are essentially differentiating neural factors, whereas non-core nccs are non-differentiating factors. the klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org explanation in the science of consciousness 5 differentiating/non-differentiating distinction is certainly useful – not least because it maps on to the practice of consciousness science – but arguably it undermines the official definition of content-specific nccs. recall that the ncc for c is defined as the minimally sufficient basis for c, but we’ve seen that the focus of consciousness science concerns a component of that ncc that isn’t minimally sufficient for it. given that the science of consciousness actually focuses on core nccs, one has to ask whether the science of consciousness is well-served by a framework that has as its ostensive focus the search for minimally sufficient neural states. given the problems that we have identified, why describe the search for the mechanisms/neural substrate of consciousness in terms of the ‘correlates of consciousness’? the terminology of correlates is largely motivated by a desire to remain neutral on contested questions about the fundamental nature of consciousness (chalmers, 2000; crick, 1996). some theorists take conscious states to be identical to brain states, others regard conscious states as constituted or realized by brain states, and still others hold that conscious states are merely correlated with brain states. by not taking sides in the disputes about the ultimate metaphysical relationship between brain states and conscious states, the ncc framework is ideally suited for the science of consciousness, for it seems to be perfectly anodyne and contests no contested issues. metaphysical neutrality is indeed desirable, but it is not the only kind of neutrality we should be looking to preserve. another kind of neutrality worth preserving is structural neutrality: insofar as it is possible, the framework of consciousness science should avoid taking sides in debates about the structure of consciousness. in this regard the ncc framework is problematic, for it presupposes a particular conception of the structure of consciousness, as we will now show. 3 the lite-brite model and the neural correlates of consciousness a lite-brite is a popular children’s toy consisting of a light box with translucent coloured plastic pegs (see fig.1 for a representation). these pegs – which come in nine vibrant colours – are fitted through slots in black paper in the light box. when illuminated, a picture emerges that is a function of the colours and locations of the pegs. what we’ll call the lite-brite model of consciousness treats the structure of consciousness as similar to that of the eponymous toy. the lite-brite model has three key commitments.2 the first is the autonomy assumption: whether or not a neural event functions as the basis of a particular 2there is a close parallel between the lite-brite model and what searle refers to as the building block model of consciousness (searle, 2000; see also bayne, 2010). there is also a close parallel between our criticisms of the lite-brite model and the criticisms of the ncc framework developed by noë and thompson (2004) and neisser (2012). klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org colin klein, jakob hohwy, and tim bayne 6 figure 1: a representation of a virtual lite-brite conscious experience does not depend on the nature or presence of other experiences had by the subject at the same time.3 this assumption has a parallel in the structure of the lite-brite, for the colour that is instantiated at any particular location is independent of the colours that are instantiated at other locations (or, indeed, of whether any other colours are instantiated in the lite-brite at all). the second key commitment of the lite-brite model is the uniqueness assumption. uniqueness says that the presence/absence of a certain type of conscious state is determined by the activity of a single neural element. given domain-general enabling conditions, an experience of (say) a face will depend on activity in a single face ncc. again, there is a parallel with the structure of a lite-brite, for the colour at any one location in the figure is determined by a single element in the light box, and to change its color there’s only one place to intervene. the third commitment is the uniformity assumption: global changes to a subject’s conscious state have uniform effects on their experiences, and, crucially, do not change the relationship between a particular type of experience and its neural basis. for example, the assumption would be that a transition from wakefulness to dreaming would uniformly affect all parts of consciousness, and not differentially alter the relation between the neural substrates and different dimensions of conscious experience. here too there is a parallel with the structure of a lite-brite, for 3there is a parallel here with the pure insertion assumption in cognitive neuroscience (friston et al., 1996; sternberg, 1969). klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org explanation in the science of consciousness 7 there is a robust distinction between the specific changes made by the pegs and the general changes made by the bulb. making global changes to the lite brite (by, for example, changing the wattage of the bulb), will have uniform effects on each of its components. the conception of consciousness that emerges from these three assumptions is beguiling in its simplicity. if autonomy is correct, then one can look for the neural basis of a particular type of experience c without worrying about what else one’s participant might be experiencing. all that matters, it seems, is that they are having a c-type experience. if uniqueness is correct, then one need find only a single neural state in order to account for c. having found such a state, one’s job is done. and if uniformity is correct, then one need not worry about controlling for differences in global states. if one has found the neural basis of c as it occurs in (say) the state of alert wakefulness, then one can be confident that one has also found the neural basis of c as it occurs across every other global state. it should be evident that the lite-brite model informs the ncc framework, for all three of these assumptions are embraced by the ncc framework. although that commitment is rarely explicit, it can be seen in the language of the ncc framework and in the kinds of experiments that people conduct – or fail to conduct. consider first autonomy. theorists rarely (if ever) consider the possibility that a neural state might be sufficient for c in some contexts but not others. in the same way that the colour produced by a particular peg in the lite-brite is independent of its context, so too it is typically assumed that the experiential state produced by an ncc will be independent of its context. in other words, the ncc framework routinely ignores the possibility of interaction effects between the various components that jointly constitute the neural basis of consciousness. instead, the ncc for an individual’s total conscious state (their overall subjective perspective) is assumed to be a simple conjunction of the nccs of each of the experiences that make up that total state. the ncc framework is also committed to uniqueness. consider the following representative quotation: the content-specific ncc are the neurons (or, more generally, neuronal mechanisms), the activity of which determines a particular phenomenal distinction within an experience. for example, the ncc for experiencing the specific content of a face are the neurons that fire, on a trial-by-trial manner, whenever a person observes, imagines or dreams a face, and are silent in other circumstances. (koch et al., 2016, p. 308) because nccs are understood to be merely sufficient for the corresponding conscious state (and not also necessary), the ncc framework allows that different kinds of neural states could qualify as nccs for a particular kind of conscious state in creatures of different kinds. that being said, it is widely assumed that for each state of consciousness there will be a single ncc in neurotypical humans. klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org colin klein, jakob hohwy, and tim bayne 8 this assumption of uniqueness is reflected in the standard phrasing of the ncc for an experience, rather than an ncc or one of many nccs. finally, the ncc framework is committed to uniformity. theorists who operate within the ncc framework rarely (if ever) consider the possibility that there might be one set of neural correlates for (say) c-as-it-occurs-in-wakefulness and another set of neural correlates for c-as-it-occurs-in-dreaming. note that this is an analogue of autonomy in the case of global states rather than contents. however, it is important for the ncc framework to preserve this distinction alongside uniformity, lest the core/total distinction break down. we have suggested, then, that the ncc framework is tightly wedded to the litebrite model, such that the failure of the lite-brite model would raise real questions about the suitability of the ncc framework. of course, that implication would be unproblematic if we had good reason to embrace the lite-brite model. do we? we don’t believe so. for one, there are numerous examples of complex scientific systems which would not fit the corresponding lite-brite model in their domains. sciences like genetics advanced precisely because they gave up on a general commitment to a simple 1-to-1 mapping between genes and phenotypic traits. cognitive neuroscience has increasingly moved away from a simple view of cognitive ontology (anderson, 2014), and we should expect the underpinnings of consciousness to be no less complex than cognitive domains. we will return to this point shortly. furthermore, few of the leading theories of consciousness are committed to – or even support – the lite-brite model. for example, one of the leading accounts of phenomenal character equates it with representational content of a certain kind (e.g., dretske, 1997; tye, 1995). although one could equate the relevant kind of representational content with the content of a specific neural state (the ‘vehicle’ of that experience), there is nothing in the representationalist treatment of phenomenal character which mandates that position. arguably, the only influential theory of consciousness that is committed to the lite-brite model is the identity theory, famously championed by j.j.c. smart in the 1950s and occasionally still defended by philosophers (e.g., hill, 1991; mclaughlin, 2007; polger, 2006). the advocate of the lite-brite model might be tempted to brush aside these points and insist that we already have robust evidence in favour of the lite-brite approach. don’t lesion, tms and neuroimaging studies show that the ffa is the ncc for visual experiences of faces? that v4 is the ncc for colour? that mt/v5 is the ncc for motion perception? we don’t deny that certain areas of the brain are more centrally involve in some types of conscious states than others. but acknowledging that point falls a long way short of endorsing the lite-brite model as a general account of consciousness. furthermore, it is perhaps not surprising that the successes of the science of consciousness (such as they are) have a ‘lite-brite’ flavour, for few theorists have explored the empirical merits of the interactionist alternative. the fundamental point here is not that the lite-brite model is false, klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org explanation in the science of consciousness 9 but that the ncc framework is systematically blind to the possibility of the interactionist picture. in sum, there is ample reason to develop a framework for the science of consciousness that is not committed to the lite-brite model in the way in which the ncc framework is. that is the task of the next section. 4 explanation and difference-making relations as we have noted, the search for the neural basis of consciousness is often identified with the search for the neural mechanisms of consciousness. yet with a few exceptions (hohwy & frith, 2004; irvine, 2013; neisser, 2012), there has not been much attempt to say what is distinctive and useful about mechanistic explanation for the science of consciousness. we suggest that the attractive component of mechanistic explanations – and the one that consciousness science ought to adopt – is the idea that explanation is done by citing not correlations but differencemaking relations (dmrs). at the core of a dmr is the idea that manipulating one aspect of a system enables one to manipulate other aspects of it. it is the identification of these manipulability relations, and not the identification of correlations, which enables one to understand, predict and intervene on the operations of mechanisms (craver, 2007). evidence about difference-making relations can be the result of direct (e.g. tms), indirect (e.g. having different treatment conditions), or serendipitous (e.g. lesion studies) manipulations.4 moving towards a difference-making account obviates the need to posit 1-to-1 relations between activities of parts and explananda. this has been explicitly noted as an advantage of difference-making accounts in mature sciences (klein, 2017; sterelny & kitcher, 1988). discussing the practice of labelling genes for phenotypic traits, sterelny and kitcher remark: consider the vast number of loci in drosophila melanogaster which are labeled for eye-color traits – white, eosin, vermilion, raspberry, and so forth. nobody who subscribes to this practice of labeling believes that a pair of appropriately chosen stretches of dna, cultured in splendid isolation, would produce a detached eye of the pertinent color. rather, the intent is to indicate the effect that certain changes at a locus would make against the background of the rest of the genome. (sterelny & kitcher, 1988, p. 348) 4the mechanistic account is thus facilitated by the difference-making associated with manipulations and interventions. note that it is not necessary to appeal to mechanisms in order to formulate explanations in terms of difference-making via interventions. that is, difference-making can be described without appeal to mechanistic notions of parts and processes underlying phenomena of interest. klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org colin klein, jakob hohwy, and tim bayne 10 dmrs between different genes and phenotypic traits provide an important initial source of data for genetic theory, but the mapping from gene to trait is rarely oneto-one. many traits are affected by multiple genes, and many genes simultaneously affect multiple traits. dmrs are not typically found by looking for simple correlations between phenomena of interest and underlying mechanisms. instead, one first catalogues the myriad and complicated dmrs that obtain, and then builds a mechanistic theory that accounts for them. consider, for example, primary ciliary dyskinesia. this is a genetic disorder associated with upper respiratory infections, infertility, and a 50% chance of having the heart on the right side of the body. this surprising cluster of defects is due to a single mutation that affects the development of cilia, and so sheds light on the mechanisms of symmetry-breaking in the developing foetus and the development of the tails of sperm. the fact that a single gene affects a number of otherwise distinct variables also suggests unexpected mechanisms common to each. two further features of dmrs are worth noting. first, dmrs make a difference relative to a contrast class. different contrast classes will result in different explanatory factors. to explain why someone has a fever rather than not we might best cite the presence of the influenza virus; why they have a severe fever rather than a mild one by reference to features of the individual’s immune system; why they have a remittent fever rather than a constant fever by reference to the hypothalamic regulatory system, and so on. second, dmrs can vary in strength across several axes. some differencemaking relations are basically switch-like: they toggle their target between two possible states. other dmrs are more specific: they connect numerous states of the difference-maker with numerous states of the target in a roughly one-to-one fashion, like the tuning dial on a radio (woodward, 2010). similarly, some difference-makers give a systematic handle on their targets, as they allow for well-defined responses to change (klein, 2017). contrast this with relationships such as ‘being minimally sufficient for’, which are basically binary: either the relationship obtains or it doesn’t, and there’s little more to say. 5 difference-makers for consciousness given the success of the search for dmrs in other fields, we suggest a similar approach to consciousness. we’ll call this the ‘dmc framework’, for it places the search for the difference-makers of consciousness (dmcs) at the centre of consciousness science. at the heart of the dmc framework is the search for control variables whose settings have systematic effects on consciousness. if we change these settings, then we should expect certain effects on consciousness. more formally, let 𝐸 be a variable which corresponds to some particular conscious state of interest, which can take values {𝐸1, 𝐸2, … , 𝐸𝑛}. intuitively, if 𝐸 designates a particular determinable klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org explanation in the science of consciousness 11 experience, the different possibilities correspond to different determinates of that determinable.5 so for example, if 𝐸 corresponds to a visual experience of color, the different options {𝐸1, 𝐸2, … , 𝐸𝑛} might correspond to different hues or levels of brightness or saturations of that color. similarly, let 𝑅 correspond to the pattern of neural activity in some region of interest, which can take values {𝑅1, 𝑅2, … , 𝑅𝑛}. (“pattern of neural activity” and “region of interest” are placeholders here; the framework is compatible with most things you might want to plug in on the brain side). then we can formalize the picture as follows: 𝑅 is a dmc with respect to 𝐸 just in case there is an intervention on the value of 𝑅 which changes the value of 𝐸. thus, intervening on v4 (while keeping everything else relevant fixed) in a certain way might (e.g.) change the character of an aspect of one’s visual experience from one hue to another. the fundamental difference between the ncc and dmc frameworks thus stems from the difference between a correlate and a difference-maker. correlates are two-place relations (𝑅 is a correlate of 𝐸), whereas difference-makers are four-place relations: 𝑅1, rather than {𝑅2, … , 𝑅𝑛} makes it that 𝐸1, rather than {𝐸2, … , 𝐸𝑛}. the added flexibility of the dmc framework comes in part because of the additional argument places and the possibilities they create. one place where the difference between the two frameworks is most stark is when properties of a single perceptual experience might depend on combinations of underlying states. this is often the case with neural population codes. so suppose, for example, that faces are represented neurally as a weighted combination of eigenfaces (tsao & livingstone, 2008). this sort of population coding is one in which the whole population is necessary for the representation, but in which different gestalt aspects of the represented face are affected by different weightings. so for example, tweaking along one dimension of the representation might make the face look younger or older, tweaking along another dimension might make the face look more stereotypically masculine or feminine, and so on. on the ncc approach we might note that activity 𝑅1 correlates with face experience 𝐸1. we might also note that different patterns might correlate with different experienced faces. the dmc approach, by contrast, allows us to capture the combinatorial aspect of the population coding. we might show, for example, that pattern 𝑅1, rather than 𝑅2, means that the face looks young rather than mature; 𝑅1, rather than 𝑅3 makes the face look heavyset rather than slender; 𝑅1, rather than 𝑅4 makes the face look friendly rather than suspicious, and so on. in short, the same population can support various difference-makers for different aspects of an experience. similar combinatorial principles can also apply across brain re5woodward discusses the determinable-determinate relation as holding between variables and their values (woodward, 2003, p. 39). in the cases we discuss, that relationship holds because of a determinable-determinate relationship between what is designated by the terms and values. following funkhouser (2006) we will also assume that a single determinable might have more than one dimension of determination. klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org colin klein, jakob hohwy, and tim bayne 12 gions, giving straightforward violations of uniqueness. so for example we might find that activity in 𝑅 makes the difference between seeing a face rather than a house, while activity in 𝑆 makes the difference between seeing a face rather than an arm.6 the extra argument places afforded by a move to dmcs thus allow for a theory which respects combinatorial relationships between neural processes and conscious experiences. in other realms where similar combinatorial relationships hold – such as the relationship between genotype and phenotype – the search for dmrs brings needed clarity. given the ubiquity of combinatorial coding schemes in the brain, one ought to expect similar benefits for the science of consciousness. a further advantage of the dmc framework is that it permits a more nuanced picture of the relationship between global states and particular conscious contents. first, any dmr holds only relative to some background conditions and over some range of possible interventions (woodward, 2003, pp. 239–315). hence the dmc framework is fine with blockers and context-sensitive difference-makers: there is nothing more remarkable here than the fact that tuning the knob on the radio manipulates the station between 88 and 108 mhz if the selector is flipped to ‘fm’ and between 540 and 1600 khz if the selector is on ‘am’. indeed, context-sensitive relations of this kind are central to understanding the structure of an underlying mechanism. hence the dmc framework is compatible with cases where uniqueness or autonomy do hold. on the other hand, the dmc is well-positioned to capture true interaction effects: that is, cases where the conjunction of two states can have effects that neither conjunct alone can have. so for example, suppose that a certain experience of suffocating paranoia e arises only when someone believes the police are at the door (content-bearing brain state r) and is under the influence of marijuana (global state s). the conjoined state r&s is a difference-maker for e, and systematic investigation will show how e can be varied or eliminated by changing either r or s. however, both autonomy and uniformity are violated. autonomy is violated because although r is a content-bearing state, its relationship to e depends on, and mediated by, s. uniformity is violated because changes to s will also generate widespread changes of varying kinds to other phenomenal states. the ncc approach handles state-dependence of various sorts by insisting on a distinction between state and content nccs, or between core and total nccs, or more generally between nccs for particular experiences and very general global enabling conditions for those particular nccs to be in place. we have suggested that such a sharp distinction might be misplaced, and that in any case the science 6this notion of a combinatorial realizer should be distinguished from the simpler variety of combination present in conscious binding. if activity at 𝑁2 generates blue experience and activity at 𝑁3 generates motion, then these two forms of activity might be combined in a binding process to generate an experience of a moving blue square. binding is compatible with the ncc project, whereas the type of combination we describe here is not. the combinatorial principles we describe do not pick out distinct aspects of an experience that can be added together – instead, the same aspect is the result of a combination of different underlying factors. klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org explanation in the science of consciousness 13 of consciousness need not and should not assume such a distinction. the extra argument places afforded by the dmc relationship give it the flexibility to deal with relationships in interactive, combinatorial systems – and we have every reason to believe that the neural underpinnings of consciousness constitute such a system. 6 the difference difference-makers of consciousness make we turn now to consider a number of possible objections to the dmc framework. first objection: “isn’t the contrast between the ncc framework and the dmc framework relatively superficial, and really just a matter of notation? sure, we could talk in terms of dmcs rather than nccs – but what precisely would we gain by so doing? as far as the science of consciousness is concerned, adopting the dmc framework would surely be business as usual.” reply: we think that this objection arises because of a tacit commitment to a lite-brite structure for consciousness. although difference-making relations are always potentially plural and contrastive, in a lite-brite this point is academic. the colour at any point is solely determined by the peg (if any) that is fitted: contrasts, contingently, do not come apart. so if consciousness has a lite-brite structure, then the dmc framework collapses into the ncc one – the extra argument places, while present, do no interesting work. however, we have suggested that the lite-brite model is false and, more importantly, that consciousness science ought not assume that it is true. one cannot build a theory of interactionist systems by studying single contrasts. instead, one must study a range of contrasts, including sophisticated ones, and studying more sophisticated contrasts and intermediate cases between causal factors and background conditions. that in turn should affect the types of experiments we run and the experimental conditions we consider. there is a grain of truth to this objection insofar as adopting the dmc framework needn’t demand a revolution in the methods of consciousness science. however, this isn’t because the contrast between the ncc and dmc frameworks has no practical implications, but because the vast majority of studies that consciousness scientists conduct are best understood as attempts to identify difference-makers. for example, although the contrastive method – comparing the brain of someone in conscious state 𝐸1 with someone in 𝐸2 – is clearly unable to tell us what the minimally sufficient neural basis of either 𝐸1 or 𝐸2 is, it is able to reveal an important difference-maker for 𝐸1-rather-than-𝐸2. hence if it seems like there is little daylight between dmcs and nccs, that is likely because much of consciousness science is already looking for dmcs rather than nccs. conversely, we suspect that any attempts to clarify the ncc approach in ways that make it consistent with actual practice will end up simply re-introducing a version of dmcs. hence the dmc approach is not merely a notational variant of the ncc approach: rather, klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org colin klein, jakob hohwy, and tim bayne 14 the ncc framework is an inadequate attempt to capture what the dmc approach does well. second objection: “doesn’t the dmc framework make consciousness science too easy? after all, difference-makers are thick on the ground. an appropriately illuminated banana is a difference-maker for experiences of yellow, is it not? if we’re only looking for difference-makers then surely consciousness science is trivially easy.” reply: there are two responses to this objection: one conservative, the other liberal. to start, note that we have so far remained neutral between differencemaking considered in the contexts of causal relationships and in the context of constitutive relationships. the two relations are distinct (the former is diachronic, the latter synchronic) but the logic of difference-making is similar in each. the extension of difference-making from causal to constitutive relationships has not been without controversy; we are on the side of those who think that it is ultimately unproblematic.7 the conservative response is simply to restrict the science of consciousness to constitutive (and so presumably neural) dmcs. external states such as the presence of a well-lit banana don’t count – the thought runs – because they are merely causal difference-makers. miracchi’s (2017) arguments in favour of ‘generative’ relationships for conscious realizers could be read as an instance of this strategy. the liberal response would be to place no restrictions on the type of differencemaker, and instead appeal to differences between difference-makers to guide consciousness science. recall that difference-making relationships vary along a number of axes: for example, they can vary in specificity, systematicity, and robustness. these are not variations in quality per se: however, in general the more specific, systematic, invariant, robust difference-makers give more information about underlying mechanisms. so the presence of a banana is a dmc for color, but it is not a particularly informative one: many variations in the banana don’t affect the color experience, the same banana gives rise to different experiences given different illuminants, the banana may make no difference at all if the subject is distracted, and so on. the reason to focus on (say) v4 is that we assume that it has a tighter relationship to experienced color under the same circumstances: we could change color experience by manipulating it even if everything else were kept fixed. on the whole, we find the liberal response more attractive. we are convinced that part of the power of the dmc framework comes from its inclusive breadth. this may include both causal (neisser, 2015) and even cross-level (janssen, 2019) relationships. here, as elsewhere, it is to the advantage of the dmc framework that it begs no contested questions about the location of the physical substrates of 7craver (2007) notably used a variety of constitutive intervention to defend mechanistic explanation. there has been a substantial debate about the degree to which causal and constitutive explanations differ. some of this debate is about the metaphysics of constitutive interventions as such, but much of it is about whether these are compatible with other assumptions of the mechanistic framework. kästner and andersen (2018) provide an excellent overview of the state of the debate. klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org explanation in the science of consciousness 15 consciousness (c.f. kaplan, 2012, for a similar argument). further, consciousness science can and does benefit from the sort of careful, systematic study of stimuli done by perceptual psychophysics. so the mere fact that something is external to the head does not mean that it is not a useful dmc for building a science of consciousness. either way, it is important not to fall back into assuming a lite-brite model. the dmc framework does not insist that we look only for maximally specific (etc.) dmcs. all dmcs are potentially relevant to building a theory of conscious experience. that may include dmcs which have surprising non-specific, non-selective effects on consciousness. if, for example, the claustrum functions as a kind of master switch for consciousness (crick & koch, 2005), that is useful information to know. it is useful in part because it might guide future research into the effects on consciousness of other interventions on the claustrum, such as those due to psychedelic drugs (stiefel et al., 2014). third objection: “but now the dmc framework is in danger of making the science of consciousness too hard! given the complexity of the brain, there is no possibility whatsoever of being able to carry out ideal/surgical interventions. if that’s the holy grail of the dmc program then the science of consciousness is doomed to failure.” reply: it is important to distinguish dmcs themselves from evidence concerning dmcs. claims about difference-makers are claims about what would happen were we to make an ideal intervention: that is (roughly speaking) an intervention on a brain region r which has an effect on e only via the r-to-e pathway, while everything else is kept fixed. a host of practical problems makes it difficult to see how an ideal intervention could ever be done. constitutive dmcs in particular require special treatment if they are to meet some plausible analogue of this condition.8 yet the search for dmcs is the search for difference-makers that would behave in the right way were we able to make ideal interventions. it is not the search for ideal interventions, and does not require them. all kinds of evidence might be relevant to establishing claims about dmcs. in the happiest scientific cases, the evidential relation is simple and direct: we establish a dmr by making an ideal intervention and seeing what happens. there are straightforward experiments in e.g. physics, or medicine, or agronomy, that admit of such manipulations. in most cases, however, we are searching for evidence about dmrs that comes from making non-ideal interventions on systems which can’t be held fixed. much of the debate is thus over how non-ideal evidence might justify claims about difference-makers. similarly, contrastive studies in neuroimaging might provide evidence about dmrs for cognition (klein, 2017), but that evidential relationships requires a fair number of background assumptions about the technical 8this is the core of the baumgartner and gebharter’s (2016) ‘fat-handedness’ objection to simple constitutive notions of difference-making. see kästner and andersen (2018) for a review of subsequent debates on fat-handedness. klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org colin klein, jakob hohwy, and tim bayne 16 details of fmri, the underlying hemodynamic response, and so on. this is not unique to the study of consciousness. economists (say) might want to know what would happen if you raise the minimum wage but kept everything else as fixed as possible. that must be extrapolated from a bunch of decidedly non-ideal policy interventions. similarly so, we suggest, in the science of consciousness. fourth objection: “the science of consciousness is in the business of providing explanations of consciousness, but difference-makers are not necessarily explanatory. how, if at all, does the dmc framework hope to explain consciousness?” reply: the question of whether dmcs are explanatory depends in part on what one requires of an explanation of consciousness. in some sense, an appeal to a difference-maker is explanatory, for difference-makers answer ‘why’ questions. q: “why did this person have a 𝐶1 experience and not a 𝐶2 experience?” a: “because they were in neural state 𝑁1. had they been in n2 then (all other things being equal) they would have had 𝐶2 experience.” of course, this answer doesn’t provide an explanation for why 𝑁1 is a difference-maker for 𝐶1 (relative to 𝐶2), for adverting to difference makers doesn’t itself explain difference-making relationships. but in this sense the dmc framework is in much the same boat as the ncc framework, for although the ncc framework invokes correlations to explain conscious states those correlations themselves remain unexplained. what the objector really has in mind when she asks for an explanation of consciousness is an explanation of why consciousness is characterized by certain dmrs rather than others. why is it the case that activation of the ffa is a difference-maker for experiences of faces (and is not a difference-maker for auditory experience)? as we see it, answering this question is not the job of the dmc framework as such but is rather the job of a theory of consciousness. deploying the dmc framework will be pivotal in the development of such a theory, but the framework is distinct from a theory of consciousness itself.9 how might the science of consciousness progress on the dmc picture? what is required is a two-step process. very broadly, stage one involves cataloguing the extent and type of difference-makers for a variety of conscious phenomena. stage two involves introducing a theory about why those difference-making relationships hold. these two stages are conceptually distinct, but are likely to be temporally intertwined in practice: theories guide our search for difference-makers, which in turn refine theories, and so on. the first stage involves cataloguing difference-makers in as many different and useful ways as possible. the dmc approach encourages breadth in at least four different ways. first, it encourages researchers to look for multiple ways to manipulate different aspects of the same experience. second, it encourages the search for 9there are intriguing remarks which suggest that generalizations themselves might be explained (see e.g. woodward, 2003, pp. 13–14) via an interventionist process. as difference-making generalizations aren’t completely invariant, one might have a picture on which the generalizations themselves can be varied in systematic ways. whether one explains a generalization or merely parameterizes an existing one is a complex question which in part on the difficult problem of variable choice (woodward, 2016). this issue is explored further in klein & barron (2020). klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org explanation in the science of consciousness 17 interventions that change a variety of conscious contents in correlated ways (as, for example, might happen with psychedelics). third, it encourages the search for more complex relationships (such as specificity and systematicity) between neural states and conscious contents. and fourth, it encourages breadth in the search for types of difference-makers on the neural end, including difference-makers at a variety of levels of organisation. stage two of the process is abductive. it involves building a theory which explains the difference-making relations discovered in stage one. there is comparatively little in the difference-making literature about how difference-making relations themselves are to be explained, and this is properly a project for a different paper. the important thing is that the shape of this explanatory story will look very different to that of stage one of the project. this is a general feature of mechanistic explanations. consider again the analogy with genetics. stage one would involve cataloguing a variety of genes for various phenotypic traits, along with their linkages and potential variants. other evidence about broader difference-making traits might be relevant here – the effects of radiation and various chemicals on overall mutationrates, for example, or coordinated influences in development. stage two involves giving a broad mechanistic story in which these individual relationships can plausibly be embedded. the story about chromosomes, dna, transcription, and so on explains why the myriad dmrs identified in stage one might plausibly hold. a similar process, we suggest, can be expected for consciousness. finding nccs is often presented as an end in itself. dmcs, by contrast, are a useful first step in building up a theory of why those dmcs hold – a theory-building process that we should expect to culminate in a systematic scientific theory of consciousness. 7 conclusion: how to run with the dmc framework consciousness has some kind of compositional structure. a total experience involves multiple phenomenal elements, many of which can occur in the context of other total experiences. (one can taste coffee whilst working in one’s office or whilst lying on the beach.) the neural substrates of consciousness also possess some kind of compositional structure, for interventions on different components of the brain can affect consciousness in different ways. it is thus tempting to suppose that there must be a straightforward mapping from the phenomenal components of consciousness to the components that make up its neural substrate. the lite-brite model endorses that thought, holding that the relationship between consciousness and its neural substrates is characterized by autonomy, uniqueness and uniformity. and, as we have seen, the ncc framework also endorses that thought insofar as it embraces the lite-brite model. klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org colin klein, jakob hohwy, and tim bayne 18 in this paper we have suggested that the relationship between the structure of consciousness and that of its neural substrates is much more likely to follow the interactionist pattern that characterizes other complex systems, in which straightforward mappings between the activity of any one component (e.g., a gene) and the presence of a high-level property (e.g., a behavioural trait) are rare. attempts to understand such systems are invariably driven not by the quest for correlations but by the search for difference-making relations. moving from nccs to dmcs not only allows us to better understand what consciousness scientists actually do, it also suggests new avenues for trying to understand how the activity of brains generates the wonder of experience. acknowledgments thanks to andrew barron for helpful discussion. work on this paper was supported by australian research council grants ft140100422 (to ck), ft 150100266 (to tb), and dp190101805 (to tb & jh). tb is also indebted to the canadian institute for advanced research (cifar) for its support. references anderson, m. l. (2014). after phrenology: neural reuse and the interactive brain. mit press. aru, j., bachmann, t., singer, w., & melloni, l. (2012). distilling the neural correlates of consciousness. neuroscience & biobehavioral reviews, 36(2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.12.003 baumgartner, m., & gebharter, a. (2016). constitutive relevance, mutual manipulability, and fat-handedness. the british journal for the philosophy of science, 67(3), 731–756. https://doi.org/10.1093/bjps/axv003 bayne, t. (2007). conscious states and conscious creatures: explanation in the scientific study of consciousness. philosophical perspectives, 21(1), 1–22. https://doi.org/10.1111/j.1520-8583.2007.00118.x bayne, t. (2010). the unity of consciousness. oxford university press. bayne, t., hohwy, j., & owen, a. m. (2016). are there levels of consciousness? trends in cognitive sciences, 20(6), 405–413. https://doi.org/10.1016/j.tics.2016.03.009 block, n. (2005). two neural correlates of consciousness. trends in cognitive sciences, 9(2), 46–52. https://doi.org/10.1016/ j.tics.2004.12.006 boly, m., massimini, m., tsuchiya, n., postle, b. r., koch, c., & tononi, g. (2017). are the neural correlates of consciousness in the front or in the back of the cerebral cortex? clinical and neuroimaging evidence. journal of neuroscience, 37(40), 9603–9613. https://doi.org/10.1523/jneurosci.3218-16.2017 chalmers, d. j. (2000). what is a neural correlate of consciousness. in t. metzinger (ed.), neural correlates of consciousness: empirical and conceptual questions (pp. 17–39). mit press. craver, c. f. (2007). explaining the brain. oxford university press. craver, c., & kaplan, d. m. (2011). towards a mechanistic philosophy of neuroscience. in s. french & j. saatsi (eds.), the bloomsbury companion to the philosophy of science (pp. 268–292). a&c black. crick, f. (1996). visual perception: rivalry and consciousness. nature, 379(6565), 485–486. https://doi.org/10.1038/379485a0 crick, f. c., & koch, c. (2005). what is the function of the claustrum? philosophical transactions. biological sciences, 360(1458), 1271–1279. https://doi.org/10.1098/rstb.2005.1661 cummins, r. (2000). “how does it work?” versus “what are the laws?”: two conceptions of psychological explanation. in f. c. keil & r. a. wilson (eds.), explanation and cognition (pp. 117–144). mit press. de graaf, t. a., hsieh, p.-j., & sack, a. t. (2012). the “correlates” in neural correlates of consciousness. neuroscience & biobehavioral reviews, 36(1), 191–197. https://doi.org/10.1016/j.neubiorev.2011.05.012 dretske, f. i. (1997). naturalizing the mind. mit press. fink, s. b. (2016). a deeper look at the “neural correlate of consciousness”. frontiers in psychology, 7, 1044. https://doi.org/ 10.3389/fpsyg.2016.01044 klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neubiorev.2011.12.003 https://doi.org/10.1093/bjps/axv003 https://doi.org/10.1111/j.1520-8583.2007.00118.x https://doi.org/10.1016/j.tics.2016.03.009 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1523/jneurosci.3218-16.2017 https://doi.org/10.1038/379485a0 https://doi.org/10.1098/rstb.2005.1661 https://doi.org/10.1016/j.neubiorev.2011.05.012 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org explanation in the science of consciousness 19 friston, k. j., price, c. j., fletcher, p., moore, c., frackowiak, r. s., & dolan, r. j. (1996). the trouble with cognitive subtraction. neuroimage, 4(2), 97–104. https://doi.org/10.1006/nimg.1996.0033 funkhouser, e. (2006). the determinable-determinate relation. noûs, 40(3), 548–569. https://doi.org/10.1111/j.14680068.2006.00623.x hill, c. s. (1991). sensations: a defense of type materialism. cambridge university press. hohwy, j. (2009). the neural correlates of consciousness: new experimental approaches needed? consciousness and cognition, 18(2), 428–438. https://doi.org/10.1016/j.concog.2009.02.006 hohwy, j., & frith, c. (2004). can neuroscience explain consciousness? journal of consciousness studies, 11(7-8), 180–198. irvine, e. (2013). consciousness as a scientific concept: a philosophy of science perspective (vol. 5). springer. janssen, a. (2019). explanations in neuroimaging research [phd thesis]. radboud university nijmegen. kaplan, d. m. (2012). how to demarcate the boundaries of cognition. biology & philosophy, 27(4), 545–570. https://doi.org/ 10.1007/s10539-012-9308-4 kästner, l., & andersen, l. m. (2018). intervening into mechanisms: prospects and challenges. philosophy compass, 13(11), e12546. https://doi.org/10.1111/phc3.12546 klein, c. (2017). brain regions as difference-makers. philosophical psychology, 30(1–2), 1–20. https://doi.org/10.1080/0951 5089.2016.1253053 klein, c., & barron, a. b. (2020). how experimental neuroscientists can fix the hard problem of consciousness. neuroscience of consciousness, 06(1). https://doi.org/10.1093/nc/niaa009 koch, c., massimini, m., boly, m., & tononi, g. (2016). neural correlates of consciousness: progress and problems. nature reviews neuroscience, 17(5), 307. https://doi.org/10.1038/nrn.2016.22 laureys, s. (2005). the neural correlate of (un) awareness: lessons from the vegetative state. trends in cognitive sciences, 9(12), 556–559. https://doi.org/10.1016/j.tics.2005.10.010 mclaughlin, b. (2007). type materialism for phenomenal consciousness. in s. schneider & m. velmans (eds.), the blackwell companion to consciousness (pp. 431–444). wiley online library. metzinger, t. (2000). neural correlates of consciousness: empirical and conceptual questions. mit press. miller, s. m. (2007). on the correlation/constitution distinction problem (and other hard problems) in the scientific study of consciousness. acta neuropsychiatrica, 19(3), 159–176. https://doi.org/10.1111/j.1601-5215.2007.00207.x miracchi, l. (2017). generative explanation in cognitive science and the hard problem of consciousness. philosophical perspectives, 31(1), 267–291. https://doi.org/doi.org/10.1111/phpe.12095 neisser, j. (2012). neural correlates of consciousness reconsidered. consciousness and cognition, 21(2), 681–690. https: //doi.org/10.1016/j.concog.2011.03.012 neisser, j. (2015). the science of subjectivity. springer. noë, a., & thompson, e. (2004). are there neural correlates of consciousness? journal of consciousness studies, 11(1), 3–28. https://www.ingentaconnect.com/content/imp/jcs/2004/00000011/00000001/1400 odegaard, b., knight, r. t., & lau, h. (2017). should a few null findings falsify prefrontal theories of conscious perception? journal of neuroscience, 37(40), 9593–9602. https://doi.org/10.1523/jneurosci.3217-16.2017 pearl, j. (2000). causality: models, reasoning and inference. cambridge university press. polger, t. w. (2006). natural minds. mit press. searle, j. r. (2000). consciousness. annual review of neuroscience, 23(1), 557–578. https://doi.org/10.1146/annurev.neuro. 23.1.557 shoemaker, s. (1981). some varieties of functionalism. philosophical topics, 12(1), 93–119. https://doi.org/10.5840/philtopi cs198112145 sporns, o. (2011). networks of the brain. mit press. sterelny, k., & kitcher, p. (1988). the return of the gene. the journal of philosophy, 85(7), 339–361. https://doi.org/10.2307/ 2026953 sternberg, s. (1969). the discovery of processing stages: extensions of donders’ method. acta psychologica, 30, 276–315. https://doi.org/10.1016/0001-6918(69)90055-9 stiefel, k. m., merrifield, a., & holcombe, a. o. (2014). the claustrum’s proposed role in consciousness is supported by the effect and target localization of salvia divinorum. frontiers in integrative neuroscience, 8. https://doi.org/10.3389/fnin t.2014.00020 tsao, d. y., & livingstone, m. s. (2008). mechanisms of face perception. annual review of neuroscience, 31, 411–437. https://doi.org/10.1146/annurev.neuro.30.051606.094238 tye, m. (1995). ten problems of consciousness: a representational theory of the phenomenal mind. mit press. klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1006/nimg.1996.0033 https://doi.org/10.1111/j.1468-0068.2006.00623.x https://doi.org/10.1111/j.1468-0068.2006.00623.x https://doi.org/10.1016/j.concog.2009.02.006 https://doi.org/10.1007/s10539-012-9308-4 https://doi.org/10.1007/s10539-012-9308-4 https://doi.org/10.1111/phc3.12546 https://doi.org/10.1080/09515089.2016.1253053 https://doi.org/10.1080/09515089.2016.1253053 https://doi.org/10.1093/nc/niaa009 https://doi.org/10.1038/nrn.2016.22 https://doi.org/10.1016/j.tics.2005.10.010 https://doi.org/10.1111/j.1601-5215.2007.00207.x https://doi.org/doi.org/10.1111/phpe.12095 https://doi.org/10.1016/j.concog.2011.03.012 https://doi.org/10.1016/j.concog.2011.03.012 https://www.ingentaconnect.com/content/imp/jcs/2004/00000011/00000001/1400 https://doi.org/10.1523/jneurosci.3217-16.2017 https://doi.org/10.1146/annurev.neuro.23.1.557 https://doi.org/10.1146/annurev.neuro.23.1.557 https://doi.org/10.5840/philtopics198112145 https://doi.org/10.5840/philtopics198112145 https://doi.org/10.2307/2026953 https://doi.org/10.2307/2026953 https://doi.org/10.1016/0001-6918(69)90055-9 https://doi.org/10.3389/fnint.2014.00020 https://doi.org/10.3389/fnint.2014.00020 https://doi.org/10.1146/annurev.neuro.30.051606.094238 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org colin klein, jakob hohwy, and tim bayne 20 woodward, j. (2003). making things happen. oxford university press. woodward, j. (2010). causation in biology: stability, specificity, and the choice of levels of explanation. biology & philosophy, 25(3), 287–318. https://doi.org/10.1007/s10539-010-9200-z woodward, j. (2016). the problem of variable choice. synthese, 193(4), 1047–1072. https://doi.org/10.1007/s11229-0150810-5 open access this article is distributed under the terms of the creative commons attributionnoncommercial 4.0 international license (https://creativecommons.org/licenses/bync/4.0/), which permits noncommercial use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. klein et al. (2020). explanation in the science of consciousness: from the neural correlates of consciousness (nccs) to the difference makers of consciousness (dmcs). philosophy and the mind sciences, 1(ii), 4. https://doi.org/10.33735/phimisci.2020.ii.60 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s10539-010-9200-z https://doi.org/10.1007/s11229-015-0810-5 https://doi.org/10.1007/s11229-015-0810-5 https://doi.org/10.33735/phimisci.2020.ii.60 https://creativecommons.org/licenses/by-nc/4.0/ https://philosophymindscience.org introduction the ncc framework the lite-brite model and the neural correlates of consciousness explanation and difference-making relations difference-makers for consciousness the difference difference-makers of consciousness make conclusion: how to run with the dmc framework on the dangers of conflating strong and weak versions of a theory of consciousness. on the dangers of conflating strong and weak versions of a theory of consciousness matthias michela (matthias.michel.curtil@gmail.com) hakwan laub (hakwan@gmail.com) abstract some proponents of the integrated information theory (iit) of consciousness profess strong views on the neural correlates of consciousness (ncc), namely that large swathes of the neocortex, cerebellum, basal ganglia, thalamus, olfactory bulb, and the so-called limbic system, are all not essential for any form of conscious experiences. we argue that this claim derives from a strong version of the theory, according to which the ncc is supposed to constitute conscious experiences. on a weaker version of the theory, iit only provides what we call a marker of consciousness. we surmise that a conflation between strong and weak versions of the theory has led these researchers to adopt definitions of ncc that are inconsistent with their own previous definitions, inadvertently betraying the promises of an otherwise fruitful empirical endeavour. keywords consciousness ∙ integrated information theory ∙ neural correlates of consciousness this article is part of a special issue on “the neural correlates of consciousness”, edited by sascha benjamin fink. 1 nccs, markers, and constituents we are not receptive to physicists trying to apply exotic physics to the brain, about which they seem to know very little, and even less about consciousness. – (crick & koch, 2003) identifying the neural correlates of consciousness (ncc) has been a central research program in consciousness science for decades. in common neurobiological acentre for philosophy of natural and social science, london school of economics and political science bdepartment of psychology & brain research institute, ucla; department of psychology & state key laboratory of brain and cognitive sciences, university of hong kong michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-5780-5702 https://orcid.org/0000-0001-8433-4232 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michel matthias and lau hakwan 2 language, the key term would have been ‘mechanisms’ rather than ‘correlates.’1 the term ‘correlates’ was carefully chosen to remain neutral on conceptual issues regarding the exact metaphysical relation between consciousness and the nccs (crick & koch, 1990). it was recognized early on that such questions are difficult, and best set aside until we have clearer answers on the ncc. in this article, we introduce a conceptual distinction between nccs, markers of consciousness, and constituents of consciousness.2 we argue that a failure to distinguish between these different concepts is limiting the progress of the ncc project. we will illustrate this point with the case of the integrated information theory (iit) of consciousness. an ncc is “the minimal set of neuronal events and mechanisms jointly sufficient for a specific conscious percept” (chalmers, 2000; koch, 2004, p. 16). that is, nccs are the minimal neural difference makers that are jointly sufficient for a mental state to be conscious, rather than unconscious. this is just another way of saying that, all other things being equal, if a mental state is unconscious, activation of the ncc should be sufficient for making it conscious. we distinguish the ncc from the constituents of consciousness. if they exist, constituents of consciousness are neural, or physical states that are identical with consciousness. this means that the constituents of consciousness should be sufficient – and necessary – for consciousness, just as h2o is the constituent of water. research has focused on finding nccs rather than constituents of consciousness because the necessity condition is generally considered too strong (chalmers, 2000). indeed, if a neural state is not only sufficient, but also absolutely necessary for a specific conscious experience to occur, it means that that experience is not multiply realizable (chalmers, 2000; michel et al., 2018; morales & lau, 2018). that is, as in the case of water being identical to h2o, we cannot replace hydrogen or oxygen with something else to produce water. h2o is the only possible recipe. this claim would, in effect, be contradictory with the widely acknowledged phenomenon of degeneracy in biology (edelman & gally, 2001; tononi et al., 1999) – which states that given a context, the same biological function can be carried out by different substrates, giving the impression of redundancy. if the same conscious experience could be realized with slightly different neuronal ensembles which may 1following illari & williamson’s ‘consensus concept’ of mechanism, a mechanism for a phenomenon is defined here as a set of “entities and activities organized in such a way that they are responsible for the phenomenon” (illari & williamson, 2012, p. 120). identifying consciousness mechanisms requires one to understand what the mechanisms do for consciousness exactly (i.e. their functions). identifying markers, correlates, or even constituents, does not require one to do so. we will not focus on mechanisms in this article (for more on mechanistic explanations in consciousness science, see miracchi (2017)). 2a constituent of consciousness is different from what is sometimes called a ‘core ncc’. a core ncc is “the part of the total ncc that distinguishes one conscious state from another – the rest of the total ncc being considered as the enabling conditions for that conscious experience” (block, 2005, p. 47). a core ncc is not necessary and sufficient for a given conscious experience – it is only sufficient when combined with some enabling conditions. by contrast, a constituent of a conscious experience is necessary and sufficient for that conscious experience to occur. michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org on the dangers of conflating strong and weak versions of a theory of consciousness 3 largely overlap but are not exactly identical, an ncc at this level cannot be considered strictly necessary for a particular experience. we will come back to this point below. lastly, we distinguish constituents and nccs from the markers of consciousness. by markers we mean general evidence that can be used to determine whether subjects have conscious mental states or not. in that sense, nccs can be used as markers of consciousness, because if a person is able to entertain a specific conscious experience, it demonstrates that the person is conscious in at least some limited sense. but not all markers of consciousness are nccs, or constituents of consciousness. these markers may be very general and might not specifically reflect the mechanisms directly responsible for consciousness. some markers of consciousness might correspond to ‘pre-requisites’ or ‘consequences’ of consciousness (aru et al., 2012). to illustrate this distinction, let us consider the fact that when one is conscious rather than in a coma, one is more capable of producing behavior, thoughts, and memory. therefore, a neural marker for sophisticated cognition may generally be used as a marker for consciousness. but such a marker may not be the minimally sufficient condition for specific conscious experiences; one may need much less for a specific single experience to arise. and such a marker may also not be necessary for conscious experiences; one might be able to entertain a specific experience without sophisticated cognition. still, it could be that, in general, one can pragmatically use a neural marker for sophisticated cognition as a marker for consciousness, as consciousness and cognition might generally correlate, at least in humans. clearly distinguishing between markers, nccs, and constituents of consciousness is important for all theories of consciousness: conflating these three notions might create confounds. for instance, the p3b wave was considered an ncc by proponents of the global workspace theory (dehaene & changeux, 2011). recent experiments, however, have shown that the p3b is neither necessary (cohen et al., 2020; pitts et al., 2014) nor sufficient (silverstein et al., 2015) for consciousness. while one can interpret the p3b as a reliable marker of consciousness in some cases, it cannot be interpreted as an ncc, or a constituent of consciousness. clearly distinguishing between markers and nccs in this case shows that an electrophysiological signal, while not being mechanistically relevant for consciousness, can remain practically useful (e.g. in clinical settings) as a marker of consciousness (faugeras et al., 2012). while we maintain that the distinction is important for all theorizing about consciousness, what follows is a case study of how the distinction plays out for a specific theory: the integrated information theory of consciousness (iit). we argue that a failure to carefully distinguish between nccs, markers, and constituents of consciousness, is at the heart of a conflation between a strong version of iit, and a weaker, empirical version. michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michel matthias and lau hakwan 4 2 empirical vs fundamental iit the recent ‘rise’ of iit is somewhat paradoxical. on the one hand, the theory has been promoted with an unusual level of enthusiasm. for example, it has been claimed that iit is “currently accepted as one of the most compelling explanations about what consciousness is”3 (italics ours), that it is a “gigantic step in the final resolution of the ancient mind-body problem” (koch, 2004), and even that it yields “a new kind of scientific spirituality.”4 on the other hand, arguments and criticisms against the theory abound (barrett & mediano, 2019; bayne, 2018; cerullo, 2015; doerig et al., 2019; pautz, 2019). to answer one of those criticisms, proponents of iit ended up acknowledging that, according to the theory, even a set of ‘inactive’ logic gates would be conscious (aaronson, 2014b, 2014a; tononi, 2014). such possibilities have been openly denounced as “untestable” by many (michel et al., 2019)5. overall, the theory has not impressed active researchers as much as non-experts outside the field (michel et al., 2018). we suspect that this striking difference in opinions is due to a conflation between two versions of the theory, which, following mediano et al. (2019), we call empirical iit and fundamental iit. while the former may have some merits, we are unsure about the latter. empirical iit is the view that measures of integrated information in brain networks – a specific subtype of complexity measure – can be used to detect states of consciousness, i.e., whether subjects have subjective experiences (such as during wakefulness, or dreams), or not (such as being in a coma, anesthetized, or dreamless-sleep). that is, empirical iit takes integrated information to be a marker of consciousness. according to this view, the choice of an exact quantitative measure of integrated information is an empirical matter. for instance, if it turns out that some measure of integrated information does a better job at categorizing states of consciousness accurately, this would count as empirical data in favor of using that measure. overall, despite some contradictory evidence (noel et al., 2019; sasai et al., 2016; tajima et al., 2015),6 there is support for empirical iit (barttfeld et al., 2015; bodart et al., 2017; casali et al., 2013; casarotto et al., 2016; demertzi et al., 2019; ferrarelli et al., 2010; rosanova et al., 2012; ruiz de miras et al., 2019; sarasso et al., 2015; tagliazucchi et al., 2013). we can consider empirical iit scientifically plausible. 3see: https://qz.com/709969/2300-years-later-platos-theory-of-consciousness-is-being-backed-upby-neuroscience/ 4https://www.huffpost.com/entry/post_b_8160914 5this point was also made in a recent letter to the nih, by dozens of researchers in the field, available online here: https://tinyurl.com/y5wokv9g. 6see http://inconsciousnesswetrust.blogspot.com/2017/08/how-to-make-iit-and-other-theoriesof.html for discussion as to why sasai et al. (2016) and tajima et al. (2015), and some other studies are considered potentially contradictory evidence against empirical iit. michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 http://inconsciousnesswetrust.blogspot.com/2017/08/how-to-make-iit-and-other-theories-of.html http://inconsciousnesswetrust.blogspot.com/2017/08/how-to-make-iit-and-other-theories-of.html https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org on the dangers of conflating strong and weak versions of a theory of consciousness 5 on the other hand, fundamental iit is the view that a specific form of complexity (integrated information) is identical with consciousness (oizumi et al., 2014). that is, fundamental iit posits that integrated information is the constituent of consciousness. specific conscious experiences are also exactly identified with specific states of a network with particular patterns of integrated information (tsuchiya et al., 2015; tsuchiya, 2017). as such, fundamental iit does not simply posit that integrated information is necessary or sufficient for consciousness, or that integrated information is a marker of consciousness. instead, consciousness is identified with a system having the features described by the ‘postulates’ of the theory. this identity claim is explicit in the writings of proponents of iit: according to iit, there is an identity between phenomenological properties of experience and informational/causal properties of physical systems (see [11] and [19] for the importance of identities for the mindbody problem). the central identity is the following: the maximally irreducible conceptual structure (mics) generated by a complex of elements is identical to its experience. (oizumi et al., 2014, p. 3) importantly, iit claims that a quale in the broad-sense is identical to a mics, generated by a particular subset of the neural system, for example, a thalamo-cortical system excluding cerebellum. in other words, iit proposes a mapping between a certain mathematical structure, which is derived from connectivity and a state of a certain subset of the neurons in the brain, and the particular quale that a subject of the brain is experiencing. (tsuchiya et al., 2015, p. 3) as fallon (2015) notes, proponents of iit also routinely offer “analogies to other fundamental physical properties. consciousness is fundamental to integrated information in the same way as it is fundamental to mass that space-time bends around it”. in summary, proponents of iit subscribe to an identity between types of experiences (shapes in ‘qualia-space’) and types of patterns of integrated information (shapes in a ‘cause-effect space’), realized in physical causal structures (tsuchiya, 2017). it follows that identifying the causal structure of a physical system, and deriving its integrated information, is sufficient to determine which type of experiences that system has, and vice versa. importantly, causal structures can be realized in various physical substrates. physical systems that are composed of different stuff can have the same causal structures and thus realize the same patterns of integrated information. as tsuchiya (2017) notes: essential relationships in iit are those between consciousness and mathematical structures derived from the physical substrates, not between consciousness and matter as is usually debated in philosophy. this means that two distinct physical substrates can generate identical consciousness. michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michel matthias and lau hakwan 6 some could argue that this aspect of iit makes the theory more similar to functionalism than identity theory, given that functionalism has historically been associated with multiple realizability (polger & shapiro, 2016). however, iit is very far from functionalism, for at least two reasons. first, there is a crucial sense in which iit does not accept multiple realizability: a given type of experience cannot be realized by different types of physical causal structures, even if those causal structures themselves can be realized in different substrates. for instance, the causal structure realizing the experience of seeing something red can be realized in a brain or in swiss cheese, and the substrate will have the experience of seeing something red as long as the right causal structure is maintained. but the same experience cannot be realized by different causal structures, even with the same substrate. two brains cannot realize the same experience by instantiating different causal structures. in that last sense, conscious experiences are not multiply realizable, according to iit. second, and more importantly, conscious experiences are not tied in any way to the realization of psychological functions. this is made clear by the fact that feedforward networks could realize the exact same functions as the human brain without being conscious, according to iit (doerig et al., 2019). what matters for consciousness is not what functions are realized – since even a set of inactive logic gates can have conscious experiences (aaronson, 2014b), but only how functions are physically realized. for this reason, proponents of fundamental iit are antifunctionalists, as tononi & koch (2015) write: “in sharp contrast to widespread functionalist beliefs, iit implies that digital computers, even if their behaviour were to be functionally equivalent to ours, and even if they were to run faithful simulations of the human brain, would experience next to nothing.”. as such, in contrast with empirical iit, which considers integrated information as a marker of consciousness, fundamental iit postulates an identity between patterns of integrated information and conscious experiences. in addition, according to fundamental iit, the exact measure of integrated information to adopt is not a matter to be determined by the data. rather, it stems from some assumptions, or axioms, that are meant to be self-evidently true (oizumi et al., 2014; tononi et al., 2016; tononi & koch, 2015). from there, one is supposed to mathematically derive the measure of integrated information that is relevant for consciousness. we will argue that fundamental iit is not in line with current scientific knowledge and practice. it may also jeopardize the ncc project. 3 conflating general markers and constituents let’s assume that some versions of the integrated information measure can be successfully used as markers of states of consciousness, as claimed by empirical iit. importantly, we have to recognize that this would be entirely compatible with the rejection of fundamental iit. michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org on the dangers of conflating strong and weak versions of a theory of consciousness 7 that is to say, data in favor of empirical iit may be compatible with other theories, such as the global workspace theory (gwt)7. according to gwt, consciousness results from the broadcast of information to a wide variety of neuro-cognitive modules through a “global workspace”. if one supposes that this global broadcast is best achieved in a network of high complexity, gwt can account for currently available evidence in support of empirical iit, without invoking the controversial conjectures of fundamental iit. also, that some measures of integrated information can be used as markers of states of consciousness does not directly support the claim that integrated information is identical with consciousness. because the exact calculation of the degree of integrated information in real biological systems is computationally challenging (barrett & mediano, 2019),8 studies currently rely on proxy measures. due to the approximate nature of the measurements, the relevant results cannot support claims concerning constituents, rather than markers. that is, they support empirical, not fundamental iit. on this note, we remark there are many different approximated measures of integrated information, and they do not necessarily empirically converge (mediano et al., 2019). given the same data fundamental iit can both be supported or falsified by these measures, depending on which approximation we choose to adopt.9 relatedly, there is currently a debate as to whether fundamental iit is empirically falsifiable at all (doerig et al., 2019). finally, and perhaps most importantly, in order to interpret the successful results of integrated information measures as support for the claim that consciousness is identical with integrated information, proponents of fundamental iit would have to rule out important confounds. for instance, a wide variety of cognitive capacities also differ between wakeful and unconscious subjects. as such, it could be that integrated information measures reflect these differences, instead of differences in consciousness per se. as a result, to use data obtained with integrated information measures of wakeful versus unconscious subjects to support fundamental iit, proponents of fundamental iit would have to provide good reasons for believing that integrated information indexes consciousness and not cognition, or any other factor that varies between wakeful and unconscious individuals. to be clear, again, this latter argument applies to fundamental iit, but not to empirical iit. as long as some integrated information measures do correlate with 7this is apparently acknowledged by some proponents of iit (koch says that “the global workspace theory and integrated information theories are not mutually exclusive”, see: https://www.livescience.com/47096-theories-seek-to-explain-consciousness.html) 8as remarked by barrett & mediano (2019), this is because “the computation time required to compute [integrated information] grows faster than exponentially with the number of system components” (p.1). 9see https://jakerhanson.weebly.com/blog/my-graduate-experience-with-integrated-informationtheory-iit for a useful discussion on the difficulty in pinning down a complexity measure for this purpose. michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.livescience.com/47096-theories-seek-to-explain-consciousness.html https://jakerhanson.weebly.com/blog/my-graduate-experience-with-integrated-information-theory-iit https://jakerhanson.weebly.com/blog/my-graduate-experience-with-integrated-information-theory-iit https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michel matthias and lau hakwan 8 differences in states of consciousness, whatever the reason behind this correlation, those measures can be used as markers of consciousness. it is an entirely different matter, however, to claim that integrated information should be identified with consciousness, as do proponents of fundamental iit. therefore, if we do not confuse markers and constituents of consciousness, it should be clear that reasons for finding empirical iit appealing generally do not extend to fundamental iit. 4 conflating constituents and nccs to make claims about the constituents of consciousness, we need to do more than just ruling out some confounds. identity relationships like that between water and h2o are very strong statements to make. to do so, we first need to have a very good understanding of the relevant substrates (hydrogen and oxygen in this case), their causal properties, how they interact, as well as to make sure that they are necessary. however, in the case of fundamental iit, it is unclear what kind of physical states are supposed to be identical to consciousness. if the physical substrate itself is not clearly identified, iit is very far from establishing a convincing identity statement. indeed, what substrate is supposed to be identical with consciousness, according to fundamental iit? according to the theory, the measure of integrated information to be assessed concerns some ‘nodes’ of a physical system that are connected with each other. these nodes are either on, or off. they are either connected to other nodes or not. of course, we already know that neurons are much more complex than simple on/off nodes, as they show degrees of intensity of firing (rate), different dynamics in firing, different types of connections with each other (kandel et al., 2012). as such, proponents of fundamental iit acknowledge that these ‘nodes’ likely operate at a “different level.”10 that is, they are most probably not to be identified with necessarily neurons, as one may intuitively think. as a result, it is unclear what the substrate of consciousness is supposed to be, according to fundamental iit. this has the downside that it is not clear where researchers should look when trying to validate or invalidate the claims of fundamental iit. but let us assume that some – possibly sub-neuronal – substrate identified by fundamental iit does correlate with consciousness. even so, such correlation will 10there is a sense in which the theory captures more than simple on/off states in the nodes, because formally the theory applies to general stochastic systems that are parameterised by probability distributions and transition matrices. but ultimately, even with stochasticity, the possible outcomes are still on versus off. this does not capture the complex known dynamics of neurons. accordingly, the intuitive examples given to explain the theory also tend to concern simple binary on/off nodes. the fact that these nodes are not really intended to plausibly model neurons has been confirmed in personal communications with the proponents of the theory. personal communications with andrew haun, and masafumi oizumi. see a conversation between hakwan lau, masafumi oizumi, and richard brown, available here: https://www.youtube.com/watch?v=8ap6yqrwan0. discussion of this point is roughly between 21m38s – 31m28s. michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.youtube.com/watch?v=8ap6yqrwan0 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org on the dangers of conflating strong and weak versions of a theory of consciousness 9 only support that these constituents may be candidates for the ncc, but not that they are identical to conscious experiences. to make the latter claim, one would need to show that they are absolutely necessary, just like we cannot replace oxygen or hydrogen with something else if we want to produce water. this kind of strong identity relationships are unlikely to be found in most areas of biology, and neuroscience in particular, in which we generally try to identify mechanisms (craver, 2007). indeed, if a given neural state n is necessary for a given conscious experience c, c cannot obtain unless n does. this implies that c could not be realized by a very slightly different neural state, n2. let’s say that n involves tens of thousands of ‘nodes’. n2 would be considered different in this context even if only the connection between two out of these nodes were modified, with everything else being identical to n.11 this would be broadly inconsistent with the phenomenon of biological degeneracy, which is particularly prominent in highly complex systems – a point with which proponents of iit should be familiar (edelman & gally, 2001; tononi et al., 1999). we therefore advocate maintaining the traditional definition of ncc. according to this definition, an ncc is a minimal set of neural activity sufficient for having a conscious experience (chalmers, 2000). it differs from the ‘non-traditional’ notion used by proponents of iit – as we show in the next section, which implies identity between physical causal structures and conscious experiences. if we are to make such a radical modification of the definition, to move from mere correlation to absolute identity, we should do so explicitly. unfortunately, we suspect this change in definition has recently slipped into current debates tacitly. 5 ncc confused one of the most striking claims made by proponents of iit, beside panpsychism (tononi, 2014), may be their views on the ncc. in particular, christof koch and colleagues have made the strong claim that most areas outside of a putative region in the ‘posterior cortex’ are not home to the ncc. the excluded regions presumably include the insula, amygdala, olfactory bulb, basal ganglia, thalamus, and different parts of the prefrontal cortex (pfc). in particular, this is meant to concern almost all conscious experiences, not just conscious perception (koch et al., 2016; tononi et al., 2016). 11this is not to say for any n and n2, the two must always lead to the same outcome on consciousness. the point about degeneracy is not that it happens 100% of the time. rather, that it can sometimes happen at all means that the identity claim involving necessity cannot be logically correct. chalmers (2000) discussed related cases, as well as other types of ‘redundancy’. overall, there may be more complexity to these issues here. for example, when we usually have two neural states, each independently sufficient for a subjective experience, if one state is abolished, one can consider the ‘background conditions’ changed, in which case, no strong definite predictions could be made. but in any case, it is clear that any reasonable and consistent interpretation of the original definition of ncc is inconsistent with the quote given in footnote 12. michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michel matthias and lau hakwan 10 we remark that this claim may seem grossly incompatible with standard textbook knowledge, especially regarding conscious experiences of hunger, emotions, pain, intentions, thoughts, etc. but perhaps analyzing the case of the role of pfc for conscious visual perception is more illuminating still, because koch himself used to hold exactly the opposite view. here, we suggest that his more recent view on this matter is likely the result of a new commitment to the search for the constituents of consciousness, instead of the search for the nccs. curiously, koch’s new justification for excluding pfc from the ncc is largely based on the very same data that previously led him to accept that the pfc was an ncc for conscious visual experiences. indeed, crick & koch (crick & koch, 1998, p. 103) hypothesized that patients with bilateral lesions to the pfc may be able to respond unconsciously to visual stimuli, without being conscious of those stimuli. however, some well-known cases of bilateral lesions to the pfc had long been taken as indicating that patients with these lesions were not blind (brickner, 1952). at that time, koch concluded that these cases were not decisive for settling the debate on whether pfc was a neural correlate of consciousness (crick & koch, 1998, p. 103; koch, 2004). to our minds, these cases remain as indecisive as they were two decades ago, in part because the completeness of these lesions remains a matter of dispute (odegaard et al., 2017). meanwhile, it is now known that even in unilateral pfc lesions, there are in fact specific perceptual deficits (fleming et al., 2014). but more importantly, in recent correspondence, koch insists that the lack of complete abolishment of conscious perception in these unilateral cases is evidence for writing off the entire pfc. specifically, according to his current thinking, disruption of any part of the ncc should “necessarily” cause changes in conscious experience.12 this ‘new’ definition of the ncc, involving necessity, is in direct contradiction with the original definition, as we explained in the opening section. for example, a group of neurons in the right pfc may be on their own minimally sufficient for a conscious experience to occur. but once lesioned, neurons in the left pfc may take over to perform the same function. this kind of dynamic reorganization of function in the pfc has also been empirically demonstrated (voytek et al., 2010). however, if one is committed to fundamental iit, specifically the claim of neural identity, one may well hold that degeneracy does not apply to consciousness. this seems to be the most charitable interpretation of the inconsistency between koch’s previous and current views on the ncc. so, conflating the constituents and the nccs probably has already had an impact on how current research and debates are conducted. 12 in an email correspondence dated june 24, 2018, koch wrote: “any change in this ncc (via a different stimulus, or causal intervention such as tms, optogenetics, drugs etc) will, of necessity, change the character of the experience (including having no experience). if the background conditions change but the ncc does not, the experience will likewise not change.” (italics ours). quoted with explicit permission. michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org on the dangers of conflating strong and weak versions of a theory of consciousness 11 6 concluding remarks we have compared two versions of a theory of consciousness. while there is some empirical support for the weaker version (empirical iit), going from there to a much stronger version (fundamental iit) seems to require an unscientific leap of faith. the points we have made here are not specific to iit per se. in general we should not make weighty claims beyond what is warranted by evidence. in analyzing the case of iit, we also illustrate two culprits limiting progress in consciousness science. the first mistake is to interpret evidence in favor of using a neural state as a marker of consciousness as evidence for the mechanistic relevance of this neural state for consciousness. this point applies specifically to the study of states of consciousness, in which we might often be too quick in identifying markers of consciousness with nccs. the second mistake is to confound nccs with neural constituents of consciousness. as we have seen, doing so could lead to the rejection of the well-known biological phenomenon of degeneracy, and to controversial interpretations of lesion studies. we would do well not to let our theoretical commitments derail the ncc project. acknowledgments we thank two anonymous reviewers for their comments. we thank david chalmers, christof koch, and masafumi oizumi for helpful discussions. references aaronson, s. (2014a). giulio tononi and me: a phi-nal exchange. https://www.scottaaronson.com/blog/?p=1823 aaronson, s. (2014b). why i am not an integrated information theorist (or, the unconscious expander). www.scottaaronson. com/blog/?p=1799 aru, j., bachmann, t., singer, w., & melloni, l. (2012). distilling the neural correlates of consciousness. neuroscience and biobehavioral reviews, 36(2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.12.003 barrett, a. b., & mediano, p. a. m. (2019). the phi measure of integrated information is not well-defined for general physical systems. journal of consciousness studies, 26(1-2). barttfeld, p., uhrig, l., sitt, j. d., sigman, m., & jarraya, b. (2015). signature of consciousness in the dynamics of restingstate brain activity. proceedings of the national academy of sciences, 112(37), e5219–e5220. https://doi.org/10.1073/pn as.1515029112 bayne, t. (2018). on the axiomatic foundations of the integrated information theory of consciousness. neuroscience of consciousness, 2018(1), 1–8. https://doi.org/10.1093/nc/niy007 block, n. (2005). two neural correlates of consciousness. trends in cognitive sciences, 9(2), 46–52. https://doi.org/10.1016/ j.tics.2004.12.006 bodart, o., amico, e., wannez, s., gomez, f., casarotto, s., rosanova, m., casali, a. g., gosseries, o., laureys, s., massimini, m., & martens, g. (2017). global structural and effective connectivity in patients with chronic disorders of consciousness. brain stimulation: basic, translational, and clinical research in neuromodulation, 10(2), 353. https: //doi.org/10.1016/j.brs.2017.01.035 brickner, r. m. (1952). brain of patient a. after bilateral frontal lobectomy; status of frontal-lobe problem. archives of neurology and psychiatry, 68(3), 293–313. https://doi.org/10.1001/archneurpsyc.1952.02320210003001 casali, a. g., gosseries, o., rosanova, m., boly, m., sarasso, s., casali, k. r., casarotto, s., bruno, m.-a., laureys, s., tononi, g., & massimini, m. (2013). a theoretically based index of consciousness independent of sensory processing and michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.scottaaronson.com/blog/?p=1823 www.scottaaronson.com/blog/?p=1799 www.scottaaronson.com/blog/?p=1799 https://doi.org/10.1016/j.neubiorev.2011.12.003 https://doi.org/10.1073/pnas.1515029112 https://doi.org/10.1073/pnas.1515029112 https://doi.org/10.1093/nc/niy007 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1016/j.brs.2017.01.035 https://doi.org/10.1016/j.brs.2017.01.035 https://doi.org/10.1001/archneurpsyc.1952.02320210003001 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michel matthias and lau hakwan 12 behavior. science translational medicine, 5(198), 198ra105 lp–198ra105. http://stm.sciencemag.org/content/5/198/198 ra105.abstract casarotto, s., comanducci, a., rosanova, m., sarasso, s., fecchio, m., napolitani, m., pigorini, a., g. casali, a., trimarchi, p. d., boly, m., gosseries, o., bodart, o., curto, f., landi, c., mariotti, m., devalle, g., laureys, s., tononi, g., & massimini, m. (2016). stratification of unresponsive patients by an independently validated index of brain complexity. annals of neurology, 80(5), 718–729. https://doi.org/10.1002/ana.24779 cerullo, m. a. (2015). the problem with phi: a critique of integrated information theory. plos computational biology, 11(9), 1–12. https://doi.org/10.1371/journal.pcbi.1004286 chalmers, d. (2000). what is a neural correlate of consciousness ? in t. metzinger (ed.), neural correlates of consciousness: empirical and conceptual questions (pp. 17–39). mit press. cohen, m. a., ortego, k., kyroudis, a., & pitts, m. (2020). distinguishing the neural correlates of perceptual awareness and postperceptual processing. journal of neuroscience, 40(25), 4925–4935. https://doi.org/10.1523/jneurosci.012020.2020 craver, c. (2007). explaining the brain: mechanisms and the mosaic unity of neuroscience. oxford university press. crick, f., & koch, c. (1990). towards a neurobiological theory of consciousness. seminars in the neurosciences, 2(263-275), 203. crick, f., & koch, c. (1998). consciousness and neuroscience. cerebral cortex, 8(2), 97–107. crick, f., & koch, c. (2003). a framework for consciousness. nature neuroscience, 6(2), 119–126. https://doi.org/10.1038/nn 0203-119 dehaene, s., & changeux, j. p. (2011). experimental and theoretical approaches to conscious processing. neuron, 70(2), 200–227. https://doi.org/10.1016/j.neuron.2011.03.018 demertzi, a., martial, c., demertzi, a., tagliazucchi, e., dehaene, s., deco, g., barttfeld, p., raimondo, f., martial, c., fernández-espejo, d., rohaut, b., voss, h. u., schiff, n. d., owen, a. m., laureys, s., naccache, l., & sitt, j. d. (2019). human consciousness is supported by dynamic complex patterns of brain signal coordination. science, 5(2), 1–12. https://doi.org/10.1126/sciadv.aat7603 doerig, a., schurger, a., hess, k., & herzog, m. h. (2019). the unfolding argument: why iit and other causal structure theories cannot explain consciousness. consciousness and cognition, 72, 49–59. edelman, g. m., & gally, j. a. (2001). degeneracy and complexity in biological systems. proceedings of the national academy of sciences, 98(24), 13763 lp–13768. https://doi.org/10.1073/pnas.231499798 fallon, f. (2015). the integrated information theory of consciousness. in j. f. & b. dowden (ed.), the internet encyclopedia of philosophy. https://iep.utm.edu/int-info/ faugeras, f., rohaut, b., weiss, n., bekinschtein, t., galanaud, d., puybasset, l., bolgert, f., sergent, c., cohen, l., dehaene, s., & naccache, l. (2012). event related potentials elicited by violations of auditory regularities in patients with impaired consciousness. neuropsychologia, 50(3), 403–418. https://doi.org/10.1016/j.neuropsychologia.2011.12.015 ferrarelli, f., massimini, m., sarasso, s., casali, a., riedner, b. a., angelini, g., tononi, g., & pearce, r. a. (2010). breakdown in cortical effective connectivity during midazolam-induced loss of consciousness. proceedings of the national academy of sciences, 107(6), 2681 lp–2686. https://doi.org/10.1073/pnas.0913008107 fleming, s. m., ryu, j., golfinos, j. g., & blackmon, k. e. (2014). domain-specific impairment in metacognitive accuracy following anterior prefrontal lesions. brain, 137(pt 10), 2811–2822. https://doi.org/10.1093/brain/awu221 illari, p. m., & williamson, j. (2012). what is a mechanism? thinking about mechanisms across the sciences. european journal for philosophy of science, 2(1), 119–135. https://doi.org/10.1007/s13194-011-0038-2 kandel, e. r., schwartz, j. h., jessell, t. m., siegelbaum, s. a., & hudspeth, a. j. (2012). principles of neural science. mcgrawhill education. koch, c. (2004). the quest for consciousness: a neuroscientific approach. roberts & co. koch, c., massimini, m., boly, m., & tononi, g. (2016). neural correlates of consciousness: progress and problems. nature reviews neuroscience, 17(5), 307–321. https://doi.org/10.1038/nrn.2016.22 mediano, p. a. m., seth, a. k., & barrett, a. b. (2019). measuring integrated information: comparison of candidate measures in theory and simulation. entropy, 21(1), 1–30. https://doi.org/10.3390/e21010017 michel, m., beck, d., block, n., blumenfeld, h., brown, r., carmel, d., carrasco, m., chirimuuta, m., chun, m., cleeremans, a., dehaene, s., fleming, s. m., frith, c., haggard, p., he, b. j., heyes, c., goodale, m. a., irvine, l., kawato, m., … yoshida, m. (2019). opportunities and challenges for a maturing science of consciousness. nature human behaviour, 3(2), 104–107. https://doi.org/10.1038/s41562-019-0531-8 michel, m., fleming, s. m., lau, h., lee, a. l. f., martinez-conde, s., passingham, r. e., peters, m. a. k., rahnev, d., sergent, c., & liu, k. (2018). an informal internet survey on the current state of consciousness science. frontiers in psychology, 9, 2134. https://www.frontiersin.org/article/10.3389/fpsyg.2018.02134 michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 http://stm.sciencemag.org/content/5/198/198ra105.abstract http://stm.sciencemag.org/content/5/198/198ra105.abstract https://doi.org/10.1002/ana.24779 https://doi.org/10.1371/journal.pcbi.1004286 https://doi.org/10.1523/jneurosci.0120-20.2020 https://doi.org/10.1523/jneurosci.0120-20.2020 https://doi.org/10.1038/nn0203-119 https://doi.org/10.1038/nn0203-119 https://doi.org/10.1016/j.neuron.2011.03.018 https://doi.org/10.1126/sciadv.aat7603 https://doi.org/10.1073/pnas.231499798 https://iep.utm.edu/int-info/ https://doi.org/10.1016/j.neuropsychologia.2011.12.015 https://doi.org/10.1073/pnas.0913008107 https://doi.org/10.1093/brain/awu221 https://doi.org/10.1007/s13194-011-0038-2 https://doi.org/10.1038/nrn.2016.22 https://doi.org/10.3390/e21010017 https://doi.org/10.1038/s41562-019-0531-8 https://www.frontiersin.org/article/10.3389/fpsyg.2018.02134 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org on the dangers of conflating strong and weak versions of a theory of consciousness 13 miracchi, l. (2017). generative explanation in cognitive science and the hard problem of consciousness. philosophical perspectives, 31(1), 267–291. https://doi.org/10.1111/phpe.12095 morales, j., & lau, h. (2018). the neural correlates of consciousness. in u. kriegel (ed.), the oxford handbook of the philosophy of consciousness. oxford university press. noel, j.-p., ishizawa, y., patel, s. r., eskandar, e. n., & wallace, m. t. (2019). leveraging non-human primate multisensory neurons and circuits in assessing consciousness theory. biorxiv, 584516. https://doi.org/10.1101/584516 odegaard, b., knight, r. t., & lau, h. (2017). should a few null findings falsify prefrontal theories of conscious perception? the journal of neuroscience, 37(40), 9593–9602. https://doi.org/10.1101/122267 oizumi, m., albantakis, l., & tononi, g. (2014). from the phenomenology to the mechanisms of consciousness: integrated information theory 3.0. plos computational biology, 10(5). https://doi.org/10.1371/journal.pcbi.1003588 pautz, a. (2019). what is the integrated information theory of consciousness? a catalogue of questions. journal of consciousness studies, 26(1-2). pitts, m. a., padwal, j., fennelly, d., martinez, a., & hillyard, s. a. (2014). gamma band activity and the p3 reflect postperceptual processes, not visual awareness. neuroimage, 101, 337–350. polger, t. w., & shapiro, l. a. (2016). the multiple realization book. oxford university press. rosanova, m., gosseries, o., casarotto, s., boly, m., casali, a. g., bruno, m.-a., mariotti, m., boveroux, p., tononi, g., laureys, s., & massimini, m. (2012). recovery of cortical effective connectivity and recovery of consciousness in vegetative patients. brain, 135(4), 1308–1320. https://doi.org/10.1093/brain/awr340 ruiz de miras, j., soler, f., iglesias-parro, s., ibáñez-molina, a. j., casali, a. g., laureys, s., massimini, m., esteban, f. j., navas, j., & langa, j. a. (2019). fractal dimension analysis of states of consciousness and unconsciousness using transcranial magnetic stimulation. computer methods and programs in biomedicine, 175, 129–137. https://doi.org/10.1 016/j.cmpb.2019.04.017 sarasso, s., boly, m., napolitani, m., gosseries, o., charland-verville, v., casarotto, s., rosanova, m., casali, a. g., brichant, j.-f., boveroux, p., rex, s., tononi, g., laureys, s., & massimini, m. (2015). consciousness and complexity during unresponsiveness induced by propofol, xenon, and ketamine. current biology, 25(23), 3099–3105. https://doi.org/10.1 016/j.cub.2015.10.014 sasai, s., boly, m., mensen, a., & tononi, g. (2016). functional split brain in a driving/listening paradigm. proceedings of the national academy of sciences, 113(50), 14444 lp–14449. https://doi.org/10.1073/pnas.1613200113 silverstein, b. h., snodgrass, m., shevrin, h., & kushwaha, r. (2015). p3b, consciousness, and complex unconscious processing. cortex, 73, 216–227. https://doi.org/10.1016/j.cortex.2015.09.004. tagliazucchi, e., wegner, f. von, morzelewski, a., brodbeck, v., jahnke, k., & laufs, h. (2013). breakdown of long-range temporal dependence in default mode and attention networks during deep sleep. proceedings of the national academy of sciences, 110(38), 15419 lp–15424. https://doi.org/10.1073/pnas.1312848110 tajima, s., yanagawa, t., fujii, n., & toyoizumi, t. (2015). untangling brain-wide dynamics in consciousness by crossembedding. plos computational biology, 11(11), 1–28. https://doi.org/10.1371/journal.pcbi.1004537 tononi, g. (2014). why scott should stare at a blank wall and reconsider (or, the conscious grid). https://www.scottaaronson. com/blog/?p=1823 tononi, g., boly, m., massimini, m., & koch, c. (2016). integrated information theory: from consciousness to its physical substrate. nature reviews neuroscience, 17(7), 450–461. https://doi.org/10.1038/nrn.2016.44 tononi, g., & koch, c. (2015). consciousness: here ,there ,and everywhere? philosophical transactions of the royal society of london. series b, biological sciences, 31(1), 12–19. https://doi.org/10.1111/var.12057.digital tononi, g., sporns, o., & edelman, g. m. (1999). measures of degeneracy and redundancy in biological networks. proceedings of the national academy of sciences, 96(6), 3257 lp–3262. https://doi.org/10.1073/pnas.96.6.3257 tsuchiya, n. (2017). ”what is it like to be a bat ?” — a pathway to the answer from the integrated information theory. philosophy compass, 12, 1–13. https://doi.org/10.1111/phc3.12407 tsuchiya, n., taguchi, s., & saigo, h. (2015). using category theory to assess the relationship between consciousness and integrated information theory. neuroscience research, 107, 1–7. https://doi.org/10.1016/j.neures.2015.12.007 voytek, b., davis, m., yago, e., barcelo, f., vogel, e. k., & knight, r. t. (2010). dynamic neuroplasticity after human prefrontal cortex damage. neuron, 68(3), 401–409. https://doi.org/10.1016/j.neuron.2010.09.018 michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1111/phpe.12095 https://doi.org/10.1101/584516 https://doi.org/10.1101/122267 https://doi.org/10.1371/journal.pcbi.1003588 https://doi.org/10.1093/brain/awr340 https://doi.org/10.1016/j.cmpb.2019.04.017 https://doi.org/10.1016/j.cmpb.2019.04.017 https://doi.org/10.1016/j.cub.2015.10.014 https://doi.org/10.1016/j.cub.2015.10.014 https://doi.org/10.1073/pnas.1613200113 https://doi.org/10.1016/j.cortex.2015.09.004. https://doi.org/10.1073/pnas.1312848110 https://doi.org/10.1371/journal.pcbi.1004537 https://www.scottaaronson.com/blog/?p=1823 https://www.scottaaronson.com/blog/?p=1823 https://doi.org/10.1038/nrn.2016.44 https://doi.org/10.1111/var.12057.digital https://doi.org/10.1073/pnas.96.6.3257 https://doi.org/10.1111/phc3.12407 https://doi.org/10.1016/j.neures.2015.12.007 https://doi.org/10.1016/j.neuron.2010.09.018 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michel matthias and lau hakwan 14 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. michel, m., & lau, h. (2020). on the dangers of conflating strong and weak versions of a theory of consciousness. philosophy and the mind sciences, 1(ii), 8. https://doi.org/10.33735/phimisci.2020.ii.54 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.54 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org nccs, markers, and constituents empirical vs fundamental iit conflating general markers and constituents conflating constituents and nccs ncc confused concluding remarks generality and content-specificity in the study of the neural correlates of perceptual consciousness. generality and content-specificity in the study of the neural correlates of perceptual consciousness tomáš marvana (marvan@flu.cas.cz) michal polákb (miha@kfi.zcu.cz) abstract the present paper was written as a contribution to ongoing methodological debates within the ncc project. we focus on the neural correlates of conscious perceptual episodes. our claim is that the ncc notion, as applied to conscious perceptual episodes, needs to be reconceptualized. it mixes together the processing related to the perceived contents and the neural substrate of consciousness proper, i.e. mechanisms making the perceptual contents conscious. we thus propose that the perceptual ncc be divided into two constitutive subnotions. the main theoretical idea that emerges as a consequence of this reconceptualization is that the neural correlate of a perceptual episode is formed in the neural interaction between content-processing and consciousness-conferring mechanisms. the paper elaborates this distinction, marshals some initial arguments in its favour, and tests it against some of the most debated theories of consciousness. keywords content-specific ncc ∙ general neural correlate of consciousness ∙ neural correlates of consciousness ∙ perceptual experience ∙ theories of perceptual consciousness this article is part of a special issue on “the neural correlates of consciousness”, edited by sascha benjamin fink. 1 introduction the last two decades have brought an avalanche of empirical studies of the neural correlates of consciousness (nccs). important new findings are reported almost on a daily basis (for good overviews of recent developments see koch et al., 2016; ainstitute of philosophy of the czech academy of sciences, prague, czech republic bdepartment of philosophy, university of west bohemia, pilsen, czech republic marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-8487-5394 https://orcid.org/0000-0001-9243-1753 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tomáš marvan and michal polák 2 förster et al., 2020; michel & morales, 2020; tsuchiya et al., 2015). parallel to this burgeoning research programme runs a string of methodological and conceptual studies that are informed by this programme and strive to make its central notions precise and its methods more fruitful. to name just the most influential of them, chalmers (2000) laid down the canonical definition of an ncc as the minimal neural system the activation of which is sufficient for conscious experience, and addressed a number of important methodological issues of the ncc studies. miller (2015) highlighted the distinction between correlation and constitution in ncc studies and pointed out the limitations of the framework of minimally sufficient nccs. hohwy (2009) emphasized the differences between contentand level-based approaches to ncc research. bayne et al. (2016) found the traditional unidimensional level-based approaches to consciousness untenable and sketched a multidimensional account of global states of consciousness. aru, bachmann, et al. (2012) urged the need to distinguish the prerequisites and consequences of conscious perception from the ncc proper. finally, haynes (2009) and sandberg et al. (2014) argued for investigating the nccs with multivariate decoding of neural activity patterns. this method, they claim, may establish a tight link between types of conscious experiences and neural processing in distributed brain locations. the present paper was written as a contribution to this ongoing methodological debate within the ncc project. the notion of ncc is equally applied in research on levels of consciousness (more precisely called “global states of consciousness”) and in the studies of contentful conscious mental states (hohwy, 2009). in the research on levels of consciousness, attempts are made to identify the neural underpinnings of various levels of consciousness such as being alert, dreaming, or being in a minimal conscious state (but see mckilliam (2020), this volume). the other main usage of the ncc concept concerns neural processes correlating with conscious mental states, especially with conscious perceptual states such as seeing a bright sunset. our focus in this paper is entirely on the neural correlates in this latter sense. we are interested in the neural correlates of conscious perceptual episodes that come and go when the level of consciousness remains more or less constant. our claim is that the ncc notion, as applied to conscious perceptual states, needs to be reconceptualized. it mixes together the processing related to the perceived contents and the neural substrate of consciousness proper, i.e. mechanisms making the perceptual contents conscious. we thus propose, in section 2 of the paper, that the perceptual ncc be divided into two constitutive subnotions. the first subnotion covers the content-specific side of the perceptual ncc. this subnotion harbors some of the most vexing puzzles of consciousness science concerning the nature and material realization of consciously perceived contents. the central topic of this paper, though, is the second subnotion, that of the neural process or processes making the perceptual contents, clustered in a perceptual state, conscious. the division of the perceptual ncc into the two subnotions is not a priori but inspired and informed by perceptual ncc research. that being said, not evmarvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org generality and content-specificity in ncc research 3 ery theorist of consciousness will endorse the reconceptualization we propose. in section 3 we selectively review the theories of consciousness that we believe are supporting the division as well as those that do not sit easily with it. section 4 puts forward two central arguments for endorsing the division. the section further motivates the idea that from the neural point of view, an episode of conscious perception arises in the neural interaction between the content-processing and consciousness-conferring mechanisms. the concluding section 5 sketches some advantages of the proposed reconceptualization. 2 distilling the general neural correlate of perceptual consciousness from the neural correlates of perceptual contents applied to episodes of conscious perception, the concept of ncc typically concerns neural processes systematically correlating with a particular type of perceptual experience. that is to say, the perceptual ncc is typically a content-specific correlate. we may speak of the ncc of seeing a face or the ncc of the smell of a rose. the centrality of this content-specific ncc concept can be illustrated by numerous examples. for instance, crick (1994, p. 9) speaks about the neural correlate of seeing red. the same emphasis on content-specificity is present in the following formulations: “a neural correlate of consciousness is a specific pattern of brain activity that correlates with particular conscious experiences” (rees et al., 2002, p. 261); “the minimal set of neuronal events and mechanisms jointly sufficient for a specific conscious percept” (koch, 2004, p. 16); “content-specific ncc are the neural mechanisms specifying particular phenomenal contents within consciousness, such as colors, faces, places, or thoughts” (boly et al., 2017, p. 9604; see also koch et al., 2016, p. 308; aru, bachmann, et al., 2012, p. 738). even one of chalmers’ two famous definitions of ncc, commonly accepted in the field as the definitions of ncc, targets content-specific mechanisms expressing the particular types of conscious percepts: “an ncc (for content) is a minimal neural representational system n such that representation of a content in n is sufficient, under conditions c, for representation of that content in consciousness” (chalmers, 2000, p. 31). because the notion of ncc is also used by researchers working on levels of consciousness, it would be desirable to have a specific label for the content-specific perceptual ncc. neural correlate of conscious (perceptual) content (“nccc” for short) suggests itself. the ncccs for particular types of conscious perceptual states differ in their neural details. these differences are reflected in the experienced differences between the percepts. there is an indefinite variety of ncccs, corresponding to the variety of perceptual contents we may consciously experience. in all this diversity, though, the search for the neural correlates of consciousness strives to marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tomáš marvan and michal polák 4 identify a common thread. the pivotal question of consciousness studies is what makes any particular perceptual content conscious. we believe that to address the pivotal question is to take seriously the possibility that the nccc is a composite. at its core are the mechanisms preparing the contents we experience. let us call these content mechanisms neural correlates of (perceptual) content (“ncc” for short). the ncc is the part of the nccc that is entirely responsible for the differences between experienced contents – for example, between the conscious seeing of green and the conscious seeing of blue.1 the second part of the nccc is the topic of this paper. it is the neural mechanism making the perceptual contents conscious. in contrast to the ever-varying processes subsumed under the ncc label, this mechanism is shared across different perceptual ncccs. whenever this mechanism is activated, and the content-specific processing is in place, conscious perceptual states are present as well. neural activity correlating with a conscious perception of an apple includes both the activation of content mechanisms presenting the visual features of an apple (ncc) and the modality nonspecific process(es) making the bundle of content features of the apple conscious. similarly for all other particular conscious experiences across all sensory modalities. we may as well call such common mechanism the mechanism of perceptual consciousness itself. let us label this mechanism the general neural correlate of (perceptual) consciousness (“gncc” for short). this content-nonspecific, domain-general neural process is no less important than the content-preparing mechanisms. in fact, the search for this universal neural mechanism is arguably the ultimate goal of perceptual ncc research.2 it is plausible that the science of consciousness will only isolate the gncc by first amassing a great deal of information about the contentspecific ncccs – both in a single sensory modality as well as across modalities. but mere gathering of more and more content-specific data won’t do. the next step, we submit, is to isolate the gncc. this additional step, though, is rarely made by consciousness theorists. on the contrary, as the definitions of the neural correlates of perceptual consciousness cited above demonstrate, the nccc is usually 1in this respect, ncc is similar to block’s (2005, p. 47) notion of core phenomenal ncc, which in turn was inspired by chalmers’ core ncc (2000, p. 26). core ncc contains the presumed core processes correlating with experiences. core ncc is a part of the total ncc, the overall brain activity sufficient for any conscious perceptual episode. as the total ncc contains more neural activity than is directly relevant for a particular perceptual episode (bayne & hohwy, 2013, p. 25; chalmers, 2000, p. 26), ncc research is predominantly focused on the core ncc. the notion of the core ncc emphasizes the qualitative differences between manifestations of such core processes in various conscious states. this is clear not just in block (2005) and chalmers (2000), but also in other authors writing about the core ncc, such as keaton (2015). 2for reasons of space, we restricted our attention here to perceptual contents and do not make a parallel case for other types of experiential contents. however, we suppose that the idea of a gncc could be extended to other types of phenomenal states, such as emotions, thoughts, metacognition, dreams, working-memory contents etc. all conscious experiences, whether perceptual, cognitive, or affective, could share the same type of the gncc mechanism (implemented in different parts of the brain). marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org generality and content-specificity in ncc research 5 pictured as an undifferentiated whole in which content mechanisms are fused with consciousness conferring mechanism. the gncc notion will hopefully become more perspicuous when compared with two somewhat related notions. the first is what koch et al. (koch et al., 2016) call full ncc. the full ncc is located in the “posterior hot zone” comprising parietal, occipital and temporal areas. it is “the neural substrate supporting experience in general, irrespective of its specific content” (koch et al., 2016, p. 308). the term appears to have the same meaning as gncc. however, on closer look, the two concepts are quite different. the full ncc is defined as a “union of the sets of content-specific ncc for all possible contents of experience” (ibid.; see also the same claim in boly et al. (2017), esp. in fig. 1 on p. 9604). the full ncc is thus a union of all possible content-specific neural patterns, all possible ncccs. the gncc is not a union at all. not being involved in the production of perceptual contents, the gncc does not implement representational aspects. further, even if the content mechanisms were all located in the posterior hot zone, the gncc might extend into other regions. the second notion, fink’s (2016, p. 4) “neural correlate of general consciousness”, comes closer to what we have in mind. consider the following passage: [w]e may talk of broader and broader types, e.g., red experiences, color experiences, visual experiences, sensory experiences, experience while being awake, and so on. the most general correlate we may be interested in is what distinguishes all conscious mental activity from unor preconscious mental activity. (fink, 2016, p. 4) we agree that the general correlate of consciousness is what distinguishes conscious from unconscious mental activity. the difference is that fink’s idea of the neural correlate of general consciousness seems broader than the gncc. it seems to involve the neural correlates of global states of consciousness, the processes differentiating between being conscious or unconscious in the global sense. in contrast, the gncc is limited to episodes of conscious perception. the mechanisms of global states of consciousness are presupposed by the gncc, but do not form part of it. in the following, we will further explore the idea that the nccc, the minimal neural activation sufficient for a particular type of experience (chalmers, 2000, p. 31; koch, 2004, p. 16), is a composite of the ncc and gncc. more precisely, since the spatial metaphor might be misleading, the idea is that ncccs are formed in the neural interactions between the contentand consciousness-mechanisms. we will continue to focus predominantly on the gncc. so far we have been speaking only about an abstract notion of gncc. in the next section 3, we will introduce some concrete proposals about how to construe the gncc. marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tomáš marvan and michal polák 6 abbreviation type of the neural correlate description notes ncc neural correlate of consciousness comes in two types: level-based ncc and content-based ncc. level-based ncc is the neural underpinning of levels of consciousness such as being awake, being in a coma, or being anaesthetised. the content-based ncc is the minimal neural system whose activation is sufficient for a particular type of conscious experience (chalmers, 2000). the content-based ncc as defined by chalmers (2000) has the same extension as the nccc below. core ncc core neural correlate of consciousness core neural processes correlating with experiences. term introduced in chalmers (2000) and further applied in block (2005) and keaton (2015). core ncc equals content-based ncc. nccc neural correlate of conscious (perceptual) contents neural correlate of a type of a conscious perceptual state (such as hearing an airplane in the distance) a dynamic combination of ncc and gncc ncc neural correlate of (perceptual) content neural mechanism/pattern creating the content of a perceptual state. the ncc is that part of the neural correlate of a conscious perceptual state that distinguishes one perceptual content from another, e.g., the seeing of blue from the seeing of green. gncc general neural correlate of (perceptual) consciousness neural mechanism conferring consciousness on perceptual contents. free from all content-specific (representational) aspects. a somewhat related concept of the neural correlate of general consciousness can be found in fink (2016). full ncc full neural correlate of consciousness union of the sets of contentspecific ncc for all possible contents of experience; situated in the posterior hot zone. term introduced in koch et al. (2016). total ncc total neural correlate of consciousness the totality of neural processes sufficient for a given conscious state. term introduced in chalmers (2000). the total ncc includes the core ncc together with all other neurally enabling conditions (relegated to the background of the core processes). 1 figure 1: kinds of neural correlates of consciousness (the yellow fields contain the proposed reconceptualization; the remaining concepts were introduced in seminal papers of the ncc debate.) marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org generality and content-specificity in ncc research 7 3 gncc and some prominent contemporary theories of perceptual consciousness although the gncc/ncc distinction does not depend on any particular theory of consciousness, it is not formulated independently of empirical research. on the contrary, some of the most prominent current theories of consciousness permit to distinguish between the content-constituting mechanisms and the gncc. in the following, we will selectively review five examples of three types of theories of consciousness: those that support, those that probably do not support and those that almost certainly do not support the gncc notion. prinz (2012) explicitly recommends to distinguish between these two mechanisms in his attentional theory of consciousness (attended intermediate-level representation theory – air). the air theory’s basic claim is that phenomenal consciousness arises if and only if the intermediate-level representations are modulated by attention (prinz, 2012, p. 89). the intermediate level of representations resides in between the too detailed and pointilistic representations in the early visual areas and the too abstract, categorial representations in the higher visual areas. at the cellular level, the intermediate-level visual representations, the contents of experience, are implemented in the pyramidal neurons within the visual cortex. combining population and temporal coding, prinz coins the term “vectorwaves”. a vectorwave is a pattern of activity in a population of neurons, a pattern with distinctive temporal dynamics. vectorwaves implement perceptual contents: red colour is implemented by a particular vectorwave, and so is blue or yellow colour. for colours to become conscious, though, more is needed than just the population’s firing in a vectorwave pattern. it is necessary that a synchronous firing of the vectorwave neurons in the gamma range (25 to 70 hz oscillations) takes place. neurally, this synchronization is triggered by interneurons connected to the visual pyramidal cells and modulating their activity by inhibitory mechanisms (prinz, 2012, p. 135); both pyramidal neurons and the interneurons are localized in the same cortex area. putting the two distinct neural processes, vectorwaves and gamma synchronization, together, prinz speaks about “gamma vectorwaves”. in our terms, a gamma vectorwave is the nccc of a particular colour experience, and it is composed of the vectorwave fixing the perceptual content (ncc) and the synchronous gamma oscillation (gncc) allowing the content to become conscious. the global neuronal workspace theory (gnwt) comports with our conceptualization, too. the theory identifies the neural substrate of consciousness itself with activation of the “global workspace” consisting mainly of long-distance axonal connections of large pyramidal neurons in cortical layers ii and iii (dehaene & changeux, 2011; mashour et al., 2020; sergent & naccache, 2012). first, the signal travels through visual areas and is further amplified by local feedback loops. visual contents generated by this processing remain unconscious. provided the signal is sufficiently strong, it travels onwards to the prefrontal cortex. longdistance axons of the pyramidal neurons then secure feedback connections from marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tomáš marvan and michal polák 8 the prefrontal cortex to lower areas, especially to associative areas in the posterior parietal cortex and also all the way back to the visual areas. this “ignition” (dehaene et al., 2003) of the neuronal workspace results in distributed yet synchronized oscillatory activity and makes the visual signals available for other modules connected by the neuronal workspace. such global availability, arising between 200 and 280ms after stimulus onset, marks a transition towards conscious perception. indeed, according to the gnwt, the global availability of contents simply is what it means for them to be perceived consciously. the sustained activation of the neuronal workspace thus can be taken to constitute the gncc according to the gnwt. the distinction between gncc and ncc does not smoothly apply to all major theories that currently compete to explain conscious perception, though. lamme’s recurrent processing theory (rpt) is a case in point. on a straightforward reading of rpt, visual contents are made conscious when, after the “feedforward sweep” of visual processing, the signal is sent back from higher visual areas to lower ones. the feedforward sweep through the hierarchy of visual areas activates the visual content mechanisms and gives the perceptual state its particular form. during this initial phase of activation, visual features such as orientation, contrast, colour, or direction of motion are extracted and visual objects are categorized. but all this happens unconsciously. it is the feedback from higher to lower areas that turns unconscious vision into conscious seeing. feedback from higher to lower areas thus can be taken to be the gncc according to the rpt. this straightforward reading is supported by lamme (2003, 2006). at one place lamme openly states that “[w]e could even define consciousness as recurrent processing” (lamme, 2006, p. 499). however, the story in lamme’s more recent article (lamme, 2020) takes a different turn. here lamme attributes complex visual functions to recurrent processing: the function of perceptual organization, mainly of the figure-ground segregation, and of object integration. none of these functions can be implemented by the feedforward sweep itself. although recurrent processing in this account remains to be tightly associated with consciousness (because none of the complex visual functions can be realized unconsciously), it now has a distinctly visual form. feedback from higher to lower areas does not constitute a content-nonspecific gncc, but is just a final stage of visual processing. on such a view, vision has no need for a true gncc: visual consciousness is the automatic outcome of the latest stages of visual processing. the content mechanisms cannot be separated from recurrent processing because recurrent processing has a visual representational nature, taking care of the visual figure-ground segregation and visual object integration. rpt in its latest form thus does not support the gncc idea. the straightforward reading, taking recurrent processing to be the gncc, would now be rejected by lamme. still, let us try to offer an argument in its favour. in a recent intermodality comparison study of neural substrates of conscious perception, auksztulewicz et al. (2012) compared the neural correlates of visual and tactile percepts. marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org generality and content-specificity in ncc research 9 their presumed gncc was recurrent cortical loops, the same gncc as in the rpt (on the straightforward reading). of the three models auksztulewicz et al. used to interpret their data, feedforward, feedback and recurrent, the recurrent model best fitted the somatosensory detection task data for data segments longer than 140ms. if recurrent loops between somatosensory regions are responsible for making tactile contents conscious, there is reason to suppose that recurrent processing is a domain-neutral mechanism, just as would be expected from a true gncc. treating recurrent processing as a content mechanism, as lamme (2020) does, is inconsistent with such inter-domain evidence. the integrated information theory of consciousness (iit; tononi, 2004; oizumi et al., 2014; tononi et al., 2016) is even less compatible with the gncc/ncc distinction. iit states that both conscious and unconscious processing lead to the integration of information within the brain. the difference between the two kinds of processing lies in the degree of integration. if the level of information integration during perception is too low to change the contents into conscious contents (clustered in a mental state), perception remains subliminal. a higher degree of information integration leads to conscious perception. iit further suggests that a system with a purely feedforward architecture cannot generate conscious states (oizumi et al., 2014, p. 19). if a feedforward architecture is not enough to turn contents into conscious contents, might not some sort of feedback processing be this additional gncc mechanism? although oizumi et al. (2014) appreciate the value of feedback processing for visual perception, the answer is no. feedback processing cannot be the iit version of gncc because even without actual feedback processing a neuronal informational “complex” can be conscious. from the perspective of the theory, the potential for local interactions among the parts of the complex is more important for consciousness than actual recurrent processing among its parts (oizumi et al., 2014, p. 20); “if all the neurons in a main complex were inactive (or active at a low baseline rate), they would still generate consciousness as long as they are ready to respond to incoming spikes” (oizumi et al., 2014, p. 17). this is one reason why iit does not support the gncc idea. the second one is that at least on some versions of iit, perceptual content mechanisms are inextricably interwoven with consciousness-conferring processes. this is the case in tsuchiya’s account, for example. tsuchiya (2017, p. 4) comments on the role of global availability of information and of recurrent processing in explanations of conscious perception. these processes, tsuchiya believes, cannot account for the phenomenological differences between experiences. why does vision feel like vision? within vision, why does color feel different from shape, despite both being generated in the visual cortex? whatever is critical for consciousness should be specific for the cortico‐thalamic system during the awake and the dreaming state and should be differentiable in ways that allow us to understand different modalities and their particular characteristics. (tsuchiya, 2017, p. 4) marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tomáš marvan and michal polák 10 indeed, recurrent processing is shared among qualitatively different experiences, and the same holds for the global availability of information (in the gnwt story). but notice that if the distinction between gncc and ncc is accepted, explaining the phenomenological differences between experiences stops being the job of either the rpt (on the straightforward reading) or of the gnwt. their job is merely to explicate how various perceptual contents become conscious – to describe the gncc. the phenomenological questions tsuchiya raises, although central for the science of consciousness, must be tackled by a radically different type of a theory. finally, no application may be given to the gncc idea in pure sensory processing theories of consciousness such as in zeki and bartels’ theory of essential nodes (zeki & bartels, 1999). the authors claim that some perceptual processing sites in the visual cortex are at the same time the sites of conscious processing. when the neural signal reaching these “essential nodes” is sufficiently strong, the nodes generate a full-blown content of the particular feature that is processed in the site, such as colour (in the areas v4 and v4α) or motion perception (in v5/mt). at the same time, a (micro)consciousness for each of these features is produced in the same node as well. the theory pictures consciousness as disunified at first because all the distinct visual features of a mental state are processed and made conscious independently of each other. all this happens within the 500ms after stimulus onset (zeki, 2003) without our noticing it introspectively. the microconsciousnesses are then bound together to form a unified conscious percept. clearly, the unifying mechanism putting the individual contents together cannot be the gncc. when it starts its work, consciousness, although distributed and not yet unified, is already present. given that no recurrence, feedback or any other additional mechanism is needed for a piece of perceptual content to become conscious, we see no space for a dissociable gncc in the essential nodes theory. for a content to become conscious, increased activity within essential node is all that is needed. 4 why countenance the gncc notion? as the previous selective review shows, some theories support the gncc idea, others do not, and still for others it is unclear. it is therefore a matter of dispute whether a theory of consciousness needs to embrace the gncc notion. the theories employing (or at least permitting) the gncc/ncc distinction, especially air and gnwt, draw on considerable empirical evidence. but so do the rival theories not supporting the distinction, such as the iit and the essential nodes theory. in this section, we want to contribute to the debate on this subject by summarizing what we see as the two main theoretical advantages of the theories supporting the distinction between the general neural process of perceptual consciousness and content-related neural processing. marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org generality and content-specificity in ncc research 11 theory of consciousness gncc mechanism proponents attended intermediate-level representation (air) theory yes – gamma synchronies prinz (2012) global neuronal workspace theory (gnwt) yes – sustained synchronous oscillatory activity in the neuronal workspace dehaene & changeux (2011) sergent & naccache (2012) mashour et al. (2020) and many others recurrent processing theory (rpt) no (?) lamme (2020) vs lamme (2003) and (2006) information integration theory (iit) no (?) tononi (2004) oizumi et al. (2014) tsuchiya (2017) and many others essential nodes theory no zeki & bartels (1999) 1 figure 2: selected theories of perceptual consciousness and the gncc (i) a strength of a theory supporting the gncc notion is that it allows a mental state to be fully formed in all its perceptual aspects without becoming conscious. perceptual processing is necessary but not sufficient for consciousness. in contrast, theories such as the essential nodes theory or the iit claim that perceptual contents become conscious just by being perceptually constituted in a certain way – by having a sufficiently strong input into the essential node or reaching a sufficiently high level of information integration. in these theories, perceptual processing is both necessary and sufficient for consciousness. a problem for the latter family of theories is that even if the input signals are sufficiently strong, and remain sufficiently stable and distinct, the contents might not become conscious (cleeremans, 2014; frith, 2019). experimental paradigms such as continuous flash suppression (faivre et al., 2014; sterzer et al., 2014) or change blindness (beck et al., 2001; reddy et al., 2006; simons & rensink, 2005) allow experimenters to keep perceptual processing in place while not letting perceptual contents enter the stream of consciousness. perceptual contents are still being constituted and poised for becoming conscious; but they do not become conscious. to take an example, jiang et al. (2009) used a dichoptic masking paradigm to measure temporal dynamics of facial information processing. in one of the experiments, the subjects were perceiving a face with neutral expression in two conditions, visible and invisible. in the invisible condition, processing of face-related attributes was reduced (as documented by a decrease in amplitudes of the erp components). still, in both conditions, the experimenters recorded a significant similarity of positive (p1) and negative (n1) components of the erp waveform. (the compared waveforms started around 220ms after stimulus onset in both visible marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tomáš marvan and michal polák 12 and invisible condition). the authors consider this to be evidence for unconscious cortical processing of faces in superior temporal sites (jiang et al., 2009, p. 1176).3 from the perspective of this paper, the most interesting possibility of this sort is when the perceptual contents mechanisms are functioning well but as a result of the gncc itself being disrupted, the contents do not reach consciousness. for instance, artificial disruption of feedback loops from mt+/v5 to v1 by transcranial magnetic stimulation causes loss of visual consciousness of motion (pascual-leone & walsh, 2001). the gncc idea provides a simple explanation. feedback from higher to lower areas is the gncc. disruption of feedback loops from mt+/v5 to v1 thus amounts to the disruption of gncc itself. resulting loss of conscious perception of motion is inevitable even if the processing at perceptual site (mt+/v5) remains intact. the feedback signals are not participating in perceptual processing but act as a consciousness conferring mechanism. such disruption of the gncc should be possible – and empirically testable – in other theories of consciousness supporting the gncc/ncc distinction. recall that gamma synchronisation is the gncc according to air theory. disrupting it should stop the stream of colour consciousness even if visual content correlates – the vectorwaves – continue to process incoming signals. in fact, prinz (2015, p. 384) considers this possibility. vectorwaves can be activated without the content they carry becoming conscious. visual contents can be formed unconsciously but fail to reach awareness if gamma synchronization fails. when a particular neuronal population fires out of synchrony, we may perceive blue but only unconsciously. similarly, according to the gnwt, the long-range axonal connections between prefrontal and parieto-temporal cortex secure wide distribution of sensory contents. disrupting these long-range connections (i.e., disrupting the gnwt’s gncc) should disrupt conscious perception. meanwhile, the sensory areas processing visual and other stimuli can still be working; processing in the sensory areas can go deep and significantly influence behaviour even when the signals ultimately do not reach consciousness (dehaene et al., 1998; dehaene, 2014; sergent et al., 2005). again, sensory contents could be completely processed at the perceptual level while not entering into conscious stream.4 (ii) being strict about keeping the processing of perceptual contents apart from consciousness-related processing in one’s theory is also an important pre3it has to be noted that the extent of unconscious processing under various experimental paradigms is the subject of ongoing controversy. in particular, while there is a general agreement that lower level visual features such as stimulus orientation can be perceived unconsciously under continuous flash suppression, unconscious perception of higher-level features, such as face identity or facial expression, was not demonstrated beyond reasonable doubt. for the most comprehensive recent review of continuous flash suppression studies, and the controversies surrounding the results, see pournaghdali & schwartz (2020). 4the signals in sensory areas during unconscious processing are weaker then during conscious processing. this does not contradict our hypothesis. the weakening of the signals in unconscious conditions could be explained by the absence of the gncc activation. for more details, (see marvan & polák, 2017). marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org generality and content-specificity in ncc research 13 requisite for building a truly general, pan-modality theory of conscious perception. although most of the theories of consciousness are applied to visual data, this limitation must be overcome. it would be desirable to have a systematic and unifying theory of what makes visual, auditory, gustatory and other forms of contents conscious. the gncc idea is a perfect match for this ambitious project, being completely modality-neutral. of the theories discussed in the previous section 3, the gnwt is already formulated in a domain-general way. the neuronal workspace globally distributes visual, auditory, tactile or olfactory contents (dehaene, 2014, pp. 13, 119–120, 158, 217, and 221). there is no specialized gncc for vision, distinct from gncc for audition etc.5 unfortunately, there are almost no other attempts to identify a shared gncc in different sensory modalities. apart from the promising visual-tactile comparison study mentioned in section 3 (auksztulewicz et al., 2012), initial comparisons of visual and auditory correlates of conscious perception were performed by snyder et al. (2015). according to them, auditory awareness-related negativity, a hypothesized marker of conscious hearing, is the auditory analogue of the visual awareness negativity, a presumed signature of arising visual consciousness. awarenessrelated negativity typically peaks around 100ms after the stimulus onset, visual awareness negativity around 50–100ms later. both negativities consistently show up in eeg and meg studies of conscious perception in both domains, displaying robust invariance under different experimental paradigms and changing task demands. this is not just an interesting result in its own right, but indirectly points to the role of recurrent processing, which is often taken to be the correlate of visual awareness negativity (see förster et al., 2020). recurrent processing theories of consciousness (such as the rpt in the straightforward reading) thus gain further evidential support. more comparative inter-modality research will show whether the results generalize and vindicate the idea of a unique gncc mechanism. the guiding principle of any such research must be to carefully disentangle the presumed gncc process from an almost indefinite variety of content endowing mechanisms. in the words of snyder and colleagues: paradigms designed to isolate neural correlates of the contents of consciousness are likely to find different neuroanatomical sources for auditory versus visual nccs. by contrast, studies that are able to isolate neural mechanisms that bring these contents into awareness, which may be a distinct process, could reveal overlapping neural sources between auditory and visual modalities. (snyder et al., 2015, p. 296) the only difference between this view and our proposal is that we do not insist on literally the same neurophysiological resources of consciousness-conferring mech5in fact, gnwt has even broader ambitions in some of its formulations. influential gnw theorists claim that the workspace idea can be extended to “felt emotions, visual imagery, working memory, and executive control” (baars et al., 2013). marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tomáš marvan and michal polák 14 anism being deployed in different modalities. this would indeed be the case of gnwt where different ignitions of the global workspace may literally use overlapping neuronal coalitions to broadcast contents from different sensory modalities. but this is not required for a gncc. what matters is merely that the same type of neural mechanism is deployed across perceptual episodes in different modalities. the gncc mechanism for different modalities may be localized non-overlapping areas. in fact, this is the picture behind air. the theory would probably locate gamma synchronies in different dedicated sensory cortices for different modalities – in the visual cortex for conscious vision, in the auditory cortex for conscious audition etc.6 5 concluding remarks while the search for neural correlates of conscious perception is an empirical enterprise, it benefits from theoretical-conceptual and methodological reflections. the hypothesis put forward in this article is that the standard notion of perceptual ncc conflates the processing of perceptual contents with consciousness conferring mechanisms. whereas the content mechanisms (ncc) secure all the richness of experience, gncc is a content-free, domain-neutral mechanism. the resulting neural correlate of a specific type of experienced content (nccc) arises in the dynamic interaction between these two distinct types of processes. although we propose a revision of the content-based ncc concept, the proposal is motivated by theoretical and empirical considerations and does not have the form of a conceptual truth. consequently, our goal in this paper was modest: to open the discussion on the possibility of the gncc/ncc distinction and to marshal some of the initial arguments in its favour. in the remainder, let us point out two advantages of endorsing it. first, the theories supporting the gncc/ncc distinction are amenable to direct testing in terms of the hypothesized gncc. if during testing the presumed gncc is either detected in conditions of subliminal perception, or if conscious perception occurs in the absence of the gncc, the tested theory is deficient.7 on the other hand, evidence against the gncc according to a particular theory does not disqualify the gncc idea as such. the gncc might be implemented elsewhere in the brain, or in a different way altogether. secondly, the gncc/ncc distinction helps to appreciate that tracing the content mechanisms and elucidating how they work may be a different project from 6we say “would probably locate” because prinz (2012) does not extend his gamma synchrony theory to non-visual modalities. 7for the first possibility see, e.g., the gnwt-contradicting evidence in silverstein et al. (2015) that p3b, an erp index of workspace ignition (dehaene et al., 2003), may be present under rigorously subliminal conditions. for the second possibility, see the air-contradicting evidence that increase in local gamma synchronisation might not be sufficient for conscious face perception in aru et al. (2012). marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org generality and content-specificity in ncc research 15 the effort to isolate the general mechanism by which the contents are made conscious. in our estimate, the first of the two research projects will be much more difficult to carry through. the questions falling to the content side belong to the most vexing issues of the science of consciousness. why are perceptual contents constituted in one way and not another? why does hearing the sound of trumpet feel so different from smelling fresh strawberries, given the similarity of the underlying neural processing in both cases? where are the contents located? do only striate and extrastriate areas have the neural resources to express the visual contents down to their very fine details, or may these contents be expressed elsewhere in the brain? are the consciously experienced contents the very same as those formed unconsciously? or is the activity in sensory areas a necessary prerequisite but not a proper neural substrate of conscious perception? in comparison with these riddles, the search for the gncc may turn out to be a more manageable enterprise. as the proposals discussed in section 3 indicate, the gncc could be a mechanism having the role of boosting or distributing the sensory signals. acknowledgments we would like to thank the anonymous reviewers for thoughtful comments and to sascha benjamin fink, wanja wiese and jennifer windt for helpful suggestions and outstanding editorial work. this article is the outcome of the czech science agency (gačr) project n. 20-14445s ”dual models of phenomenal consciousness”. references aru, j., axmacher, n., do lam, a. t. a., fell, j., elger, c. e., singer, w., & melloni, l. (2012). local category-specific gamma band responses in the visual cortex do not reflect conscious perception. the journal of neuroscience, 32(43), 14909– 14914. https://doi.org/10.1523/jneurosci.2051-12.2012 aru, j., bachmann, t., singer, w., & melloni, l. (2012). distilling the neural correlates of consciousness. neuroscience and biobehavioral reviews, 36(2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.12.003 auksztulewicz, r., spitzer, b., & blankenburg, f. (2012). recurrent neural processing and somatosensory awareness. the journal of neuroscience, 32(3), 799–805. https://doi.org/10.1523/jneurosci.3974-11.2012 baars, b. j., franklin, s., & ramsoy, t. z. (2013). global workspace dynamics: cortical ”binding and propagation” enables conscious contents. frontiers in psychology, 4, 200. https://doi.org/10.3389/fpsyg.2013.00200 bayne, t., & hohwy, j. (2013). consciousness: theoretical approaches. in a. e. cavanna, a. nani, h. blumenfeld, & s. laureys (eds.), neuroimaging of consciousness (pp. 23–35). springer. bayne, t., hohwy, j., & owen, a. m. (2016). are there levels of consciousness? trends in cognitive sciences, 20(6), 405–413. https://doi.org/10.1016/j.tics.2016.03.009 beck, d. m., rees, g., frith, c. d., & lavie, n. (2001). neural correlates of change detection and change blindness. nature neuroscience, 4(6), 645–650. https://doi.org/10.1038/88477 block, n. (2005). two neural correlates of consciousness. trends in cognitive sciences, 9(2), 46–52. https://doi.org/10.1016/ j.tics.2004.12.006 boly, m., massimini, m., tsuchiya, n., postle, b. r., koch, c., & tononi, g. (2017). are the neural correlates of consciousness in the front or in the back of the cerebral cortex? clinical and neuroimaging evidence. the journal of neuroscience, 37(40), 9603–9613. https://doi.org/10.1523/jneurosci.3218-16.2017 chalmers, d. j. (2000). what is a neural correlate of consciousness? in t. metzinger (ed.), neural correlates of consciousness (pp. 17–39). mit press. cleeremans, a. (2014). prediction as a computational correlate of consciousness. international journal of computing anticipatory systems, 29, 3–13. https://axc.ulb.be/uploads/2016/01/14-casys11.pdf crick, f. h. c. (1994). the astonishing hypothesis: the scientific search for the soul. charles scribner’s sons. marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1523/jneurosci.2051-12.2012 https://doi.org/10.1016/j.neubiorev.2011.12.003 https://doi.org/10.1523/jneurosci.3974-11.2012 https://doi.org/10.3389/fpsyg.2013.00200 https://doi.org/10.1016/j.tics.2016.03.009 https://doi.org/10.1038/88477 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1523/jneurosci.3218-16.2017 https://axc.ulb.be/uploads/2016/01/14-casys11.pdf https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tomáš marvan and michal polák 16 dehaene, s. (2014). consciousness and the brain: deciphering how the brain codes our thoughts. penguin books. dehaene, s., & changeux, j.-p. (2011). experimental and theoretical approaches to conscious processing. neuron, 70(2), 200–227. https://doi.org/10.1016/j.neuron.2011.03.018 dehaene, s., naccache, l., le clec’h, g., koechlin, e., mueller, m., dehaene-lambertz, g., moortele, p. f. van de, & le bihan, d. (1998). imaging unconscious semantic priming. nature, 395(6702), 597–600. https://doi.org/10.1038/26967 dehaene, s., sergent, c., & changeux, j.-p. (2003). a neuronal network model linking subjective reports and objective physiological data during conscious perception. proceedings of the national academy of sciences, 100(14), 8520–8525. https://doi.org/10.1073/pnas.1332574100 faivre, n., berthet, v., & kouider, s. (2014). sustained invisibility through crowding and continuous flash suppression: a comparative review. frontiers in psychology, 5, 475. https://doi.org/10.3389/fpsyg.2014.00475 fink, s. b. (2016). a deeper look at the neural ”correlate of consciousness”. frontiers in psychology, 7, 1044. https://doi.org/ 10.3389/fpsyg.2016.01044 förster, j., koivisto, m., & revonsuo, a. (2020). erp and meg correlates of visual consciousness: the second decade. consciousness and cognition, 80, 102917. https://doi.org/10.1016/j.concog.2020.102917 frith, c. d. (2019). the neural basis of consciousness. psychological medicine, 1–13. https://doi.org/10.1017/s00332917190 02204 haynes, j.-d. (2009). decoding visual consciousness from human brain signals. trends in cognitive sciences, 13(5), 194–202. https://doi.org/10.1016/j.tics.2009.02.004 hohwy, j. (2009). the neural correlates of consciousness: new experimental approaches needed? consciousness and cognition, 18(2), 428–438. https://doi.org/10.1016/j.concog.2009.02.006 jiang, y., shannon, r. w., vizueta, n., bernat, e. m., patrick, c. j., & he, s. (2009). dynamics of processing invisible faces in the brain: automatic neural encoding of facial expression information. neuroimage, 44(3), 1171–1177. https: //doi.org/10.1016/j.neuroimage.2008.09.038 keaton, d. (2015). constitution, realization and identity. in s. miller (ed.), the constitution of phenomenal consciousness: toward a science and theory (pp. 372–399). john benjamins. koch, c. (2004). the quest for consciousness: a neurobiological approach. roberts & company. koch, c., massimini, m., boly, m., & tononi, g. (2016). neural correlates of consciousness: progress and problems. nature reviews. neuroscience, 17(5), 307–321. https://doi.org/10.1038/nrn.2016.22 lamme, v. a. (2003). why visual attention and awareness are different. trends in cognitive sciences, 7(1), 12–18. https: //doi.org/10.1016/s1364-6613(02)00013-x lamme, v. a. f. (2006). towards a true neural stance on consciousness. trends in cognitive sciences, 10(11), 494–501. https://doi.org/10.1016/j.tics.2006.09.001 lamme, v. a. f. (2020). visual functions generating conscious seeing. frontiers in psychology, 11, 83. https://doi.org/10.3 389/fpsyg.2020.00083 marvan, t., & polák, m. (2017). unitary and dual models of phenomenal consciousness. consciousness and cognition, 56, 1–12. https://doi.org/10.1016/j.concog.2017.09.006 mashour, g. a., roelfsema, p., changeux, j.-p., & dehaene, s. (2020). conscious processing and the global neuronal workspace hypothesis. neuron, 105(5), 776–798. https://doi.org/10.1016/j.neuron.2020.01.026 mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 1–17. https://doi.or g/https://doi.org/10.33735/phimisci.2020.ii.58 michel, m., & morales, j. (2020). minority reports: consciousness and the prefrontal cortex. mind and language, 35(4), 493–513. https://doi.org/10.1111/mila.12264 miller, s. (2015). the correlation/constitution distinction problem: foundations, limits and explanation in consciousness science. in s. miller (ed.), the constitution of phenomenal consciousness: toward a science and theory (pp. 104–154). john benjamins. oizumi, m., albantakis, l., & tononi, g. (2014). from the phenomenology to the mechanisms of consciousness: integrated information theory 3.0. plos computational biology, 10(5), e1003588. https://doi.org/10.1371/journal.pcbi.1003588 pascual-leone, a., & walsh, v. (2001). fast backprojections from the motion to the primary visual area necessary for visual awareness. science, 292(5516), 510–512. https://doi.org/10.1126/science.1057099 pournaghdali, a., & schwartz, b. l. (2020). continuous flash suppression: known and unknowns. psychonomic bulletin & review, 27(6), 1071–1103. https://doi.org/10.3758/s13423-020-01771-2 prinz, j. (2012). the conscious brain. oxford university press. prinz, j. j. (2015). unconscious perception. in m. matthen (ed.), the oxford handbook of philosophy of perception. (pp. 371–389). oxford university press. marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neuron.2011.03.018 https://doi.org/10.1038/26967 https://doi.org/10.1073/pnas.1332574100 https://doi.org/10.3389/fpsyg.2014.00475 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.1016/j.concog.2020.102917 https://doi.org/10.1017/s0033291719002204 https://doi.org/10.1017/s0033291719002204 https://doi.org/10.1016/j.tics.2009.02.004 https://doi.org/10.1016/j.concog.2009.02.006 https://doi.org/10.1016/j.neuroimage.2008.09.038 https://doi.org/10.1016/j.neuroimage.2008.09.038 https://doi.org/10.1038/nrn.2016.22 https://doi.org/10.1016/s1364-6613(02)00013-x https://doi.org/10.1016/s1364-6613(02)00013-x https://doi.org/10.1016/j.tics.2006.09.001 https://doi.org/10.3389/fpsyg.2020.00083 https://doi.org/10.3389/fpsyg.2020.00083 https://doi.org/10.1016/j.concog.2017.09.006 https://doi.org/10.1016/j.neuron.2020.01.026 https://doi.org/https://doi.org/10.33735/phimisci.2020.ii.58 https://doi.org/https://doi.org/10.33735/phimisci.2020.ii.58 https://doi.org/10.1111/mila.12264 https://doi.org/10.1371/journal.pcbi.1003588 https://doi.org/10.1126/science.1057099 https://doi.org/10.3758/s13423-020-01771-2 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org generality and content-specificity in ncc research 17 reddy, l., quiroga, r. q., wilken, p., koch, c., & fried, i. (2006). a single-neuron correlate of change detection and change blindness in the human medial temporal lobe. current biology, 16(20), 2066–2072. https://doi.org/10.1016/j.cub.2006 .08.064 rees, g., kreiman, g., & koch, c. (2002). neural correlates of consciousness in humans. nature reviews. neuroscience, 3(4), 261–270. https://doi.org/10.1038/nrn783 sandberg, k., andersen, l. m., & overgaard, m. (2014). using multivariate decoding to go beyond contrastive analyses in consciousness research. frontiers in psychology, 5, 1250. https://doi.org/10.3389/fpsyg.2014.01250 sergent, c., baillet, s., & dehaene, s. (2005). timing of the brain events underlying access to consciousness during the attentional blink. nature neuroscience, 8(10), 1391–1400. https://doi.org/10.1038/nn1549 sergent, c., & naccache, l. (2012). imaging neural signatures of consciousness: “what”, “when”, “where” and “how” does it work? archives italiennes de biologie, 150(2-3), 91–106. https://doi.org/10.4449/aib.v150i2.1270 silverstein, b. h., snodgrass, m., shevrin, h., & kushwaha, r. (2015). p3b, consciousness, and complex unconscious processing. cortex, 73, 216–227. https://doi.org/10.1016/j.cortex.2015.09.004 simons, d. j., & rensink, r. a. (2005). change blindness: past, present, and future. trends in cognitive sciences, 9(1), 16–20. https://doi.org/10.1016/j.tics.2004.11.006 snyder, j. s., yerkes, b. d., & pitts, m. a. (2015). testing domain-general theories of perceptual awareness with auditory brain responses. trends in cognitive sciences, 19(6), 295–297. https://doi.org/10.1016/j.tics.2015.04.002 sterzer, p., stein, t., ludwig, k., rothkirch, m., & hesselmann, g. (2014). neural processing of visual information under interocular suppression: a critical review. frontiers in psychology, 5, 453. https://doi.org/10.3389/fpsyg.2014.00453 tononi, g. (2004). an information integration theory of consciousness. bmc neuroscience, 5, 42. https://doi.org/10.1186/14 71-2202-5-42 tononi, g., boly, m., massimini, m., & koch, c. (2016). integrated information theory: from consciousness to its physical substrate. nature reviews neuroscience, 17(7), 450–461. https://doi.org/10.1038/nrn.2016.44 tsuchiya, n. (2017). “what is it like to be a bat?”—a pathway to the answer from the integrated information theory. philosophy compass, 12(3), e12407. https://doi.org/10.1111/phc3.12407 tsuchiya, n., wilke, m., frässle, s., & lamme, v. a. f. (2015). no-report paradigms: extracting the true neural correlates of consciousness. trends in cognitive sciences, 19(12), 757–770. https://doi.org/10.1016/j.tics.2015.10.002 zeki, s. (2003). the disunity of consciousness. trends in cognitive sciences, 7(5), 214–218. https://doi.org/10.1016/s13646613(03)00081-0 zeki, s., & bartels, a. (1999). toward a theory of visual consciousness. consciousness and cognition, 8(2), 225–259. https: //doi.org/10.1006/ccog.1999.0390 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii), 5. https://doi.org/10.33735/phimisci.2020.ii.61 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.cub.2006.08.064 https://doi.org/10.1016/j.cub.2006.08.064 https://doi.org/10.1038/nrn783 https://doi.org/10.3389/fpsyg.2014.01250 https://doi.org/10.1038/nn1549 https://doi.org/10.4449/aib.v150i2.1270 https://doi.org/10.1016/j.cortex.2015.09.004 https://doi.org/10.1016/j.tics.2004.11.006 https://doi.org/10.1016/j.tics.2015.04.002 https://doi.org/10.3389/fpsyg.2014.00453 https://doi.org/10.1186/1471-2202-5-42 https://doi.org/10.1186/1471-2202-5-42 https://doi.org/10.1038/nrn.2016.44 https://doi.org/10.1111/phc3.12407 https://doi.org/10.1016/j.tics.2015.10.002 https://doi.org/10.1016/s1364-6613(03)00081-0 https://doi.org/10.1016/s1364-6613(03)00081-0 https://doi.org/10.1006/ccog.1999.0390 https://doi.org/10.1006/ccog.1999.0390 https://doi.org/10.33735/phimisci.2020.ii.61 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction distilling the general neural correlate of perceptual consciousness from the neural correlates of perceptual contents gncc and some prominent contemporary theories of perceptual consciousness why countenance the gncc notion? concluding remarks finding the neural correlates of consciousness will not solve all our problems. finding the neural correlates of consciousness will not solve all our problems morten s. overgaarda,b,c (morten.storm.overgaard@cfin.au.dk) asger kirkeby-hinrupd (asger.kirkeby-hinrup@fil.lu.se) abstract subjective experience has often taken center stage in debates between competing conceptual theories of the mind. this is also a central object of concern in the empirical domain, and especially in the search for the neural correlates of consciousness (nccs). by now, most of the competing conceptual theories of consciousness have become aligned with distinct hypotheses about the nccs. these hypotheses are usually distinguished by reference to their proposed location of the nccs. this difference in hypothesized location of the nccs has tempted participants in these debates to infer that evidence indicating the location of the nccs in one or the other brain region can be taken as direct evidence for or against a given conceptual theory of consciousness. we argue that this is an overestimation of the work finding the nccs can do for us, and that there are principled reasons to resist this kind of inference. to show this we point out the lack of both an isomorphism and a homomorphism between the conceptual frameworks in which most theories are cached, and the kind of data we can get from neuroimaging. the upshot is that neural activation profiles are insufficient to distinguish between competing theories in the conceptual domain. we suggest two ways to go about ameliorating this issue. keywords behavioural methods ∙ correlations ∙ homomorphism ∙ isomorphism ∙ localization∙ neural correlates of consciousness ∙ theories of consciousness this article is part of a special issue on “the neural correlates of consciousness,” edited by sascha benjamin fink and ying-tung lin. aprofessor, head of cognitive neuroscience research unit (cnru), aarhus university bdepartment of clinical medicine, center of functionally integrative neuroscience (cfin), aarhus university caarhus institute of advanced studies, aarhus university, denmark ddepartment of philosophy and cognitive science, theoretical philosophy, lund university, sweden overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-1215-5355 https://orcid.org/0000-0003-1330-8680 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org morten s. overgaard and asger kirkeby-hinrup 2 1 introduction by now, most researchers concerned with the study of consciousness believe that the mind can be naturalized, for instance in the sense that consciousness depends on activity in the brain. for this reason, the project of identifying the neural correlates of consciousness (nccs) has attracted a lot of attention and precipitated interdisciplinary interactions to clarify the mesh between competing conceptual theories of consciousness, particularly (but not exclusively) from the domain of philosophy1, and findings from the empirical domain. subjective experience—the peculiar first-person aspects of being conscious, the nature of which often has taken centre stage in debates between competing conceptual theories of the mind—is also a central object of concern in the empirical domain, and especially in the search for the nccs. one upshot of this shared explanandum is that most of the competing conceptual theories of consciousness have become aligned with distinct hypotheses about the nccs. for instance, many proponents of higher-order thought (hot) theories of consciousness have embraced the hypothesis that the nccs are located in the prefrontal cortex (pfc). furthermore, proponents of hot theories argue that conscious access and phenomenal consciousness may coincide (a sentiment shared by proponents of other theories, see e.g. naccache, 2018; see also overgaard, 2018). a main competitor to the hot theories, the so-called recurrency theory, argue that the pfc is required only for conscious access, whereas the nccs for phenomenal consciousness rely on recurrent activations in early sensory areas, such as the primary visual cortex (v1) in the case of visual phenomenology (lamme, 2018; pinto et al., 2017). this difference in views has sparked extensive debate (importantly, this debate is not limited to proponents of hot and recurrency theories, see e.g. peters et al., 2017; railo et al., 2015). furthermore, the difference in hypothesized location of the nccs has tempted participants in these debates to infer that evidence indicating the location of the nccs in one or the other brain region can be taken as direct evidence for or against a given conceptual theory of consciousness. arguing in this way seems to be an overestimation of the work the nccs can do for us, when it comes to theories in the conceptual domain, and we will argue that there are principled reasons to resist these kinds of argument. importantly, and to be clear, finding the nccs almost certainly will allow us to make significant headway in our understanding of consciousness, and how it may arise from the brain. this is not the object of our concern. instead, what we highlight is that 1throughout, we use ‘conceptual theories’ and ‘the conceptual domain’ broadly to refer to the various conceptual frameworks aiming to explain consciousness, such as higher-order thought theories, reflexive theories, panpsychic theories, workspace theories, to name a few. importantly, while many existing conceptual theories are now closely tied to empirical work, almost all contain theory-crafting in their deployment of concepts (e.g. ‘mental state’) and posits (e.g. ‘recurrent processing are necessary for phenomenal consciousness’) that are not essentially empirical concepts (as opposed to e.g. ‘spiketrains’ or the ‘prefrontal cortex’). it is the relation between these concepts and posits, on the one hand, and empirical data on the other hand, that is our main concern here. overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org finding the neural correlates of consciousness will not solve all our problems 3 there are reasons to doubt that finding the nccs—in and of itself—may resolve the conceptual debates by supporting the empirically plausibility of one conceptual theory over the other. this problem turns on two issues. the first issue concerns a missing isomorphism—the lack of a 1-to-1 correspondence—between theoretical posits and neural data.2 the second concerns the lack of homomorphism—the lack of a structure-preserving mapping—between conceptual frameworks deployed in the theoretical and empirical domain respectively. as we will argue, the consequence of the missing isomorphism is that findings or predictions about activity localized to specific brain regions, by themselves, are insufficient to distinguish between competing theories in the conceptual domain. to boot, the consequence of the missing homomorphism is that the concepts deployed in the conceptual debates do not map straightforwardly onto the kind of data generated by the brain sciences, and vice versa.3 therefore, significant liberties are often taken when mapping an empirical finding onto a conceptual theory. it is worth noting that the way we operationalize our conception of the nccs may 2the notion of isomorphism we deploy throughout the paper is similar to the one found in pessoa et al. (1998), which they call analytic isomorphism. pessoa and colleagues conceive of analytic isomorphism as “essentially a conceptual or methodologic doctrine about the proper form of explanation in cognitive neuroscience.” (pessoa et al., 1998, p. 726). however, while pessoa and colleagues seek to apply the isomorphism relation to neural data and phenomenal experience, we instead (for reasons related to correlation explained in the next section) apply the isomorphism relation to neural data and theoretical posits. one reviewer suggested isomorphic relations cannot hold between entities of different cardinality since this would be a category mistake. we do not think this is the case and note that application of the isomorphism relation to entities of different cardinality is rife within all areas of the scientific literature. for instance, isomorphism is discussed as a relation between thought and reality (glock, 1997) and in models in the philosophy of science (da costa & french, 1990). 3clearly, some theories from the conceptual domain are better equipped to cater to a conceptual mapping between their theoretical posits and the kind of data generated by the brain sciences. to some extent, it seems that theories developed at the intersection between philosophy and the empirical sciences, fare somewhat better in this regard. however, we do not put much weight on this observation, and nothing hinges on it, since we believe what we have to say is pertinent to most if not all theories. thus, our use of hot theory and recurrent processing theory as examples in this paper does not mean that these are the only conceptual frameworks that have issues with the conceptual mapping task (see also footnote 1). a secondary question pertains to whether the mapping issues we address in this paper find their root cause in the conceptual frameworks themselves or in the conceptual background assumptions of the varying proponents of the competing conceptual frameworks. while our concern here is mainly about the conceptual frameworks themselves, it seems obvious that there is an interplay between these two factors. certainly, the (pre-theoretic) predilections and conceptual commitments of researchers are a key factor in the development of theories. consequently, the interpretations of the concepts deployed by a given theory influences the conceptual mapping to the empirical domain. additionally, we acknowledge that much debate in the theoretical domain is driven by longstanding disagreements about what the central concepts actually mean (“consciousness” is a clear cut example), and there is an overhanging risk (and some evidence) that these disagreements bleed into the work in the empirical domain. we see this as all the more reason to meticulously and critically scrutinize the practices involved in conceptual mapping, which is exactly our objective with this paper. we thank an anonymous reviewer to prompting us to clarify this. overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org morten s. overgaard and asger kirkeby-hinrup 4 compound (or, in the best case, alleviate) these issues (see e.g. discussion in fink, 2016). be that as it may, we here advertise two ways to go about ameliorating these issues. before we move on, it is necessary to briefly summarize the foundation for our argument, starting with a reminder about the nature of empirical investigations of consciousness. 2 correlations, subjective and objective measures the most fundamental distinction in empirical methods applied in consciousness research is between subjective and objective measures (koch et al., 2016; timmermans & cleeremans, 2015). subjective methods tap into the assumed privileged access an individual has to her own conscious states. usually, subjective methods rely on subjects to introspectively determine what they are consciously experiencing. objective methods, on the other hand, rely on behavioural data (e.g. eye movements, reaction times) or measurements of events in the brain (e.g. blood oxygen level, event related potentials, spike trains). strictly speaking, objective methods are not measuring consciousness per se. rather, objective measures concern behaviour or neural events that we think are somehow related to the presence or absence of conscious states. so, our objective data is not about consciousness per se. subjective measures, conversely, are (at least prima facie) about conscious states. however, subjective measures suffer from a range of other issues. first and foremost, subjective methods do not sit well with standard conceptions of scientific rigor such as third person access. in contrast to this, objective methods can cater to this requirement. thus, it seems a combination of subjective and objective methods may be the best solution to satisfy, one the one hand, the desideratum that our data are about consciousness, and on the other hand, the desideratum that our data are third person accessible, i.e., can be verified by an external observer. still, it is worth remembering that, even in a best case scenario, neither consciousness nor the neural processes presumably underpinning it actually are what we observe. the aim is to establish a correlation between subjective measures (that we assume reflect consciousness more or less accurately), and measurements of neural (or behavioural) events (that we assume are related to the presence or absence of conscious states by proxy of the neural states or processes that underpin them). importantly, the ways we establish these correlations themselves depend on the assumption that other correlations have been reliably established (e.g. that activity in lgn and v1 is correlated with the presentation of a visual stimulus). this means that reaching our objective consists in establishing a correlation between two distinct correlations. on the one hand, we assume that our neural or behavioural data correlate with the nccs. on the other hand, we assume that our subjective measures correlate with consciousness. the result, then, is a correlation between our neural/behavioural data and our data from our subjective measures (overgaard, 2015). overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org finding the neural correlates of consciousness will not solve all our problems 5 figure 1: a schematic view of the role of correlations in the search for the nccs suppose we establish the desired correlation between subjective and objective measures, how will this impact competing theories at the conceptual level? theories in the conceptual domain predict more or less the same with respect to subjective measures and are mute on objective measures. objective measures, by virtue of being nothing more than visual or mathematical representations of measurements, reciprocally are mute on conceptual theories. finally, to the extent that subjective measures have anything to say about conceptual theories, it is by appeal to how things appear to a single individual through introspection. such introspective reports, as noted above, have trouble with third person verification, and, somewhat worse, appear arbitrary with respect to the conclusions individuals take them to warrant with respect to conceptual theories (something that should be obvious from the last 50 or so years of philosophical debate on consciousness) (overgaard, 2017). 3 the missing isomorphism the idea that—when identified—the neural correlates of consciousness (nccs) will be sensitive to the nuances of competing conceptual theories, in a way that allows us to distinguish between them, and confirm one over the others, presupposes an isomorphism between the theoretical posits and our empirical data about the nccs. that two entities are isomorphic means that there is a 1:1 correspondence between the two entities with respect to the aspect(s) under consideration. the kind of isomorphism relevant in the current context is supposed to hold between posits of conceptual theories of consciousness and our (eventual) evidence for the nccs. as mentioned above, the work discovering the nccs is supposed to do for overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org morten s. overgaard and asger kirkeby-hinrup 6 us with respect to the debate between competing theories requires it to be sensitive to the nuances of competing theories. effectively, this means that, in order for the nccs to be sensitive to the nuances of distinct theories, these nuances need to show up unequivocally in the conceptual mapping. isomorphism between a conceptual framework and the kind of evidence we can get for the nccs would constitute the strongest possible mapping. anything short of isomorphism will leave significant conceptual wiggle room and likely fail to settle the debate. to boot, issues stemming from conceptual wiggle room, are among the problems that finding the nccs is supposed to solve for us. as mentioned above, one prominent debate pitches higher-order thought (hot) theories of consciousness against recurrency theory. the two theories are now closely aligned with two distinct hypotheses about the neural correlates of consciousness (ncc). on the one hand, hot theories are aligned with the idea that consciousness is generated by late processes, tentatively attributed to the prefrontal cortex (pfc). on the other hand, recurrency theories have aligned themselves with the hypothesis that consciousness depends on early processes, for instance recurrent processing in the primary visual cortex (v1) in the case of conscious visual awareness. however, there is no isomorphism supporting evidential transitivity between activations in this or that brain region and one or another theory. for example, higher-order theories hypothesize that consciousness generating higher-order thoughts may involve prefrontal activity, but this does not entail that findings where conscious subjects do not exhibit significant prefrontal activations is evidence against the higher-order theories as conceived of in the conceptual domain. nevertheless, many seem to engage with this kind of argument (kozuch, 2014; odegaard et al., 2017; sebastián, 2014). in one recent example of this kind of endeavour, michel and morales, in a nice paper (2020), provide extensive arguments to the effect that in many studies prefrontal activations correlate with consciousness, as opposed to reports. according to michel and morales, if this can be shown to be the case, it would support so-called prefrontal theories against their competitors. importantly, (at least in the present context) we are not contending the content of this paper. what we are sceptical about is the premise of the paper that links certain theories to certain regions of the brain. monikers such as ‘prefrontal theories’ referring to a group of theories in the conceptual domain (e.g. higher-order thought theory, global workspace and others) and ‘local recurrency theory’ generally used to refer to the first-order reflexive theory associated with ned block, tacitly indicate that somehow the viability of these theories from the conceptual domain depends on where in the brain an eventual discovery locates the nccs. it is important to remember that even if the nccs are eventually located in the early sensory regions, this does not entail the falsification of hot theory. similarly, the reflexive theories are not automatically wrong, in case the nccs happen to be in the pfc. it is exactly the implicit assumption (as found in e.g. michel and morales’ paper) that evidence of the nccs being in one brain region or another automatically supports this or that theory in the conceptual domain of which we are sceptical. overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org finding the neural correlates of consciousness will not solve all our problems 7 4 the missing homomorphism in addition to the missing isomorphism between brain regions and conceptual theories, there are a plethora of technical terms from the conceptual debates that have no obvious and/or measurable empirical counterpart. prominent examples are the notions of phenomenality and subjectivity. these notions are prevalent in theoretical debates, but we have no good idea of how to cash them out in measurements amenable to third person scrutiny. thus, in addition to the missing isomorphism presenting a problem on the general level when connecting theories with brain regions (a significant problem in itself) there is a lack of homomorphism between much of the conceptual and empirical terminology and overall frameworks. for instance, a central theoretical construct scaffolding both hot and reflexive theories is mental states. to illustrate, the hot theory posits that a mental state, such as a sensation, is conscious when it is the intentional object of another (higher-order) mental state. similarly, reflexive theories hold that a mental state has the property of being conscious by instantiating a special reflexive relation to itself. both theories are similar in the sense that they both seek to explain consciousness by reference to properties of—or relations between—mental states. however, when it comes to the brain, the notion of mental state does not straightforwardly apply. it is true that neuroscientists may deploy the concept of a mental state, for instance when referring to representations in the brain incurred by the introduction of a stimulus. however, this usage is merely a proxy referring to a collection of signals and processes propagating through different brain areas. so, while the notion of mental states, conceived of as ontologically isolatable entities is conversationally useful and theoretically harmless, it is—strictly speaking— empirically mistaken. on the neural level states are signals or patterns of neural activation, and this is what we measure empirically. signals move through brain regions—and patterns unfold—over time. we can trace a signal and its interactions as it propagates through the brain and reaches different stages of processing. and speaking loosely, we may agree that the signal somehow corresponds to, say, a perceptual representation of a visual stimulus. such loose talk is useful, and for most purposes more than sufficient. but, in virtue of being a general gesture toward an underspecified group of phenomena, loose speak fails to provide a specific answer to what and where the mental state actually is. and this is what matters when we need to individuate a mental state. there is no non-arbitrary way to fixate upon an exact point in time and space to delineate the boundaries of a mental state, which is what is necessary for individuation. to see why this is important, suppose the signal from a visual stimulus arrives at v1, after which the signal propagates through the visual system as normal, and makes its way around the frontoparietal network, and eventually the subject reports being conscious of the stimulus. according to reflexive theories, the signal, through processing, acquired some reflexive relationship to itself (supposedly from recurrent processing). according to hot theory, the signal arrived at some location in the pfc (supposedly the dorsoovergaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org morten s. overgaard and asger kirkeby-hinrup 8 lateral prefrontal cortex; dlpfc), at which point it got represented by a hot. so, in the first case, we have one mental state (with fuzzy boundaries), the processing of which yields consciousness. whereas in the latter case we have a mental state (with fuzzy boundaries) arriving at some location and triggering another mental state (the hot). but how do we determine from the neural data, whether there are one or two states? in light of the debate between competing theories, it matters a great deal, whether we are talking about one or two states. there are, of course, plenty of timeslices we can pick out and lines we can draw around the signal. for instance, one could argue that a particular development in the signal, such as a change in distribution, route, amplitude, or events such as arrival at specific cortical areas, changes in encoding, transfer to different kinds of memory-caches and so, should form the basis for the individuation. however, such a decision is ultimately arbitrary. the upshot of this is that there is no non-arbitrary structure preserving mapping, (i.e., homomorphism) between many theoretical frameworks and empirical data. 5 activation profiles are insufficient to distinguish between theories as noted above, most of the competing conceptual theories of consciousness have become aligned with distinct hypotheses about the nccs. essentially, what distinguishes one hypothesis about the nccs from another is the location and the nature of neural processes it posits as necessary and/or sufficient for consciousness to occur. in particular, the localization of the nccs has been the topic of much debate. probably, the hypothesis that early processes in occipital cortex constitute the nccs is sufficiently distinct from the hypothesis that the nccs are to be found in late processes in the front of the brain, to allow us to distinguish empirically between the two. however, the possibility of distinguishing empirically between competing hypotheses of the nccs is not the object of concern here. what we doubt is whether the empirical confirmation of any particular ncc hypothesis allows the confirmation of any particular conceptual theory (e.g. hot) over another. to understand this worry, it is informative to reflect on how conceptual theories have come to be aligned with respective ncc hypotheses. deploying hot theory as an example, the consensus that the pfc is the brain region most likely to underpin higher-order thoughts is derived from perceived similarities between how hots are conceived of in the conceptual domain and a range of metacognitive abilities which evidence suggests depend on prefrontal processes (beeckmans, 2007; brown, 2012; kozuch, 2014; kriegel, 2007; lau, 2007; lau & brown, 2019; lau & rosenthal, 2011). for instance, some theoretical posits of hot theory are conceptually or functionally similar to metacognitive abilities that are thought to be located in the pfc, such as self-monitoring and judgements of (own) performance (fleming & dolan, 2012), cognitive control and planning overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org finding the neural correlates of consciousness will not solve all our problems 9 (lau, 2011) and theory of mind (frith & frith, 2006). however, we came to associate each of these metacognitive abilities with the pfc because we have empirical reasons—from behavioural measures—to think they depend on pfc activity (see e.g. kirkeby-hinrup, 2020, p. 142). in contrast to this, the putative reasons to think that the viability of e.g. hot theory depends on whether the nccs are found in the pfc, is extrapolated from data obtained from investigations not of consciousness, but of other cognitive phenomena. 6 facing up to the missing isomorphism our first advice to ameliorate the troubles incurred by the missing isomorphism concerns how to establish a better mesh between our conceptual theories and the kind of data we can collect about the brain. the physical and functional nature of the brain constrains the kinds of data we can obtain from it. we currently have a decent preliminary grasp of these kinds of data. even if new neuroscientific methods become available, we would not expect these to fundamentally solve the conceptual divide between theory and data. and (barring the unlikely discovery of non-physical causally efficacious phenomena) the data we can obtain about the brain will be constrained by physics and will reflect the physical nature of the processes and mechanisms in the brain. this means that the empirical data we collect, while possibly crude compared to future technologies, will retain certain shared fundamental characteristics with data collected in the future. given the constraints on the nature of the empirical data we can collect, it does not appear we can remedy the missing isomorphism through changes in the data we treat or the way we collect it. this is because the missing isomorphism, essentially, is a conceptual disconnect between the way we conceive of our theories and the nature of the data about neural instantiations supposed to exemplify or instantiate the theoretical posits.4 therefore, if we cannot bridge this disconnect by altering the data, it seems the only option is to revise the conceptual frameworks in which theories are cached, i.e., bottom up data driven theoretical revision based on the kind of data we can collect (a similar sentiment can be gleaned from genon et al., 2018). in short: because we cannot make the data fit our theories, we should make our theories fit the data. this of course does not mean throwing out the baby with the bathwater such as a wholesale discarding of theories or giving up on the central explanandum (consciousness). rather, it means reformulating or updating theories to be sensitive to the kind of data we can collect, and—importantly—be amenable to testing by behavioural methods, as discussed in the next section.5 4we appreciate that there are historical parallels to this problem, for instance in the philosophical literature on identity theory. given the prevalence of this kind of argument we see value in restating the problem and applying it in the context of contemporary debates. we thank an anonymous reviewer for reminding us of this. 5to elaborate, the goal here is neither to engage in conceptual revision of the kind proposed by eliminativists (churchland, 1981), nor do we encourage putting our hopes in the discovery of overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org morten s. overgaard and asger kirkeby-hinrup 10 to illustrate these issues, we offer the following examples. the first example concerns localization. in relation to the discussion of higher order theories above, some versions of hot (lau & rosenthal, 2011) posit that specific neural regions are necessary for conscious experience. particularly, the dorsolateral prefrontal cortex (dlpfc) has received a lot of attention. however, the proposal is not merely underdetermined by empirical evidence, it is challenged by seemingly obvious counterevidence. for instance, during rem sleep, prefrontal cortex activity is often low compared to awake states, even though the experienced content may be just as vivid (for some discussion see e.g. sebastián, 2014; weisberg, 2014). similarly, patients undergoing generalized seizures have increased blood flow in prefrontal areas, even though they are considered unconscious. an energetic debate as of recent stems from the advent of so-called no-report paradigms (e.g. block, 2019). this debate concerns attempts—without the use of reports—to contrast conditions in which the subject is (presumed to be) conscious of a stimulus with conditions where the subject is (presumed to be) unconscious of the stimulus. a range of issues have been raised with this methodology. one of the more significant issues, is that, in virtue of the lack of reports, no-report paradigms have no methodological control for whether a stimulus was in fact consciously seen (overgaard & fazekas, 2016). nevertheless, prima facie, they appear to provide strong evidence that the frontal activations found in many experiments in fact are correlates of reporting, task preparation and execution rather than the conscious experience per se. to illustrate that issues of this kind are not exclusive to higher-order theory, let us consider another example. the main opponents of higher-order theory generally endorse the hypothesis that consciousness is related to early activity in temporo-parietal-occipital networks, rather than late activity in the fronto-parietal networks. one obvious issue with so-called early theories is that the varying theoretical proposals suggesting a temporo-parietal-occipital network—while not necessarily mutually exclusive—are not particularly compatible. to boot, it is very unclear that even if we succeeded in locating the nccs in early temporo-parietaloccipital processes, it would be possible to distinguish further between the competing theories available. to illustrate, victor lamme argues that all brain regions are unconscious during a feed-forward sweep of information, whereas conscious experience happens when information feeds back to occipital regions (lamme, 2006). other views, e.g. the visual awareness negativity hypothesis, makes no an adequate vocabulary sometime in the future as nagel (1974) ponders. what we suggest is for theories to be reformulated or translated so as to be sensitive to the kind of data we can collect, while maintaining their core hypotheses. to exemplify, we need an explanation of what the difference is between an information carrying signal (mental state) to reach, say, the pfc and become higher-order represented, as opposed to the signal merely reaching the pfc and becoming transformed. we need an answer to why we should think there are two states as opposed to just one that changes, and we need a way to discern this difference either from a dataset that depicts continuous activations or from behavioural measures. we thank an anonymous reviewer for pushing us to clarify this. overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org finding the neural correlates of consciousness will not solve all our problems 11 such claim, but argues that consciousness correlates with activity in the “van range” (koivisto & revonsuo, 2010). although the feed-back view and the van view both posit that consciousness is to be found in primary sensory regions, they disagree for instance with regards to timing. the feed-back view argues that information must first travel through a feed-forward sweep before it “returns,” whereas the van view expects the feed-forward sweep to be conscious. very few experiments have directly attempted to disentangle these views (crouzet et al., 2014, but see, e.g., 2017), and whether this can be done conclusively is very much an open question. these problems stem from the fact that localizing the ncc can only tell us so much. simply pointing to a brain region does not move the needle much. in order properly to bridge the gap between the conceptual domain and our neurological data, we need to at least know about implementation. localization is not enough in itself. that this is the case has already been acknowledged in discussions of attempts to localize (other) cognitive functions, where merely aiming for localization has been criticised repeatedly as being insufficient in order to provide a deeper understanding (e.g. carandini, 2012; mogensen & overgaard, 2018; overgaard & mogensen, 2011). according to this criticism, it is insufficient to point to a correspondence between a ‘function’ at the mental level and a given structure at the neural level. there is a need for an ‘intermediate’ computational level. without a computational level (or understanding the implementation), pinpointing the location of an ncc does little to improve our understanding of consciousness. it is worth mentioning that a few of the current models of the ncc are closer to addressing the issue regarding these missing computational processes than others. among these are the integrated information theory (e.g. tononi et al., 2016) and the reorganization of elementary functions (ref) framework in the form of the refcon and refgen models proposed by overgaard and mogensen (mogensen & overgaard, 2018; overgaard, 2015). in these models the suggested computational processes may point to mechanisms which eventually may yield new empirically testable predictions and thereby contribute to answers in the context of problems of consciousness. the examples we have provided here are just a sample to illustrate some of the issues that arise from the current practice of trying to fit the evidence to the theory rather than the other way around. although the debate is far from resolved, we take the issues presented here to be good reason to at least consider alternatives to this practice. one such alternative concerns behavioural methods. we will consider that next. overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org morten s. overgaard and asger kirkeby-hinrup 12 7 back to behavioural methods our second advice concerns shifting focus away from topography. the focus on topography is a trend that has been prevalent in the early-late debate, i.e., the debate concerning whether the ncc are to be found in early processes in the back of the head, or late processes in the front of the head (boly et al., 2017). indeed, the debate itself is often framed in terms of topography (either as the occipital-frontal debate, the front-back debate, or by proxy of early-late). as we highlighted above, there is no necessary connection between the theoretical posits of higher-order theory and the pfc. similarly, recurrent processing in the early visual system is, absent radical interpretation, not isomorphic with the ideas of reflexive theories of consciousness. nevertheless, pace disclaimers (brown et al., 2019; morales & lau, forthcoming) activations in specific brain regions have come to be seen as reliable indicators of whether a given empirical finding meshes better with higher-order theory or reflexive theories of consciousness. for instance, arguing that if subjects with lesions to the pfc retained a capacity for consciousness, this would count against the higher-order theories reflects this focus on topography, as does the defence mounted against the lesion data by proponents of higher-order theories (odegaard et al., 2017) in spite of their own disclaimers (brown et al., 2019). in our opinion, focusing on topography is premature. it is premature because the way various brain regions contribute to the processes underpinning consciousness has yet to be determined with any inkling of consensus. that this is true can be seen straightforwardly from the fact that there is still widespread disagreement about where to even locate the nccs. this disagreement itself is part of what forms the foundation for the debate in the first place. some may here object that everyone involved is well aware of this and investigating the competing hypotheses involves exactly making predictions about their location. that is standard scientific practice. it is the testing of these predictions that in the end (sometime far into the future) will allow us to distinguish between the hypotheses empirically and determine the real nccs and the nature of consciousness. the extrapolation of topographical regions of interest based on conceived similarity in functional characteristics from data obtained in behavioural paradigms is not per se misguided. at least to the extent that this practice merely serves to inform new paradigms. what, on the other hand, is misguided is thinking that topographical data obtained through behavioural studies of other phenomena warrants bypassing further behavioural studies when it comes to consciousness, i.e., we cannot let the default be reasoning directly from our (perceived) similarity in functional characteristics of one process to the claim that another process is likely to (or must) reside in the same brain region. this problem is further aggravated if we allow ourselves to count empirical findings as evidence for or against a conceptual theory simply because some or other brain region is involved. if we want to truly deploy empirical sciences to establish the nccs and distinguish between competing theories of consciousness, we must do so on the basis of differing behavioural predictions rather than predicting activation profiles in brain regions. overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org finding the neural correlates of consciousness will not solve all our problems 13 experiments that attempt to replace subjective reports with objective behaviour, and behaviour with neural activation, face specific problems. if one is to avoid using subjective reports, one is forced to answer the obvious question: without relying on reports, how do we know that a particular no-report paradigm measures consciousness and not something else? what reasons could one have to think of, say, a binocular rivalry experiment without report as a method to investigate conscious experience per se in the first place (frässle et al., 2014; overgaard & fazekas, 2016)? arguably, the only reason is that one introspectively attends to what it feels like to experience binocular rivalry. in other words, the only way to invent a measure of consciousness without asking others to report seems to depend on the scientist’s own intuitions, which could hardly be said to be any less based on introspection. another answer could be that if a particular type of behaviour or phenomenon has been found to be associated with subjective experience in previous experiments using subjective reports, this behaviour or phenomenon can be used as an ‘objective measure’ to replace the report. however, this would carry any methodological weakness in reporting forward—as the identified measure is only a reliable measure under the condition that the correlation with the subjective report was reliable in the first place. ironically, this requires the subjective measure to be a reliable type of measure, which would eliminate the reason for finding other measures to begin with. if two competing theories of consciousness have different predictions about neural correlates of consciousness yet identical predictions about behaviour, the two theories will be difficult to compare. this is in fact typically the case in current debates between first and higher-order theories of consciousness. based on the reasoning above, consciousness theories should be able to propose specific outcomes of behavioural experiments to be empirically compared directly. 8 concluding remarks given the shortcomings discussed above, changes in the interdisciplinary practice are warranted to ensure a fit between neural evidence about the brain and conceptual theories about consciousness. we have offered advice indicating two different ways to go about this. our first advice concerned how to bridge the missing isomorphism between theories in the conceptual domain and neural data. we argued that to bridge the conceptual disconnect between the empirical and conceptual domains, bottom-up revision of the conceptual theories is required. this is captured in the dictum because we cannot make the data fit our theories, we should make our theories fit the data. our second piece of advice concerns how to test competing conceptual theories. with regard to this, we advocated the necessity of competing conceptual theories delivering differing behavioural predictions. overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org morten s. overgaard and asger kirkeby-hinrup 14 we do not purport that these are the only ways forward and acknowledge that each requires further refinement than we can provide in this limited space. however, the two different ways may serve as the foundation for further development and facilitate future research in this area. acknowledgments akh is funded by the swedish research council (grant # 2018-06595). morten overgaard was supported for this work by a jens christian skou fellowship from aarhus institute of advanced studies. references beeckmans, j. (2007). can higher-order representation theories pass scientific muster? journal of consciousness studies, 14(9-1), 90–111. block, n. (2019). what is wrong with the no-report paradigm and how to fix it. trends in cognitive sciences, 23(12), 1003– 1013. https://doi.org/10.1016/j.tics.2019.10.001 boly, m., massimini, m., tsuchiya, n., postle, b. r., koch, c., & tononi, g. (2017). are the neural correlates of consciousness in the front or in the back of the cerebral cortex? clinical and neuroimaging evidence. journal of neuroscience, 37(40), 9603–9613. https://doi.org/10.1523/jneurosci.3218-16.2017 brown, r. (2012). the brain and its states. in s. edelman, t. fekete, & n. zach (eds.), being in time: dynamical models of phenomenal experience (vol. 88, pp. 211–238). https://doi.org/10.1075/aicr.88.10bro brown, r., lau, h., & ledoux, j. e. (2019). understanding the higher-order approach to consciousness. trends in cognitive sciences, 23(9), 754–768. https://doi.org/10.1016/j.tics.2019.06.009 carandini, m. (2012). from circuits to behavior: a bridge too far? nature neuroscience, 15(4), 507–509. https://doi.org/10.1 038/nn.3043 churchland, p. m. (1981). eliminative materialism and the propositional attitudes. journal of philosophy, 78(2), 67–90. https://doi.org/10.2307/2025900 crouzet, s. m., kovalenko, l. y., del pin, s. h., overgaard, m., & busch, n. a. (2017). early visual processing allows for selective behavior, shifts of attention, and conscious visual experience in spite of masking. consciousness and cognition, 54, 89–100. https://doi.org/10.1016/j.concog.2017.01.021 crouzet, s. m., overgaard, m., & busch, n. a. (2014). the fastest saccadic responses escape visual masking. plos one, 9(2), e87418. https://doi.org/10.1371/journal.pone.0087418 da costa, n., & french, s. (1990). the model-theoretic approach in the philosophy of science. philosophy of science, 57(2), 248–265. https://doi.org/10.1086/289546 fink, s. b. (2016). a deeper look at the “neural correlate of consciousness.” frontiers in psychology, 7. https://doi.org/10.338 9/fpsyg.2016.01044 fleming, s. m., & dolan, r. j. (2012). the neural basis of metacognitive ability. philosophical transactions of the royal society b: biological sciences, 367(1594), 1338–1349. https://doi.org/10.1098/rstb.2011.0417 frässle, s., sommer, j., jansen, a., naber, m., & einhäuser, w. (2014). binocular rivalry: frontal activity relates to introspection and action but not to perception. journal of neuroscience, 34(5), 1738–1747. https://doi.org/10.1523/jneuro sci.4403-13.2014 frith, c. d., & frith, u. (2006). the neural basis of mentalizing. neuron, 50(4), 531–534. https://doi.org/10.1016/j.neuron.2 006.05.001 genon, s., reid, a., langner, r., amunts, k., & eickhoff, s. b. (2018). how to characterize the function of a brain region. trends in cognitive sciences, 22(4), 350–364. https://doi.org/10.1016/j.tics.2018.01.010 glock, h. (1997). kant and wittgenstein: philosophy, necessity and representation. international journal of philosophical studies, 5(2), 285–305. https://doi.org/10.1080/09672559708570857 kirkeby-hinrup, a. (2020). a higher-order faculty and beyond. in overgaard, morten and mogensen, jesper and kirkebyhinrup, asger (ed.), beyond the neural correlates of consciousness (pp. 131–152). psychology press, routledge. koch, c., massimini, m., boly, m., & tononi, g. (2016). neural correlates of consciousness: progress and problems. nature reviews neuroscience, 17, 307. https://doi.org/10.1038/nrn.2016.22 overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.tics.2019.10.001 https://doi.org/10.1523/jneurosci.3218-16.2017 https://doi.org/10.1075/aicr.88.10bro https://doi.org/10.1016/j.tics.2019.06.009 https://doi.org/10.1038/nn.3043 https://doi.org/10.1038/nn.3043 https://doi.org/10.2307/2025900 https://doi.org/10.1016/j.concog.2017.01.021 https://doi.org/10.1371/journal.pone.0087418 https://doi.org/10.1086/289546 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.1098/rstb.2011.0417 https://doi.org/10.1523/jneurosci.4403-13.2014 https://doi.org/10.1523/jneurosci.4403-13.2014 https://doi.org/10.1016/j.neuron.2006.05.001 https://doi.org/10.1016/j.neuron.2006.05.001 https://doi.org/10.1016/j.tics.2018.01.010 https://doi.org/10.1080/09672559708570857 https://doi.org/10.1038/nrn.2016.22 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org finding the neural correlates of consciousness will not solve all our problems 15 koivisto, m., & revonsuo, a. (2010). event-related brain potential correlates of visual awareness. neuroscience & biobehavioral reviews, 34(6), 922–934. https://doi.org/10.1016/j.neubiorev.2009.12.002 kozuch, b. (2014). prefrontal lesion evidence against higher-order theories of consciousness. philosophical studies, 167(3), 721–746. https://doi.org/10.1007/s11098-013-0123-9 kriegel, u. (2007). a cross-order integration hypothesis for the neural correlate of consciousness. consciousness and cognition, 16(4), 897–912. https://doi.org/10.1016/j.concog.2007.02.001 lamme, v. a. f. (2006). towards a true neural stance on consciousness. trends in cognitive sciences, 10(11), 494–501. https://doi.org/10.1016/j.tics.2006.09.001 lamme, v. a. f. (2018). challenges for theories of consciousness: seeing or knowing, the missing ingredient and how to deal with panpsychism. philosophical transactions of the royal society b: biological sciences, 373(1755). https: //doi.org/10.1098/rstb.2017.0344 lau, h. (2007). a higher order bayesian decision theory of consciousness. progress in brain research, 168, 35–48. https: //doi.org/10.1016/s0079-6123(07)68004-2 lau, h. (2011). theoretical motivations for investigating the neural correlates of consciousness. wiley interdisciplinary reviews: cognitive science, 2(1), 1–7. https://doi.org/10.1002/wcs.93 lau, h., & brown, r. (2019). the emperor’s new phenomenology? the empirical case for conscious experiences without first-order representations. in a. pautz & d. stoljar (eds.), blockheads! essays on ned block’s philosophy of mind and consciousness (pp. 171–197). mit press. lau, h., & rosenthal, d. (2011). empirical support for higher-order theories of conscious awareness. trends in cognitive sciences, 15(8), 365–373. https://doi.org/10.1016/j.tics.2011.05.009 michel, m., & morales, j. (2020). minority reports: consciousness and the prefrontal cortex. mind & language, 493–513. https://doi.org/10.1111/mila.12264 mogensen, j., & overgaard, m. (2018). reorganization of the connectivity between elementary functions as a common mechanism of phenomenal consciousness and working memory: from functions to strategies. philosophical transactions of the royal society b: biological sciences, 373(1755), 20170346. https://doi.org/10.1098/rstb.2017.0346 morales, j., & lau, h. (forthcoming). the neural correlates of consciousness. in u. kriegel (ed.), the oxford handbook of the philosophy of consciousness. oxford university press. naccache, l. (2018). why and how access consciousness can account for phenomenal consciousness. philosophical transactions of the royal society b: biological sciences, 373(1755), 20170357. https://doi.org/10.1098/rstb.2017.0357 nagel, t. (1974). what is it like to be a bat. philosophical review, 83(4), 435–450. https://doi.org/10.2307/2183914 odegaard, b., knight, r. t., & lau, h. (2017). should a few null findings falsify prefrontal theories of conscious perception? journal of neuroscience, 37(40), 9593–9602. https://doi.org/10.1523/jneurosci.3217-16.2017 overgaard, m. (2015). behavioral methods in consciousness research. oxford university press. overgaard, m. (2017). the status and future of consciousness research. frontiers in psychology, 8(1719). https://doi.org/10.3 389/fpsyg.2017.01719 overgaard, m. (2018). phenomenal consciousness and cognitive access. philosophical transactions of the royal society b: biological sciences, 373(1755), 20170353. https://doi.org/10.1098/rstb.2017.0353 overgaard, m., & fazekas, p. (2016). can no-report paradigms extract true correlates of consciousness? trends in cognitive sciences, 105(20(4)), 241–242. https://doi.org/10.1016/j.tics.2016.01.004 overgaard, m., & mogensen, j. (2011). a framework for the study of multiple realizations: the importance of levels of analysis. frontiers in psychology, 2, 79. https://doi.org/10.3389/fpsyg.2011.00079 pessoa, l., thompson, e., & noë, a. (1998). finding out about filling-in: a guide to perceptual completion for visual science and the philosophy of perception. behavioral and brain sciences, 21(6), 723–748. https://doi.org/10.1017/s0140525x9 8001757 peters, m. a., kentridge, r. w., phillips, i., & block, n. (2017). does unconscious perception really exist? continuing the assc20 debate. neuroscience of consciousness, 3(1). https://doi.org/10.1093/nc/nix015 pinto, y., vandenbroucke, a. r., otten, m., sligte, i. g., seth, a. k., & lamme, v. a. (2017). conscious visual memory with minimal attention. journal of experimental psychology: general, 146(2), 214–226. https://doi.org/10.1037/xge0000255 railo, h., revonsuo, a., & koivisto, m. (2015). behavioral and electrophysiological evidence for fast emergence of visual consciousness. neuroscience of consciousness, 2015(1). https://doi.org/10.1093/nc/niv004 sebastián, m. á. (2014). not a hot dream. in r. brown (ed.), consciousness inside and out: phenomenology, neuroscience, and the nature of experience (vol. 6, pp. 415–432). springer netherlands. timmermans, b., & cleeremans, a. (2015). how can we measure awareness? an overview of current methods. in m. overgaard (ed.), behavioral methods in consciousness research (pp. 21–46). oxford university press. overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neubiorev.2009.12.002 https://doi.org/10.1007/s11098-013-0123-9 https://doi.org/10.1016/j.concog.2007.02.001 https://doi.org/10.1016/j.tics.2006.09.001 https://doi.org/10.1098/rstb.2017.0344 https://doi.org/10.1098/rstb.2017.0344 https://doi.org/10.1016/s0079-6123(07)68004-2 https://doi.org/10.1016/s0079-6123(07)68004-2 https://doi.org/10.1002/wcs.93 https://doi.org/10.1016/j.tics.2011.05.009 https://doi.org/10.1111/mila.12264 https://doi.org/10.1098/rstb.2017.0346 https://doi.org/10.1098/rstb.2017.0357 https://doi.org/10.2307/2183914 https://doi.org/10.1523/jneurosci.3217-16.2017 https://doi.org/10.3389/fpsyg.2017.01719 https://doi.org/10.3389/fpsyg.2017.01719 https://doi.org/10.1098/rstb.2017.0353 https://doi.org/10.1016/j.tics.2016.01.004 https://doi.org/10.3389/fpsyg.2011.00079 https://doi.org/10.1017/s0140525x98001757 https://doi.org/10.1017/s0140525x98001757 https://doi.org/10.1093/nc/nix015 https://doi.org/10.1037/xge0000255 https://doi.org/10.1093/nc/niv004 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org morten s. overgaard and asger kirkeby-hinrup 16 tononi, g., boly, m., massimini, m., & koch, c. (2016). integrated information theory: from consciousness to its physical substrate. nature reviews neuroscience, 17(7), 450–461. https://doi.org/10.1038/nrn.2016.44 weisberg, j. (2014). sweet dreams are made of this? a hot response to sebastián. in r. brown (ed.), consciousness inside and out: phenomenology, neuroscience, and the nature of experience (vol. 6, pp. 433–443). springer netherlands. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. overgaard, m. s., & kirkeby-hinrup, a. (2021). finding the neural correlates of consciousness will not solve all our problems. philosophy and the mind sciences, 2, 5. https://doi.org/10.33735/phimisci.2021.37 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1038/nrn.2016.44 https://doi.org/10.33735/phimisci.2021.37 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction correlations, subjective and objective measures the missing isomorphism the missing homomorphism activation profiles are insufficient to distinguish between theories facing up to the missing isomorphism back to behavioural methods concluding remarks you can't always get what you want: predictive processing and consciousness. you can’t always get what you want predictive processing and consciousness tobias schlichta (tobias.schlicht@rub.de) krzysztof dołęgaa (krzysztof.dolega@rub.de) abstract the predictive processing framework has gained significant popularity across disciplines investigating the mind and brain. in this article we critically examine two of the recently made claims about the kind of headway that the framework can make in the neuroscientific and philosophical investigation of consciousness. firstly, we argue that predictive processing is unlikely to yield significant breakthroughs in the search for the neural correlates of consciousness as it is still too vague to individuate neural mechanisms at a fine enough scale. despite its unifying ambitions, the framework harbors a diverse family of competing computational models which rely on different assumptions and are under-constrained by neurological data. secondly, we argue that the framework is also ill suited to provide a unifying theory of consciousness. here, we focus on the tension between the claim that predictive processing is compatible with all of the leading neuroscientific models of consciousness with the fact that most attempts explaining consciousness within the framework rely heavily on external assumptions. keywords bayesian brain hypothesis ∙ consciousness ∙ instrumentalism ∙ overflow ∙ predictive processing ∙ realism this article is part of a special issue on “the neural correlates of consciousness,” edited by sascha benjamin fink and ying-tung lin. 1 introduction the predictive processing (pp) framework has caused a lot of excitement among philosophers and cognitive scientists in the last decade. its main proponents take it to provide a “unified account” of brain function (friston, 2010), integrating recent developments in computational neuroscience with the insights of bayesian aruhr-university bochum, institute of philosophy ii schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-1829-9253 https://orcid.org/0000-0001-6859-6179 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 2 psychology. moreover, it is allegedly “the first truly unifying account of perception, cognition and action” (clark, 2016, p. 2), having the potential to “explain perception and action and everything mental in between” (hohwy, 2013, p. 1). this is because accounts of cognitive phenomena formulated within this framework posit only one fundamental type of information processing, i.e., prediction error minimization, reiterated throughout the brain on many levels in a complex computational hierarchy. the core idea is that the brain harbours an internal model of the world, the cognitive agent, and the agent’s biological interior milieu, which constantly generates expectations (or guesses) about incoming stimuli and their sources on all levels of the hierarchy. these top-down predictions about (the causes of) sensory input are constantly compared to the actual incoming sensory inputs, yielding (bigger or smaller) prediction errors. the internal model and world can be matched in one of two ways: either the input is used to update the model that yielded the predictions in the first place (perceptual inference) or the sensory input is changed via action in order to match the model (active inference).1 proponents of such approaches have already demonstrated its strong potential to accommodate various mental phenomena surrounding perception, cognitive penetration, perceptual binding, and attention (cf. hohwy, 2013, pp. 101–138). in line with their impressive prior contributions to pp, hohwy and seth (2020) as well as clark, friston, and wilkinson (2019), and solms (2021) among others, have recently stressed the potential of this framework to address consciousness. hohwy and seth issued a programmatic statement trying to make good on the claim that predictive processing can also inform us about different facets of conscious experience and guide the search for systematic neural correlates of consciousness (nccs). they suggest that pp offers the most promising approach for embedding the on-going search for the nccs within a unifying framework, one which can even motivate and operationalize “closer links between phenomenological properties of conscious experience and mechanistic properties of underlying neural substrates” (hohwy & seth, 2020, p. 25). although they admit that the “pp approach is not, in itself, a theory of consciousness” (hohwy & seth, 2020, p. 2), they claim that it is this feature that makes it uniquely fit for providing unification, while remaining optimistic for “the prospect that, eventually, aspects of pp may themselves coalesce into a theory of consciousness in its own right” (hohwy & seth, 2020, p. 24). hohwy and seth’s enthusiasm for pp’s ability to facilitate consciousness research echoes that of authors like clark, friston, and wilkinson (2019) as well as clark (2019), who have proposed ways in which pp can contribute to the philosophical side of explaining consciousness. in particular, they address the so-called hardand meta-problems of consciousness (chalmers, 1996, and 2018 1although perceptual and active inference are often distinguished in the literature, the former is formally subsumed by the latter in more recent versions of the framework which assume that the system is minimizing its long-term expected free energy. minimizing this quantity is taken to be equivalent to maximizing expected utility while simultaneously reducing the uncertainty about the possible causes of valuable outcomes (cf. friston et al., 2015). schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 3 respectively) which are concerned with the subjective qualities of experience (commonly referred to as qualia) and the judgements we make and intuitions we have about consciousness. similarly, solms (2021) commits explicitly to friston’s freeenergy principle in his defense of a “free energy theory of consciousness” (2021, p. 2), a variant of the pp framework, but combines it with suggestions from his work on freudian neuro-psychoanalysis. however, it is still unclear whether pp can deliver on all of its promises. while we share much of hohwy and seth’s optimism, we are also keenly aware of the framework’s many limitations as well as outstanding problems, which are often downplayed by its proponents. for example, pp’s generality and concomitant compatibility with different existing theories of consciousness could be seen as an obstacle for making a substantial contribution to any specific debate in this area, rather than as an advantage. as we will show, in order to make such contributions, pp must be enriched with further, often externally motivated, assumptions which raise the question whether it is the framework itself or these additional postulates that carry the explanatory burden. in this paper, we will critically evaluate the explanatory potential of the pp framework with respect to determining the nccs as well as providing an account of conscious experience more generally. some of our points are general concerns about the features and the explanatory status of the framework, others are directed at specific variants of pp and can be seen as direct responses to what we see as the most formidable contributions made by hohwy and seth, clark et al., and others. we assume the reader to be familiar with the basics of this explanatory framework for reasons of space.2 nevertheless, we begin the paper by introducing the ncc research program and rehearsing the main points of hohwy and seth’s contribution, in order to clarify the stakes of the debate about nccs as well as to ensure we do not misrepresent their position (section 2). the remainder of the paper is then divided into two general parts. in section 3, we turn to the general issue of scientific realism with respect to the bayesian brain, discuss methodological issues surrounding the pp framework, and call into question its supremacy over competing modelling frameworks. section 4 turns to the philosophical issues surrounding pp and consciousness. first, we point to the fact that the framework cannot answer the hard problem without making additional assumptions. we then evaluate the legitimacy of these assumptions, showing that although the framework is compatible with multiple competing theories of consciousness, the assumptions which allow for this are either motivated by considerations external to the framework or tend to lead to puzzling predictions about phenomenal experience. this is then illustrated using the familiar example of the müller-lyer illusion. in sum, we demonstrate that the optimistic attitude towards the prospects of pp contributing to scientific work on consciousness is quite limited, when focusing only on its central theoretical commitments. thus, in order to contribute to some of the most 2see metzinger and wiese (2017) for an accessible overview of its main claims and assumptions. schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 4 interesting and controversial on-going debates surrounding conscious experience, pp must either be amended with additional assumptions or remain silent. 2 pp and the search for the ncc 2.1 determining neural correlates of consciousness despite many methodological challenges and philosophical objections, one of the core projects of current scientific research on consciousness is the search for the ‘neural correlates of consciousness’ or nccs (see chalmers, 2000). the neural correlates in question are usually “defined as the minimal neuronal mechanisms jointly sufficient for any specific conscious percept” (tononi & koch, 2008, p. 239; see also fazekas & overgaard, 2018; fink, 2016; and noë & thompson, 2004 for sceptical looks at this research program). in practice, the aim of this research program is to find brain states or processes which reliably correlate with different aspects of consciousness, which here is used as an umbrella term covering a diverse range of phenomena related to how we experience the world with us in it. ncc researchers commonly distinguish between consciousness in the sense of a global state of vigilance and specific conscious experiences of something in various sensory or cognitive modalities. the former allows an organism to receive sensory stimulation and process information at all and comes in degrees, ranging from full wakefulness via various levels of sleep, anaesthesia, vegetative state, all the way down to deep coma (dehaene et al., 2006; tononi et al., 2016; tononi & koch, 2008). consciousness in this sense is sometimes also referred to as ‘creature consciousness’ (bayne, 2007; hohwy, 2009). researchers have already made substantial progress on this aspect of consciousness and its nccs (see e.g., fernándezespejo & owen, 2013 for a helpful overview). for example, the state of vigilance has been associated with activity in the brainstem, as one’s ability to stay alert diminishes after sustaining damage to that area (damasio, 2011). the other use of the term ‘consciousness’ refers to particular experiences such as that of pain, color, objects, and so on. these states of consciousness are distinguished by their (phenomenal) contents and can be considered as modifications of our overall unified conscious state at a particular time (bayne & chalmers, 2003; hohwy, 2009; searle, 1992). for example, looking outside my open window, i may simultaneously see trees in the garden, hear birds singing, and smell the fresh air, accompanied by a mixture of different feelings and thoughts. the assumption of the ncc project is that for every conscious percept there will be a corresponding neural correlate such that artificially inducing the same activity in the correlated brain structure will induce the percept while disrupting it will eliminate the percept. the ncc research program emphasises identifying the minimal neuronal mechanisms since presumably not all brain areas or neural processes will play a role in generating conscious experience. for example, although it contains many more neurons than the cerebral cortex, the cerebellum does not seem to contribute schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 5 to the generation of conscious experience (tononi & koch, 2015). moreover, the areas which do contribute, need not necessarily do so in the same way. one goal of this research is thus to determine the neuronal mechanism sufficient for producing one particular percept instead of another one. this minimal mechanism is sometimes called the ‘core’ ncc, and it will always be embedded within a ‘total’ ncc, i.e., a much more encompassing neural activation pattern that also comprises various other mechanisms, e.g., mechanisms in the brainstem necessary for sustaining global consciousness (or vigilance) as described above (see also marvan & polák, 2020). in this way, searching for these more specific mechanisms holds promise, ultimately, to arrive at a mechanistic explanation of consciousness. 2.2 pp and the search for systematic nccs in their recent article, hohwy and seth follow chalmers (2000) in arguing that the search for the ncc of any particular content of experience should be interlocked systematically with research into the correlates of other aspects of experience, such as levels of consciousness as well as phenomenology. otherwise, any identified ncc would be arbitrary, as merely observing patterns of neural activation does not, by itself, allow for uncovering neural mechanisms underlying consciousness. merely registering that consecutive activity in different neural structures, say x, y, and z, is correlated with some type of experience would not explain what each of these areas contributes to the experience, or why these and not other neural pathways are responsible for that type of experience. the way forward, in their view, is to rely on a larger unified theory of brain function that could yield better predictions about the connections between brain activity and phenomenal states, since being a systematic ncc boils down to “being systematically guided by theoretical considerations of some sort” (hohwy & seth, 2020, p. 4). in this vein, hohwy and seth identify three main challenges of ncc research. the first one is that of relating local nccs for particular contents to global states of consciousness in terms of the overall functioning of the subject’s neural system. the fact that any phenomenal state is only a modification of the subject’s overall conscious state must inform the search for the nccs. the second and related issue is that of teasing apart neural activations which are causally contributing to a given experience from those that merely coincide with it. at any given time, there will be many patterns of neural activation differing in resolution and specificity (see e.g., aru et al., 2012). finding the minimal relevant process requires making inferences about the causal structure of information processing in the brain. finally, hohwy and seth stress that the search for nccs should integrate action as causally or constitutively relevant for conscious experience. this constraint is in line with recent enactive and embodied approaches to consciousness and cognition (kirchhoff & kiverstein, 2018; see, e.g., noë, 2005) as well as the methodological need for integrating behavioral measures into neuroscience (krakauer et al., 2017). schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 6 in addition to the above challenges, hohwy and seth also identify two constraints which any promising framework for guiding the search for systematic nccs should meet. even though the theoretical considerations which guide this search should not be specifically tied to a particular theory of consciousness, they should, the authors argue, respect the insights of the current neuroscientific consensus about the phenomenon under investigation. by analyzing multiple prominent neuro-centric theories of consciousness, hohwy and seth manage to identify two key features which most views hold to be crucial for conscious experience – uncertainty reduction and top-down signalling.3 unsurprisingly, these key features also happen to be the core elements of the pp framework. while the bulk of hohwy and seth’s paper is devoted to demonstrating how pp can successfully address the three challenges and fit within their constraints in order to provide a theoretical basis for systematic nccs, we think that the core argument of their paper rests on the assumption that pp is not only a correct description of the functions carried out by the brain, but that it is aimed at uncovering the actual neural mechanisms (in the sense of illari & williamson, 2010) underpinning consciousness. as the authors state in their recent paper: “the language of pp enables new and productive mappings to be drawn between phenomenological descriptions and underlying neurocognitive processes” (hohwy & seth, 2020, p. 18). it is this assumption that allows them to argue for the explanatory and predictive power of the framework when it comes to the search for nccs. simply put, if pp can uncover the real organization and causal interactions between neural components at different spatial and temporal levels, then it can guide meaningful interventions on the mechanism’s components, answering to the problem of inferring which patterns of neural activity are causally relevant to a given experience. furthermore, hohwy and seth argue that by offering a complete framework for describing neural mechanisms, pp provides a useful way to systematize and conceptualize conscious experience in a way that will help to solve remaining problems. they illustrate this point by noting that pp already recognizes the features shared by most theoretical proposals to consciousness (reduction of uncertainty and the importance of top-down signalling mentioned above). again, the pp approach turns out to not only be well positioned to account for these two features, but also offers a unification of the relevant aspects of different neuroscientific accounts of consciousness in a way not afforded by “any other theoretical framework” (hohwy & seth, 2020, p. 24). in this paper we want to question what we see as the two claims made by hohwy and seth which also drive much of the enthusiasm for the contributions 3examples of uncertainty reduction provided by hohwy and seth include the evidence accumulation threshold for ignition postulated by the global neuronal workspace (dehaene, 2014), the signal-to-noise ratio criterion postulated by versions of higher-order theories based on signal detection (lau, 2008), and the exclusion of possibilities used to define conscious states in the integrated information theory (oizumi et al., 2014). similarly, the authors point out that some form of top-down modulation is not only included by the views listed above, but applies also to the recurrent processing theory of consciousness (lamme, 2006). schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 7 that the framework can make in explaining consciousness. the first of these claims is the presupposition that pp delivers a mechanistic description of brain function, while the second is the claim that the framework offers a complete basis for developing a mature theory of consciousness. as we see it, the kind of work that pp can do is best understood in terms of providing a high-level conceptual/computational framework, but this kind of abstract description does not supply a detailed enough account of how its algorithms can be implemented in neural tissue to deliver the details necessary for guiding the search for systematic nccs. moreover, as we show in the later part of this article, such a high-level conceptualization of neural function can be used for formulating an account of consciousness, but it will also need to rely on externally motivated assumptions to achieve this, and even then it may fail to account for some of the aspects of conscious experience. 3 bayes in the brain? like many of the authors working within pp, hohwy and seth are committed to a realist attitude, which can be summarized in the expression that “the brain is a bayesian mechanism” (hohwy, 2013, p. 25). a crucial element of this account is that the brain is equipped with “a single type of mechanism, reiterated throughout the brain” (hohwy, 2013, p. 2), which can account for all cognitive functions. this is often (and perhaps erroneously, see colombo & hartmann, 2017) thought to be the source of the framework’s explanatory power. in its strongest form, the conviction about the unifying and explanatory credentials of the framework is expressed in friston’s claim that “if one looks at the brain as implementing this [i.e., pp] scheme, nearly every aspect of its anatomy and physiology starts to make sense” (2009, p. 293). however, pp’s ontological and explanatory status is one of the more contentious issues in philosophy of mind and cognitive science (see e.g., aitchison & lengyel, 2017; bruineberg et al., 2020; colombo & wright, 2017; heilbron & chait, 2018). hohwy and seth acknowledge this when they point out that “the pp framework can be cast at different levels of abstraction which make different claims about the underlying mechanism” (hohwy & seth, 2020, p. 15). they go on to distinguish between ‘strong’ and ‘weak’ interpretations of the framework. the former are committed to the brain “in fact utilizing generative models and approximate bayesian inference to accomplish prediction error minimization” (2020, p. 15), while the latter are mere re-descriptions of cognitive systems “as if they are utilizing such models and inferential processes, though the underlying physical-causal mechanism may be different” (2020, p. 15, emphasis added). unfortunately, this distinction between strong and weak interpretations is not as clear as it might seem at first. on the one hand, any model that recapitulates the joint probability distribution of some observable variable x and a latent variable y can be formally described as a generative model (bishop, 2006; ng & jordan, 2001). on the other hand, psychologically plausible approximations to bayesian inschlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 8 ference can be achieved without the calculation of explicit probabilities (sanborn & chater, 2016). hohwy and seth refine the strong approach to pp by pointing out that their commitment to a general, computational-level description provided by the framework constrains the space of viable process theories that can be “implemented mechanistically” through biologically plausible algorithms, which in the case of pp commonly include predictive coding or gradient descent on variational free energy (hohwy & seth, 2020, p. 15). however, this way of elaborating the strong interpretation of pp creates the possibility that the framework might offer an accurate description of the computational function (i.e., the abstract mathematical operation) carried out by the cognitive system (in this case approximate bayesian inference) and still be wrong about the details of how this function is implemented or approximated by the brain (in this case whether it is done via bottom-up propagation of prediction error). in the next two sections we will show that neither a commitment to a particular algorithm nor a commitment to a more coarse-grained view of pp as a high-level conceptual/computational framework are enough to clear some of the challenges for securing systematic nccs. 3.1 the issue of fine-grained realism about pp recall from section 2.2 that in order to establish a systematic ncc, one needs to have means of directing interventions on the brain in such a way as to tease apart which patterns of neural activity are relevant to a given experience. thus, systematic nccs can only be found if we have some form of a blueprint about how pp can be implemented in the brain. however, as we will now show, at least one of the ways in which the framework has been claimed to be realized in the brain suffers from severe problems of underdetermination. let us turn to predictive coding (pc). although pc is a popular algorithmic solution among the framework’s proponents, there are many ways in which this kind of algorithm could be realized by neural structures (spratling, 2017). for example (and without getting into too much detail), the most prevalent version of pc originating with rao and ballard (1999) and further developed by friston (2010) and colleagues assumes that neural networks perform inferences through the iterative process of generating predictions and minimizing prediction error – the error in the reconstruction of the input – by updating the coefficients (i.e., the hypotheses about the possible causes of sensory stimulation) so as to minimize the error of future predictions.4 importantly, pc schemes growing out of rao and ballard’s formulation assume prediction errors to be coded as a difference between the expected and actual inputs, and signalled by inhibitory feedback connections. such connections, as spratling explains, are usually assumed to be realized by the axon projections of a sub-population of pyramidal cells. however, the problem is that only a relatively minor number of such connections (10–20%) terminate in 4effectively the network is utilizing a common machine learning technique known as gradient descent on residual error (bishop, 2006, p. 240). schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 9 inhibitory neurons, while for the vast majority “the primary targets are other pyramidal cells […], where cortical feedback has an excitatory (modulatory) effect on response; […] at least in the short term […]” (spratling, 2017, p. 6). furthermore, rao and ballard’s algorithm involves several other biologically implausible idealizations, such as allowing for the modelled neurons to have negative firing rates (cf. spratling, 2017). to deal with these issues, spratling has devised a competing pc algorithm that aims to alleviate some of the pitfalls of the rao and ballard model by utilizing division rather than subtraction for calculating errors (spratling, 2008a, 2008b; spratling et al., 2009). this pc algorithm based on divisive input modulation (we will call it pc-dim for short) has proven successful in not only being able to account for neural phenomena like biased competition, but has also allowed for simulations as many as tens of millions of neurons with hundreds of billions of connections (cf. spratling, 2017). however, the biggest and most significant difference between these two families of pc algorithms is that spratling proposes a different grouping of neural populations into processing stages than the one used by rao and ballard. this is the most empirically salient of the changes as it means that the two algorithms will map onto cortical structures differently (see spratling, 2017). the pc-dim algorithm is consistent with the data about the neural feedback pathways because it groups neural populations in a way that requires all intercortical connections to be excitatory. this implies that all inhibitory connections projecting from error neurons terminate within the processing stage/cortical area in which they are embedded. it should be noted that, while this complicates the typical pc story – since the inter-stage bottom-up channels are no longer communicating an arithmetic difference between the predictions and the input –, spratling’s algorithm still uses them to communicate the way in which one probability distribution diverges from another, effectively implementing an approximation to bayesian inference. thus, even though the pc-dim algorithm does not utilize the forward pathways to propagate prediction errors, but rather the information about the most likely cause, it is still consistent with the more abstract computational scheme postulated by pp. the example of spratling’s pc-dim algorithm is helpful in bringing out the first problem with the underdetermination of fine-grained realism about pp and its posits. it shows that process models employed in the framework allow for different kinds of idealizations, meaning that two competing realizations of the same algorithm that postulate different mappings onto cortical structures and, consequently, different ways in which information flows can be consistent with the same neuroanatomical evidence. however, this is not the only problem of underdetermination that the framework is facing. rosa cao (2020) has recently cautioned that the underdetermination of the algorithmic solutions by the empirical data goes beyond different varieties of the same computational scheme and that the same neuroanatomical evidence might be equally consistent with pp as well as other, more traditional message-passing schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 10 algorithms. she shows that – as far as the flow of information is concerned – pp models of neural signalling can be “relabelled in traditional, non-predictive terms, with no empirical consequences relevant to existing or future data” (cao, 2020, p. 517).5 labelling a neural activity ‘predictive’ or not is not constrained by the empirical data gathered by scientists unless further assumptions about the coding scheme or implementation relations connecting neural contents to neural signals can be made (cao, 2020, p. 522). if we assume that pp proponents take notions like ‘model,’ ‘prediction,’ ‘error’ etc. to be representational, i.e., about something, then what is missing is a specification of how to yield a “mapping from informational contents to measurable neural activities” (cao, 2020, p. 524). as cao points out, the way to establish this kind of mapping is by exploiting some degree of correlation (and perhaps structural resemblance, see e.g., shea, 2018) between the elements of the model and the brain. one commonly used framework is that of shannon’s communication theory under which some neural populations are assumed to act as sources of information for other neural systems which act as receivers. cao argues that what matters is whether a receiving system can decode and react to that information. however, in order for the receiver to decode information, the probability of a signal given the source needs to be specified. in other words, the receiver needs to know what the prior probability of a signal is, in order to react to it appropriately. thus, cao’s main point is that both the traditional framework and the predictive framework must specify this set of possible states for the system in question which can then be narrowed down by the incoming signal. but then, “given the prior, sending error and sending stimulus have the same effects from the point of view of the system” (cao, 2020, p. 526). therefore, it does not add anything of importance whether we call the statistical information ‘error signal’ or ‘bottom-up input.’ in this information-theoretic respect, which is crucial for the impact and originality of the framework (as stressed by hohwy and seth), pp is demonstrably equivalent to traditional frameworks, not competitive. another way of putting it is to say that pp adds a different ‘gloss’ (in the sense proposed in egan, 2014, 2020) to the same familiar information-theoretic story. this conclusion seems to be difficult to swallow for pp realists like clark, hohwy, or seth as it means that proponents of pp “have little new to say about the causal structure of the system” in question (cao, 2020, p. 517). but if this is the case, then how can pp clear the requirement for systematic nccs? the two ways in which pp is underdetermined by its implementation present a significant 5spratling (2008a, 2008b) and issa et al. (2018) make similar points, while summerfield and de lange (2014) insist that, under some assumptions, pc and evidence accumulation algorithms can be shown to be formally equivalent. finally, aitchison and lengyel end their analysis of different neural implementations of bayesian and predictive coding schemes by pointing out that “while predictive coding is an attractive algorithmic idea that accounts for a remarkable range of phenomena, the experimental evidence for it seems inconclusive in the sense that it does not rule out bayesian inference with a direct variable code, potentially in combination with a variety of non-probabilistic processes including attention and adaptation.” (aitchison & lengyel, 2017, p. 224). schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 11 difficulty in determining which neural signals or cell populations should be interpreted as carrying predictions within a hierarchical model and which should be interpreted as prediction errors. this is especially relevant to the neuroscience of consciousness, since part of the controversy surrounding the potential nccs is whether local recurrent signals (e.g., in the primary visual areas) are sufficient for (visual) consciousness or whether recurrent signals originating in prefrontal cortex must also be involved (dehaene et al., 2006; lamme, 2006). in light of the multiplicity of available algorithms as well as cao’s misgivings about the framework’s ability to offer specific predictions about the causal organization of the brain, it is unclear whether pp could adjudicate between these competing hypotheses. 3.2 the issue of coarse-grained realism about pp the issues discussed so far have a direct implication for the ways in which pp can guide investigations into the neurological bases of different cognitive phenomena, including the search for nccs. however, hohwy and seth could claim that even though the details of the framework’s algorithmic implementations are vague, a coarse-grained version of pp can still make progress in elucidating systematic nccs. in this section we will argue that the framework is still deficient even when viewed as a more abstracted description of neural mechanisms. this problem can be better illustrated with an example of a neuroimaging study on the significance of early visual processing in illusory contour perception done by kok and de lange (2014). the study used fmri and population receptive field mapping to investigate the neural correlates of illusory shape perception and perceptual grouping in the primary visual cortex. the researchers presented twenty subjects with a series of similar visual arrays consisting of four ‘pac-man’ shapes or inducers, arranged so that each would be equidistant from two others. during test trial presentations of the arrays, three of the four inducer shapes were aligned to form an illusory kanizsa triangle. these inducing configurations were presented for 500ms, separated by a similarly arranged mask image of full circles displayed over the same time interval. each of the trial sequences lasted 14.4 seconds, and the same positioning of the illusory triangle was used throughout each series. in control configurations, inducers were not aligning into an illusory shape. additionally, all trials were controlled for attentional spread and modulation by presenting subjects with one of two peripheral detection tasks which did not overlap with the illusory shapes. the results of comparing the reconstruction of the neural responses to illusory figure configurations with the responses to the non-inducing condition proved more than surprising. v1 regions corresponding to the surface of the illusory figure showed a significant increase in response to illusion inducing configurations, despite the absence of bottom-up input. concurrently, the response of v1 regions corresponding to the inducer shapes was suppressed in the presence of the illuschlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 12 sory shape. both results were also visible in v2, due to the well-known scaling of receptive field sizes in higher levels of the visual cortex (kastner et al., 2001). finally, the fixation point falling within or outside of the illusory shape did not significantly modulate these effects. similarly, subjects’ performance in either of the attention control tasks did not show dependence on the presence or absence of an illusory shape. kok and de lange interpret the results of their study by appealing to the topdown modulation of lower cortical areas by higher-level ones. the importance of feedback connections for early visual areas has already been widely confirmed by empirical research on figure-ground segmentation (zhang & heydt, 2010), as well as shape and contour detection (zhou et al., 2000). most previous studies have stressed the excitatory role of these connections, postulating that they result from higher-level areas passing down information about features relevant for the grouping process. although this hypothesis explains the observed increase in response during the illusory figure condition, it fails to explain why the responses of v1 areas corresponding to non-absent stimulus (here, the inducers) were inhibited in this and in other studies. to accommodate both of the discussed phenomena, kok and de lange turn to an interpretation of the results via the lense of pc algorithms (friston & kiebel, 2009; e.g., rao & ballard, 1999) which assume that the mismatch between the information carried in the feedback signal (e.g., neural activity where there was no activity predicted, or vice versa) will result in a surge of feed forward ‘prediction error’ signals. however, without a clear commitment to a particular fine-grained model of the neural populations, this interpretation falls short of disambiguating different possibilities for how the observed activity fits systematically with the working of the wider visual system as well as the phenomenology of the illusory percept. kok and de lange’s assessment of their results echoes the observation that it is difficult to pinpoint the ‘neural signature’ of error signals in sensory neurons, which in turn means that it is not clear whether the suppression of activity in cells with receptive fields corresponding to inducer regions is observed because “(1) they are predicted to be active by higher-order areas, (2) this prediction is violated, or (3) both.” (kok & lange, 2014, p. 1535). such uncertainty regarding how to interpret these findings is not just a case of limitations in our empirical methods or lack of additional data (though of course, both issues are present here), but also of the unclear empirical predictions made by the framework. kok and de lange’s uncertainty about how to interpret their data even after assuming a coarse-grained reading of pp casts a shadow on hohwy and seth’s optimism about the ways in which pp can facilitate the search for systematic nccs. according to them, the framework should offer a high-level description which would allow for relating evidence about different neural correlates and how they correspond to and change with conscious experience. however, it is difficult to see how much new information can be gained from relating different empirical findings in this way, if the interpretations of such findings are rendered ambiguous schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 13 by the pp framework. hohwy and seth can reply that it is precisely by locating different empirical findings in a common framework that we can corroborate the available data and resolve such ambiguities, but this strategy can only take us so far. after all, we have already shown that there is more than one way to implement pp, and that competing algorithmic implementations may lead to different interpretations of the data. in short, different pp algorithms may lead to different systematic nccs. notice that all of the above worries are especially problematic when we consider hohwy and seth’s claim that the framework can solve the challenge of teasing apart neural activity that causally contributes to particular experiences from activity which merely co-occurs with them (i.e., the second challenge mentioned in section 2.2). as they note, the crux to solving this challenge is having the capacity for “intervening on causal chains in neural systems that in a reasonable sense ‘carve nature at the joints,’ that is, are informed by a mechanistic framework for brain function, and which are formulated at a level that can capture phenomenological distinctions” (hohwy & seth, 2020, p. 17). while we have some worries about pp’s ability to effectively capture phenomenology (see next sections), we would like to draw attention to the fact that, given what has been said so far, pp simply does not deliver on the promise of guiding causal interventions in a way that the authors envisage. however, our skepticism about pp’s inability to meet the challenge of uncovering the mechanisms behind systematic nccs does not stem only from the framework’s underdetermination by the data, but from a further methodological worry, namely that models of cortical activity formulated within pp are usually heavily reliant on the use of causal analysis methods such as dynamic causal modelling (dcm) (friston et al., 2003). dcm is a statistical method of analysing neuroimaging data which allows for estimating the causal influence between different brain regions by treating them as coupled dynamical systems. the method relies on constructing possible models of cortical interaction and using bayesian model selection to fit these models to the available measurements. dcm is in many ways an inspiration and a precursor to pp, but the two should not be confused. the former is closer to what philosophers of science have called a model of data – a way of processing and extracting relevant information from measurement (suppes, 1962), while the latter is more akin to a theoretic model meant to explain why that configuration of data has been observed (bailer-jones, 1999; hartmann, 1995). most pp accounts of cognitive function based on neuroimaging data not only rely on the use of dcm and similar methods for validation, but also are often built on the results of prior dcm analysis. what this means is that models of neural processing offered by the framework are fitted to the causal analysis of neural connectivity which is largely independent from the pp story.6 we think that this worry presents a particular problem for those who would like to avoid 6the problem of post-hoc fitting of pp models to causal analyses has recently resulted in a controversy regarding which kinds of neurodynamic models support pp descriptions of cognitive functions (litwin & miłkowski, 2020). schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 14 the problem of underdetermination of particular implementations of pp by claiming that the framework still captures some causal relationships between different brain regions on a less-detailed level of description. simply put, the problem we see here is that the methods of causal analysis which capture causal relations between different brain areas provide the frame which is then glossed over using pp terminology. therefore, it is those models and not the framework itself that do most of the heavy lifting with regards to explaining the contributions that activity in different neural populations make to particular conscious experiences. consequently, it is the causal modelling and not pp that seems to help in the search for systematic nccs. that being said, pp may, in the long run, offer a refined view of why some neural processes causally contribute to particular changes in consciousness over others; but, for now, most of the explanatory burden is carried by auxiliary methods rather than by the framework itself. having dealt with the ontological and methodological issues that pp is facing on the road to elucidating consciousness, it is finally time to turn to some of the most famous problems arising for putative explanations of consciousness. here we think that pp’s very general nature prevents it from solving the hard problem without invoking additional assumptions. however, as we will show such assumptions are usually motivated by reasoning and evidence external to the framework. this, in turn, casts doubt on whether it is pp that does the explanatory work in accounts of consciousness formulated within the framework. 4 pp and the hard problem one of the crucial conceptual issues faced by all proponents of strong pp who want to tackle the mystery of conscious experience is the hard problem of consciousness. although hohwy and seth do not claim that pp can explicitly attack the hard problem, they do think that the framework can offer a piecemeal approach focusing on the search for nccs that will eventually make the hard problem disappear. as they note, such a “strategy will deliver greater predictive and explanatory power regarding factors that shape and modulate consciousness,” while “a focus on systematic nccs recognizes that ‘consciousness’ is not a singular explanatory target” (2020, p. 10). this, in turn, is claimed to also provide a better grasp of the overall structure as well as individual differences in phenomenology. in other words, a sensible mapping – achieved by satisfying the systematicity constraint – of brain states and phenomenology may make the hard problem less pressing, or perhaps even dissolve it altogether. (hohwy & seth, 2020, p. 11) while the proposed approach can definitely make inroads into the nature of consciousness, it is doubtful that even a complete knowledge of systematic nccs will yield a satisfying answer to the hard problem. the reason for this is that having complete knowledge about correlations between neural events and phenomenolschlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 15 ogy would likely not yield, on its own, an explanation of why the states are correlated in these ways. after all, the hard problem of consciousness is not the problem of understanding what psychological states are correlated with which physical states, but why any physical states should give rise to any phenomenal experiences. it is worth noting that hohwy and seth’s optimism is striking in light of hohwy’s previous statements about the inroads that pp can make in answering to the hard problem. in his book, hohwy lays the claim that pp “tracks characteristics of conscious perception nicely: binding, penetrability, reality testing, illusions, inextricably rich experience, and first-person perspective,” but “is not intended as a proposal that can explain why perceptual states are phenomenally conscious rather than not.” (hohwy, 2013, pp. 201–202). indeed, (and as the authors note) there are several proposals on the market, which aim to amend pp to solve this problem. importantly, the existence of such proposals shows that pp cannot solve the hard problem unless it is equipped with auxiliary metaphysical assumptions. but what kind of conceptual and methodological assumptions might those be? 4.1 pp and cognitivism about consciousness perhaps the most common proposal in the literature (most famously defended by clark, 2019; but see dołęga & dewhurst, 2019; dołęga & dewhurst, 2020) is the idea that pp does not respond to the hard problem directly, but rather does so indirectly, by addressing the meta-problem of consciousness first (chalmers, 2018).7 unlike the hard problem, the meta-problem is concerned with our intuitions and judgements about consciousness. this problem concerns answering the question why a cognitive agent may come to make judgments about conscious experience and/or become puzzled by its subjectivity and mysterious nature, rather than try7solms (2021, pp. 238–269) diverges from this mainstream by outlining a theory of consciousness in terms of friston’s free energy principle, maintaining that “in order to solve the hard problem of consciousness, science needs to discern the laws governing the mental function of ‘feeling’ ” (solms, 2021, p. 265). on his view, “everything springs from a system’s drive to exist” (solms, 2021, p. 268) which is explained in terms of the free energy principle, in particular its contribution to explaining self-organization: “starting with thermodynamics, we arrive – surprisingly easily – at a qualified and agentic subjectivity, one whose most urgent priorities are weighed for a moment, felt, and then transformed with (one hopes) due circumspection into ongoing action.” (solms, 2021, p. 268) solms objects to chalmers’ panpsychist suggestion that all information has a phenomenal aspect on the grounds that, if that were the case, we would not need (or get) an explanation of consciousness. thus, he calls for further constraints on the sort of information that has a phenomenal aspect. these come in the form of information processing and possession of a markov blanket which together entail the system under investigation is minimising its own entropy (solms, 2021, p. 263). although refreshingly direct and relying heavily on the free energy formulation of the pp framework, seth (2021) does not accept solms’ overall story, especially his take on the hard problem and his identification of the neural origin of consciousness in the brain stem which diverges significantly from the mainstream view that its spring is in the cortex. this disagreement is interesting in light of our criticism since it illustrates the shortcomings of the pp framework – with its diverse formulations – of guiding the search for systematic nccs. schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 16 ing to close the stipulated metaphysical gap between consciousness and the physical world. proponents of tackling consciousness via the meta-problem are, effectively, following dennett’s (1991, 1996, 2015) approach according to which a satisfactory account of consciousness should only need to explain judgements about qualitative feels, irrespective of whether there is or is not anything that the notion of phenomenal consciousness picks out. thus, philosophers and scientists following this approach only accept consciousness in the sense of conscious access to information (e.g., dehaene, 2014). clark and colleagues’ account of consciousness in pp aims to “take the metaphysical sting out of the quale’s tail” (clark et al., 2019, p. 16) by likening qualia to inferred contents, on a par with our projections of other internal and external causes of our sensory inputs. by likening qualia to inferred contents and rejecting any position on which they would be admitted as data to be explained (chalmers, 2013), clark in effect sides with dehaene (2014) and dennett (1996) in rejecting a notion of phenomenal consciousness that asks for explanation in addition to cognitive access to information. thus, the notion of consciousness that gives rise to the hard problem in the first place does not play a role in clark’s account. that it seems otherwise to proponents of phenomenal consciousness is not a problem if the pp account can make plausible why it should seem that way. and this is where clark’s pp story marries the familiar illusionist approach to consciousness, adding “engineering, neuroscientific, and information-theoretic flesh to the familiar dennett-style picture” (clark, 2019, p. 16; see also dennett, 1991; frankish, 2016; kammerer, 2016; kammerer, 2018 for details on illusionism). without making it explicit, clark’s take on pp assumes a cognitivist account of consciousness, addressing only cognitively accessible contents that can potentially yield perceptual judgements and lead to action. but the crucial explanatory work here is imported from illusionism, rather than being provided by the pp story itself. clark takes it as an advantage that his version of pp is consistent with global workspace theory (dehaene, 2014), information integration theory (tononi, 2008), and lamme’s (2006) recurrent processing approach. hohwy and seth make analogous comments in praise of pp’s flexibility and compatibility with a variety of otherwise competing theories of consciousness. but these competing theories take different stances (and make differing predictions) on important issues, e.g., on whether or not phenomenal consciousness ‘overflows’ access, on criteria for consciousness (block, 2011; dehaene et al., 2006), and on potentially different nccs for phenomenality and cognitive access (block, 2005; tsuchiya et al., 2015). thus, rather than taking this alleged compatibility as an advantage of the pp account, we submit that it shows the limitations of pp with respect to its ability to make standalone contributions in this area. pp is compatible with the leading neuroscientific theories of consciousness because it fails to make predictions which would rule out any of these theories. to illustrate our reservations about pp’s limited ability to further debates surrounding the assumption of cognitivism brought on board by clark and colleagues, schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 17 let us briefly turn to the debate about overflow that is also concerned with the localization of conscious processes in the brain. is the ncc to be found in sensory areas ‘in the back of’ the neocortex or does the minimally sufficient ncc involve (pre-)frontal circuits of the brain, which are also associated with cognition and attention (block, 2019)? as it turns out, this question is closely related to the conceptual question about the relation between phenomenology and cognitive accessibility (block, 2007, 2011) as well as a methodological issue of operationalizing consciousness in a way that would prevent conflating neural correlates of consciousness with correlates of cognitive abilities or unconscious processes (see schlicht, 2018). cognitivists like dehaene (2014, p. 8) have famously defended the idea that “conscious perception must […] be evaluated by subjective report, preferably on a trial-by-trial basis” (dehaene et al., 2006, p. 206). relying on reports presupposes that the contents of experience can be cognitively accessed and that a subject can exhaustively access and report what she phenomenally experiences. this, however, involves a conceptual decision on the part of the experimenter which outright dismisses the issue of whether the capacity of consciousness coincides with (or outstrips) the capacity of the cognitive system that is involved in providing access to and enabling report of conscious information. relying on reports alone prejudges this empirical issue despite evidence to the contrary (block, 2007, 2011). thus, since subjective reports not only involve cognitive access but also other cognitive phenomena like attention, working memory, and possibly metacognition, this approach is in danger of confounding a potential ncc with processes underlying these cognitive phenomena. to overcome this restriction, several researchers have suggested ways to bypass reports by using no-report paradigms in which neural activity is measured without asking subjects to report what they see, for example (frässle et al., 2014; tsuchiya et al., 2015). one such strategy exploits the finding that certain automatic eye movements show a high correlation with conscious reports of perceptual dominance during binocular rivalry studies. in their ingenious study, frässle et al. (2014) made use of the optokinetic nystagmus and pupillary reflex. the nystagmus consists of a slow phase in which the eye follows the stimulus, and a fast phase in which the eye quickly reorients in the opposite direction. we are unaware of these movements but they seem to be highly correlated with the percept rather than just the stimulus. although subjects’ patterns of neural activation were superficially similar to the ones measured in the report condition, the intensity in the no-report condition was somewhat reduced and the predominantly frontal activation was largely missing. the contrast between putative nccs in the two paradigms is striking. this result suggests that frontal activation typical for report studies (used by dehaene et al., 2006 and others) is due to the involvement of cognitive access to the information in giving the report, not necessarily with phenomenology. on the one hand, proponents of no-report paradigms rightly suggest that their approach is more promising for determining the ncc without confounding it with the neuschlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 18 ral mechanisms correlated with cognitive functions. on the other hand, this move is in danger of including neural processes within the ncc that are actually underlying non-conscious processing instead of conscious processing (schlicht, 2018). after all, the objective measures used – although tested regarding their correlation with the conscious percept rather than the stimulus – are reflexes outside voluntary control and outside conscious experience. yet, they are presumably supported by neural mechanisms enabling them. tsuchiya et al. (2015) acknowledge this possibility but do not elaborate on how to overcome this challenge.8 we bring up this example to illustrate that any reconceptualization of the sides of the debate in terms of pp does not seem to get us any closer to a solution. on the one hand, the generality of pp and its compatibility with dehaene’s global workspace, tononi’s integration information, and lamme’s recurrent processing approach, prevent it from contributing any detail which could adjudicate between these theories. on the other hand, clark et al.’s (2019) commitment to a cognitivist stance on consciousness is too specific and does not allow proponents of pp to take the possibility of phenomenal overflow seriously. to drive home this point we now want to look at a different assumption brought in by proponents of pp accounts of consciousness. the final point we wish to make concerns phenomenology, or more specifically, the contents of phenomenal experience. using a familiar example, we illustrate that pp not only fails to settle empirical debates in research on consciousness, but can also make false predictions about phenomenal experience. 4.2 pp, mid-levels, and phenomenology according to hohwy, the pp story implies that “perceptual content is the predictions of the currently best hypothesis about the world” (hohwy, 2013, p. 48). in other words, perceptual content is equated with the content of the probabilistic representations that make up the model of the world constructed by the brain. clark agrees by claiming that “what we experience as qualia are simply content items in our best generative model of the world […]; conscious contents emerge as precision-driven best-estimates of organism-salient (potentially action-driving) states of affairs” (clark, 2019, p. 17, n. 43). as already mentioned, clark, friston, and wilkinson (2019) aim to dispel concerns about the hard problem by responding to the meta-problem. this response holds that the brain, viewed as a kind of “inference machine will be led to conclude that it is home to some very puzzling states that have many of the hallmarks of ‘qualia’ ” (clark et al., 2019, p. 20). more specifically, given that the brain is supposed to harbour a multi-layered hierarchical generative model of (the causes of) its sensory inputs, any creature equipped with such a brain, “will represent some of its mid-level inferences as especially certain. these mid-level states confidently re-code raw sensory stimulation in ways that […] fall short of fully de8see block (2019) and michel and morales (2020) for further discussion of these issues. schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 19 termining how properties and states of affairs are arranged in the distal world.” (clark et al., 2019, p. 19). thus, the authors appeal to the level of certainty associated particularly with mid-level hypotheses within the complex hierarchy of top-down and bottom-up processing in the brain. that is, particular qualia are assumed to be identical to inferred causes “that are also represented as especially certain” (clark et al., 2019, p. 21).9 it should be noted that, despite the disagreements between hohwy’s and clark’s interpretations of the pp framework, hohwy also endorses a similar view (marchi & hohwy, 2020), though his way of arriving at this position is somewhat different. from his perspective, what specifies which levels of the hierarchy will be privileged in relation to consciousness is dependent not only on the complexity of models that the organism’s cognitive system is capable of harbouring, but also on the range of actions that the organism can engage in its environment. therefore, what determines the privileged role played by the intermediate levels with regard to human consciousness are the contingent ecological and evolutionary facts about the temporal scales at which humans can act. thus, on this view, “intermediateness is not an essential feature of consciousness” (marchi & hohwy, 2020, p. 1). while hohwy and marchi develop this position in greater nuance, what is relevant here is that, like clark, they defend the intermediate level view of consciousness as particularly important. there are (at least) three problems with this approach. firstly, one should ask why the representational states that are important to consciousness should be located on the mid-level? clark et al. (2019) emphasize the amount of certainty, which we allegedly assign to particular mid-level hypotheses (encodings). why? what is special about them in contrast to hypotheses on other levels? and what does certainty have to do with phenomenology? why should there be a principled connection, let alone an explanatory one? generally speaking, all hypotheses in the hierarchical generative model are constrained and determined in two directions, namely, by lower-level sensory feedback and by higher-level hypotheses. this ubiquitous dependency and “entanglement” (clark, 2019) sheds doubt on singling out a particular level which could be held responsible for the “puzzlement … concerning the ‘explanatory gap,’ where we are almost fooled into believing that there’s something special about qualia – that they are not simply highly certain mid-level encodings” (clark et al., 2019, p. 24). to evaluate this, it is important to note that ‘mid-level’ is shorthand for a multiplicity of 9of course, clark et al.’s (2019) account is more refined. following dennett (2013, 2015), clark elaborates that on the basis of its interaction with the world, the embodied agent’s brain is prone to infer qualitative features like cuteness, redness, painfulness etc. as objective (internal or external) features just like it infers the existence of babies, tomatoes, and toothaches as potential causes of sensory input. but the agent does not get cognitive or conscious access to the on-going process of inference, only to the “final estimations” (2019, p. 11), which clark considers to be adaptively reasonable and economical. to the extent that interoceptive features of the agent’s internal bodily milieu (see damasio, 1999; seth, 2013) drive unconscious inference to qualia, the inferred worldly objects themselves involve predictions that track our own reactive dispositions. schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 20 levels in the hierarchy. this raises the question where in the hierarchy exactly the so-called mid-level starts and ends. in which direction is certainty supposed to increase or decrease, in the direction of greater specificity or greater generality? as it turns out, most of the authors appealing to the mid-level view look outside of the framework to answer these questions. firstly, they seem to motivate their view either by an appeal to evidence in favour of previously available midlevel accounts (jackendoff, 1987; prinz, 2012)10 or other factors external to pp, like the perspective from the embodied and ecological research programs adopted by marchi and hohwy (2020). thus, the mid-level hypothesis requires not only further motivation from within the pp frameworks, but also a more detailed treatment. this is especially important since, as has been previously discussed by dołęga and dewhurst (2020), the current globally best hypothesis must be cashed out in terms of many hypotheses pertaining to the multiple levels of the hierarchy, where each of them simultaneously makes predictions about the level below and is constrained by that level. but if higher-level hypotheses constrain and determine lower-level (thus including mid-level) hypotheses, then this leads to a further problem which we illustrate below. consider the well-known müller-lyer illusion in which two horizontal lines of equal length appear to us as being different in length. now, we can take our hands (engage in active inference) and cover the arrows that induce the illusion, just to be sure and reduce uncertainty. once we do that, we will see that the lines are of equal length. however, the illusion returns once we remove our hands.11 the peculiarity of the müller-lyer effect is that it persists even once we know that the lines are of equal length.12 let’s assume then, following clark et al. (2019) and hohwy (2013), that ‘what it is like’ for me to see the lines (appearing different in length) is determined by the currently best guess coded in the hierarchical model. in clark’s case this will be the mid-level hypothesis, which is taken to be especially certain. the problem then is that, intuitively, we should expect that the sensory evidence gathered by covering the arrows or by measuring them (i.e., via active inference), and the subsequent updating, should increase the brain’s confidence in the belief that they are equal in length. after all, this evidence should be taken more seriously than the evidence based on looking alone (i.e., perceptual 10these views usually stress the importance of working memory as one of the crucial vehicles for conscious contents. we see this as problematic for pp because no substantial treatment of the nature and role of working memory has been offered within the framework. 11some proponents of pp may object that there is a significant difference between the scenario in which the lines are perceived with the inducers visible and the one where they are covered, since the latter involves an action which effectively changes the visual stimulus and the state of the environment that the system is trying to infer. this is right, but notice that in both cases the perceptual system is inferring the length of the same two lines, i.e., the content of the hypothesis about the length of the lines should be the same after the inducer arrows are made visible. 12this has led philosophers to argue for the encapsulation of perception, see fodor (1983), starting a debate with churchland (1988) about the penetrability of perception by cognitive processes, which lasts until today (see mccauley & henrich, 2006; and nes & chan, 2020 for helpful overviews and a summary of recent developments). schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 21 inference). nevertheless, we continue to experience the lines as being different in length, our perceptual faculties seem to be cognitively impenetrable to our perceptual beliefs. while the pp approach that aligns phenomenality with assigned certainty to a given hypothesis predicts that one should experience the lines being equal in length, this does not seem to fit the fact that one remains subject to the illusion that they are unequal in length. if the reasoning above is correct, then the hypothesis that determines phenomenal character (different in length) differs from the hypothesis with the highest certainty (equal in length). this also creates a problem of defining what it means for a hypothesis (i.e., an internal representation of the external world) to be ‘mid-level’ in the sense relevant for the meta-problem response relying on the certainty of mid-level hypotheses, further undermining its persuasive strength towards dispelling the hard problem. 5 conclusion in this paper, we critically assessed the proposal that the popular pp approach to perception, cognition, and action can be also successfully applied to consciousness and thus guide the search for the neural correlates of consciousness. as we have shown, it is questionable whether or not the pp framework can, in fact, offer a componential analysis of the nervous system which could guide empirical investigations aimed at uncovering the mechanisms underlying conscious experiences. we have also demonstrated that the accounts of consciousness currently utilizing pp are usually dependent on assumptions external to the framework for conceptualizing and accounting for consciousness. such accounts, furthermore, often lead to additional problems and misleading predictions. despite the critical voice of this paper, we do not want to suggest that the pp framework cannot be useful to neuroscientific research into consciousness on principle. our criticism is only meant as a voice of caution which highlights that the hype surrounding the framework often outstrips the support that current research lends to it. it is simply too early in the development of the framework to reliably judge whether or not it can deliver on many of its promises. we should remain cautious about the grand claims about pp’s explanatory potential and focus on filling in the details as well as overcoming conceptual and methodological obstacles that the framework is facing. acknowledgments work on this paper was generously supported by the volkswagen foundation (grant no. 87 105 1 for our research project ‘situated cognition. perceiving the world and understanding other minds.’) we are grateful for this support and for the helpful comments on an earlier draft by our team members marco facchin, paola gega, françois kammerer, nina poth, bartosz radomski, tobias starzak and elmarie venter, two anonymous reviewers and the editors of this journal. schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 22 references aitchison, l., & lengyel, m. (2017). with or without you: predictive coding and bayesian inference in the brain. current opinion in neurobiology, 46, 219–227. https://doi.org/10.1016/j.conb.2017.08.010 aru, j., bachmann, t., singer, w., & melloni, l. (2012). distilling the neural correlates of consciousness. neuroscience & biobehavioral reviews, 36(2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.12.003 bailer-jones, d. m. (1999). tracing the development of models in the philosophy of science. in l. magnani, n. j. nersessian, & p. thagard (eds.), model-based reasoning in scientific discovery (pp. 23–40). springer us. bayne, t. (2007). conscious states and conscious creatures: explanation in the scientific study of consciousness. philosophical perspectives, 21(1), 1–22. https://doi.org/https://doi.org/10.1111/j.1520-8583.2007.00118.x bayne, t., & chalmers, d. j. (2003). what is the unity of consciousness? in a. cleermans (ed.), the unity of consciousness: binding, integration, dissociation (pp. 23–58). oxford university press. bishop, c. (2006). pattern recognition and machine learning. springer-verlag. block, n. (2005). two neural correlates of consciousness. trends in cognitive sciences, 9(2), 46–52. https://doi.org/10.1016/ j.tics.2004.12.006 block, n. (2007). consciousness, accessibility, and the mesh between psychology and neuroscience. behavioral and brain sciences, 30(5-6), 481–499. https://doi.org/10.1017/s0140525x07002786 block, n. (2011). perceptual consciousness overflows cognitive access. trends in cognitive sciences, 15(12), 567–575. https: //doi.org/10.1016/j.tics.2011.11.001 block, n. (2019). what is wrong with the no-report paradigm and how to fix it. trends in cognitive sciences, 23(12), 1003– 1013. https://doi.org/10.1016/j.tics.2019.10.001 bruineberg, j., dołęga, k., dewhurst, j., & baltieri, m. (2020). the emperor’s new markov blankets. http://philsci-archive.pi tt.edu/18467/ cao, r. (2020). new labels for old ideas: predictive processing and the interpretation of neural signals. review of philosophy and psychology, 11(3), 517–546. https://doi.org/10.1007/s13164-020-00481-x chalmers, d. j. (1996). the conscious mind: in search of a fundamental theory. oxford univeristy press. chalmers, d. j. (2000). what is a neural correlate of consciousness? in t. metzinger (ed.), neural correlates of consciousness (pp. 17–39). the mit press. chalmers, d. j. (2013). how can we construct a science of consciousness? annals of the new york academy of sciences, 1303(1), 25–35. https://doi.org/https://doi.org/10.1111/nyas.12166 chalmers, d. j. (2018). the meta-problem of consciousness. journal of consciousness studies, 25(9-10), 6–61. churchland, p. m. (1988). perceptual plasticity and theoretical neutrality: a reply to jerry fodor. philosophy of science, 55(june), 167–187. https://doi.org/10.1086/289425 clark, a. (2016). surfing uncertainty: prediction, action, and the embodied mind. oxford university press. clark, a. (2019). consciousness as generative entanglement. journal of philosophy, 116(12), 645–662. https://doi.org/10.584 0/jphil20191161241 clark, a., friston, k., & wilkinson, s. (2019). bayesing qualia: consciousness as inference, not raw datum. journal of consciousness studies, 26(9-10), 19–33. colombo, m., & hartmann, s. (2017). bayesian cognitive science, unification, and explanation. the british journal for the philosophy of science, 68, 451–484. colombo, m., & wright, c. (2017). explanatory pluralism: an unrewarding prediction error for free energy theorists. brain and cognition, 112, 3–12. https://doi.org/10.1016/j.bandc.2016.02.003 damasio, a. r. (1999). the feeling of what happens: body and emotion in the making of consciousness. harcourt brace; co. damasio, a. r. (2011). self comes to mind: constructing the conscious brain. pantheon. dehaene, s. (2014). consciousness and the brain. viking. dehaene, s., changeux, j.-p., naccache, l., sackur, j., & sergent, c. (2006). conscious, preconscious, and subliminal processing: a testable taxonomy. trends in cognitive sciences, 10(5), 204–211. https://doi.org/10.1016/j.tics.2006.03.007 dennett, d. c. (2013). expecting ourselves to expect: the bayesian brain as a projector. the behavioral and brain sciences, 36(3), 209–210. https://doi.org/10.1017/s0140525x12002208 dennett, d. c. (1991). real patterns. journal of philosophy, 88(1), 27–51. https://doi.org/10.2307/2027085 dennett, d. c. (1996). facing backwards on the problem of consciousness. journal of consciousness studies, 1(3), 4–6. dennett, d. c. (2015). why and how does consciousness seem the way it seems? (w. wanja & t. metzinger, eds.). open mind. schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.conb.2017.08.010 https://doi.org/10.1016/j.neubiorev.2011.12.003 https://doi.org/10.1111/j.1520-8583.2007.00118.x https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1017/s0140525x07002786 https://doi.org/10.1016/j.tics.2011.11.001 https://doi.org/10.1016/j.tics.2011.11.001 https://doi.org/10.1016/j.tics.2019.10.001 http://philsci-archive.pitt.edu/18467/ http://philsci-archive.pitt.edu/18467/ https://doi.org/10.1007/s13164-020-00481-x https://doi.org/10.1111/nyas.12166 https://doi.org/10.1086/289425 https://doi.org/10.5840/jphil20191161241 https://doi.org/10.5840/jphil20191161241 https://doi.org/10.1016/j.bandc.2016.02.003 https://doi.org/10.1016/j.tics.2006.03.007 https://doi.org/10.1017/s0140525x12002208 https://doi.org/10.2307/2027085 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 23 dołęga, k., & dewhurst, j. (2019). bayesian frugality and the representation of attention. journal of consciousness studies, 26(3-4), 38–63. dołęga, k., & dewhurst, j. (2020). fame in the predictive brain: a deflationary approach to explaining consciousness in the prediction error minimization framework. synthese. https://doi.org/10.1007/s11229-020-02548-9 egan, f. (2014). how to think about mental content. philosophical studies, 170(1), 115–135. https://doi.org/10.1007/s11098013-0172-0 egan, f. (2020). a deflationary account of mental representation. in j. smortchkova, k. dołęga, & t. schlicht (eds.), what are mental representations? (pp. 26–53). oxford university press. fazekas, p., & overgaard, m. (2018). perceptual consciousness and cognitive access: an introduction. philosophical transactions of the royal society of london. series b, biological sciences, 373(1755). https://doi.org/10.1098/rstb.2017.0340 fernández-espejo, d., & owen, a. m. (2013). detecting awareness after severe brain injury. nature reviews. neuroscience, 14(11), 801–809. https://doi.org/10.1038/nrn3608 fink, s. b. (2016). a deeper look at the “neural correlate of consciousness.” in frontiers in psychology (vol. 7, p. 1044). https://doi.org/10.3389/fpsyg.2016.01044 fodor, j. a. (1983). the modularity of mind: an essay on faculty psychology. mit press. frankish, k. (2016). illusionism as a theory of consciousness. journal of consciousness studies, 23(11-12), 11–39. frässle, s., sommer, j., jansen, a., naber, m., & einhäuser, w. (2014). binocular rivalry: frontal activity relates to introspection and action but not to perception. the journal of neuroscience: the official journal of the society for neuroscience, 34(5), 1738–1747. https://doi.org/10.1523/jneurosci.4403-13.2014 friston, k. (2009). the free-energy principle: a rough guide to the brain? trends in cognitive sciences, 13(7), 293–301. https://doi.org/10.1016/j.tics.2009.04.005 friston, k. (2010). the free-energy principle: a unified brain theory? nature reviews neuroscience, 11(2), 127–138. https: //doi.org/10.1038/nrn2787 friston, k., harrison, l., & penny, w. (2003). dynamic causal modelling. neuroimage, 19(4), 1273–1302. https://doi.org/10 .1016/s1053-8119(03)00202-7 friston, k., & kiebel, s. (2009). predictive coding under the free-energy principle. philosophical transactions of the royal society of london. series b, biological sciences, 364(1521), 1211–1221. https://doi.org/10.1098/rstb.2008.0300 friston, k., rigoli, f., ognibene, d., mathys, c., fitzgerald, t., & pezzulo, g. (2015). active inference and epistemic value. cognitive neuroscience, 6(4), 187–214. https://doi.org/10.1080/17588928.2015.1020053 hartmann, s. (1995). models as a tool for theory construction: some strategies of preliminary physics. in w. herfel, w. krajewski, i. niiniluoto, & r. wójcicki (eds.), theories and models in scientific processes (pp. 26–53). rodopi. heilbron, m., & chait, m. (2018). great expectations: is there evidence for predictive coding in auditory cortex? neuroscience, 389, 54–73. https://doi.org/10.1016/j.neuroscience.2017.07.061 hohwy, j. (2009). the neural correlates of consciousness: new experimental approaches needed? consciousness and cognition, 18(2), 428–438. https://doi.org/10.1016/j.concog.2009.02.006 hohwy, j. (2013). the predictive mind. oxford: oxford university press. hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii). https://doi.org/10.33735/phimisci.2020.ii.64 illari, p. m., & williamson, j. (2010). function and organization: comparing the mechanisms of protein synthesis and natural selection. studies in history and philosophy of biological and biomedical sciences, 41(3), 279–291. https://doi.or g/10.1016/j.shpsc.2010.07.001 issa, e. b., cadieu, c. f., & dicarlo, j. j. (2018). neural dynamics at successive stages of the ventral visual stream are consistent with hierarchical error signals. elife, 7, e42870. https://doi.org/10.7554/elife.42870 jackendoff, r. (1987). consciousness and the computational mind. mit press. kammerer, f. (2016). the hardest aspect of the illusion problem — and how to solve it. journal of consciousness studies, 23(11–12), 124–139. kammerer, f. (2018). can you believe it? illusionism and the illusion meta-problem. philosophical psychology, 31(1), 44–67. https://doi.org/10.1080/09515089.2017.1388361 kastner, s., de weerd, p., pinsk, m. a., elizondo, m. i., desimone, r., & ungerleider, l. g. (2001). modulation of sensory suppression: implications for receptive field sizes in the human visual cortex. journal of neurophysiology, 86(3), 1398– 1411. https://doi.org/10.1152/jn.2001.86.3.1398 kirchhoff, m. d., & kiverstein, j. (2018). extended consciousness and predictive processing: a third-wave view. routledge. kok, p., & lange, f. p. de. (2014). shape perception simultaneously upand downregulates neural activity in the primary visual cortex. current biology: cb, 24(13), 1531–1535. https://doi.org/10.1016/j.cub.2014.05.042 schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11229-020-02548-9 https://doi.org/10.1007/s11098-013-0172-0 https://doi.org/10.1007/s11098-013-0172-0 https://doi.org/10.1098/rstb.2017.0340 https://doi.org/10.1038/nrn3608 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.1523/jneurosci.4403-13.2014 https://doi.org/10.1016/j.tics.2009.04.005 https://doi.org/10.1038/nrn2787 https://doi.org/10.1038/nrn2787 https://doi.org/10.1016/s1053-8119(03)00202-7 https://doi.org/10.1016/s1053-8119(03)00202-7 https://doi.org/10.1098/rstb.2008.0300 https://doi.org/10.1080/17588928.2015.1020053 https://doi.org/10.1016/j.neuroscience.2017.07.061 https://doi.org/10.1016/j.concog.2009.02.006 https://doi.org/10.33735/phimisci.2020.ii.64 https://doi.org/10.1016/j.shpsc.2010.07.001 https://doi.org/10.1016/j.shpsc.2010.07.001 https://doi.org/10.7554/elife.42870 https://doi.org/10.1080/09515089.2017.1388361 https://doi.org/10.1152/jn.2001.86.3.1398 https://doi.org/10.1016/j.cub.2014.05.042 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org tobias schlicht and krzysztof dołęga 24 krakauer, j. w., ghazanfar, a. a., gomez-marin, a., maciver, m. a., & poeppel, d. (2017). neuroscience needs behavior: correcting a reductionist bias. neuron, 93(3), 480–490. https://doi.org/10.1016/j.neuron.2016.12.041 lamme, v. a. f. (2006). towards a true neural stance on consciousness. trends in cognitive sciences, 10(11), 494–501. https://doi.org/10.1016/j.tics.2006.09.001 lau, h. (2008). a higher order bayesian decision theory of consciousness. in r. banerjee & b. k. chakrabarti (eds.), models of brain and mind: physical, computational, and psychological approaches (pp. 35–48). elsevier. litwin, p., & miłkowski, m. (2020). unification by fiat: arrested development of predictive processing. cognitive science, 44(7), e12867. https://doi.org/https://doi.org/10.1111/cogs.12867 marchi, f., & hohwy, j. (2020). the intermediate scope of consciousness in the predictive mind. erkenntnis. https://doi.or g/10.1007/s10670-020-00222-7 marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii). https://doi.org/10.33735/phimisci.2020.ii.61 mccauley, r. n., & henrich, j. (2006). susceptibility to the müller-lyer illusion, theory-neutral observation, and the diachronic penetrability of the visual input system. philosophical psychology, 19(1), 79–101. https://doi.org/10.1080/09 515080500462347 metzinger, t., & wiese, w. (2017). philosophy and predictive processing. mind group. michel, m., & morales, j. (2020). minority reports: consciousness and the prefrontal cortex. mind & language, 35(4), 493–513. https://doi.org/https://doi.org/10.1111/mila.12264 nes, a., & chan, t. (2020). inference and consciousness. london: routledge. ng, a. y., & jordan, m. i. (2001). on discriminative vs. generative classifiers: a comparison of logistic regression and naive bayes. 841–848. https://doi.org/10.5555/2980539.2980648 noë, a. (2005). action in perception. mit press. noë, a., & thompson, e. (2004). are there neural correlates of consciousness? journal of consciousness studies, 1(11), 3–28. oizumi, m., albantakis, l., & tononi, g. (2014). from the phenomenology to the mechanisms of consciousness: integrated information theory 3.0. plos computational biology, 10(5), e1003588. https://doi.org/10.1371/journal.pcbi.1003588 prinz, j. j. (2012). the conscious brain: how attention engenders experience. oxford university press. rao, r. p., & ballard, d. h. (1999). predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive-field effects. nature neuroscience, 2(1), 79–87. https://doi.org/10.1038/4580 sanborn, a. n., & chater, n. (2016). bayesian brains without probabilities. trends in cognitive sciences, 20(12), 883–893. https://doi.org/10.1016/j.tics.2016.10.003 schlicht, t. (2018). a methodological dilemma for investigating consciousness empirically. consciousness and cognition, 66, 91–100. https://doi.org/10.1016/j.concog.2018.11.002 searle, j. (1992). the rediscovery of the mind. mit press. seth, a. k. (2013). interoceptive inference, emotion, and the embodied self. trends in cognitive sciences, 17(11), 565–573. https://doi.org/10.1016/j.tics.2013.09.007 seth, a. k. (2021). mixed feelings about a hard problem: review of the hidden spring. https://neurobanter.com/2021/02/18/m ixed-feelings-about-a-hard-%0aproblem-review-of-the-hidden-spring shea, n. (2018). representation in cognitive science. oxford university press. solms, m. (2021). the hidden spring: a journey to the source of consciousness. london, uk: profile books. spratling, m. w. (2008a). predictive coding as a model of biased competition in visual attention. vision research, 48(12), 1391–1408. https://doi.org/10.1016/j.visres.2008.03.009 spratling, m. w. (2008b). reconciling predictive coding and biased competition models of cortical function. frontiers in computational neuroscience, 2. https://doi.org/10.3389/neuro.10.004.2008 spratling, m. w. (2017). a review of predictive coding algorithms. brain and cognition, 112, 92–97. https://doi.org/10.1016/ j.bandc.2015.11.003 spratling, m. w., de meyer, k., & kompass, r. (2009). unsupervised learning of overlapping image components using divisive input modulation. computational intelligence and neuroscience, 2009, e381457. https://doi.org/10.1155/2009/3 81457 summerfield, c., & lange, f. p. de. (2014). expectation in perceptual decision making: neural and computational mechanisms. nature reviews. neuroscience, 15(11), 745–756. https://doi.org/10.1038/nrn3838 suppes, p. (1962). models of data. in e. nagel, p. suppes, & a. tarski (eds.), logic, methodology and philosophy of science: proceedings of the 1960 international congress (pp. 252–261). stanford university press. tononi, g. (2008). consciousness as integrated information: a provisional manifesto. the biological bulletin, 215(3), 216– 242. https://doi.org/10.2307/25470707 schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neuron.2016.12.041 https://doi.org/10.1016/j.tics.2006.09.001 https://doi.org/10.1111/cogs.12867 https://doi.org/10.1007/s10670-020-00222-7 https://doi.org/10.1007/s10670-020-00222-7 https://doi.org/10.33735/phimisci.2020.ii.61 https://doi.org/10.1080/09515080500462347 https://doi.org/10.1080/09515080500462347 https://doi.org/10.1111/mila.12264 https://doi.org/10.5555/2980539.2980648 https://doi.org/10.1371/journal.pcbi.1003588 https://doi.org/10.1038/4580 https://doi.org/10.1016/j.tics.2016.10.003 https://doi.org/10.1016/j.concog.2018.11.002 https://doi.org/10.1016/j.tics.2013.09.007 https://neurobanter.com/2021/02/18/mixed-feelings-about-a-hard-%0aproblem-review-of-the-hidden-spring https://neurobanter.com/2021/02/18/mixed-feelings-about-a-hard-%0aproblem-review-of-the-hidden-spring https://doi.org/10.1016/j.visres.2008.03.009 https://doi.org/10.3389/neuro.10.004.2008 https://doi.org/10.1016/j.bandc.2015.11.003 https://doi.org/10.1016/j.bandc.2015.11.003 https://doi.org/10.1155/2009/381457 https://doi.org/10.1155/2009/381457 https://doi.org/10.1038/nrn3838 https://doi.org/10.2307/25470707 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org you can’t always get what you want 25 tononi, g., boly, m., gosseries, o., & laureys, s. (2016). the neurology of consciousness. in g. tononi, m. boly, o. gosseries, & s. laureys (eds.), the neurology of conciousness (pp. 407–461). tononi, g., & koch, c. (2008). the neural correlates of consciousness: an update. annals of the new york academy of sciences, 1124, 239–261. https://doi.org/10.1196/annals.1440.004 tononi, g., & koch, c. (2015). consciousness: here, there and everywhere? philosophical transactions of the royal society of london. series b, biological sciences, 370(1668). https://doi.org/10.1098/rstb.2014.0167 tsuchiya, n., wilke, m., frässle, s., & lamme, v. a. f. (2015). no-report paradigms: extracting the true neural correlates of consciousness. trends in cognitive sciences, 19(12), 757–770. https://doi.org/10.1016/j.tics.2015.10.002 zhang, n. r., & heydt, r. von der. (2010). analysis of the context integration mechanisms underlying figure-ground organization in the visual cortex. journal of neuroscience, 30(19), 6482–6496. https://doi.org/10.1523/jneurosci.516809.2010 zhou, h., friedman, h. s., & heydt, r. von der. (2000). coding of border ownership in monkey visual cortex. the journal of neuroscience: the official journal of the society for neuroscience, 20(17), 6594–6611. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. schlicht, t. & dołęga, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1196/annals.1440.004 https://doi.org/10.1098/rstb.2014.0167 https://doi.org/10.1016/j.tics.2015.10.002 https://doi.org/10.1523/jneurosci.5168-09.2010 https://doi.org/10.1523/jneurosci.5168-09.2010 https://doi.org/10.33735/phimisci.2021.80 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction pp and the search for the ncc determining neural correlates of consciousness pp and the search for systematic nccs bayes in the brain? the issue of fine-grained realism about pp the issue of coarse-grained realism about pp pp and the hard problem pp and cognitivism about consciousness pp, mid-levels, and phenomenology conclusion what is a global state of consciousness? what is a global state of consciousness? andy mckilliama (andrew.mckilliam@monash.edu) abstract the notion of a global state (or level) of consciousness is an increasingly important construct in the science of consciousness. however, exactly what a global state of consciousness is remains poorly understood. in this paper i offer an account of global states of consciousness as consciousness-related capacity modulations. on this view global states are not themselves phenomenal states – they are not occurring experiences. rather, they are states that specify which of a creature’s overall consciousness-related capacities are currently online. i argue that a mature mechanistic science of consciousness needs to account for global states of consciousness and that doing so requires moving beyond the current focus on occurrent experiences and integrating consciousness research with research in other areas of cognitive neuroscience. keywords consciousness ∙ disorders of consciousness ∙ global state ∙ level of consciousness ∙ ncc this article is part of a special issue on “the neural correlates of consciousness”, edited by sascha benjamin fink. 1 introduction consciousness is a multifaceted phenomenon. most research into the neural basis of consciousness focusses on occurring conscious experiences: the visual experience of a face, the taste of pineapple, the feeling of pain, and so on. with regard to occurrent experiences there are two questions that researchers are interested in. by virtue of what is a conscious experience of a face of a face? and by virtue of what is a conscious experience of a face conscious? to answer the first question researchers seek neural representational systems with states that systematically correlate with the contents of consciousness (chalmers, 2000; crick & koch, 2003; koch et al., 2016). given that the brain harbours both conscious and unconscious representations, revealing these neural representational systems does not immediately answer the second question. to answer the second question, researchers seek amonash university mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-8223-3817 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org andy mckilliam 2 the “difference makers that transform unconscious content into conscious content” (michel & morales, 2020). they seek the neural mechanism, perhaps common to all conscious experiences, by virtue of which conscious contents are conscious (bayne, 2007; hohwy, 2007, 2009; hohwy & bayne, 2015). in this paper i focus on another feature of consciousness that has variously been referred to as “global states” (bayne et al., 2016), “background states” (chalmers, 2000), “levels” (boly et al., 2013; laureys, 2005), “modes” (bayne & hohwy, 2016), “global modes” (fortier-davy & millière, 2020), and “aspects” (klein, 2017) of consciousness. as is often the case with features of consciousness, global states of consciousness are best introduced by way of example. alert wakefulness, dreaming, dreamless sleep, mild anaesthesia, and coma are all examples of global states of consciousness. they are ways in which a creature may be conscious (or unconscious). in addition to answering the two questions above pertaining to occurrent experiences, we might also ask, ‘by virtue of what is a creature in one global state rather than another?’ however, before we can answer that question we first need to know what a global state of consciousness is. perhaps the closest we have to a definition of global states is that offered by bayne, hohwy and owen when they state that global states are “states of consciousness that characterize an organism’s overall conscious condition” (bayne et al., 2016). but this definition cries out for elaboration. in what sense are global states states of consciousness? and in what ways do they characterize an organism’s overall conscious condition? there are a wide range of suggested answers to these questions in the literature. in sections 2 and 3, i argue that none of them adequately capture the concept of a global state of consciousness. the absence of an adequate, agreed upon analysis of global states of consciousness might lead one to think that the concept doesn’t really refer to anything interesting or substantive and that we should eliminate global states of consciousness from our taxonomy of phenomena that the science of consciousness must contend with. however, to do so on the grounds that no suitable analysis currently exists would be premature. for it may be that a suitable analysis can be given. my objective in this paper is to provide one. section 2 discusses the problems with the widespread convention of thinking of global states in terms of ‘conscious level’. section 3 considers three ways in which one might try to reduce global states of consciousness to occurrent experiences and finds them all unsatisfactory. section 4 provides a positive account of global states as states that regulate which of a creature’s consciousness-related capacities are currently online. and section 5 serves as a reminder that we want more from a mature mechanistic science of consciousness than just an account of occurrent experiences. accounting for global states of consciousness requires integrating research into the mechanisms supporting occurrent experiences with research into the mechanisms supporting attention, working memory, expectation, and so on. mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org what is a global state of consciousness? 3 2 are global states levels of consciousness? in the scientific literature it is common practice for researchers to characterize the difference between alert wakefulness, dreaming, anaesthesia, coma and so on in terms of ‘conscious level’ (aru et al., 2019; boly et al., 2013; koch et al., 2016; laureys, 2005). the notion of ‘conscious level’ stems from clinical tools for gauging the conscious capacities of brain damage patients. in the glasgow coma scale subjects are scored on a number of separate behavioural tests: eye opening, motor response, verbal response, and pupil reactivity (brennan et al., 2018; teasdale et al., 2014). for example, a patient who can move their limbs (give a thumbs up) in response to command scores higher on the motor response test than a subject who can only be coaxed to move in response to a painful stimulus. and a patient who spontaneously opens her eyes scores higher than one who only opens her eyes to command, or painful stimulus, or not at all. the scores for each test are tallied, and used to determine the conscious level – coma, vegetative state, minimally conscious state, normal wakefulness – of the patient. the thought is that the higher the score, the ‘more conscious’ the subject. there is something intuitively appealing about the idea that some global states are more conscious than others. in some sense, a creature is more conscious when alert and awake than when sleeping or in a coma. but what sense of consciousness is at play here? an initial suggestion might be that a creature is more ‘generically conscious’ when alert and awake than when dreaming. as i use the term, ‘generic consciousness’ is just the most general property of ‘being conscious’. a creature is generically conscious when it has a subjective perspective from which it is like something to be.1 this interpretation of ‘conscious level’ as ‘degree of generic consciousness’ is common throughout the literature. for example, tononi and koch introduce the notion of ‘conscious level’ by suggesting that “when you fall asleep […] the degree to which you are conscious (of anything) becomes less and less” (tononi & koch, 2008). but interpreting ‘conscious level’ as ‘degree of generic consciousness’ is ultimately untenable. generic consciousness isn’t the sort of property that can come in degrees. this is a point that has been recently stressed by a number of philosophers (bayne et al., 2016; carruthers, 2019). to see why generic consciousness can’t come in degrees it will be helpful to introduce some metaphysics: the determinate-determinable relation (funkhouser, 2006; wilson, 2017). the determinate-determinable relation describes the relationship between a certain class of abstract entities and their less-abstract manifestations. a paradigm example is the relationship between being coloured, being red, and being fire engine red. being coloured is an abstract (determinable) property. an entity can’t be coloured simpliciter. in order to be coloured an entity has to be a particular colour: 1others have referred to this notion as ‘creature consciousness’ (bayne, 2007, 2010; chalmers, 2000). i avoid that term here since ‘creature consciousness’ is also used to refer to a creature being awake rather than asleep (carruthers, 2019; rosenthal, 2005). mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org andy mckilliam 4 red, or green, or blue, or some other determinate colour. of course, an entity can’t be red simpliciter either, but must be red by being a particular shade of red: fire engine red for example. maximally specific shades of colour, such as fire engine red, are super-determinates. at the other end of the specificity spectrum is the super-determinable property, being coloured. super-determinable properties are maximally abstract. knowing that an entity is coloured tells us nothing specific about the way in which it is coloured. it merely specifies that it is coloured in some determinate way or other. following bayne (2007) we can think of generic consciousness as the superdeterminable property associated with being conscious. knowing that a creature is generically conscious tells us nothing about what it is like to be that creature. it merely specifies that it is like something to be that creature. specific conscious experiences, such as the visual experience of a face, the taste of pineapple, and so on are determinate ways of being generically conscious. intuitively, just as an entity can’t be coloured simpliciter but must be coloured in a determinate way, so too an entity can’t be generically conscious simpliciter but must be generically conscious in a determinate way. the reason this is relevant here is that the super-determinable property of being coloured, is not the sort of property that can come in degrees. although we often speak of colourfulness as if it comes in degrees – bob’s shirt is more colourful than jane’s – what we mean here is that bob’s shirt is brighter, or more saturated, or a more vibrant hue. we don’t mean that the colour of jane’s shirt is somehow less of a determinate of being coloured than the colour of bob’s. even the drabbest of colours is a determinate of being coloured. generic consciousness, like being coloured, is a super-determinable property that does not come in degrees. an entity either instantiates a determinate experience and is generically conscious, or does not and it is not generically conscious. it seems that ‘conscious level’ can’t be as simple as ‘degree of generic consciousness’. others have used ‘conscious level’ to refer to ‘level of wakefulness’ or ‘level of arousal’ (laureys, 2005; pal et al., 2018, 2020). however, if conscious level just means level of wakefulness/arousal then it is not an adequate analysis of global states of consciousness. the issue here is that level of wakefulness/arousal dissociates from consciousness in interesting ways. dreaming subjects are asleep but certainly conscious (revonsuo, 2006). and vegetative state patients exhibit a relatively normal sleep/wake cycle but are typically thought to be unconscious (stender et al., 2017). indeed, this is precisely why laureys takes level of wakefulness to be only one of the components of global states of consciousness; the other being level of awareness/conscious contents. a third suggestion is that ‘conscious level’ refers to the range of conscious contents a subject is currently experiencing. this appears to be what melanie boly and colleagues have in mind in places (boly et al., 2013, see especially the y-axis of their figure 1). but, as they recognize, this analysis is not entirely satisfactory. the problem, as they put it, is that: mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org what is a global state of consciousness? 5 typically, high conscious levels are associated with an increased range of conscious contents. whether or not high level of consciousness without any conscious contents is possible remains unclear. (boly et al., 2013, p. 2) what i take the authors to be grappling with here is the thought that it may be possible for a creature to be in a high global state of consciousness without experiencing much in the way of conscious contents. the states reached by experienced meditators might be an example here (thompson, 2014). but if it is an open question whether global states dissociate from range of conscious contents, and conscious level just means range of conscious contents, then conscious level is not a suitable analysis of global states of consciousness. rather than appealing to the range of contents currently being experienced, one might try to unpack ‘conscious level’ in terms of the range of contents a creature is capable of experiencing. anil seth suggests something along these lines when he notes that “conscious level and content are related inasmuch as the range of possible contents increases with increasing conscious level” (seth, 2009, p. 51). however, note that seth is careful not to say that ‘conscious level’ is ‘range of possible contents’. rather he just points out that the range of contents a creature is capable of experiencing tends to be correlated with its conscious level. of course, one could go further and stipulate that ‘conscious level’ just is ‘range of possible contents’. however, doing so would undermine its utility for analysing global states of consciousness – coma, vegetative state, dreaming, alert wakefulness, and so on. the consciousness-related differences between a patient in a minimally conscious state and an alert awake subject cannot be captured purely in terms of the range of contents each is capable of experiencing. while it is true that an alert neurotypical subject is capable of experiencing more content, they are also capable of endogenously directing the focus of their attention onto specific conscious contents and using these conscious contents to guide their intentional actions in ways that a vegetative state patient cannot. by manifesting these capacities, the alert neurotypical subject plays an active role in how their stream of consciousness unfolds over time. a subject in a minimally conscious state arguably lacks these capacities. moreover, it is also possible that two subjects may have the same range of possible conscious contents, but differ in terms of their capacity to endogenously direct attention and their capacity to use their conscious contents to guide intentional action. in this case, while they have the same range of possible contents, only one would play an active role in orchestrating how their conscious life unfolds over time. one might wonder whether these intentional capacities are really relevant to global states of consciousness. that they are becomes clear when we consider the objectives that clinicians have when treating disordered global states of consciousness. they want to do more than just restore to a subject her capacity for a wide range of possible conscious contents. a treatment that restored to a patient the full range of conscious contents, but left them incapable of guiding their intentional mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org andy mckilliam 6 actions and engaging in everyday conscious life would not be entirely satisfactory. indeed, one might question whether such a treatment is worth doing at all. similar points can be made regarding other areas of research in which the concept of a global state of consciousness can be applied. a central distinction within dream research is between lucid and non-lucid dream states. the main dimension of difference here is not the range of possible contents dreamers can experience, but the degree to which they have intentional control over their actions in the dream world (windt, 2015a). moreover, it is not just intentional capacities that are relevant. cognitive capacities such as the ability to lay down new long-term memories, as well as connectedness to the environment are also relevant. for example, in evaluating cases of interoperative awareness – conscious awareness during surgery – researchers are interested not only in the range of contents subjects experience while anaesthetised, but also their ‘connectedness’ to the environment, and the degree to which their experiences are remembered (mashour & avidan, 2015; sanders et al., 2012). the really distressing cases of interoperative awareness are those in which subjects have explicit recall of events that took place during the surgery. it is these cases that tend to have a higher instance of long-term negative effects such as post-traumatic stress disorder. as a result, if ‘conscious level’ means ‘range of possible contents’ then it is not a suitable analysis of global states of consciousness as it is applied in many areas of research into global states of consciousness. a final suggestion is that level of consciousness can be cashed out in terms of the quality or vividness of conscious content (aru et al., 2019; bachmann & hudetz, 2014; overgaard & overgaard, 2010). for example, aru et al suggest that sleep, vegetative state, and the minimally conscious state are “intermediate states” between fully conscious and fully unconscious “with varying levels of clarity or degrees of experience” (aru et al., 2019, p. 2). here we need to distinguish the application of the term ‘level of consciousness’ to the investigation of conscious contents, with its application in research on global states of consciousness. in content-based research researchers are interested in what is required for a mental representation to be conscious. one of the questions that researchers are currently grappling with is how best to explain the fact that some of our conscious contents are clear and vivid while others we only glimpse. one view is that there are two more or less independent variables here: richness of representational content, and degree of awareness of that content (fazekas & overgaard, 2018; but see michel, 2019; rosenthal, 2019 for challenges to this view). in this discussion, degrees of awareness are sometimes referred to as ‘levels of consciousness’. however, we need to be careful to distinguish ‘conscious level’ as it applies to degrees of awareness of a content, and ‘conscious level’ as it applies to global states of conscious creatures. these are not the same construct. one applies to mental states. the other applies to whole creatures. and they can vary independently. for example, in masking experiments, researchers seek to manipulate the degree to which subjects are conscious of a stimulus. but doing so mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org what is a global state of consciousness? 7 does not change their global state. subjects remain alert and awake throughout the experiment. what is manipulated is the degree to which they are conscious of the stimulus. one attempt to bridge the gap between conscious level as it applies to conscious contents and conscious level as it applies to whole creatures is to define a creature’s level of consciousness in terms of its “most clearly experienced contents” (overgaard & overgaard, 2010, p. 3). on this view a vegetative state patient with one crystal clear experience of pain would be just as conscious as an alert awake subject. but if this is what ‘level of consciousness’ means when applied to whole creatures, then it is not an appropriate analysis of global states of consciousness. even if a patient in a vegetative state did have a conscious experience that was as clear as an alert awake neurotypical subject, there are consciousness-related differences between these two subjects that go beyond the quality of their most vivid conscious content. not only do we assume that the alert neurotypical subject has a much wider range of conscious contents available to her than the vegetative state patient, she can endogenously direct the focus of her attention onto specific conscious contents and use these conscious contents to guide her intentional actions in ways that the vegetative state patient cannot. to recapitulate, ‘conscious level’ can’t mean ‘degree of generic consciousness’ since, as i have argued, generic consciousness does not come in degrees. if ‘conscious level’ means ‘level of wakefulness/arousal’ then it is not an adequate analysis of global states since wakefulness/arousal dissociate from global states of consciousness. if ‘conscious level’ means ‘range of conscious contents’ then it is not a suitable analysis of global states since it is an open question whether global state and range of conscious contents dissociate. and if ‘conscious level’ just means ‘range of possible contents’ or ‘vividness of conscious content’ then it is not an adequate analysis of global states of consciousness since there is more to global states of consciousness than just range and vividness of conscious contents. so, while there is something intuitively appealing about the idea that some global states are more conscious than others, by more conscious we can’t mean more generically conscious, or more awake/arousable, or conscious of more content, or capable of being conscious of more content, or experiencing more vivid conscious contents. 3 are global states phenomenal states? according to bayne, hohwy and owen, global states are “states of consciousness that characterise an organism’s overall conscious condition” (bayne et al., 2016). how should we understand ‘states of consciousness’ in this context? perhaps the most natural way to think about states of consciousness is that they are phenomenal states: experiential states that are characterized in terms of what it is like to be in them. on this reading, global states are occurrent experiences that characterise an organism’s overall conscious condition. this appears to be how chalmers thinks of global states. he maintains that the set of global states form a “phemckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org andy mckilliam 8 nomenal family” – which is “a set of mutually exclusive phenomenal properties” (chalmers, 2000, p. 22). on this view, global states of consciousness are states with characteristic phenomenal properties. bayne, hohwy and owen also suggest that this is the right way to think about some (but not all) global states. dreaming is an example of what they call an “occurrent global state” a global state “defined in terms of the individual’s current conscious experiences” (bayne et al., 2016, p. 412). there are a number of ways one might try to characterize global states in terms of current experiences. one might try to account for global states in terms of the unified totality of fine-grained conscious contents that a creature has at a time. on this view global states are to be characterized in terms of the totality of what it is like to be a creature at a time. alternatively, one might follow chalmers in thinking that while global states are phenomenal states, they will “not be defined in terms of specific contents” (chalmers, 2000, p. 18). there are two ways one might go here. global states might have distinct residual phenomenal signatures even once all the conscious content is subtracted away. alternatively, one might appeal to the structural relations between conscious contents in order to account for global states of consciousness. in this section, i argue that none of these options is satisfactory. 3.1 are global states total states? for the purposes of investigating consciousness, it is often useful to focus on local states of consciousness that are characterized in terms of conscious content: states such as the visual experience of a face, the taste of pineapple or the feeling of pain. at any given time, conscious creatures are conscious of a wide range of contents, and these contents are bound together (or at least appear to be bound together which may amount to the same thing) into a unified conscious experience. following bayne we can call this unified conscious experience a “total state” of consciousness (bayne, 2010). an initially appealing thought is that we might be able to account for global states in terms of total states. however, there are two problems here. first, one and the same global state can give rise to radically different total states. second, it is conceivable that two distinct global states can give rise to indistinguishable total states. i’ll elaborate on these in turn. to motivate the idea that one and the same global state can give rise to radically different total states, contrast the totality of what it is like to snorkel around a coral reef with the totality of what it is like to wait in line at the department of motor vehicles. in the first instance you experience a rich visual scene with vibrant colours and intricate shading due to the motion of the waves above. you feel the cool of the water and the familiar weightlessness that accompanies swimming. if you are not a frequent snorkeler, you may experience a mix of excitement and anxiety about breathing through your mouth while your face is submerged. in the mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org what is a global state of consciousness? 9 second instance your visual experience is drab by comparison. your body heavy and feet sore from standing stationary for too long. and in place of excitement, frustration colours your affective experience. in terms of their phenomenal character, these two total states are strikingly different. and yet, they are both cases of a single global state: alert wakefulness. one problem with attempting to account for global states in terms of total states is that total states are much too finely specified. it is not clear what principle could be appealed to in order to account for why these two strikingly different total states are both instances of the same kind of global state. the second issue is that there is reason to think that distinct global states might yield indistinguishable total states. false awakenings – in which you dream that you have woken up – might be an example here. false awakenings are often very convincing. in many cases, the dream experiences during false awakenings appear to be subjectively indistinguishable from waking experience (windt, 2015a). further, it may be that certain dream experiences in neurotypical subjects are indistinguishable from the waking experience of subjects in a minimally conscious state. the problem here is that the totality of what it is like to be a subject at a time may be incapable of distinguishing between distinct global states. the problems that arise when trying to account for global states in terms of the totality of what it is like to be a creature in that state do not have obvious solutions. a more plausible approach could be to follow chalmers’ suggestion that while global states are phenomenal states, they will “not be defined in terms of specific contents” (chalmers, 2000, p. 18). as mentioned earlier, there are two ways in which this might go. it might be that global states have a distinctive phenomenological signature in addition to specific contents. or, it might be that global states structure a subject’s phenomenal field in distinctive ways. i will address these in turn. 3.2 do global states have residual phenomenal signatures? according to the view that global states of consciousness have residual phenomenal signatures it would still be like something to be alert and awake, or to dream, or to have a seizure, even if all the specific contents had been subtracted away. and that this ‘phenomenal residue’ is how we should think about global states of consciousness. at first pass, evidence from sleep research might be taken to support this view. upon waking from sleep, subjects are sometimes able to confirm that they were dreaming even though they are unable to recall any of the details of the dream: so-called “white dreams” (fazekas et al., 2019; strauch & meier, 1996). a natural interpretation of white dreams is that subjects had normal, contentful dreams but failed to encode these dream contents into memory or are unable to retrieve them (cohen, 1974; goodenough, 1991). an alternative interpretation that has recently sparked considerable interest is that subjects are accurately reporting “contentless” mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org andy mckilliam 10 or “imageless” dream experiences (windt, 2015b; windt et al., 2016). on this view, during white dreams,”subjects experience only a minimal form of conscious presence with no […] specific percepts, bodily sensations or thoughts occurring during such experiences” (fazekas et al., 2019, p. 85). it is tempting to infer from this that there is some phenomenal signature to dreaming independent of conscious contents. after all, subjects are able to report that they were dreaming even though they are not able to report any dream contents. however, such an inference would be a mistake. for one thing there is an argument to be made that white dreams are not in fact contentless. rather they are dreams with “low-quality” contents (fazekas et al., 2019). but even if genuinely contentless dreams do occur, this would not support the idea that global states can be accounted for by subtracting conscious content and revealing some residual phenomenal signature distinct to those states. supposedly, all that remains of a dream state once all the conscious content is subtracted away is a “minimal form of conscious presence” (fazekas et al., 2019, p. 85). but it is not clear how this ‘minimal form of conscious presence’ could be used to distinguish one global state from another. presumably, all that is left of any global state once all the conscious content is subtracted away will be a minimal form of conscious presence. so even if there is a residual phenomenal experience once all conscious content is subtracted away, it is far from obvious that global states should have distinct residual phenomenal signatures. if they don’t then the subtraction method is not capable of differentiating between global states of consciousness. 3.3 are global states distinct phenomenal structures? this leaves the view that global states are distinct ways a subject’s phenomenal field can be structured. here the operation is not subtraction but abstraction. rather than subtracting away conscious content and seeing what is left, we abstract away from the specific details concerning the contents of consciousness and focus on structural features that characterise the relations between conscious contents. perhaps what first comes to mind when we think about the structure of a phenomenal field are the relations between conscious contents at a time. if we abstract away the specific content of a total state – recall the total state is the totality of what it’s like to be a creature at a time – we reveal that these contents are unified (bayne, 2010) into a seemingly 3-dimensional field (revonsuo, 2006, pp. 265–281), with a distinctive centre and periphery (watzl, 2017). but when analysing global states, it’s also important to consider temporal structure as well. for example, a structural feature of certain dream states is that they lack narrative coherence. at one stage in a dream you may be at dinner with your family, the very next basejumping off the eiffel tower. in fact, dreams can be so disjointed that researchers are typically unable to distinguish genuine dream reports from artificial reports constructed by splicing together segments of distinct dream reports (stickgold et mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org what is a global state of consciousness? 11 al., 1994).2 narrative coherence characterizes the relationship between conscious contents across time. so, we should think of phenomenal structure here as encompassing both the relations between conscious contents at a time and the relations between conscious contents across time. can phenomenal structure provide a way of characterizing global states? while it is plausible that global states do determine phenomenal structure, it is a mistake to think that global states are distinct phenomenal structures. there are two problems. first, what we take to be a single global state can, at different times, structure the phenomenal field in different ways. second, it is conceivable that distinct global states can nonetheless give rise to phenomenal fields with indistinguishable phenomenal structure. i will address these in turn. consider the following scenario. it’s evening and you are watching a thriller on tv. the suspense generated by the film has your undivided attention. the movie cuts to a commercial break and your attention dissolves away from what’s happening on the screen into features of the surrounding environment. you notice the faint sound of a truck passing outside, remember that you had boiled the kettle to make tea during the previous commercial break, and feel guilty about the neglected pile of ironing on the couch. the structure of your phenomenal field is strikingly different at the end of this scenario than at the beginning. at the beginning, your attention is acutely focussed on what is happening on the screen, resulting in a sharp center-periphery structure to your phenomenal field (watzl, 2017). you have vivid, high-quality contents at the centre of your phenomenal field, but only faint experiences of what is in your phenomenal periphery. at the end, your attention is much more diffusely distributed and the contrast between center and periphery is much less acute. if we identify global states with distinct phenomenal structures, then despite the fact that you are alert and awake the whole time, it looks like we are committed to saying that these are distinct global states. second, i argued earlier that it is conceivable that paradigmatically distinct global states can nonetheless give rise to indistinguishable total states of consciousness. it is also conceivable that two distinct global states could nonetheless give rise to phenomenal fields with indistinguishable phenomenal structure. again, false awakenings – where dream experience appears to replicate waking experience – may provide an example (windt, 2015a). in order to avoid these issues, we need to consider not just the experiences that a subject actually has and the relations between them, but the set of possible experiences that the subject can have while in that state. this reveals perhaps the most convincing reason for abandoning the project of trying to account for global states in terms of occurring experi2moreover, there are candidate explanations for this lack of narrative coherence. lateral prefrontal inactivity may result in diminished capacity for working memory, which in turn “may prevent reflection upon immediately past events leading to an unquestioned, forward progression of the plot as well as frequent narrative divergence” (pace-schott, 2013, p. 2). mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org andy mckilliam 12 ences: global states of consciousness and occurring experiences belong to different ontological categories. global states are complex capacities. part of what it is for me to be alert and awake is for me to be able to experience semantically-rich content. i can see faces and recognise them as faces not merely as bundles of shape, colour, and texture. but i don’t need to be currently experiencing semantically-rich content in order to be alert and awake. similarly, part of what it is to be alert and awake is to be able to make use of conscious contents to guide intentional action and provide verbal reports. but i don’t need to be currently guiding my intentional actions and providing verbal reports in order to be alert and awake. what matters is that i can do these things.3 phenomenal states – occurring experiences – are not capacities. they are occurrents. to illuminate the difference here compare ‘being in pain’ with ‘having the capacity to be in pain’. the former hurts. the latter doesn’t. in the next section i provide an account of global states in terms of consciousnessrelated capacities. 4 the capacities account of global states before explicating the capacities account of global states, it will be helpful to introduce another distinction: the distinction between ‘a system’ and ‘a system in a state.’4 my washing machine has a range of settings: cottons, quick wash, delicates, off. common sense tells us that the washing machine that is currently making a racket during its final spin cycle of a cottons wash is the same washing machine that was quietly sitting in off two hours ago. here, the distinction that i want to draw out is that between ‘my washing machine’ (a system) and ‘my washing machine in the final spin cycle of a cottons wash’ (a system in a state). creatures, like washing machines, are systems. they persist beyond the current states they are in. the alert awake (and highly caffeinated) creature currently writing these sentences is the same creature that was asleep in bed three hours ago. there are a number of ways in which a creature (a system) may be said to have, or fail to have, a capacity. consider the capacity to speak a language. a creature may lack the capacity to speak a language because it has never learned one (as is the case with infants). a creature may lack the capacity to speak a language because it lacks the capacity to learn one (as is the case with most non-human animals). alternatively, a creature may lack the capacity to speak a language because 3the idea that global states of consciousness are sets of capacities is not new. in his account of waking consciousness as “capacitation”, crowther (2018) traces the idea back to aristotle. more recently, fortier-davy & millière (2020) also discuss global states in terms of capacities. 4in the philosophical literature the distinction is more commonly put as continuants and occurrences, or things and events (lowe, 2005). i opt for ‘system’ and ‘state’ here as it is likely to be more intuitive to non-philosophical readers and it is less likely to get us bogged down in long standing metaphysical issues about how to unpack this distinction that are tangential to the current project. mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org what is a global state of consciousness? 13 its capacity to speak a language is currently off-line (as is the case with anaesthetized adult humans). it is the third sense that i am interested in here. in order to understand the capacities account of global states we need to appreciate the distinction between the total set of capacities that a creature (a system) has, and the set of capacities that creature has while in a particular global state. an example can help to make this more intuitive. take my capacity to bake cookies. i am not always able to manifest this capacity (i can’t bake cookies in my sleep i am sorry to say). so, while ‘andy’ (a system) has the capacity to bake cookies, ‘sleeping andy’ (a system in a particular state) does not have the capacity to bake cookies. while andy is asleep, his capacity to bake cookies is off-line. according to the capacities account of global states, global states of consciousness are states of creatures that characterize the range of consciousness-related capacities that are currently online. in order for this account of global states of consciousness to be informative we need to know what makes a capacity a consciousness-related capacity. how are we to distinguish consciousness-related capacities from capacities that are not consciousness-related? at first glance, given that consciousness is realized in biological organisms, many of the capacities involved in maintaining biological organisms are related to consciousness in some way or other. in some sense, being able to breathe is a capacity that is related to consciousness. if you can’t breathe you will quickly lose consciousness. but being able to breathe is not a consciousnessrelated capacity as i am using that term. rather, being able to breathe is a background condition that enables the possibility of consciousness. we might call it a consciousness-enabling capacity. i reserve the term ‘consciousness-related capacity’ to refer to those capacities that cannot be specified without reference to consciousness.5 consider once more the capacity to breathe. in order to specify what it is to have this capacity we don’t need to refer to consciousness. you can breathe when you have the capacity to control the muscles in your chest in such a way that it draws air into your lungs. as such, the capacity to breathe is not a consciousnessrelated capacity. by contrast, the capacity to visually experience colour cannot be specified without reference to consciousness. here it is important to point out that the capacity to visually experience colour is not just the capacity to perform better than chance on forced choice tests. it is the capacity to have phenomenally conscious mental states with colours as their content. similarly, the capacity to verbally report your conscious experience of colour cannot be specified without reference to consciousness. some aspects of attention offer another example. the capacity to focus one’s attention on the colour of an object so as to bring a vivid conscious experience of colour into the center of one’s phenomenal field cannot be specified without reference to consciousness. each of the above count as consciousness-related capacities. 5thanks to tim bayne for this suggestion. mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org andy mckilliam 14 specifying the full range of consciousness-related capacities is beyond the scope of this paper, but we might begin with the three mentioned above. following bayne, hohwy and owen we can construct a multi-dimensional account of global states with capacities along three dimensions (or sets of dimensions): a content dimension, an attentional dimension, and a cognitive dimension (bayne et al., 2016; bayne & hohwy, 2016). the content dimension specifies the range and quality of contents a subject is capable of being conscious of while in that state. the attentional dimension specifies the extent to which attentional resources can be focussed on a particular target and whether or not that focus can be endogenously directed as well as exogenously captured. the cognitive dimension specifies the range of cognitive systems into which the contents of consciousness can be mobilized. the capacities account of global states of consciousness is as follows: global states of consciousness: states of creatures (systems) that regulate (i) the range and quality of conscious contents the creature is capable of experiencing while in that state, (ii) the range of cognitive systems into which those contents can be mobilized while in that state, and (iii) the range of attentional capacities the creature has while in that state. contrasting alert wakefulness with dreaming can shed light on what this might look like. while alert and awake, the full range of a creature’s consciousnessrelated-capacities are online. alert awake adults can be conscious of a wide range of high-quality, semantically-rich content in addition to low-level features. they can see faces as particular people, not merely as bundles of shape, colour and texture. they can hear speech as meaningful, not merely as sounds, and so on. those conscious contents can be mobilized into a wide range of cognitive systems. alert awake adults can make use of conscious contents to guide intentional action, provide verbal reports, consolidate memories, and so on. and, alert awake adults have a wide range of attentional capacities. in addition to having their attention exogenously captured by features of the environment they can endogenously direct their attention to particular items, and by modulating the intensity with which they direct the focus of their attention they have the capacity to modulate the centre/periphery structure of their phenomenal field. when dreaming some of these consciousness-related capacities are offline. although dreaming subjects are still capable of experiencing a wide range of semantically-rich contents, these may be lower-quality. dreams often involve vague contents: contents “whose identity or precise nature is indeterminate, unknown, or obscure in a way which does not occur in waking life” (revonsuo & salmivalli, 1995, p. 174). this may be because the low-level features (colour, shape, and texture) that typically accompany semantically rich conscious content in alert wakefulness are less vivid (fazekas et al., 2019). in some cases, high-level content may even dissociate entirely from low-level features. it is not uncommon mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org what is a global state of consciousness? 15 for subjects in dream studies to report “just knowing” that the person that they were “seeing” in their dream was their mother, even though they cannot provide specific information about low-level features (windt, 2015a, p. 267). while dreaming attentional capacities may also be substantially limited. in most cases dreaming subjects lack the endogenous control of attention that they have while awake – lucid dreams, where dreamers are aware they are dreaming and have a degree of dream control being the notable exception (voss & hobson, 2014). and there also tends to be limitations to the range of cognitive systems into which the contents of consciousness can be mobilized. conscious contents may be available for memory consolidation, but they are typically not available for reasoning and guiding intentional action. there is a question as to whether global states can be characterised as fixed sets of capacities or need to be indexed to individual subjects. disorders of consciousness – minimally conscious state, vegetative state, and coma – tend to be characterised in absolute terms. however, for some global states, alert wakefulness for example, it seems more appropriate to index global states to specific subjects. both infants and adults can be alert and awake, but their consciousness-related capacities are very different. alert wakefulness may be best characterized as a state in which a subject’s full range of consciousness-related capacities are online – to be “fully switched on”, as crowther (2018) puts it. this may have considerable implications for taxonomies of global states. however, my primary goal here is not taxonomic, so i will leave this issue for another time. there are a number of advantages of the capacities view of global states. first, it can make sense of the idea that global states admit of some degree of ordering. roughly, the more consciousness-related capacities a creature has the more conscious that creature is. however, a few qualifications are in order. we need to be clear that the notion of consciousness at play here is not generic consciousness. moreover, as bayne et al. (2016) stress, this approach will not admit of a strict ordering. two global states may have the same number of consciousness-related capacities online but be very different in terms of which consciousness-related capacities are online. there may be no fact of the matter as to which is the ‘higher’ global state of consciousness. second, this framework provides the conceptual tools to make sense of a number of claims about consciousness that at first glance appear to be conceptual contradictions. i mentioned earlier that it is an open question whether or not high level of consciousness without conscious content is possible. if conscious level is thought of as generic consciousness – the property of having a subjective perspective from which it is like something to be – it is not obvious how to make sense of this claim. a creature’s being generically conscious entails that there is something it is like to be that creature. a creature’s having no conscious contents appears to entail that there is nothing it is like to be that creature. this looks like a contradiction. the capacities account of global states can make sense of this claim. a creature might be in a high level of consciousness by virtue of having a wide range mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org andy mckilliam 16 of consciousness-related capacities online, and yet have no conscious contents by virtue of failing to manifest any of its content-related capacities. the capacities account of global states can also account for certain meditative states such as “witness consciousness” (albahari, 2009) or “pure awareness” (thompson, 2014) as well as contentless dream states (windt, 2015a) in which there is conscious awareness or presence in the absence of any conscious contents. in such states, subjects might have the full range of attentional capacities online, but none of their content-related capacities online. while the capacities account has the resources to make sense of these states, it is not committed to thinking that these are states that ever actually occur. it may be that the mechanisms underlying a creature’s consciousness-related capacities are such that attentional capacities are only ever online when at least some range of cognitive capacities and content-related capacities are also online. 5 implications for the science of consciousness although a number of researchers have recently begun to explore global states of consciousness as i have presented them here (bayne & carter, 2018; bonhomme et al., 2019), for much of the past 30 years they have seen relatively little attention from consciousness scientists. theories of consciousness have said little about global states of consciousness. the integrated information theory of consciousness (iit) has focussed exclusively on conscious contents and the structural relationships between them. in its current form, it appears to lack the resources to account for global states of consciousness as conceived under the capacities account (tononi et al., 2016). cognitive theories of consciousness appear better equipped in this regard. however, they have primarily been developed in the context of investigating what is required for a mental representation to be conscious. as a result, they have said little about the global states of conscious creatures. for example, according to the global neuronal workspace theory, in order for a neural representation to be conscious, its content needs to be mobilized into a global workspace rendering it available for a wide range of cognitive systems (baars, 2005; dehaene et al., 2011). here we should take care not to confuse the contents of the global workspace with a creature’s global state of consciousness. rather, a creature’s global state of consciousness might be taken to correspond to the current state of the perceptual, cognitive, and attentional systems that are connected to the workspace, and the ways in which these connections are gated. perhaps the closest any theory of consciousness has come to addressing global states of consciousness is the ref-con model proposed by mogensen and overgaard (mogensen & overgaard, 2017). according to this theory it is only when the content of a perceptual representation is integrated into a “dynamic network” combining elements from sensory areas with “a broad spectrum of other dimensions relevant to the current status of the individual” that it become available for cognitive access, behavioural control and conscious awaremckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org what is a global state of consciousness? 17 ness (mogensen & overgaard, 2017, p. 17). while there may be some overlap here between this “dynamic network” and a creature’s global state of consciousness, it remains unclear exactly what these authors have in mind. considerable work remains to be done in clarifying how and indeed whether the various theories of consciousness can accommodate global states of consciousness. moreover, the central project in the science of consciousness – the search for the nccs – has been explicitly defined as the search for the minimal set of neural mechanisms that are jointly sufficient for the occurrence of a conscious experience (chalmers, 2000; fink, 2016; koch, 2004). revealing the minimal mechanisms supporting occurrent experiences is an important task for the science of consciousness. doing so will go some way towards extending our ability to make reliable inferences about (1) which creatures and systems are conscious and (2) what they are conscious of. though it is worth noting some complications here. for the most part, nccs are defined as minimally sufficient in the context of a conscious neurotypical adult rather than minimally sufficient simpliciter (block, 2005; chalmers, 2000; koch, 2004). if this is the case then revealing the ncc won’t straightforwardly extend our ability to make inferences about which creatures/systems are conscious and which not. it is within the context of investigating the minimal neural correlates of consciousness in alert neurotypical adults that current debates between cognitive and sensory theories of consciousness are situated (aru et al., 2012; block, 2007, 2019; boly et al., 2017; cohen & dennett, 2011; koch & tsuchiya, 2007; odegaard et al., 2017; phillips, 2018). cognitive theorists argue that the processes in prefrontal cortex that support attention and cognition are part of the minimal neural correlates of consciousness in alert neuro typical adults. sensory theorists argue that they are confounds that need to be screened-off. this final section serves as a reminder that we want a mature mechanistic science of consciousness to do more than just reveal the minimal mechanisms that support consciousness. we want it to account for the full suite of capacities that consciousness imbues us with. and in the context of this broader project the neural mechanism supporting attention and cognition are clearly relevant. i will briefly consider two cases to motivate this. first, consider the use of anaesthesia as a tool for controlling consciousness for surgical purposes. what anaesthetists want to ensure is that patients don’t experience the surgery. in order to achieve this it is not essential that patients are rendered completely unconscious. all that is required is that their consciousness be disconnected from the environment. dissociative anaesthetics such as ketamine work in just this way (sanders et al., 2012). under dissociative anaesthetics, subjects remain conscious, but the capacities that enable them to experience the external world are off-line. it is reasonable to expect that a mature mechanistic science of consciousness facilitate prediction, explanation, and fine-grained control over conscious creatures’ connectedness to the environment. this goes beyond revealing the minimal mechanisms that support occurrent experiences. mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org andy mckilliam 18 a second example comes from considering disorders of consciousness. compare the conscious life of a creature who has the capacity to endogenously direct her attention with that of a creature who lacks that capacity. there is a very real sense in which the conscious experiences of the former are “part of her own making” (watzl, 2017, p. 44). by actively directing her attention to the sound of the wind through the trees or the feeling of grass under her feet, she can bring those contents to the center of her conscious field and thereby ensure that she enjoys high-quality representations of those conscious contents. similarly, by intentionally moving her body, she can alter the sensory input she receives and by virtue of doing so, play an active role in the production of her future experiences. by contrast, in disordered states of consciousness such as the minimally conscious state subjects lack, or have greatly diminished capacities to orchestrate the unfolding of their conscious life. even if it turns out that attentional and cognitive consciousness-related capacities are not necessary constituents of the minimal mechanisms sufficient for supporting conscious experiences, they are central to the endogenous formation of intentions, goals, and desires by virtue of which conscious creatures play an active causal role in orchestrating their conscious life. the take home message here is that a mature mechanistic science of consciousness needs to account for more than just occurring experiences. it also needs to account for (i) the ways in which conscious experiences unfold over time, (ii) the active role the subject of those experiences plays in how they unfold, and (iii) the ways in which these can be compromised in disordered global states of consciousness. in order to achieve these goals, research into the neural mechanisms underlying occurrent experiences needs to be integrated with the research from other areas of cognitive neuroscience such as attention, working memory, and so on. to be clear, the idea is not that occurrent experiences should be reduced to these other constructs. rather, the idea is that a mature mechanistic science of consciousness needs to make explicit the causal/constitutive relationships between them. it needs to reveal how changes in attention, working memory, expectation and so on effect consciousness, and it needs to reveal how changes in consciousness can in turn effect these cognitive processes. 6 conclusion in this paper i have argued that global states of consciousness cannot be fruitfully understood in terms of ‘conscious level’, nor can they be reduced to occurring experiences. i have offered an account of global states of consciousness as states of creatures that characterize the range of consciousness-related capacities that are currently online. i have finished with a reminder that a mature mechanistic science of consciousness needs to do more than reveal the minimal neural mechanisms that support consciousness. it needs to integrate these findings with cognitive neuroscience more broadly in order to account for the full suite of consciousness-related capacities that determine a creature’s global state of conmckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org what is a global state of consciousness? 19 sciousness. considerable work remains to be done in considering the implications that the capacities account of global states has for our taxonomy of global states of consciousness and for extending consciousness research so as to investigate the full range of consciousness-related capacities. acknowledgments i would like to thank tim bayne, jakob hohwy, and members of the cognition and philosophy lab at monash university for helpful feedback on early versions of this paper. this research was supported by an australian government research training program (rtp) scholarship. references albahari, m. (2009). witness-consciousness: its definition, appearance and reality. journal of consciousness studies, 16(1), 62–84. https://www.ingentaconnect.com/content/imp/jcs/2009/00000016/00000001/art00003 aru, j., bachmann, t., singer, w., & melloni, l. (2012). distilling the neural correlates of consciousness. neuroscience & biobehavioral reviews, 36(2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.12.003 aru, j., suzuki, m., rutiku, r., larkum, m. e., & bachmann, t. (2019). coupling the state and contents of consciousness. frontiers in systems neuroscience, 13, 43. https://doi.org/10.3389/fnsys.2019.00043%20 baars, b. j. (2005). global workspace theory of consciousness: toward a cognitive neuroscience of human experience. in s. laureys (ed.), the boundaries of consciousness: neurobiology and neuropathology (vol. 150, pp. 45–53). elsevier. https://doi.org/10.1016/s0079-6123(05)50004-9 bachmann, t., & hudetz, a. g. (2014). it is time to combine the two main traditions in the research on the neural correlates of consciousness: c= l× d. frontiers in psychology, 5, 940. https://doi.org/10.3389/fpsyg.2014.00940%20 bayne, t. (2007). conscious states and conscious creatures: explanation in the scientific study of consciousness. philosophical perspectives, 21, 1–22. https://doi.org/10.1111/j.1520-8583.2007.00118.x%20 bayne, t. (2010). the unity of consciousness. oxford university press. bayne, t., & carter, o. (2018). dimensions of consciousness and the psychedelic state. neuroscience of consciousciousness, 2018(1), niy008. https://doi.org/10.1093/nc/niy008 bayne, t., & hohwy, j. (2016). modes of consciousness. in w. sinnott-armstrong (ed.), finding consciousness: the neuroscience, ethics and law of severe brain damage (pp. 57–80). oxford university press. https://doi.org/10.1093/acprof: oso/9780190280307.001.0001%20 bayne, t., hohwy, j., & owen, a. m. (2016). are there levels of consciousness? trends in cognitive sciences, 20(6), 405–413. https://doi.org/10.1016/j.tics.2016.03.009 block, n. (2005). two neural correlates of consciousness. trends in cognitive sciences, 9(2), 46–52. https://doi.org/10.1016/ j.tics.2004.12.006 block, n. (2007). consciousness, accessibility, and the mesh between psychology and neuroscience. behavioral and brain sciences, 30(5-6), 481–499. https://doi.org/10.1017/s0140525x07002786 block, n. (2019). what is wrong with the no-report paradigm and how to fix it. trends in cognitive sciences, 23(12), 1003– 1013. https://doi.org/10.1016/j.tics.2019.10.001%20 boly, m., massimini, m., tsuchiya, n., postle, b. r., koch, c., & tononi, g. (2017). are the neural correlates of consciousness in the front or in the back of the cerebral cortex? clinical and neuroimaging evidence. journal of neuroscience, 37(40), 9603–9613. https://doi.org/10.1523/jneurosci.3218-16.2017 boly, m., seth, a. k., wilke, m., ingmundson, p., baars, b., laureys, s., edelman, d., & tsuchiya, n. (2013). consciousness in humans and non-human animals: recent advances and future directions. frontiers in psychology, 4, 625. https: //doi.org/10.3389/fpsyg.2013.00625%20 bonhomme, v., staquet, c., montupil, j., defresne, a., kirsch, m., martial, c., vanhaudenhuyse, a., chatelle, c., larroque, s. k., raimondo, f., demertzi, a., bodart, o., laureys, s., & gosseries, o. (2019). general anesthesia: a probe to explore consciousness. frontiers in systems neuroscience, 13(36). https://doi.org/10.3389/fnsys.2019.00036 brennan, p. m., murray, g. d., & teasdale, g. m. (2018). simplifying the use of prognostic information in traumatic brain injury. part 1: the gcs-pupils score: an extended index of clinical severity. journal of neurosurgery, 128(6), 1612– 1620. https://doi.org/10.3171/2017.12.jns172780 mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.ingentaconnect.com/content/imp/jcs/2009/00000016/00000001/art00003 https://doi.org/10.1016/j.neubiorev.2011.12.003 https://doi.org/10.3389/fnsys.2019.00043%20 https://doi.org/10.1016/s0079-6123(05)50004-9 https://doi.org/10.3389/fpsyg.2014.00940%20 https://doi.org/10.1111/j.1520-8583.2007.00118.x%20 https://doi.org/10.1093/nc/niy008 https://doi.org/10.1093/acprof:oso/9780190280307.001.0001%20 https://doi.org/10.1093/acprof:oso/9780190280307.001.0001%20 https://doi.org/10.1016/j.tics.2016.03.009 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1017/s0140525x07002786 https://doi.org/10.1016/j.tics.2019.10.001%20 https://doi.org/10.1523/jneurosci.3218-16.2017 https://doi.org/10.3389/fpsyg.2013.00625%20 https://doi.org/10.3389/fpsyg.2013.00625%20 https://doi.org/10.3389/fnsys.2019.00036 https://doi.org/10.3171/2017.12.jns172780 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org andy mckilliam 20 carruthers, p. (2019). human and animal minds: the consciousness questions laid to rest. oxford university press. chalmers, d. j. (2000). what is a neural correlate of consciousness? in t. metzinger (ed.), neural correlates of consciousness (pp. 17–39). mit press. cohen, d. (1974). toward a theory of dream recall. psychological bulletin, 81(2), 138. https://doi.org/10.1037/h0037616%20 cohen, m. a., & dennett, d. c. (2011). consciousness cannot be separated from function. trends in cognitive sciences, 15(8), 358–364. https://doi.org/10.1016/j.tics.2011.06.008 crick, f., & koch, c. (2003). a framework for consciousness. nature neuroscience, 6(2), 119–126. https://doi.org/10.1038/nn 0203-119 crowther, t. (2018). experience, dreaming, and the phenomenology of wakeful consciousness. in f. dorsch, f. macpherson, & m. nida-rümelin (eds.), phenomenal presence (pp. 252–282). oxford university press. https://doi.org/10.1093/oso/ 9780199666416.003.0011 dehaene, s., changeux, j.-p., & naccache, l. (2011). the global neuronal workspace model of conscious access: from neuronal architectures to clinical applications. in d. stanislas & c. y. (eds.), characterizing consciousness: from cognition to the clinic? (pp. 55–84). springer. https://doi.org/10.1007/978-3-642-18015-6_4 fazekas, p., nemeth, g., & overgaard, m. (2019). white dreams are made of colours: what studying contentless dreams can teach about the neural basis of dreaming and conscious experiences. sleep medicine reviews, 43, 84–91. https: //doi.org/10.1016/j.smrv.2018.10.005%20 fazekas, p., & overgaard, m. (2018). a multi‐factor account of degrees of awareness. cognitive science, 42(6), 1833–1859. https://doi.org/10.1111/cogs.12478 fink, s. b. (2016). a deeper look at the “neural correlate of consciousness”. frontiers in psychology, 7, 1044. https://doi.org/ 10.3389/fpsyg.2016.01044%20 fortier-davy, m., & millière, r. (2020). the multi-dimensional approach to drug-induced states: a commentary on bayne and carter’s “dimensions of consciousness and the psychedelic state”. neuroscience of consciousness, 2020(1), niaa004. https://doi.org/10.1093/nc/niaa004%20 funkhouser, e. (2006). the determinable-determinate relation. noûs, 548–569. https://doi.org/10.1111/j.1468-0068.2006.00 623.x%20 goodenough, d. r. (1991). dream recall: history and current status of the field. in s. j. ellman & j. s. antrobus (eds.), the mind in sleep: psychology and psychophysiology (2nd ed., pp. 143–171). john wiley & sons. hohwy, j. (2007). the search for neural correlates of consciousness. philosophy compass, 2(3), 461–474. https://doi.org/10 .1111/j.1747-9991.2007.00086.x hohwy, j. (2009). the neural correlates of consciousness: new experimental approaches needed? consciousness and cognition, 18(2), 428–438. https://doi.org/10.1016/j.concog.2009.02.006%20 hohwy, j., & bayne, t. (2015). the neural correlates of consciousness: causes, confounds and constituents. in s. m. miller (ed.), the constitution of phenomenal consciousness: toward a science and theory (pp. 155–176). john benjamins publishing company. https://doi.org/10.1075/aicr.92.06hoh klein, c. (2017). consciousness, intention, and command-following in the vegetative state. british journal for the philosophy of science, 68(1), 27–54. https://doi.org/10.1093/bjps/axv012%20 koch, c. (2004). the quest for consciousness: a neurobiological approach. roberts & company publishers. koch, c., massimini, m., boly, m., & tononi, g. (2016). neural correlates of consciousness: progress and problems. nature reviews neuroscience, 17(5), 307–321. https://doi.org/10.1038/nrn.2016.22 koch, c., & tsuchiya, n. (2007). attention and consciousness: two distinct brain processes. trends in cognitive sciences, 11(1), 16–22. https://doi.org/10.1016/j.tics.2006.10.012%20 laureys, s. (2005). the neural correlate of (un)awareness: lessons from the vegetative state. trends in cognitive sciences, 9(12), 556–559. https://doi.org/10.1016/j.tics.2005.10.010%20 lowe, e. j. (2005). the four-category ontology: a metaphysical foundation for natural science. oxford university press. mashour, g., & avidan, m. (2015). intraoperative awareness: controversies and non-controversies. british journal of anaesthesia, 115, i20–i26. https://doi.org/10.1093/bja/aev034%20 michel, m. (2019). the mismeasure of consciousness: a problem of coordination for the perceptual awareness scale. philosophy of science, 86(5), 1239–1249. https://doi.org/10.1086/705509%20 michel, m., & morales, j. (2020). minority reports: consciousness and the prefrontal cortex. mind & language, 35(4), 493–513. https://doi.org/10.1111/mila.12264 mogensen, j., & overgaard, m. (2017). reorganization of the connectivity between elementary functions – a model relating conscious states to neural connections. frontiers in psychology, 8, 625. https://doi.org/10.3389/fpsyg.2017.00625%20 odegaard, b., knight, r. t., & lau, h. (2017). should a few null findings falsify prefrontal theories of conscious perception? journal of neuroscience, 37(40), 9593–9602. https://doi.org/10.1523/jneurosci.3217-16.2017%20 mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1037/h0037616%20 https://doi.org/10.1016/j.tics.2011.06.008 https://doi.org/10.1038/nn0203-119 https://doi.org/10.1038/nn0203-119 https://doi.org/10.1093/oso/9780199666416.003.0011 https://doi.org/10.1093/oso/9780199666416.003.0011 https://doi.org/10.1007/978-3-642-18015-6_4 https://doi.org/10.1016/j.smrv.2018.10.005%20 https://doi.org/10.1016/j.smrv.2018.10.005%20 https://doi.org/10.1111/cogs.12478 https://doi.org/10.3389/fpsyg.2016.01044%20 https://doi.org/10.3389/fpsyg.2016.01044%20 https://doi.org/10.1093/nc/niaa004%20 https://doi.org/10.1111/j.1468-0068.2006.00623.x%20 https://doi.org/10.1111/j.1468-0068.2006.00623.x%20 https://doi.org/10.1111/j.1747-9991.2007.00086.x https://doi.org/10.1111/j.1747-9991.2007.00086.x https://doi.org/10.1016/j.concog.2009.02.006%20 https://doi.org/10.1075/aicr.92.06hoh https://doi.org/10.1093/bjps/axv012%20 https://doi.org/10.1038/nrn.2016.22 https://doi.org/10.1016/j.tics.2006.10.012%20 https://doi.org/10.1016/j.tics.2005.10.010%20 https://doi.org/10.1093/bja/aev034%20 https://doi.org/10.1086/705509%20 https://doi.org/10.1111/mila.12264 https://doi.org/10.3389/fpsyg.2017.00625%20 https://doi.org/10.1523/jneurosci.3217-16.2017%20 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org what is a global state of consciousness? 21 overgaard, m., & overgaard, r. (2010). neural correlates of contents and levels of consciousness. frontiers in psychology, 1, 164. https://doi.org/10.3389/fpsyg.2010.00164%20 pace-schott, e. f. (2013). dreaming as a story-telling instinct. frontiers in psychology, 4, 159. https://doi.org/10.3389/fpsyg. 2013.00159%20 pal, d., dean, j. g., liu, t., li, d., watson, c. j., hudetz, a. g., & mashour, g. a. (2018). differential role of prefrontal and parietal cortices in controlling level of consciousness. current biology, 28(13), 2145–2152. e5. https://doi.org/10.1101/ 262360%20 pal, d., li, d., dean, j. g., brito, m. a., liu, t., fryzel, a. m., hudetz, a. g., & mashour, g. a. (2020). level of consciousness is dissociable from electroencephalographic measures of cortical connectivity, slow oscillations, and complexity. journal of neuroscience, 40(3), 605–618. https://doi.org/10.1523/jneurosci.1910-19.2019%20 phillips, i. (2018). the methodological puzzle of phenomenal consciousness. philosophical transactions of the royal society b: biological sciences, 373(1755), 20170347. https://doi.org/10.1098/rstb.2017.0347%20 revonsuo, a. (2006). inner presence: consciousness as a biological phenomenon. mit press. revonsuo, a., & salmivalli, c. (1995). a content analysis of bizarre elements in dreams. dreaming, 5(3), 169. https: //doi.org/10.1037/h0094433%20 rosenthal, d. (2005). consciousness and mind. clarendon press. rosenthal, d. (2019). consciousness and confidence. neuropsychologia, 128, 255–265. https://doi.org/10.1016/j.neuropsych ologia.2018.01.018%20 sanders, r. d., tononi, g., laureys, s., sleigh, j. w., & warner, d. s. (2012). unresponsiveness ≠ unconsciousness. the journal of the american society of anesthesiologists, 116(4), 946–959. https://doi.org/10.1097/aln.0b013e318249d0a7%20 seth, a. (2009). explanatory correlates of consciousness: theoretical and computational challenges. cognitive computation, 1(1), 50–63. https://doi.org/10.1007/s12559-009-9007-x%20 stender, j., laureys, s., & gosseries, o. (2017). altered states of consciousness after brain injury. in s. schneider & m. velmans (eds.), the blackwell companion to consciousness (pp. 662–681). wiley online library. https://doi.org/10.1 002/9781119132363.ch47%20 stickgold, r., rittenhouse, c. d., & hobson, j. a. (1994). dream splicing: a new technique for assessing thematic coherence in subjective reports of mental activity. consciousness and cognition, 3(1), 114–128. https://doi.org/10.1006/ccog.199 4.1008 strauch, i., & meier, b. (1996). in search of dreams: results of experimental dream research. suny press. teasdale, g., maas, a., lecky, f., manley, g., stocchetti, n., & murray, g. (2014). the glasgow coma scale at 40 years: standing the test of time. the lancet neurology, 13(8), 844–854. https://doi.org/10.1016/s1474-4422(14)70120-6 thompson, e. (2014). waking, dreaming, being: self and consciousness in neuroscience, meditation, and philosophy. columbia university press. tononi, g., boly, m., massimini, m., & koch, c. (2016). integrated information theory: from consciousness to its physical substrate. nature reviews neuroscience, 17(7), 450–461. https://doi.org/10.1038/nrn.2016.44 tononi, g., & koch, c. (2008). the neural correlates of consciousness – an update. annals of the new york academy of sciences, 1124(1), 239–261. https://doi.org/10.1196/annals.1440.004 voss, u., & hobson, a. (2014). what is the state-of-the-art on lucid dreaming? recent advances and questions for future research. in t. k. metzinger & j. m. windt (eds.), open mind. mind group. https://doi.org/10.15502/9783958570306 watzl, s. (2017). structuring mind: the nature of attention and how it shapes consciousness. oxford university press. wilson, j. (2017). determinables and determinates. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (spring 2017). metaphysics research lab, stanford university. https://plato.stanford.edu/archives/spr2017/entries/determinatedeterminables/ windt, j. m. (2015a). dreaming: a conceptual framework for philosophy of mind and empirical research. the mit press. windt, j. m. (2015b). just in time—dreamless sleep experience as pure subjective temporality. in t. k. metzinger & j. m. windt (eds.), open mind. mind group. windt, j. m., nielsen, t., & thompson, e. (2016). does consciousness disappear in dreamless sleep? trends in cognitive sciences, 20(12), 871–882. https://doi.org/10.1016/j.tics.2016.09.006 mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.3389/fpsyg.2010.00164%20 https://doi.org/10.3389/fpsyg.2013.00159%20 https://doi.org/10.3389/fpsyg.2013.00159%20 https://doi.org/10.1101/262360%20 https://doi.org/10.1101/262360%20 https://doi.org/10.1523/jneurosci.1910-19.2019%20 https://doi.org/10.1098/rstb.2017.0347%20 https://doi.org/10.1037/h0094433%20 https://doi.org/10.1037/h0094433%20 https://doi.org/10.1016/j.neuropsychologia.2018.01.018%20 https://doi.org/10.1016/j.neuropsychologia.2018.01.018%20 https://doi.org/10.1097/aln.0b013e318249d0a7%20 https://doi.org/10.1007/s12559-009-9007-x%20 https://doi.org/10.1002/9781119132363.ch47%20 https://doi.org/10.1002/9781119132363.ch47%20 https://doi.org/10.1006/ccog.1994.1008 https://doi.org/10.1006/ccog.1994.1008 https://doi.org/10.1016/s1474-4422(14)70120-6 https://doi.org/10.1038/nrn.2016.44 https://doi.org/10.1196/annals.1440.004 https://doi.org/10.15502/9783958570306 https://plato.stanford.edu/archives/spr2017/entries/determinate-determinables/ https://plato.stanford.edu/archives/spr2017/entries/determinate-determinables/ https://doi.org/10.1016/j.tics.2016.09.006 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org andy mckilliam 22 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. mckilliam, a. (2020). what is a global state of consciousness? philosophy and the mind sciences, 1(ii), 7. https://doi.org/10.33735/phimisci.2020.ii.58 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.58 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction are global states levels of consciousness? are global states phenomenal states? are global states total states? do global states have residual phenomenal signatures? are global states distinct phenomenal structures? the capacities account of global states implications for the science of consciousness conclusion breaking the self: radical disruptions of self-consciousness and impossible conscious experiences breaking the self radical disruptions of self-consciousness and impossible conscious experiences wanja wiesea (wawiese@uni-mainz.de) abstract are there logically possible types of conscious experience that are nomologically impossible, given independently justified assumptions about the neural underpinnings of consciousness in human beings? in one sense, this is trivial: just consider the fact that the types of perceptual experiences we can have are limited by our sensory organs. but there may be non-trivial types of conscious experience that are impossible. for instance, if there is a basic type of self-consciousness, corresponding to a phenomenal property that is nomologically necessary for consciousness, then experiences lacking this phenomenal property will be (nomologically) impossible. more generally, it may be that there are causal dependencies between the neural mechanisms that are required to instantiate distinct phenomenal properties (in human beings). if this is the case, instantiating one of these phenomenal properties without certain others may be impossible, which means there are non-trivial cases of nomologically impossible types of conscious experience. this paper clarifies this hypothesis, outlines a general methodology for its investigation, and relates it to research on radical disruptions of self-consciousness. keywords consciousness ∙ impossible conscious experiences · self-consciousness · sense of agency · sense of boundaries · sense of time this article is part of a special issue on “radical disruptions of self-consciousness”, edited by thomas metzinger and raphaël millière. 1 introduction this paper explores the hypothesis that some logically possible configurations of self-consciousness are nomologically impossible, given certain assumptions about consciousness and its neural underpinnings in human beings. put differently, an impossible conscious experience, as defined here, is an experience that cannot be ajohannes gutenberg university mainz wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-3338-7398 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 2 generated by human brains.1 radical disruptions of self-consciousness involving a total loss of self-consciousness are logically possible, but some authors doubt that they are nomologically possible (given the human neural architecture). if correct, this would provide an example of an impossible conscious experience. however, as some of the other contributions to this special issue show, that claim is at least controversial (and probably false, see letheby, 2020; metzinger, 2020; millière, 2020, all in this special issue; for the opposite view, see sebastián, 2020, this special issue) – hence, i will not take a stance on the question whether conscious experiences without any kind of self-consciousness whatsoever are possible. instead, i will pursue slightly more general goals. first, i will argue that research on impossible conscious experiences is relevant to understanding consciousness. if we had a complete theory of consciousness (including detailed knowledge of the neural mechanisms of consciousness), we would be able to say which types of conscious experience are possible and which are not. i shall argue that the reverse statement is also true: impossible experiences can provide insights into the neural mechanisms of consciousness. secondly, i will propose a general methodology for finding evidence for impossible conscious experiences. in addition, i will argue that radical disruptions of self-consciousness (whether they involve a total loss of self-consciousness or not) provide particularly good opportunities for finding candidates for impossible experiences. in what follows, i will first define what an impossible conscious experience is (section 2), and why research on impossible conscious experiences is relevant (section 3). then i will propose general strategies to investigate impossible conscious experiences (section 4). finally, i will apply one of these strategies to an example, viz. the hypothesis that having a sense of agency without a sense of time is impossible (section 5). 2 what exactly is an impossible conscious experience? the brain is an extremely plastic organ that is capable of generating a seemingly unbounded number of different conscious experiences.2 at the same time, the brain is a finite organ, which has not been selected for its ability to generate the largest number of logically possible experiences, but for its ability to enable flexible control of behavior in changing environments. therefore, it is likely that some types of conscious experience are logically possible, but in fact impossible for creatures like us. this is the main hypothesis of this paper: there are logically possible 1unless indicated otherwise, i will use “possible” and “impossible” to refer to nomological possibility throughout this paper. 2i will use the terms “conscious experience”, “experience”, and “conscious state” interchangeably in this paper. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 3 types of conscious experience that are nomologically impossible,3 given independently justified assumptions about the neural underpinnings of consciousness in human beings.4 in the following, i shall use the term “impossible (conscious) experience” as a shorthand for this more complicated formulation (unless indicated otherwise). this formulation raises a few questions: 1) how do we individuate types of conscious experience? 2) what are independently justified assumptions about consciousness and its neural underpinnings? 3) what does it mean to say that a conscious experience is nomologically impossible, given certain assumptions? the first question asks for individuation criteria for types of conscious experience. the second question asks about the relationship between types of experience and types of neural activity and structure. the third asks what bearing this relationship has on whether a given type of experience is nomologically possible or not. let us address these questions in turn. 2.1 individuation criteria here, i will stipulate that types of conscious experience can be individuated in terms of their phenomenal character, i.e., by the phenomenal properties instantiated by these experiences. examples of phenomenal properties are experienced perceptual features, such as experienced redness, the experienced smell of a rose, a feeling of warmth in your body, the sense of mental ownership accompanying 3this is trivially true, in the sense that we cannot have experiences that would, say, require radically different sensory organs. i distinguish this from a different, substantial sense of impossibility in section 3.1 below. 4one could try to subsume this under a notion of “biological impossibility”, derived from tim bayne’s concept of biological necessity: many features of human beings are not necessitated by our design specifications but vary from one biologically normal human being to another. […] but other features of our nature – such as the possession of a heart – are mandated by our design specifications; they are included within our functional blueprint. […] [a] feature can be ‘biologically necessary’ without being an essential feature of all possible human beings. (bayne, 2014, pp. 526–527) biologically necessary features are features that all “biologically normal” human beings possess. as bayne emphasizes, such features need not be essential features of all possible human beings, because one could change a person’s “normal” features by intervening with their body. after the intervention, the person may continue to be a human being, without being “biologically normal”. instead of using the notion of biological normality to define biological necessity, one could use it to refer to features that no “biologically normal” human being possesses, and call these features “biologically impossible”. however, this raises the question: what counts as normal and what does not? answers to this question will, to a large extent, be a matter of debate, and there are likely to be many borderline cases. furthermore, restricting the notion biologically normal to “normal human brains” raises similar questions. for this reason, i am using the more complicated formulation “nomologically impossible, given independently justified assumptions about consciousness and its neural underpinnings in human beings”. i unpack this formulation in the main text. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 4 many conscious thoughts, or a feeling of effort while thinking about a puzzling philosophical problem. i will, furthermore, assume that phenomenal character is grounded in (phenomenally represented) intentional content. in particular, i assume that conscious experiences have accuracy conditions, and that they determine what it is like to have those experiences. the intentional content of a conscious experience is then a phenomenal content. some philosophers distinguish between phenomenal content and the phenomenal mode of presentation:5 the idea would be that the difference between, say, consciously seeing a fluffy puppy and consciously touching its fur is not exhausted by a difference in phenomenal content; as a result, such authors would distinguish between two types of phenomenal properties (see chudnoff, 2013, pp. 565–566), or would argue that phenomenal properties are not determined by pure representational properties (i.e., one would also have to specify the particular phenomenal manner in which something is represented, such as a visual or auditory phenomenal manner – see chalmers, 2004, pp. 159–161, for discussion). here, i shall assume that there are no differences in phenomenal character that cannot be accounted for in terms of differences between representational content, and that one does not have to distinguish between two types of phenomenal properties: there are only phenomenal properties corresponding to phenomenal contents, and differences in the phenomenal mode or manner of representation can be analyzed as differences in contents – but note that the claims argued for later in this paper do not depend on these background assumptions. must two (token) experiences of the same type have the same phenomenal character? or does similarity of phenomenal character suffice for membership of a given type of conscious experience (koksvik, 2014, p. 129)? relatedly, how determinate do the phenomenal properties picked out by a description of a type of experience have to be (kriegel, 2015, p. 10)? for the purposes of this paper, nothing hinges on how we answer these questions. after all, we want to find descriptions of conscious experiences that cannot be created by the human brain. the (token) conscious experiences falling under a given description may have the same or only similar phenomenal characters (for instance, because they instantiate different determinates of the same determinable phenomenal property) – and there may be reasons to regard them as tokens of the same or of different types of experience. but what counts is what they have in common: they are nomologically impossible (independently of whether they share membership of the same type of experience or not). 5just as one can distinguish between intentional content and intentional mode (see crane, 2009, p. 477; searle, 1983, p. 6), or between representative content and psychological mode (see searle, 1979, p. 82). wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 5 2.2 consciousness and the brain we cannot find impossible conscious experience (in the way in which black swans can be found), so we can at most hypothesize that a given description of a conscious experience does not refer to any nomologically possible conscious experiences. how can this hypothesis be tested? i shall suggest that we find evidence for a possible explanation as to why the type of conscious experience in question is impossible. such an explanation must make some assumptions about the neural underpinnings of consciousness, because it purports to explain why the brain cannot generate this type of conscious experience. furthermore, these assumptions should ideally either be statements for which strong empirical support already exists, or they should enable us to derive empirical hypotheses that can be tested in practice (not just in principle). in addition to this, there may also be useful assumptions that are to some extent theoretically justified (for instance, assumptions about general computational principles implemented by neural activity). these should at least be consistent with currently available empirical evidence. i shall say more about such assumptions below, in section 4.2. 2.3 nomological possibility and the brain why do we have to make further assumptions about the brain to justify the claim that a given type of experience is nomologically impossible? after all, don’t the laws of nature already put constraints on possible conscious experiences? this is certainly true, and there may be many interesting types of conscious experience that are nomologically impossible, but logically possible, regardless of how consciousness is generated in the human brain. however, as intimated in the previous paragraph, “finding” impossible experiences is methodologically difficult. adding constraints that are more specific than constraints derived from physics alone can therefore facilitate finding candidates for impossible experiences. to the extent that these additional constraints suggest empirically testable hypotheses, we do not have to rely on theoretical considerations alone to justify the claim that a given type of experience is impossible. a description of an impossible conscious experience, in the sense used in this paper, must therefore specify a phenomenal character (i.e., a phenomenal property) and suggest why the human brain is incapable of generating an experience with this phenomenal character. but why should some types of experience be impossible in the first place? here is a prima facie reason why we should expect that there are systematic limits on the types of conscious experience that the brain is capable of creating. the ability to generate consciousness is likely to serve an adaptive function, and this includes the ability to regulate consciousness or to shut it down. extreme pain can cause subjects to faint, and this may have an adaptive value. crucially, it also suggests that there is a threshold above which no pain experience is possible (because consciousness will be shut down if this threshold is crossed). assuming wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 6 that the intensity of logically possible pain experiences is unbounded (i.e., for every logically possible pain experience, there is another that is even more painful), this suggests that certain phenomenal properties (“more than extreme” painfulness) cannot be instantiated by the human brain. more generally, there may be other “extreme” phenomenal properties, or combinations of phenomenal properties, that cannot be instantiated by the human brain, because of the way in which the brain generates and regulates conscious experience. to clarify and test these ideas, we need a more systematic way of arriving at descriptions of impossible conscious experiences; this will also allow us to specify how such experiences are still logically possible. in the following section, i will motivate and describe three different strategies. 3 why are impossible conscious experiences relevant for consciousness research? any scientific theory needs a clear description of its target domain. phenomenal consciousness is particulary vexing in this respect, because defining the explanandum is notoriously difficult. although it seems we can, conceptually, distinguish phenomenal consciousness from mere tonic alertness (metzinger, 2020), attention, the capacity for integrated control of behavior, and other cognitive capacities (see chalmers, 1995; revonsuo, 1998), we cannot completely dissociate phenomenal consciousness from these phenomena, and their relationship remains debated. this points to a central feature of theory-building in consciousness research, viz. that more rigorous specifications of the target phenomenon (i.e., phenomenal consciousness) have to rely on empirical results. as research progresses, we gain deeper insights into the relationship between phenomenal consciousness and cognitive capacities. for instance, we learn more about which perceptual and cognitive tasks can be performed unconsciously (lamme, 2015) and which require consciousness.6 discovering phenomena that do not (necessarily) involve conscious processing enables one to circumscribe the concept of consciousness. in addition to this, it is also possible to constrain the concept of consciousness from within, by considering tasks that cannot even be performed consciously. for instance, although some savants are able to perform highly remarkable forms of mental arithmetic, there are limits to what is humanly possible. some people are able to consciously ex6one could object that such empirical discoveries do not tell us anything about the concept of consciousness, but merely contingent facts about how consciousness and cognitive capacities are related in actual human beings. in other words, it is conceivable that there are beings with the same cognitive capacities, but without consciousness, or beings with conscious experiences that are impossible for us. first, this objection seems to presuppose a strong separation between empirical investigations and conceptual analyses. secondly, even if our current, folk-psychological concept of consciousness is extremely general, it can be useful to develop a more specific, scientific concept of consciousness, which will mainly apply to actual human beings. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 7 perience the process of multiplying astronomically large numbers, but no human being is able to mentally multiply two arbitrary numbers with one million digits each. speaking more generally, there are limits on cognitive capacities (e.g., on the capacity of working memory7), and therefore also limits on consciousness, because one cannot8 experience the exercise of a cognitive capacity if one does not possess this capacity in the first place.9 call this the ability approach to impossible conscious experiences. a further approach is to consider phenomenal properties that vary on a continuum. a possible example is the experience of toothache, which varies from a dull feeling of pressure to a sharp pain of tremendous intensity. other possible examples include the experienced brightness of a visual stimulus, the experienced loudness of a sound, or the experienced sweetness of a piece of food.10 the core assumption underlying this approach is that there are levels of intensity that cannot be consciously experienced by human beings. call this the intensity approach.11 as a simple example (which is not related to consciousness), consider a flexible stick. the stick can be bent to some extent, but (as we shall assume) it will eventually break, if the bend becomes too extreme. bending the stick beyond the point at which, as a matter of fact, it breaks is logically possible, but nomologically impossible (given the physical nature of the stick). similarly, instantiating extreme phenomenal properties may be logically possible, but nomologically impossible (given the nature of our brains). this might also apply to some forms of self-consciousness: finding out at which point “the self breaks” can provide hints about the underlying neural mechanisms. in this paper, i will pursue a third approach. the main assumption is that there are phenomenal properties that are dependent on each other, such that it is 7as is well-known, human working memory is limited to around 7 “items” (miller, 1956). while the precise capacity of working memory may vary developmentally or be context-dependent (for a review, see cowan, 2010), it is out of the question that the capacity of working memory is limited. for instance, states of consciousness that involve imagining a number of 30 specific chunks in working memory are nomologically impossible. 8even if one imagined exercising a capacity one does not have, the experience could not be one of exercising that capacity, but would be a different experience. 9i am grateful to martin dresler for suggesting this. 10the claim that such properties vary on a continuum should be taken with a grain of salt. in particular, i do not want to claim that there are infinite different levels of intensity, nor that they can be measured in the way temperature can be measured. following a suggestion by the editors of this special issue, it may even be more appropriate to say that what differently intense feelings of pain have in common is that they pertain to the same (natural) kind of phenomenal property, even though they are distinct phenomenal properties. the idea would then be that certain members of this kind of phenomenal property cannot be instantiated by human brains. 11upon closer inspection, this approach may partly overlap with the ability approach. while the latter focuses on cognitive capacities, it may be possible to characterize the current approach in terms of perceptual abilities. for instance, the fact that there are limits on experienced loudness could also be formulated as the fact that the loudness of sounds cannot be discriminated beyond a certain threshold. this would presuppose that phenomenal differences go along with discriminatory abilities, which is not uncontested. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 8 nomologically impossible to experience one without the other. the (hypothesized) reason for this is that there are causal dependencies between the neural underpinnings of these phenomenal properties. call this the dependency approach. below, i will give the sense of time (i.e., experiencing the passage of time) and the sense of agency as a potential example: the hypothesis would be that the sense of time is causally necessary for the sense of agency, because of dependencies between the neural realizers of those phenomenal properties. more generally, self-consciousness provides a useful testbed for the dependency approach. self-consciousness is not a unitary, simple phenomenon. instead, it is characterized by a variety of features, such as de se thought, body ownership, spatial self-location (millière, 2020, this issue, p. 5), a pre-reflective self/non-self distinction, a sense of agency, patterns of affective states (gallagher & daly, 2018, p. 4), a sense of time, or a sense of center (ataria, dor-ziderman, & berkovich-ohana, 2015, p. 142). none (or few) of these features are necessary for self-consciousness (in fact, it is even controversial whether all of them exist, and whether they are really distinct). put differently, self-consciousness should be treated as a mongrel concept (just as the more general concept “consciousness” – see block, 1995). similar versions of this view have been expressed as a pattern theory of self (gallagher, 2013; gallagher & daly, 2018; newen, 2018), or as the claim that “selfconsciousness may be best construed as a multidimensional construct” (millière, carhart-harris, roseman, trautwein, & berkovich-ohana, 2018, p. 1). what i take to be central to this position is that particular types of self-consciousness are best characterized in terms of a variety of features. crucially, some of these features may be sufficient without being necessary for self-consciousness. this means that different types of self-consciousness can be instantiated (probably independently of each other). in addition, this enables the hypothesis that some combinations of features are nomologically impossible (although they are logically possible). again, evidence regarding how “the self breaks” in radical disruptions of self-consciousness can suggest hypotheses about which combinations of phenomenal properties related to self-consciousness are possible and which are not. i will elaborate on this in section 5. in sum, investigating impossible conscious experience is valuable, because it increases our understanding of consciousness (by narrowing down the concept of consciousness). but in addition to that, explanations of why certain conscious experiences are impossible may yield insights into the neural mechanisms underpinning consciousness. 3.1 two types of impossible conscious experiences before moving on, it will be important to make a distinction between two types of impossible conscious experiences. one of them is rather trivial.12 the other 12i am grateful to an anonymous reviewer for pressing me on this issue. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 9 turns the claim that there are impossible conscious experiences into an interesting hypothesis. i will introduce this distinction by using an analogy. imagine you are painting with a brush and a small set of watercolors. provided you have the three primary colors, cyan, yellow, and magenta, you can mix all other colors. if you only have cyan and magenta, you will only be able to mix a more restricted set of colors. for instance, painting in orange will be impossible for you, given your watercolor equipment. still, painting in orange is logically possible, and even nomologically possible, if we relax the constraints on watercolors that are at your disposal. human consciousness is similarly constrained. we have a certain set of sensory organs, which shape the way we experience the world. there are types of experience we will never have, for the simple reason that we lack the required sensory organs and do not have the neural machinery that would control and make sense of signals received through them. with a nod to thomas nagel (1974), one could point out that we will never have a bat’s conscious experience. but this type of limitation is trivial, because it does not reflect the way our brains create consciousness. in order to overcome this type of limitation, we would have to augment our senses (or the way we use them, cf. teng & whitney, 2011); this would enable us to have conscious experiences that were impossible beforehand (just as adding a primary color extends the set of possible colors you can paint with). we might even be able to have experiences that are similar to experiences of bats (even if this might not show us what being a bat is like for a bat). now imagine you are painting on a dark, absorbent piece of wood. you have all three primary colors, but the consistencies of the colors differ: your yellow color is nice and pasty, but your magenta and cyan are too thin to leave a visible mark on the wooden board. the yellow color, by contrast, can readily be applied to the ground. fortunately, you find out that you can still mix cyan and magenta with yellow, and the mixed color can be applied as nicely as the pure yellow color. this means you can paint in pure yellow, in green, orange, and any colors that are obtained by mixing all three primary colors. however, you will not be able to paint in pure cyan or pure magenta – because you will always have to mix those colors with a bit of yellow, in order to apply the color on your painting surface. in other words, some colors will be impossible for you, not because you lack the required primary colors, but because of how your painting equipment constrains the way in which you can paint. if consciousness is constrained in this way (and this is the hypothesis explored in this paper), then there will be dependencies between phenomenal properties (just as blue and red “depend on yellow” in the analogy). these dependencies are contingent on the neural mechanisms underlying consciousness in human brains. although there is a sense in which the first type of impossible conscious experiences (e.g., perceptual experiences based on sonar signals) is also contingent on the neural mechanisms underlying consciousness, there is a crucial difference, because the first type is not created by dependencies. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 10 let me further illustrate the second type of impossible experiences with a toy example: assume every human conscious experience is characterized by at most two phenomenal properties, and that these properties are binary (either the phenomenal property is instantiated or not; there are no differences in the degree to which the property is expressed). every conscious experience can then be described as an ordered pair, in which each element is either zero or one (zero means the respective phenomenal property is present, and one means it is absent). the set of nomologically possible experiences, given our neural machinery (which can, by assumption, only instantiate two different phenomenal properties), is then a set of four conscious experiences (or three, if at least one phenomenal property must be instantiated). the set of logically possible experiences is much greater, because there are further phenomenal properties. experiences that require instantiating further phenomenal properties are impossible experiences of the first type. now assume that there is a dependency between the two phenomenal properties (which is contingent on the way our brains create consciousness): the first phenomenal property cannot be instantiated without the second. if this is the case, then the space of nomologically possible experiences, given our neural machinery, has a hole: an experience in which the first phenomenal property is present, while the second is absent, is impossible. this is an impossible experience of the second type. this type of impossible experience is non-trivial. for, knowing that a system can have experiences with certain phenomenal properties, it is not obvious that some combinations of these phenomenal properties are impossible. 3.2 constraining models of consciousness what can impossible experiences tell us about the neural mechanisms of consciousness? note first that it is straightforward to see that the opposite is true: if we had a complete theory of consciousness (including detailed knowledge of the neural mechanisms of consciousness), we would be able to say which types of conscious experience are possible and which are not. but can impossible experiences, conversely, yield insights into the mechanisms of consciousness? i will argue that they can. we can see this most clearly by considering a strong type of dependency between phenomenal properties. if there is a phenomenal property that is nomologically necessary for consciousness, then conscious experiences without this phenomenal property will be impossible in the sense that is relevant here. furthermore, this will suggest that all other phenomenal properties depend on this property (at least causally). certain types of self-consciousness may provide controversial examples: if there is a basic sense of self that is constituted by a particular phenomenal property, viz. a self quale (or “ur quale”, as it is called in perlis, 1997, p. 509) that is necessary for consciousness, then conscious experience without this self quale would be impossible. if correct, this would suggest that there is a causal dependency between wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 11 the neural underpinnings of this basic type of self-consciousness and the underpinnings of all other phenomenal properties. hence, such a dependency between a single phenomenal property and all types of conscious experience is likely to reveal insights into the neural mechanisms of consciousness. a crucial step would be to find the neural correlate of the self quale, and then to relate it to candidates for total neural correlates of consciousness (that are involved whenever a person is conscious, but not when she is unconscious). however, as already noted, the example is controversial. what is more, proponents of the claim that consciousness necessarily involves a basic sense of self usually explicitly reject the claim that this corresponds to a self quale (contra perlis, 1997). for instance, zahavi & kriegel (2016) point out that “for-me-ness is not a detachable self quale that one could introspect in isolation from any other content of consciousness” (zahavi & kriegel, 2016, p. 39). this suggests that your visual experience of a red tomato and my visual experience of a red tomato never have any (maximally determinate) phenomenal properties in common. they may share the determinable property of seeing a red tomato, but always instantiate more determinate phenomenal properties that do not overlap. another way of interpreting this view is that there is a sense in which a basic sense of self is a (necessary) phenomenal property, but one that cannot be instantiated without instantiating at least one further phenomenal property. this would still be in line with zahavi’s and kriegel’s claim that one cannot introspect the self quale “in isolation from any other content of consciousness”. furthermore, it illustrates another interesting potential result of investigating candidates for impossible conscious experiences. it could turn out that what we described as two distinct phenomenal properties 𝐹 and 𝐺 are in fact constitutively dependent properties: 𝐺 may actually be a composite phenomenal property constituted by 𝐹 and 𝐹 ′, i.e., 𝐺 ≡ 𝐹 &𝐹 ′. the fact that 𝐺 depends on 𝐹 would then be trivial (just as 𝐹 trivially depends on itself). in a recent article, zahavi (2018) points out that the basic sense of self is not an “additional experiential object” (zahavi, 2018, p. 5). however, in line with the interpretation that all phenomenal properties constitutively depend on a self quale, one could turn this statement on its head: all other contents of consciousness, such as a visually experienced red tomato, are not additional experiential objects, but always involve self-content. the general idea of constraining models of consciousness by research on impossible experiences is as follows. instead of starting with an assumption of how consciousness is generated, we start with an assumption about the impossibility of certain types of experience. the evidence for impossibility can be preliminary (and non-conclusive). but from the assumption that certain types of experience are impossible, one can derive constraints on their neural underpinnings. in the ideal case, this will allow one to derive testable predictions. if these predictions bear out, we gain further support for the hypothesis that certain types of conscious experience are impossible, and we learn (much) about the neural underpinnings of wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 12 those types of conscious experience. if the predictions do not bear out, we should lower our confidence in the impossibility hypothesis (and we may not learn as much about the neural underpinnings of conscious experience). 3.3 matching descriptions at different levels of analysis there is another, more theoretical benefit that research on impossible conscious experiences may yield: it may increase our understanding of what phenomenal properties are in the first place, by providing support for structural theories of consciousness. if we want to understand consciousness, we have to understand the characteristic features of consciousness. features of consciousness are usually described on a personal level of analysis (as opposed to subpersonal levels – for this distinction, see dennett, 1969; drayson, 2012, 2014). the personal level of analysis has a variety of properties that make it difficult to establish links to subpersonal levels, on which functional, computational, or neurobiological features of conscious systems are described. for instance, phenomenological reports are subjective (as opposed to objective) and often make reference to phenomenal and intentional properties (as opposed to functional and physical properties). still, one can learn a lot about consciousness without worrying about these most difficult aspects of consciousness: research on consciousness usually operates under the assumption of what david chalmers (1995, 1996) calls “the principle of structural coherence”. he defines it as follows: this is a principle of coherence between the structure of consciousness and the structure of awareness. […] briefly put, we can think of awareness as direct availability for global control. […] in general, any information that is consciously experienced will also be cognitively represented. […] this principle reflects the central fact that even though cognitive processes do not conceptually entail facts about conscious experience, consciousness and cognition do not float free of one another but cohere in an intimate way. (chalmers, 1995, pp. 212–213) the principle of structural coherence, if true, alleviates the problem of matching phenomenological descriptions of consciousness with descriptions at subpersonal levels of analysis. for instance, research on the neural correlates of consciousness can be conducted by finding systematic relationships between availability for verbal or non-verbal report and neural activity in different areas, and differences in reports (e.g., reporting a house vs. a face) can be matched with differences in neural activity (e.g., in the parahippocampal place area vs. the fusiform face area, see tong, nakayama, vaughan, & kanwisher, 1998). but such research does not reveal why neural activity of a particular type is correlated with the presence of certain types of experience, and it does not by itself clarify “what it is like” (farrell, 1950, p. 181; nagel, 1974) to be in certain conscious states. part of the answer may involve explaining why the experience wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 13 has the phenomenal character it has, and not a different phenomenal character. in particular, this leads to the question whether “inverting”13 phenomenal properties (an “inverted spectrum”) is possible: not knowing why a given neural activity goes along with the instantiation of a specific phenomenal property such as experienced redness, the connection between neural activity and phenomenal properties can seem more or less arbitrary, and it may seem as if one could permute a subject’s phenomenal properties without evoking any differences in behavioral or cognitive capacities. one way to block such inversion scenarios is to point to asymmetries in the space of conscious experiences (see clark, 1996). for instance, if subjects are able to discriminate a larger variety of hues within a certain part of the color spectrum than in other parts, this could be taken as evidence that the space of consciousness (or “quality space” – see rosenthal, 2015) is asymmetrical, and that inversion scenarios are impossible. showing that certain types of conscious experience are impossible is another way of discovering asymmetries. if certain combinations of phenomenal properties are impossible, the space of nomologically possible conscious experiences will have holes, which may involve asymmetries. for instance, an asymmetry exists if a phenomenal property 𝐹 can be instantiated without instantiating phenomenal property 𝐺, whereas 𝐺 cannot be instantiated without instantiating 𝐹 . investigating impossible experiences may therefore contribute to developing structural theories of phenomenal properties (see also wiese, 2018). 4 how can impossible conscious experiences be discovered? 4.1 the general idea here, i will pursue what i called the dependency approach in section 3 above. this approach gives us at least two ways in which impossible conscious experiences can be discovered. a first starts with what we know about the neural mechanisms underpinning consciousness. to the extent that this enables us to see how contents of consciousness are determined by these mechanisms, it will also tell us about the limits to the contents that can be consciously experienced. a second approach 13in “inversion scenarios”, there are subjects with deviant phenomenal properties: for instance, whenever we would have an experience as of red, a subject with an inverted spectrum would have an experience as of green. if this inversion is systematic, it may not go along with any functional or behavioral differences, and it may be impossible to detect the presence of the inverted spectrum. block & fodor (1972) use such scenarios to argue that behavioral and functional criteria are insufficient to provide type-identity criteria for mental states – for if an inverted spectrum is possible, two subjects can have different types of conscious experience (because their phenomenal properties differ), without any functional or behavioral differences. for an overview of the debate, see byrne (2018). wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 14 starts from the other side: if a given type of conscious experience is impossible, what does this tell us about the neural mechanisms underpinnings consciousness? the dialectic is a bit more complex in this case: supposing that an experience 𝑒 is impossible, one tries to derive hypotheses about its neural underpinnings 𝑛. if these hypotheses about 𝑛 can be tested and are corroborated, this also supports the hypothesis that 𝑒 is impossible. the goal of the second approach is thus not just finding out about the limits of (contents of) consciousness, but finding phenomenological and, say, neurobiological constraints on consciousness at the same time. but why should we suppose that a particular conscious experience 𝑒 is impossible in the first place? here, i am suggesting that we start with the observation that there is a lack of counter-examples: no report of an experience of a given type is known to exist. in other words, there is no evidence to the effect that the type of experience in question is possible. the crucial step in the argument is, of course, to move from a lack of counter-examples to positive evidence for the impossibility of a type of conscious experience. one can start with a model of how it could possibly be that a given type of experience is impossible. this can then enable one to determine whether it is actually true (i.e., moving from a how-possibly to a how-actually model – see machamer, darden, & craver, 2000, p. 21). in terms of explanation, this would involve starting with a possible epistemic explanation and using it to develop an ontic explanation (craver, 2014, p. 37), e.g., by making hypotheses about the causal mechanisms involved, and then testing these hypotheses. how can this reveal that certain types of experience are impossible? on the one hand, it could turn out that the information processing involved in generating a given type of experience is too complex to be realized by a human brain. this could be due to cognitive limitations or to physical constraints on the speed of information processing.14 on the other hand, and more fundamentally, there can be causal dependencies between the neural mechanisms that instantiate different properties of experience. in this case, i shall also say that the properties are causally dependent (or that there is a causal dependency between the properties). as already noted, i will focus on this second approach in what follows. the challenge here is to find causal dependencies between properties. one could start by searching for correlations between properties. but a correlation between two properties 𝐹 and 𝐺 is compatible with different causal relations: 𝐹 may be causally necessary for 𝐺 (or vice versa), or there may be a third property 𝐻, the instantiation of which functions as a common cause, or there might be just a correlation, without any direct causal connection between 𝐹 and 𝐺. hence, it will be useful (if not even necessary) to make further assumptions about features of consciousness and their neural underpinnings. examples of such assumptions can be computational, functional, or neurobi14i am grateful to philip gerrans for suggesting thinking about constraints on the speed of processing. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 15 ological. since the exact neural underpinnings of consciousness are still unclear, it would be of little use to make specific assumptions about neural mechanisms underlying consciousness (unless they allow us to derive empirical hypotheses that can actually be tested using current neuroscientific methods). in general, the challenge here is to strike a balance between highly informative, but too specific, assumptions and non-informative, unspecific assumptions. the problem with too specific assumptions is that they are likely to be false (which, unfortunately, does not mean they can easily be falsified). however, if assumptions are too unspecific (but also more likely), they do not put any strong constraints on mechanisms underlying consciousness. background assumptions must be specific enough to enable an explanation as to why a given type of conscious experience is impossible, without making controversial claims that cannot be tested in practice. here, i will suggest assuming, first, that computations implemented by the brain are relevant to understanding consciousness (note that this does not mean that implementing certain computations is sufficient for consciousness – see schweizer, 2019), and, secondly, that these computations are (among others) characterized by the following two features: the ideomotor principle and the principle of spatio-temporal hierarchies.15 4.2 two general computational principles as background assumptions according to the ideomotor principle, representations underpinning perception and action, respectively, overlap (at least partially). some of the neural areas that are activated when i perceive someone move are therefore also activated when i conduct this movement myself. this is the core idea of the ideomotor principle, which was first formulated by johann friedrich herbart (herbart, 1825, 1834), and was later picked up by william james (for a historical overview of ideomotor ideas, see stock & stock, 2004).16 active inference provides a formal elaboration of this idea (and offers important extensions – see wiese, 2017a; friston et al., 2017a, 2017b). note that merely adopting the ideomotor principle does not require rejecting other approaches to motor learning and control. for instance, the principle is compatible with applications of optimal control theory to motor learning and control (see herbort & butz, 2012) – although researchers using optimal control theory do not typically incorporate the ideomotor principle in their models (cf. wolpert, ghahramani, & flanagan, 2001; for critical discussion, see friston, 2011). 15incidentally, these are also key ideas of active inference, a framework developed by karl friston and colleagues (see friston, samothrakis, & montague, 2012). however, the two computational principles can be accepted without presupposing that active inference provides an adequate way (let alone the only way) of modeling brain function. but i submit that the framework can still be useful to further develop ideas sketched in this paper. 16a further aspect of the principle is a particular view on how actions are learned – this will not be relevant to the discussion in this article. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 16 another feature that will be relevant here is the spatio-temporal hierarchy of information processing in the brain. the assumption behind this is that environmental features reflected in sensory signals vary with respect to their temporal and spatial grains. some features are invariant over large spatial regions in objective space (e.g., the color of the sky), and some are invariant over long temporal intervals in objective time (for instance, the size of my hands). an efficient way of representing mixtures of features of various temporal and spatial grains is by using hierarchical representations. lower processing levels represent features that change quickly and are more concrete. higher processing levels represent features that are invariant over longer periods of time or larger regions of space (they may also be more abstract). this also resonates with the idea that percepts consists of hierarchical representations, in which the same objects are represented more than once (i.e., in different ways; cf. lycan, 1996; bayne & mcclelland, 2018). so these are two computational features which, i will suggest, will be useful to find epistemic explanations as to why certain conscious experiences are impossible: the ideomotor principle and the principle of spatio-temporal hierarchies. without getting into any formal details, we can already derive some constraints on how the brain represents the world. in particular, this also puts constraints on possible conscious experience, as we shall see now. ideomotor principle. action is partially underpinned by representations that are also activated when perceiving or imagining a similar action. this means, for instance, that there are limits on what can easily be imagined and observed at the same time. there must also be interferences between action observation and execution (for a review of empirical evidence, see zwickel & prinz, 2012; for a recent supporting study see lohmann, belardinelli, & butz, 2019). spatio-temporal hierarchy. if the contents of perception are hierarchical (which is suggested by hierarchical computational models), then temporal features are also represented at different timescales (kiebel, daunizeau, friston, & sporns, 2008). if such hierarchical models not only underpin unconscious perception, but also determine the contents of conscious perception, then the experienced present is hierarchical as well (wiese, 2017b). this has subtle implications for how to understand the conscious perception of temporally extended events, such as a melody: at short time-scales, we do not first experience one note, and then another note, together with an experience of “diachronic togetherness”. rather, we experience a single, temporally extended event (the melody or succession of notes) as a whole, and at the same time experience changing properties that belong to the same temporal whole. that is, we do not experience these different properties (e.g., notes) as belonging to different temporal parts of the whole event. simon prosser puts it thus: when we experience change we do not […] have experience as of an 𝐹 temporal part of an object succeeded by a non-𝐹 temporal part of that object, with it somehow understood that both parts belong to the same composite whole. (prosser, 2016, p. 173) wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 17 this may sound puzzling, because it seems to require that we somehow experience an object as having temporal parts and as not having temporal parts. but what prosser emphasizes here is that we do not experience persisting objects as perduring, but as enduring. if they were experienced as perduring, they would be experienced as having temporal parts. instead, we experience them as enduring, as being wholly present at any instant; and changing objects are, in addition, experienced as having properties that are only transiently instantiated. hierarchical models of temporal perception suggest a way to make this idea more precise: when two successive notes of a single melody are perceived, both are experienced as part of a single extended event, which is partly underpinned by a representation at a higher level (which represents properties at a more coarsegrained timescale).17 at that level, properties are represented as being instantiated now, but also as just having been instantiated – they are represented as continuing to be instantiated. this means that those properties cannot be represented as being instantiated now without thereby also being represented as being extended into the recent past and the near future (in other words, the objects having those properties are represented as enduring). properties that are transiently instantiated (such as the pitch of a single note) are represented at lower levels, and can specify in what way enduring objects change. see figure 1 for an illustration. however, the point that is crucial for the purposes of this article is that higher levels encode objects that are temporally extended. i submit that these two computational principles, i.e., the ideomotor principle and the principle of spatio-temporal hierarchies, are general enough so as to function as relatively uncontroversial background assumptions about neural mechanisms underlying consciousness. in other words, i only assume that neural activity underpinning consciousness is consistent with the way of representing the world entailed by these two features. at the same time, we have seen that these two features already constrain the way in which the brain represents the world. for this reason, they at least promise to suggest explanations of why certain conscious experiences may be impossible. in every particular case, we will need further, ad hoc assumptions, but general background assumptions, such as the ones specified here, will serve as guiding principles. such assumptions and guiding principles specify at most epistemic explanations, in that they make it likely that certain conscious experiences are impossible. in other words, neither do they logically entail the impossibility of certain types of conscious experience, nor do they specify neural (causal) mechanisms that render certain experiences impossible. but they do specify mechanism sketches (cf. machamer et al., 2000, p. 18; piccinini & craver, 2011), and thereby also suggest ontic explanations as to why a given candidate for an impossible experience cannot be created by human brains. an ontic explanation of a phenomenon describes causal interactions between parts of a system, in a way that shows how the 17in wiese (2017b) i call this a dynamic event representation (see wiese, 2017b, p. 14). wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 18 figure 1: the yellow bar illustrates a dynamic event representation. it is a distributed representation of multiple properties (individual properties are represented by 𝑒1, … , 𝑒𝑛−1, 𝑒𝑛) that jointly specify a temporally extended event: near the top, properties that change slowly and are instantiated over a longer period of time (e.g., hundreds of milliseconds) are represented; near the bottom, properties that change quickly and are instantiated in the immediate present only (e.g., for less than one hundred milliseconds) are represented. the represented period of time during which the different properties are instantiated is illustrated with black arrows (marking the beginning and the end of each represented period during which a property is instantiated). together, 𝑒1, … , 𝑒𝑛−1, 𝑒𝑛 represent a single, enduring event that we experience “as rising from the recent past, and as continuing into the near future” (wiese, 2017b, p. 13), because some of its properties remain invariant over a period of time that is slightly longer than the short interval we experience as the present. together, they represent a specious present (james, 1890). (figure adapted from wiese, 2017b, p. 12.) activities of the parts and their organization give rise to the target phenomenon (see craver & tabery, 2017, sec. 2). the idea is thus that we can move from descriptions of computational features to descriptions of mechanisms, and this will (hopefully) enable us to derive testable predictions. 5 from radical disruptions of self-consciousness to impossible conscious experiences 5.1 general idea recall that i suggest to start with the observation that no positive evidence for the existence of a given type of conscious experience exists. with respect to selfwiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 19 consciousness, this means that there is a phenomenal configuration (a combination of features associated with different types of self-consciousness), that is not reported in the literature (or in databases such as https://erowid.org). given the many variations of phenomenal configurations reported in radical disruptions of self-consciousness, there is thus a prima facie reason to believe that the lack of evidence for the phenomenal configuration in question is not a mere coincidence. by itself, this reason is too weak to justify the claim that the phenomenal configuration at hand is indeed nomologically impossible. this is where the various approaches outlined in sections 3 and 4 come into play. as already pointed out, i will pursue the dependency approach, which focuses on causal dependencies between the neural underpinnings of phenomenal properties. when two features are mutually causally dependent then one cannot be instantiated without the other. when only one feature depends on the other (𝐴 is causally necessary for 𝐵, but not vice versa), the first cannot be instantiated without the second, but the second can be instantiated without the first. another type of causal dependency involves a third feature 𝐶, such that both 𝐴 and 𝐵 are causally dependent on 𝐶 (but 𝐶 need not be dependent on 𝐴 and 𝐵). in what follows, i will consider “the sense of boundaries” (ataria, 2014), which is constituted by different phenomenal properties. two of them are the sense of agency and the sense of time, and i will suggest that the sense of time is necessary for the sense of agency. 5.2 the sense of boundaries we usually experience a boundary between ourselves and the world. i experience the world from a first-person perspective: i am at the center of my experience, i am on the inside, within the boundaries of my body, and the rest of the world is outside. yochai ataria (2014) calls this the sense of boundaries. the sense of boundaries corresponds to a bundle of phenomenal properties. it is particularly interesting because it is likely that the phenomenal properties constituting the sense of boundaries are not completely independent: at least some of them cannot be instantiated without also instantiating some other phenomenal properties. put differently, there are configurations of phenomenal properties (associated with the sense of boundaries) that correspond to (nomologically) impossible experiences – or at least this is what existing research on the sense of boundaries suggests. but what exactly is the “sense of boundaries” (sb)? first of all, the sb is not identical to a physical boundary. instead, it is an experienced boundary between one’s lived18 body and the rest of the world. crucially, since the lived body depends on sensations in different sensory modalities, it can be modulated by changing sensory input (e.g., by closing one’s eyes or by entering a hot bath of water19), or by 18the lived body is a term used by phenomenologists to refer to the body, experienced as your own body, not experienced as an object. 19in some cases this can even affect the experienced boundary between the lived body and the environment. millière (2019) cites a written interview, in which a deafblind individual describes wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 20 focusing attention on different aspects of one’s current experience (cf. ataria, 2014). moreover, ataria (2014) distinguishes between the experience of boundaries and experiencing these boundaries as having determinate spatial locations (cf. ataria, 2014, p. 1142; martin, 1995, p. 271). since boundaries between oneself and the world can be experienced in different ways, it is best to define the sb as a bundle of phenomenal properties. ataria et al. (2015) characterize the sb in terms of several phenomenal properties (which they call “categories”), including the sense of agency and the sense of time (see ataria et al., 2015, pp. 143–144). as already noted, i shall focus on these two phenomenal properties here. by the “sense of time”, one could mean different phenomena, for instance, the experience of temporal properties such as duration or succession, consciously remembering past events or imagining future events, a conscious estimate of how long it has taken to read up to this sentence, or having a sense of which of two temporally extended events was longer.20 ataria et al. (2015) do not give a definition of the sense of time. however, quoted reports in which their subject describes how their sense of time changes indicate that the sense of time is regarded as a gradual property and that it can also fail to be instantiated.21 the absence of the sense of time is described as “timelessness” in an earlier publication, to which the authors refer in a footnote (see berkovich-ohana, dorziderman, glicksohn, & goldstein, 2013). in a more recent publication, the sense of time is characterized as follows: specifically, we are referring to a sense of a past continuing into the future; to the reduction in the sense of duration; and finally, to the sense of continuity itself which also disintegrates. (dor-ziderman, ataria, fulder, goldstein, & berkovich-ohana, 2016, p. 3) i submit that the relevant notion of “sense of time” is that of the passage of time on short timescales. that is, it is not primarily about having a sense of the duration of past events, or of events on long timescales. instead, it is about experiencing temporally extended events that are occurring now, have a duration, and reach into the very recent past and the immediate future (including internal events, such as the ongoing endeavor to regulate attention). furthermore, the sense of time does not necessarily entail a sense of self, i.e., temporal experience without any form of self-experience may be possible (see windt, 2015, pp. 18–19). the experience of a loss of spatial boundaries when they take a hot bath (see millière, 2019, p. 271). 20i am grateful to an anonymous reviewer for asking me to be more specific about the sense of time. 21“sense of time is obviously much clearer … yet i can’t describe it really” (ataria et al., 2015, p. 139) “i would need to stop meditating, totally open my eyes and be in totally ordinary consciousness, only then would the sense of time come back fully” (ataria et al., 2015, p. 140) “time is not relevant anymore” (ataria et al., 2015, p. 142) wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 21 furthermore, ataria et al. (2015) suggest that there is a causal dependency between the sense of boundaries and the sense of time: it seems that the sense of time is a kind of ‘mirror reflection’ of the sb. thus any alteration in the level of flexibility is reflected by a proportional adjustment in the sense of time. if we examine this phenomenon more broadly, we may say that the sense of self and the sense of time go hand in hand (ataria et al., 2015, p. 143) unpacking this statement requires a bit reflection, because the sense of time, as characterized by ataria et al. (2015), is not an additional feature, but actually one of the features that constitute the sb. so saying that the sb changes whenever the sense of time changes would be trivial (just as saying that the state of my computer screen changes when one of the displayed pixels changes). however, the statement does suggest more substantive implications. first, it suggests that any change in the sense of time (i.e., its becoming weaker or stronger) is reflected in corresponding changes in all other features constituting the sb. secondly, it suggests that any change in one of the other features also goes along with a corresponding change in the sense of time. perhaps the strongest interpretation would be that all of the features constituting the sb are positively correlated, such that the degree to which one of the features is expressed is predictive of how strongly (or weakly) the other features are expressed. for the purpose of illustration, let us consider the much weaker hypothesis that the sense of time is necessary for the sense of agency (which is one of the other properties constituting the sb). in a narrow sense, the sense of agency is the experience of causing a bodily or mental action (gallagher, 2000, p. 15). in addition to that, there is a more general notion that includes the experienced ability to act (cf. metzinger, 2013, p. 2). let us call this the sense of ability (cf. slaby, 2012). in what follows, “sense of agency” will refer to the general concept, which subsumes the sense of ability. i will also write “minimal sense of agency” for the sense of ability. is it possible to have a sense of agency without a sense of time? if i experience myself as causing or generating an event that is occurring now, i must, by logical necessity, experience that causing as an event that is unfolding now. so a strong sense of agency (involving the experience of causing an action) seems to require a sense of time.22 this is compatible with the hypothesis that “timeless” experiences 22what if it is possible to experience an action that is not unfolding at any particular point in time, but is an “eternal” event? in fact, this might actually be the case in some drug-induced states in which the sense of time is thoroughly disrupted. many such states may involve “thought loops” in which the same thought consistently recurs with no sense of when it begins and when it ends. a person may have the experience of causing a particular (mental) action, but without having the experience of causing it at any particular point in time, because the action does not have a determinate temporal location. here, it would be relevant to determine whether an “eternal” event is best conceived as an event that has always existed and will always exist (in that case, it would have temporal properties) – or as an event that simply lacks any temporal properties (in wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 22 exist (in which there is no conscious experience of the passage of time at all). the idea is merely that a sense of time must be there whenever there is a sense of agency. what about the sense of ability? in its minimal form, the sense of agency only requires an awareness of possibilities for action. so the hypothesis derived from this example is that being aware of possibilities for action requires a sense of time. and this is equivalent to the hypothesis that a conscious experience involving a (minimal) sense of agency, but no sense of time, is impossible. if i merely experience the ability to initiate an action, i do not have to experience the action as unfolding. still, it seems that i have to experience this possibility as referring to a future event, viz. a potential action that can be executed in the future. however, when i experience myself as having the ability to act, potential actions may be represented indeterminately,23 and they may not explicitly be experienced as future actions, but simply as something that can be done (or as something i can do) – just as i can think about potentially existing objects, without imagining that they exist at any particular point in time. but is it nomologically possible to experience myself as an agent, without experiencing the passage of time, given the background assumptions outlined in section 4 above? according to the ideomotor principle, there is an overlap between the representations involved in early action initiation (planning) and the representations involved in perceiving actual actions.24 assuming that the phenomenal representations contributing to the sense of ability include such “ideomotor” representations, we get a heuristic (epistemic) explanation of why one cannot experience oneself as being the potential author of actions, without having a sense of time. the explanation runs as follows. we are assuming that the sense of ability is underpinned by representations that can be used to initiate actions and to represent actions that are actually unfolding. as such, they must be relatively high in the processing hierarchy. this hierarchy not only reflects differences in levels of abstraction, or differences in spatial grain, but also differences in temporal grain: at higher levels, features that change slowly are represented, which underpins experiences of endurance. those features are relatively abstract and can be common to many different actions. this may explain why the sense of ability is not necessarily specific: i can experience the ability to act, without experiencing the ability to do anything in particular. but if i experience myself as the potential author of an action by experiencing myself as instantiating such an enduring feature, i thereby experience the passage of time, at least in a minimal sense, because the feature is experienced as being instantiated now, in the near future, and in the recent past. that case, a sense of agency without a sense of time would be possible. i am grateful to the guest editors of this special issue for suggesting this possibility. 23such representations may also underpin existential feelings – see ratcliffe (2012). 24as bernhard hommel, one of the proponents of the ideomotor principle puts it, the theory “is meant to provide a framework for understanding linkages between (late) perception and (early) action, or action planning.” (hommel, müsseler, aschersleben, & prinz, 2001, p. 849). wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 23 put more simply, the idea is that the sense of ability inherits experienced temporal properties from representations of potential action. and since potential actions cannot be represented without reference to temporal properties (this follows from the two computational principles), the sense of ability must also involve a sense of time. this explanation is relatively ad hoc. in line with the dependency approach, empirical evidence must be sought for a causal dependence between the neural realizers of the representations referred to above. this would require making more specific hypotheses about the underlying mechanisms. since the dependency between the sense of agency and the sense of time mainly serves as an example here, further analyses remain a task for future work. but notice that the dependency approach may yield surprising results. so far, i have talked as if one could always map phenomenal properties to distinct parts of a mechanism, and establish whether there is a causal dependency. however, it may also turn out that what has been described as a single phenomenal property (such as the sense of ability) actually involves different parts of a mechanism (which only sometimes interact and bring about a sense of ability). for instance, perhaps the sense of ability is underpinned by a structure that underpins the subject component and a structure that underpins the “object component” (potential action). the sense of time could be underpinned by a third structure, or it could be intertwined with the first two. using background assumptions as a framework, some of these possibilities may be more likely than others. and in the end, the dependency between phenomenal properties may even turn out to be more intimate than a causal dependency between different neural events. that is, it could turn out that there is actually a constitutive dependency between different phenomenal properties. conclusion the general hypothesis of this paper is that some phenomenal configurations (combinations of phenomenal properties) are logically possible but nomologically impossible, given further assumptions about how the brain generates conscious experiences. i have argued that research on impossible conscious experiences may provide insights about neural mechanisms underpinning consciousness and is not just interesting for the sake of marking out the terrain of conscious experiences that can be generated by the human brain. i have described three approaches to finding impossible conscious experiences: the ability approach (which starts from limits on cognitive abilities), the intensity approach (which starts from considerations about extreme conscious experiences, involving phenomenal properties that vary on a spectrum, but may not exceed certain thresholds), and the dependency approach (which focuses on causal dependencies between the neural underpinnings of different phenomenal properties). radical disruptions of self-consciousness provide a useful test bed for at least the dependency approach, since many types of self-consciousness can be wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 24 described in terms of combinations of phenomenal properties (such as the sense of agency or a sense of time). and if even in altered states (or radical disruptions) of self-consciousness certain configurations of phenomenal properties are not experienced, this warrants further explorations of the hypothesis that those configurations are indeed impossible. this does not presuppose that conscious experiences lacking any kind of consciousness of oneself are impossible. rather, it assumes that there are different types of self-consciousness, and that at least some of them are dependent on other phenomenal properties. i have illustrated this with the hypothesis that the sense of agency requires a sense of time (without there being a logical implication between the two). in addition to investigating the neural underpinnings of actual conscious experiences, thinking about impossible conscious experiences may sharpen the conceptual tools used to describe conscious experiences and lead to deeper explorations of their underlying mechanisms. acknowledgments i am grateful to two anonymous reviewers for constructive criticism and to thomas metzinger and raphaël millière for their detailed comments on the penultimate draft of this paper. i would like to thank jennifer windt for comments on the accepted version of the manuscript, and the reviewers and participants of the selfless minds workshop at the frankfurt institute for advanced studies in 2018 for comments and constructive criticism. i thank robin wilson for editing the manuscript. references ataria, y. (2014). where do we end and where does the world begin? the case of insight meditation. philosophical psychology, 28(8), 1128–1146. https://doi.org/10.1080/09515089.2014.969801 ataria, y., dor-ziderman, y., & berkovich-ohana, a. (2015). how does it feel to lack a sense of boundaries? a case study of a long-term mindfulness meditator. consciousness and cognition, 37, 133–147. https://doi.org/10.1016/j.concog.2015.09.002 bayne, t. (2014). replies to commentators. analysis, 74(3), 520–529. https://doi.org/10.1093/analys/anu066 bayne, t., & mcclelland, t. (2018). ensemble representation and the contents of visual experience. philosophical studies, 176(3), 733–753. https://doi.org/10.1007/s11098-018-1037-3 berkovich-ohana, a., dor-ziderman, y., glicksohn, j., & goldstein, a. (2013). alterations in the sense of time, space, and body in the mindfulness-trained brain: a neurophenomenologically-guided meg study. frontiers in psychology, 4, 912. https://doi.org/10.3389/fpsyg.2013.00912 block, n. (1995). on a confusion about a function of consciousness. behavioral and brain sciences, 18(2), 227–247. https://doi.org/10.1017/s0140525x00038188 block, n. j., & fodor, j. a. (1972). what psychological states are not. the philosophical review, 81(2), 159–181. https://doi.org/10.2307/2183991 byrne, a. (2018). inverted qualia. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (winter 2018). https://plato.stanford.edu/archives/win2018/entries/qualia-inverted/; metaphysics research lab, stanford university. chalmers, d. j. (1995). facing up to the problem of consciousness. journal of consciousness studies, 2(3), 200–219. retrieved from https://www.ingentaconnect.com/contentone/imp/jcs/1995/00000002/00000003/653 chalmers, d. j. (1996). the conscious mind: in search of a fundamental theory. oxford: oxford university press. chalmers, d. j. (2004). the representational character of experience. in b. leiter (ed.), the future for philosophy (pp. 153–181). oxford: oxford university press. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1080/09515089.2014.969801 https://doi.org/10.1016/j.concog.2015.09.002 https://doi.org/10.1093/analys/anu066 https://doi.org/10.1007/s11098-018-1037-3 https://doi.org/10.3389/fpsyg.2013.00912 https://doi.org/10.1017/s0140525x00038188 https://doi.org/10.2307/2183991 https://plato.stanford.edu/archives/win2018/entries/qualia-inverted/ https://www.ingentaconnect.com/contentone/imp/jcs/1995/00000002/00000003/653 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 25 chudnoff, e. (2013). gurwitsch’s phenomenal holism. phenomenology and the cognitive sciences, 12(3), 559–578. https://doi.org/10.1007/s11097-012-9279-6 clark, a. (1996). sensory qualities. oxford: oxford university press. cowan, n. (2010). the magical mystery four: how is working memory capacity limited, and why? current directions in psychological science, 19(1), 51–57. https://doi.org/10.1177/0963721409359277 crane, t. (2009). intentionalism. in b. mclaughlin, a. beckermann, & s. walter (eds.), the oxford handbook of philosophy of mind (pp. 479–493). oxford: oxford university press. craver, c. f. (2014). the ontic account of scientific explanation. in m. i. kaiser, o. r. scholz, d. plenge, & a. hüttemann (eds.), explanation in the special sciences: the case of biology and history. (pp. 27–52). https://doi.org/10.1007/978-94-007-7563-3_2 craver, c., & tabery, j. (2017). mechanisms in science. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (spring 2017). retrieved from https://plato.stanford.edu/archives/spr2017/entries/science-mechanisms/ dennett, d. c. (1969). content and consciousness. new york; london: routledge. dor-ziderman, y., ataria, y., fulder, s., goldstein, a., & berkovich-ohana, a. (2016). self-specific processing in the meditating brain: a meg neurophenomenology study. neuroscience of consciousness, 2016(1), niw019. https://doi.org/10.1093/nc/niw019 drayson, z. (2012). the uses and abuses of the personal/subpersonal distinction. philosophical perspectives, 26, 1–18. https://doi.org/10.1111/phpe.12014 drayson, z. (2014). the personal/subpersonal distinction. philosophy compass, 9(5), 338–346. https://doi.org/10.1111/phc3.12124 farrell, b. a. (1950). experience. mind, 59(234), 170–198. https://doi.org/10.1093/mind/lix.234.170 friston, k. (2011). what is optimal about motor control? neuron, 72(3), 488–498. https://doi.org/10.1016/j.neuron.2011.10.018 friston, k., fitzgerald, t., rigoli, f., schwartenbeck, p., & pezzulo, g. (2017a). active inference: a process theory. neural computation, 29(1), 1–49. https://doi.org/10.1162/neco_a_00912 friston, k. j., parr, t., & vries, b. de. (2017b). the graphical brain: belief propagation and active inference. network neuroscience, 1(4), 381–414. https://doi.org/10.1162/netn_a_00018 friston, k., samothrakis, s., & montague, r. (2012). active inference and agency: optimal control without cost functions. biological cybernetics, 106(8), 523–541. https://doi.org/10.1007/s00422-012-0512-8 gallagher, s. (2000). philosophical conceptions of the self: implications for cognitive science. trends in cognitive sciences, 4(1), 14–21. https://doi.org/10.1016/s1364-6613(99)01417-5 gallagher, s. (2013). a pattern theory of self. frontiers in human neuroscience, 7, 443. https://doi.org/10.3389/fnhum.2013.00443 gallagher, s., & daly, a. (2018). dynamical relations in the self-pattern. frontiers in psychology, 9(664). https://doi.org/10.3389/fpsyg.2018.00664 herbart, j. f. (1825). psychologie als wissenschaft neu gegründet auf erfahrung, metaphysik und mathematik. zweiter, analytischer teil. königsberg: unzer. herbart, j. f. (1834). lehrbuch zur psychologie (2nd ed.). königsberg: unzer. herbort, o., & butz, m. v. (2012). too good to be true? ideomotor theory from a computational perspective. frontiers in psychology, 3, 494. https://doi.org/10.3389/fpsyg.2012.00494 hommel, b., müsseler, j., aschersleben, g., & prinz, w. (2001). the theory of event coding (tec): a framework for perception and action planning. behavioral and brain sciences, 24(05), 849–878. https://doi.org/10.1017/s0140525x01000103 james, w. (1890). the principles of psychology. new york: henry holt. kiebel, s. j., daunizeau, j., friston, k. j., & sporns, o. (2008). a hierarchy of time-scales and the brain. plos computational biology, 4(11), e1000209. https://doi.org/10.1371/journal.pcbi.1000209 koksvik, o. (2014). three models of phenomenal unity. journal of consciousness studies, 21(7–8), 105–131. retrieved from https://www.ingentaconnect.com/content/imp/jcs/2014/00000021/f0020007 kriegel, u. (2015). the varieties of consciousness. oxford: oxford university press. lamme, v. (2015). the crack of dawn. in t. k. metzinger & j. m. windt (eds.), open mind. https://doi.org/10.15502/9783958570092 letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11097-012-9279-6 https://doi.org/10.1177/0963721409359277 https://doi.org/10.1007/978-94-007-7563-3_2 https://plato.stanford.edu/archives/spr2017/entries/science-mechanisms/ https://doi.org/10.1093/nc/niw019 https://doi.org/10.1111/phpe.12014 https://doi.org/10.1111/phc3.12124 https://doi.org/10.1093/mind/lix.234.170 https://doi.org/10.1016/j.neuron.2011.10.018 https://doi.org/10.1162/neco_a_00912 https://doi.org/10.1162/netn_a_00018 https://doi.org/10.1007/s00422-012-0512-8 https://doi.org/10.1016/s1364-6613(99)01417-5 https://doi.org/10.3389/fnhum.2013.00443 https://doi.org/10.3389/fpsyg.2018.00664 https://doi.org/10.3389/fpsyg.2012.00494 https://doi.org/10.1017/s0140525x01000103 https://doi.org/10.1371/journal.pcbi.1000209 https://www.ingentaconnect.com/content/imp/jcs/2014/00000021/f0020007 https://doi.org/10.15502/9783958570092 https://doi.org/10.33735/phimisci.2020.i.47 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese 26 lohmann, j., belardinelli, a., & butz, m. v. (2019). hands ahead in mind and motion: active inference in peripersonal hand space. vision, 3(2), 15. https://doi.org/10.3390/vision3020015 lycan, w. g. (1996). layered perceptual representation. philosophical issues, 7, 81–100. https://doi.org/10.2307/1522894 machamer, p., darden, l., & craver, c. f. (2000). thinking about mechanisms. philosophy of science, 67(1), 1–25. retrieved from https://www.jstor.org/stable/pdf/188611.pdf martin, m. g. (1995). bodily awareness: a sense of ownership. in j. l. bermúdez, a. marcel, & n. eilan (eds.), the body and the self (pp. 267–289). cambridge, ma: mit press. metzinger, t. (2013). the myth of cognitive agency: subpersonal thinking as a cyclically recurring loss of mental autonomy. frontiers in psychology, 4, 931. https://doi.org/10.3389/fpsyg.2013.00931 metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of ”pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 miller, g. a. (1956). the magical number seven, plus or minus two: some limits on our capacity for processing information. psychological review, 63(2), 81–97. millière, r. (2019). are there degrees of self-consciousness? journal of consciousness studies, 26(3-4), 252–282. retrieved from https://www.ingentaconnect.com/contentone/imp/jcs/2019/00000026/f0020003/art00012 millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 millière, r., carhart-harris, r. l., roseman, l., trautwein, f.-m., & berkovich-ohana, a. (2018). psychedelics, meditation, and self-consciousness. frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018.01475 nagel, t. (1974). what is it like to be a bat? the philosophical review, 83(4), 435–450. https://doi.org/10.2307/2183914 newen, a. (2018). the embodied self, the pattern theory of self, and the predictive mind. frontiers in psychology, 9, 2270. https://doi.org/10.3389/fpsyg.2018.02270 perlis, d. (1997). consciousness as self-function. journal of consciousness studies, 4(5), 509–525. piccinini, g., & craver, c. (2011). integrating psychology and neuroscience: functional analyses as mechanism sketches. synthese, 183(3), 283–311. https://doi.org/10.1007/s11229-011-9898-4 prosser, s. (2016). experiencing time. oxford: oxford university press. ratcliffe, m. (2012). the phenomenology of existential feeling. in j. fingerhut (ed.), feelings of being alive (pp. 23–53). berlin: de gruyter. revonsuo, a. (1998). how to take consciousness seriously in cognitive neuroscience. communication and cognition: an interdisciplinary quarterly journal, 30(3-4), 185–205. rosenthal, d. m. (2015). quality spaces and sensory modalities. in p. coates & s. coelman (eds.), phenomenal qualities: sense, perception, and consciousness (pp. 33–65). oxford: oxford university press. schweizer, p. (2019). triviality arguments reconsidered. minds and machines, 29, 287–308. https://doi.org/10.1007/s11023-019-09501-x searle, j. r. (1979). what is an intentional state? mind, 88(349), 74–92. retrieved from https://www.jstor.org/stable/pdf/2253065.pdf searle, j. r. (1983). intentionality: an essay in the philosophy of mind. cambridge: cambridge university press. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 slaby, j. (2012). affective self-construal and the sense of ability. emotion review, 4(2), 151–156. https://doi.org/10.1177/1754073911430136 stock, a., & stock, c. (2004). a short history of ideo-motor action. psychological research, 68, 176–188. https://doi.org/10.1007/s00426-003-0154-5 teng, s., & whitney, d. (2011). the acuity of echolocation: spatial resolution in sighted persons compared to the performance of an expert who is blind. journal of visual impairment & blindness, 105(1), 20–32. https://doi.org/10.1177/0145482x1110500103 tong, f., nakayama, k., vaughan, j. t., & kanwisher, n. (1998). binocular rivalry and visual awareness in human extrastriate cortex. neuron, 21, 753–759. https://doi.org/10.1016/s0896-6273(00)80592-9 wiese, w. (2017a). action is enabled by systematic misrepresentations. erkenntnis, 82(6), 1233–1252. https://doi.org/10.1007/s10670-016-9867-x wiese, w. (2017b). predictive processing and the phenomenology of time consciousness. a hierarchical extension of rick grush’s trajectory estimation model. in t. metzinger & w. wiese (eds.), philosophy and predictive processing. frankfurt am main: mind group. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.3390/vision3020015 https://doi.org/10.2307/1522894 https://www.jstor.org/stable/pdf/188611.pdf https://doi.org/10.3389/fpsyg.2013.00931 https://doi.org/10.33735/phimisci.2020.i.46 https://www.ingentaconnect.com/contentone/imp/jcs/2019/00000026/f0020003/art00012 https://doi.org/10.33735/phimisci.2020.i.48 https://doi.org/10.3389/fpsyg.2018.01475 https://doi.org/10.2307/2183914 https://doi.org/10.3389/fpsyg.2018.02270 https://doi.org/10.1007/s11229-011-9898-4 https://doi.org/10.1007/s11023-019-09501-x https://www.jstor.org/stable/pdf/2253065.pdf https://doi.org/10.33735/phimisci.2020.i.44 https://doi.org/10.1177/1754073911430136 https://doi.org/10.1007/s00426-003-0154-5 https://doi.org/10.1177/0145482x1110500103 https://doi.org/10.1016/s0896-6273(00)80592-9 https://doi.org/10.1007/s10670-016-9867-x https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org breaking the self 27 wiese, w. (2018). toward a mature science of consciousness. frontiers in psychology, 9, 693. https://doi.org/10.3389/fpsyg.2018.00693 windt, j. m. (2015). just in time. dreamless sleep experience as pure subjective temporality. in t. k. metzinger & j. m. windt (eds.), open mind. frankfurt am main: mind group. wolpert, d. m., ghahramani, z., & flanagan, j. r. (2001). perspectives and problems in motor learning. trends in cognitive sciences, 5, 487–494. zahavi, d. (2018). consciousness, self-consciousness, selfhood: a reply to some critics. review of philosophy and psychology. https://doi.org/10.1007/s13164-018-0403-6 zahavi, d., & kriegel, u. (2016). for-me-ness: what it is and what it is not. in d. o. dahlstrom, a. elpidorou, & w. hopp (eds.), philosophy of mind and phenomenology: conceptual and empirical approaches (pp. 36–53). new york: routledge. zwickel, j., & prinz, w. (2012). assimilation and contrast: the two sides of specific interference between action and perception. psychological research, 76, 171–182. https://doi.org/10.1007/s00426-011-0338-3 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. wiese, w. (2020). breaking the self: radical disruptions of self-consciousness and impossible conscious experiences. philosophy and the mind sciences, 1(i), 10. https://doi.org/10.33735/phimisci.2020.i.32 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.3389/fpsyg.2018.00693 https://doi.org/10.1007/s13164-018-0403-6 https://doi.org/10.1007/s00426-011-0338-3 https://doi.org/10.33735/phimisci.2020.i.32 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction what exactly is an impossible conscious experience? individuation criteria consciousness and the brain nomological possibility and the brain why are impossible conscious experiences relevant for consciousness research? two types of impossible conscious experiences constraining models of consciousness matching descriptions at different levels of analysis how can impossible conscious experiences be discovered? the general idea two general computational principles as background assumptions from radical disruptions of self-consciousness to impossible conscious experiences general idea the sense of boundaries conclusion machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. machine learning and its impact on psychiatric nosology findings from a qualitative study among german and swiss experts georg starkea,b (georg.starke@unibas.ch) bernice simone elgera,c(b.elger@unibas.ch) eva de clercqa(eva.declercq@unibas.ch) abstract the increasing integration of machine learning (ml) techniques into clinical care, driven in particular by deep learning (dl) using artificial neural nets (anns), promises to reshape medical practice on various levels and across multiple medical fields. much recent literature examines the ethical consequences of employing ml within medical and psychiatric practice but the potential impact on psychiatric diagnostic systems has so far not been well-developed. in this article, we aim to explore the challenges that arise from the recent use of anns for the old problems of psychiatric nosology. to enable an empirically supported critical reflection on the topic, we conducted semi-structured qualitative interviews with swiss and german experts in computational psychiatry. here, we report our findings structured around two themes, namely (1) the possibility of using ml for defining or refining of psychiatric classification, and (2) the desirability of employing ml for psychiatric nosology. we discuss these themes by relating them to recent debates about network theory for psychiatric nosology and show why empirical research in the field should critically reflect on its contribution to psychopathology research. in sum, we argue that beyond technical, regulatory, and ethical challenges, philosophical reflection is crucial to harness the potential of ml in psychiatry. keywords machine learning ∙ nosology ∙ psychiatry ∙ ethics ∙ expert interviews this article is part of a special issue on “models and mechanisms in philosophy of psychiatry,” edited by lena kästner and henrik walter. ainstitute for biomedical ethics, university of basel, switzerland bcollege of humanities, école polytechnique fédérale de lausanne, switzerland cuniversity center of legal medicine, university of geneva, switzerland starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-7428-2619 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org georg starke, bernice s. elger, and eva de clercq 2 1 introduction deep learning (dl) based on artificial neural networks (anns) is at the heart of many recent success stories in the field of machine learning (ml). within psychiatry, dl also promises useful tools for the diagnosis and treatment of psychiatric disorders (durstewitz et al., 2019; jacobson & bhattacharya, 2022; quaak et al., 2021; walter et al., 2019). the recent first approval of a dl-based program by the us food and drug administration to aid with the diagnosis of autism spectrum disorder in young children bears witness to this potential (dattaro, 2021). beyond diagnosis, dl-based programs could also provide complementary offers of digital psychotherapy (lui et al., 2017; martinez-martin & kreitmair, 2018), predict individual treatment outcomes (chekroud et al., 2021) or give prognostic estimates, for instance concerning psychosis (salazar de pablo et al., 2021). responding to long-standing nosological debates within the discipline and dissatisfaction with existing diagnostic criteria (cuthbert & insel, 2013; insel & cuthbert, 2015; kendler, 2016), dl is also increasingly discussed as a potential technique to arrive at novel or refined psychiatric classifications (brunn et al., 2020; eitel et al., 2021). dl-based clustering promises to provide a data-driven approach that can subdivide groups of patients automatically based on neurobiological and behavioural data, finding novel modes of representation (karim et al., 2021; schulz et al., 2020). in principle, such clustering can draw on many different kinds of data, including functional and structural neuroimaging data, eeg measurements, genetic and epigenetic data as well as clinical and neurocognitive observations (huys et al., 2016). to give an example for a neuroscience-focused approach, harnessing the advantages of dl, chang et al. recently reported to have identified subgroups of patients with major psychiatric disorders such as bipolar depression, major depressive disorder, and schizophrenia that are characterized by a frontal–posterior functional imbalance and seem to respond differently to psychopharmacological interventions (chang et al., 2021). while such findings require validation and replication, they could improve existing diagnostic criteria and provide hypothesis for future research (eitel et al., 2021). in parallel to the rise of neuroscientific and psychiatric research endeavours driven by dl, there has also been a blossoming of theoretical approaches that define psychiatric disorders in terms of clusters or networks. such approaches have been especially prominent among nonessentialist theories, i.e., theories that do not espouse a mind-independent understanding of psychiatric disorders as given natural kinds that share intrinsic natural properties. among these nonessentialist approaches, denny borsboom’s suggestion that mental disorders could best be described as complex networks of causally-linked, interconnected symptom components has been particularly influential (borsboom, 2017). symptom network theory promises to provide a non-reductionist link between biological and psychological features of mental disorders (borsboom et al., 2018) and is, as highlighted by a recent review, also supported by a large corpus of empirical results (robinaugh et al., 2020). similarly, peter zachar’s description of psychiatric disorders starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org machine learning and its impact on psychiatric nosology 3 as “imperfect communities” represents an influential nonessentialist approach, describing mental disorders as clusters of symptoms that are historically grown and reflect pragmatical interest (zachar, 2014, pp. 115–136). definitions of psychiatric disorders that are based on neuroscience more frequently represent essentialist views, i.e., theories that take reality to be mindindependent and attempt to carve nature at its joints. such definitions are, for instance, rooted in an understanding of psychiatric disorders as brain disorders (insel & cuthbert, 2015) or point to harmful impairment of natural functioning (faucher & forest, 2021; horwitz & wakefield, 2007). however, it is crucial to distinguish in this context between the ontological question what psychiatric disorders are, and the more practical question how to classify them. as zachar has noted with regard to the diagnostic and statistical manual of mental disorders (dsm), “a careful reading of the introduction to both the dsm-iv and the dsm-5 indicates that alongside the de facto essentialism about the nature of psychiatric disorders there is also a de facto nonessentialism about classification” (zachar, 2014, p. 128). distinguishing between viewpoints about the nature of psychiatric disorders and beliefs about classificatory systems, which in turn fulfil multiple functions (reed et al., 2011), is therefore important to understand how neuroscience-based essentialist views can be seen as compatible with a dimensional approach to psychiatric classification, as endorsed in the dsm-5 (regier et al., 2013). surprisingly, despite the individual prominence of each topic in recent literature, the impact of ml techniques and in particular of dl on psychiatric nosology has so far not received much systematic consideration. many authors have hinted at the potential of dl for nosology (brunn et al., 2020; durstewitz et al., 2019) and some have called for increased attention to the conceptualization of psychiatric disorders in the context of ai-based methods (winter et al., 2021). yet, the relation of a dl-based clustering of disorder subtypes to the competing models of psychiatric disorders remains to be investigated in depth. an exception to this is the paper by wanja wiese and karl friston, who have provided an insightful philosophical discussion of the transformative effects of computational methods on psychiatric nosology and warned against an unintended marginalisation of subjective experience (wiese & friston, 2021). to gain a better understanding whether this worry is shared by other researchers from neuroscience and psychiatry and in which ways ml may have an impact on psychiatric nosology, it seems crucial to explore the explicit and implicit knowledge of scholars in the field (döringer, 2021). reporting the findings from semi-structured qualitative interviews with researchers from germany and switzerland, we present the opinions and attitudes of experts in computational psychiatry with regard to the impact of ml on psychiatric nosology. to our knowledge, while there have been some qualitative findings investigating the attitudes of psychiatrists and psychologists towards ai methods (blease et al., 2020, 2021), this is the first interview-based study looking at nosology in particular. in addition, we relate our findings to debates from the philosophy of science, arguing for a non-reductionist view of mental disorders that allows for starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org georg starke, bernice s. elger, and eva de clercq 4 methodological pluralism. based on these considerations, we point to further lines of research that seem warranted. 2 methods we recruited swiss and german experts on the use of ml in psychiatry. participants were identified by systematically searching on the websites of psychiatric university hospitals in switzerland and germany for clinicians and researchers engaging with artificial intelligence or machine learning. within our narrow recruitment criteria, we aimed to include as diverse a sample as feasible, with view to the respective career stages and gender. once identified, we invited experts to participate in our study via e-mail and sent a reminder after a week in case we did not receive a response. we limited the field of experts to scholars who held at least a doctorate in a relevant field, i.e., medicine, neuroscience, or computer science. the interviews took place between april 2020 and july 2021 and were conducted by the first author, a german physician (md) with an additional degree in philosophy, research and working experience in neuroscience and psychiatry, and basic knowledge of programming and ml. the interviews formed part of his phd in bioethics, which included intensive training and supervision in qualitative data collection. to fine-tune the interview guide and review the interview quality, the first three interviews with experts served as pilots. based on their transcripts, gs and edc revised the interview guide critically, resulting in minor changes. due to the constraints of the pandemic, interviews were conducted via phone (10) or online video call (5), in german (13) or english (2), depending on the participants’ individual preferences. interviews lasted between 25 and 66 minutes. all interviews were transcribed verbatim by the first author, with help from a medical master student (see acknowledgments). all quotes used for the purpose of this paper were translated by gs and checked by edc. the interviewer was familiar with three of the participants prior to conducting the interviews, owing to earlier research activities. to identify important themes relating to psychiatric nosology, we analysed the data from our sample using reflexive thematic analysis (braun & clarke, 2006, 2019). individual codes were given to each segment of each transcribed interview, with one segment representing a unit of meaning, consisting of one or more sentences. initially, the authors conducted the coding jointly for the first four interviews, supported by a master student (see acknowledgments). after agreeing on a coding tree structure, comprising themes and subthemes, the remaining transcripts were coded by the first author, using maxqda software. this data analysis accompanied the data collection, also to monitor data saturation, conceptualized as thematic redundancy indicated by recurrent coding (given, 2015). a full description of our study design, including the informed consent sheet and the interview guide, was submitted for review to the responsible research ethics committee (ethikkommission nordwestund zentralschweiz, eknz), prior starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org machine learning and its impact on psychiatric nosology 5 to any data acquisition. the ethics committee determined that our project did not fall under restrictions that the swiss legal framework imposes on research with human subjects and issued a statement of non-objection (req-2019-00920). notwithstanding this decision, we adhered to high ethical standards, by obtaining informed consent and by ensuring confidentiality and data security: (1) prior to their participation in our study, we asked participants for their written informed consent, and confirmed this again orally at the beginning of the interview. (2) furthermore, we omitted identifying information such as names and places already at the stage of transcribing, (3) and stored the data separately from identifying data on our university servers in switzerland. a detailed analysis of our main findings concerning the ethical dimension of using ml in psychiatry is provided elsewhere (starke et al., 2022). in this manuscript, we focus on the impact of ml on psychiatric nosology, allowing for a more indepth conceptual reflection. questions from the interview guide that are relevant to the current manuscript are provided in table 1. • for which applications of machine learning do you see the greatest potential in future psychiatry? – for which particular clinical objectives? – for which psychiatric disorders? • are there, in your opinion, challenges of using medical machine learning that are specific to psychiatry? • as you know, some authors argue that machine learning, and deep learning in particular, promise a way to divide psychiatric disorders objectively into natural types and thus solve the old problems of psychiatric nosology. where would you stand on this? • how should one best deal with cases of impaired judgement, for example when it comes to a potential program to recommend a particular antipsychotic medication during a psychotic episode? table 1: relevant questions from the interview guide. 3 results semi-structured interviews were conducted with 15 participants out of 26 invited experts (57,6%; 2 women and 13 men). three experts declined due to time constraints, one did not consider themself an expert, and four did not reply. we stopped recruiting additional participants after reaching saturation on the main themes of our study, i.e., once participants reiterated ideas that had already been starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org georg starke, bernice s. elger, and eva de clercq 6 present in similar form in previously conducted interviews (saunders et al., 2018). all participants held at least a doctorate (md and/or phd), covering career stages between postdoc and retired professor (mean years since doctorate 14.4a, sd ±10.8) and were affiliated with german or swiss academic institutions pursuing research on psychiatric disease. ten participants were licensed physicians, five had degrees in psychology or neuroscience, and eight participants reported additional multidisciplinary training in mathematics, physics, engineering, and philosophy. analysing our interviews with particular focus on nosology, we related our findings to two large themes, namely (1) the possibility of using ml for defining psychiatric classifications, and (2) the desirability of employing ml to design psychiatric classificatory systems. 3.1 on the possibility of using ml for defining psychiatric classifications with view to psychiatric classification, the desire to improve current systems was shared unanimously among the interviewees. however, participants’ views on the possibility of using ml for this purpose diverged. some participants embraced an optimist outlook, hoping for new classifications through the use of dl on large data samples comprising biological and behavioural data as well as self-reported symptoms: “i think that if we manage to put together large amounts of data, which you can do with these [neural] networks, that we will then also have another possibility to find groups, subgroups in psychiatry, or perhaps new forms of groupings. i believe that this requires a lot of data that we do not yet have formatted accordingly […], but in principle i think it is possible, yes.” (p2) also others considered ml as particularly useful for psychiatric nosology since it could contribute to mapping different features of psychiatric disorders in a higher dimensional space, taking into account the complex and contingent forms of mental disorders, shaped by history, culture, and language. some participants were therefore optimistic concerning ml, if it incorporated a turn towards a dimensional diagnostic system. “i think we would have to find a dimensional system to describe psychiatric illnesses in the best possible way, similar to the way we describe personality. […] instead of dividing people somehow into diagnostic classes, one could simply describe them with a profile on these different dimensions. and if you then have to decide somehow whether you should treat someone with antidepressants or something, then you could also define a cut-off on the dimension of depressiveness.” (p13). starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org machine learning and its impact on psychiatric nosology 7 the majority of interviewees however regarded ml for nosological purposes more sceptically. some experts insisted that if we were to aim at new classifications, we would need to move beyond mapping symptoms to specific biomarkers, and turn to the underlying mechanisms instead, rooted in neurobiology. “so, if you try to do that at the level of symptoms, i think it’s hopeless. because you know only too well that with prominent examples, – that a certain symptom can be caused by completely different neurological mechanisms. and that’s why a parcellation or a delimitation of diseases can generally, in my view, not be done on the symptom level, but always only on the level of mechanisms and causes. all our claims are not at the level of data, but at the level of possible mechanisms that can explain the data we have observed.” (p4) at the same time, other sceptics frequently pointed to the lack of success in identifying univocal associations between neurobiological data and psychiatric disorders in research so far, even after decades searching for psychiatric biomarkers. “i am very suspicious, having worked in the field for quite a few years, as to whether it will really be possible – whether [machine learning] will prove so helpful to arrive at diagnostic classifications. that isn’t possible at the moment because there are no unequivocal correlations, for example, between certain brain-structural changes and a diagnosis of some kind. you do not have this for a single disorder in psychiatry. you can’t say, for example, frontal lobe grey matter reduction means someone suffers from depression. no: they might suffer from depression, or maybe schizophrenia and so on. there are no unequivocal correlations.” (p1) some interviewees stressed the additional difficulty of arriving at suitable ml models, in light of the fact that current psychiatric diagnostic classifications are not built on biological observations but on the reported phenomenological symptoms of patients. “i mean, in psychiatry in general it is also a methodological problem. because, as i said, the classifications are phenomenological, they have nothing to do with neurobiology, i think we still know far too little about it. and this whole psychiatric classification system has to do with that. that would have to be fundamentally questioned if we were to imagine a greater significance for ai.” (p7) one interviewee reasoned that our current classificatory approaches are reflected in the training data to a degree that makes it impossible to arrive at a new classificatory system. “that’s where the dragon bites its own tail. […]. at the end of the day, we feed our algorithms with pre-assumptions and pre-allocations. […] and machine learning, which forms certain substructures through starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org georg starke, bernice s. elger, and eva de clercq 8 deep learning, so to speak, must always be mapped to the outcome at the end of the day, otherwise it can’t be used. […] this is why we will fail to introduce new psychiatric dimensions now. at the end of the day, [deep learning] may provide us with hypotheses, make us reconsider certain labels and, in particular, reconsider the response to medication in the context of our diagnosis. but i don’t think machine learning itself will miraculously give us any true entities.” (p5) 3.2 on the desirability of using ml for defining psychiatric classifications the second recurring theme was whether it would be desirable to use ml to arrive at novel psychiatric classifications. optimist stances emphasized the methodological benefits of an ml-based classificatory system, allowing for hypothesis-free or more objective approaches, whereas others delineated conditions which such approaches should respect. in the view of optimists, ml could enable new ideas and move beyond existing hypotheses: “i have always been of the opinion, even before ml existed, that we need much more hypothesis-free thinking and not these prefabricated pigeonholes that we have in psychiatry. and that, in my opinion, is one of the great possibilities of such methods, that one can really recognise completely new associations, and perhaps also connections of symptoms, patterns of brain changes, patterns of other endocrine changes, patterns of causes, and thereby generate new causal ideas.” (p15) as a potential result of such hypothesis-free methods, several scholars named the ambition of moving beyond subjective symptoms and gaining a more objective model of psychiatric disorders through an automated approach. “especially in psychiatry there is the problem that many symptoms are subjective and retrospective. this already plays a big and problematic role in clinical care but also in assessments. because many things a patient says cannot be objectively affirmed or denied. it would be interesting if there were possibilities to have more objective access to the inner world of the patient. that would be of great importance for the patient.” (p8) more sceptical voices mentioned the danger that defining psychiatric disorders based on ml models could imply ignoring the history of psychiatry and may contribute to impoverishing the discipline as such. “what is not good is to postulate, as some authors do, and say: in 5 years we will have reached the point with computing power that we can simply put this 19th century thinking, schizophrenia, bipolar etc., starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org machine learning and its impact on psychiatric nosology 9 in the museum, and that’ s it. i think that’s wrong. and not because of the terms. you can abolish the terms if you like. i can also do psychiatry without the schizophrenia term, no problem. but behind the concept of schizophrenia there is a very rich tradition of thought. key words: jaspers, kurt schneider … if all that were to be stirred away because it is old, i would consider that a substantial loss for the discipline.” (p3) another objection to an ml-based nosology was raised by several experts who tied the desirability of a refined classificatory system to its clinical usefulness, providing a prognosis or predict therapeutic response for individual patients. “you can determine a lot after you have talked to the person for two minutes, because everything may already be clear. or if you just see them walking down the corridor. this means that it is certainly not so much a question of finding a diagnosis and classification, but rather the important thing is to give a prognosis or a therapy response. i think these are the important areas of application.” (p7) on a related note, several clinicians also called for a focus on the subjective perspective of the individual patient when asked about the desirability of an ml-based classificatory system. “i am convinced that the diagnosis itself is not relevant. it’s about how the person is doing; can i make them feel better? i don’t need the diagnosis for that if i have a treatment right away. diagnosis is just a vessel to get to treatment. if the biomarker says this person has depression, but the person laughs, can sleep well and says,”i am not depressed”, then he is not depressed. i.e., the diagnosis is always in the eye of the beholder – what the psychiatrist defines, what the patient feels.” (p9) another participant embedded their scepticism in a historical context, linking the history of different ml techniques and the history of modern biologically oriented psychopathology. reflecting on long-standing failures to provide a biologically grounded classification of psychiatric disorders, they were convinced, that although helpful, ml could not resolve the problem of nosology and that investing too much hope in such a project might even be harmful, by leading to another ai winter. “if kraepelin had had deep learning, he would have been using that to classify the patients. but he couldn’t. so, he just classified them with his sorting cards and everything. and then, you know, k-means and clustering algorithms came up in 1958. it was the first – one of the first introductions of the techniques. and then by the 1960s and 1970s they were already using it for psychiatry. but it hasn’t worked. and you know, it’s just an overstatement that it will solve all the problems and starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org georg starke, bernice s. elger, and eva de clercq 10 define objective groups. we have been going after that for a hundred and something years, and it hasn’t happened yet. it certainly may help. i am not denying that. […] but saying that deep learning is going to solve all these problems is exactly like what happened in the 1960s, and then the first ai winter came after that, because the claims were so ridiculously inflated.” (p14) finally, on a more clinical level, several experts reported concern that moving towards an ml-based classificatory, diagnostic system may also alter clinical symptoms. given that the themes of delusions often mirror aspects of a particular age, these clinicians reasoned that such a shift would likely also result in an increase of ml-related delusions. “paranoid experiences, delusions often reflect the times, the zeitgeist. in the past, delusions were often caused by religion. since religion no longer plays such a role, at some point this idea of being bugged came up, or of being irradiated by rays, and now the delusional contents are changing more and more in the direction of the computer.” (p1) “we very often see psychotic patients whose delusions have a lot to do with this topos, i.e., computers, artificial intelligence, who’s listening to me, is there a cia guy sitting around the corner and so on. and i could imagine that for this group of patients, for chronically psychotic people, it would […] become an issue if psychiatry were to become more and more algorithmised and mechanised. because that would somehow strengthen their suspicions, which they have due to their illness. in concrete terms, if i’m sitting here at my desk and the patient is sitting opposite of me and i have 10 computers on the table that are constantly printing out something and beeping, then you don’t have to be schizophrenic to become a bit suspicious.” (p3) 4 discussion the present study aimed to explore experts’ attitudes on the role of ml for psychiatric nosology. to our knowledge, this is the first study that reports the viewpoints of researchers in the field on this topic. with regard to both the possibility and the desirability of using ml to define mental disorders and refine classificatory system, we found optimist and sceptical stances. in the following, we draw on our findings to argue in favour of a methodologically pluralist, non-reductive approach to psychiatric disorders. in particular, we highlight how engaging with conceptual theories such as the network theory of mental disorders could help to advance research in the field, and we show how the reflexive impact of ml-based diagnostics on patients’ symptoms described by our interviewees further supports a non-reductionist approach if seen in the light of hacking’s notion of human kinds. starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org machine learning and its impact on psychiatric nosology 11 concerning the possibility of employing ml methods to solve problems of psychiatric nosology, we found conflicting voices among our interviewees. optimist stances were embraced by few scholars, pointing out potential benefits of using hypothesis-free, data-driven approaches. yet, despite interviewing only experts pursuing research in the very field, the majority of interviewees questioned such promises on a methodological basis. they stressed that available data already mirror current nosological assumptions, leading to feedback effects that prevent advancing beyond current conceptual frameworks. they also referenced the historically poor track record of searching for clinically useful biomarkers in psychiatry as well as our incomplete understanding of causal connections between neurobiology and mental phenomena, between mind and brain. this polyphony of our interviewees’ positions constitutes one of the main findings of our study. the diverse stances mirror longstanding scholarly debates, for instance whether research in psychiatry should be data-driven or theory-driven (huys et al., 2016; itani & rossignol, 2020) or how to bridge the gap between neurobiological mechanisms and phenomenological symptoms (borsboom et al., 2018). the variety of positions also seemed to reflect fundamental metaphysical disagreement about the nature of mental disorders. many of our interviewees seemed to implicitly endorse an understanding of psychiatric disorders as brain disorders that can and should be objectified, whereas others highlighted the limits of dl, stressing phenomenological and historically contingent aspects of mental disorders. wiese and friston (2021) have recently highlighted how research in computational psychiatry, while in theory metaphysically neutral, often tends to place its focus on brain function (friston et al., 2014; montague et al., 2012; stephan & mathys, 2014) and less on genetic mechanisms (rødevand et al., 2021) or clinical predictors (koutsouleris et al., 2021). our sample seems therefore quite reflective of the nosological debates that have vexed psychiatry since its inception (aftab & ryznar, 2021), and to mirror questions how to conceptualise the relation between neurobiology and mental phenomena that remain unsolved for biological psychiatry (walter, 2013). while this result is already interesting in itself as an overview of current attitudes and opinions in the field, we believe that our findings can also inform the philosophical debate on using machine learning for psychiatric nosology. in particular, the various perspectives raised by the interviewed experts highlight the multi-faceted and complex way in which mental disorders present themselves, ranging from the biological and chemical to the social and phenomenological. if some form of unsupervised ml is supposed to advance research towards a more complete account of mental disorders, it would therefore need to integrate these varying levels of explanations. a helpful model for thinking about the integration of such levels has been proposed by lena kästner (2018) in the context of mechanistic explanations: instead of conceptualizing different levels of an explanation in a hierarchical or layered manner, it may prove beneficial to our scientific understanding of complex phenomena if we assume a dimensional view of explanatory levels (kästner, 2018). such dimensions can account for the diverging epistemic starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org georg starke, bernice s. elger, and eva de clercq 12 perspectives of the involved research domains and preserve the respective richness of their descriptions, allowing for complementary and pluralist accounts (ibid.). appreciating and integrating diverging epistemic perspectives, as presented in this paper, seems also very well-suited for the analysis and conceptualisation of mental disorders: it helps to avoid forms of reductionism that promise overly simplistic explanations of psychiatric disorders but do not appreciate the complexity of the phenomenon. for as ludwik fleck provokingly admonished in his 1927 some specific features of the medical way of thinking, “the worse the physician the ‘more logical’ his therapy” (fleck, 1986, p. 42). the worry expressed here, that in medical practice overly simple explanations are hardly a sign of an experienced clinician, resonates well with the opinions of the interviewed experts that put the benefit to the patient front and center. these positions are also in line with the comprehensive literature criticising psychiatric practice for its focus on assigning labels (brinkmann, 2017; callard et al., 2013) and with positions that favour more pragmatic definitions of mental disorders (kendler et al., 2011; zachar, 2014). one proposed and much-discussed system of mental disorders that offers a non-reductionist view, accommodating different dimensions of explanations, is the symptom network theory (borsboom, 2017; borsboom et al., 2018; oude maatman, 2020). as mentioned in the introduction, this theory takes causally connected symptoms as its focal point, satisfying the call by practitioners to focus on clinically relevant features. at the same time, it allows for appreciating biological as much as social determinants of mental disorders by situating them in a complex network that can be described from different epistemic perspectives. engaging with these philosophical debates will therefore also prove useful to empirical researchers, as it provides a framework for the integration of empirical research from different research domains, to make use of the “growing body of empirical research and move the field toward its fundamental aims of explaining, predicting, and controlling psychopathology” (haslbeck et al., 2021). machine learning, and deep learning in particular, should therefore not be seen as a remedy in itself to the challenges of nosology, but rather as a computational tool that may support scientific progress by allowing an improved modelling of complexity, integrating vast amounts of different data types that represent different dimensions of a phenomenon. in this context, at least three caveats though seem crucial. first, a diagnostic system based on ml should not be mistaken to provide an objective “view from nowhere”, to borrow nagel’s phrase (1986). on the one hand, any computational model will be shaped by the type of data selected for its training, and by the context of their acquisition, as repeatedly stressed by our interviewees. in addition, insofar as computational psychiatry draws on a concept of miscomputation, it employs a value-laden and perspectival notion of normalcy for its explanations (colombo, 2021). also with the support of ml, it will therefore remain crucial to be mindful of the epistemic perspectives informing classificatory systems in psychiatry. starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org machine learning and its impact on psychiatric nosology 13 a second caveat concerns the limited possibility of arriving at causal structures with deep learning techniques. while dl may provide researchers with new hypotheses or inspiration through its ability to detect correlations in large datasets (davies et al., 2021), it usually does not provide causal scientific explanations, with very few exceptions such as explicit causal modelling (parascandolo et al., 2018). this constitutes an important difference to symptom network theory, which demands causal links between different nodes in the symptom network (borsboom, 2017). deep neural network models therefore only provide one step in the generation of scientific knowledge, offering “first steps to determining which causal mechanisms or dependency relations should be explored further” (sullivan, 2022), or as p15 put it: first steps to “recognise completely new associations, and perhaps also connections of symptoms, patterns of brain changes, patterns of other endocrine changes, patterns of causes, and thereby generate new causal ideas.” a third caveat is that also with the use of ml, psychiatric classificatory systems will not carve nature at its joints but will remain dynamic and open to change. evidence for this claim can be found in the anecdotal clinical reports of psychotic symptoms being shaped by the real or feared integration of ml into psychiatry that came up repeatedly in many of our interviews, despite not corresponding to any item in our interview guide. assuming that these reports are not isolated concerns, this unintended impact of ml on psychiatric diagnostic seems to fit well with what ian hacking has described as the looping effect of human kinds where human classifications and their social environment are causally intertwined through feedback mechanisms (hacking, 1999). hacking’s work on natural and human kinds has informed the past decades of debate in psychiatric research. in hacking’s view, natural kinds are supposed to offer a unique taxonomy “that represents nature as it is, and reflects the network of causal laws” (hacking, 1991, p. 111), whereas human kinds are the subject of the social sciences, providing “classifications that could be used to formulate general truths about people” (hacking, 1996, p. 352). while the debate about this distinction’s conceptual bearings is vast and controversial (bird & tobin, 2022; cooper, 2004; craver, 2009; tsou, 2007; van riel, 2016), some authors have also used it to design empirical research, investigating for instance the way in which young adolescents interact and transform psychiatric concepts (lindholm & wickström, 2020). here, our point is much more modest though: if the use of an ml-based diagnostic regime does indeed shape the symptom of patients, and if said symptoms are used as training data to the diagnostic model, this would imply the need to regularly update the classificatory model. this observation alone may therefore be seen as a reason to not harbour a machine-learning based “aspiration to automatically segregate brain disorders into natural kinds” (bzdok & meyer-lindenberg, 2018). our study has several limitations. since our purposive sampling was highly targeted on a specific research field within psychiatry in germany and switzerland, our results are not representative, neither for psychiatry in general nor for starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org georg starke, bernice s. elger, and eva de clercq 14 other cultural contexts. as is the case for all qualitative research, our results are therefore not generalizable. for this reason and to safeguard the anonymity of our participants, we can therefore not provide insights into quantifiable relations between, e.g., the experts’ years of experience or their success in publishing and their attitudes towards ml, but believe that such inquiry would constitute a valuable route for future research. in addition, the close involvement of the interviewer in the field as well as his medical background may have influenced his interactions with the interviewees. yet, since our study aimed at exploring different facets of an emerging research field, not at representative descriptions, we believe that these limitations do not diminish the value of our findings. 5 conclusion this study provides the first qualitative insights into the impact of ml on psychiatric nosology. it highlights how ml and dl in particular does seemingly not provide a solution to problems of defining psychiatric disorders but instead mirrors existing disagreements. our findings should therefore be read as an exhortation to scholars working in the field of computational psychiatry to engage more deeply with philosophical debates and bridge the gaps between research employing dl and the philosophy of mind. doing so may support the development of non-reductionist research programs that appreciate the complexity of mental disorders by integrating empirical findings from different research domains. acknowledgments first and foremost, we would like to thank our interviewees for their time and willingness to participate in our study, despite their multiple obligations during the pandemic. we are also grateful for valuable comments on an earlier draft by christopher poppe and two anonymous reviewers and indebted to benedikt schmidt for help with transcribing the interviews. gs would also like to acknowledge funding from the era-net neuron project hybridmind (swiss national science foundation’s grant number: 32ne30_199436) to work on the revision of this paper. references aftab, a., & ryznar, e. (2021). conceptual and historical evolution of psychiatric nosology. international review of psychiatry, 33(5), 486–499. https://doi.org/10.1080/09540261.2020.1828306 bird, a., & tobin, e. (2022). natural kinds. in e. n. zalta (ed.), the stanford encyclopedia of philosophy, spring 2022 ed. metaphysics research lab, stanford university. https://plato.stanford.edu/archives/spr2022/entries/natural-kinds/ blease, c., kharko, a., annoni, m., gaab, j., & locher, c. (2021). machine learning in clinical psychology and psychotherapy education: a mixed methods pilot survey of postgraduate students at a swiss university. frontiers in public health, 9, 623088. https://doi.org/10.3389/fpubh.2021.623088 blease, c., locher, c., leon-carlyle, m., & doraiswamy, m. (2020). artificial intelligence and the future of psychiatry: qualitative findings from a global physician survey. digital health, 6, 2055207620968355. https://doi.org/10.1177/20 55207620968355 borsboom, d. (2017). a network theory of mental disorders. world psychiatry, 16(1), 5–13. https://doi.org/10.1002/wps.20375 starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1080/09540261.2020.1828306 https://plato.stanford.edu/archives/spr2022/entries/natural-kinds/ https://doi.org/10.3389/fpubh.2021.623088 https://doi.org/10.1177/2055207620968355 https://doi.org/10.1177/2055207620968355 https://doi.org/10.1002/wps.20375 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org machine learning and its impact on psychiatric nosology 15 borsboom, d., cramer, a., & kalis, a. (2018). brain disorders? not really… why network structures block reductionism in psychopathology research. behavioral and brain sciences, 1–54. https://doi.org/10.1017/s0140525x17002266 braun, v., & clarke, v. (2006). using thematic analysis in psychology. qualitative research in psychology, 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa braun, v., & clarke, v. (2019). reflecting on reflexive thematic analysis. qualitative research in sport, exercise and health, 11(4), 589–597. https://doi.org/10.1080/2159676x.2019.1628806 brinkmann, s. (2017). perspectives on diagnosed suffering. nordic psychology, 69(1), 1–4. https://doi.org/10.1080/19012276.2016.1270404 brunn, m., diefenbacher, a., courtet, p., & genieys, w. (2020). the future is knocking: how artificial intelligence will fundamentally change psychiatry. academic psychiatry, 44, 461–466. https://doi.org/10.1007/s40596-020-01243-8 bzdok, d., & meyer-lindenberg, a. (2018). machine learning for precision psychiatry: opportunities and challenges. biological psychiatry: cognitive neuroscience and neuroimaging, 3(3), 223–230. https://doi.org/10.1016/j.bpsc.2017.11.007 callard, f., bracken, p., david, a. s., & sartorius, n. (2013). has psychiatric diagnosis labelled rather than enabled patients? bmj, 347, f4312. https://doi.org/10.1136/bmj.f4312 chang, m., womer, f. y., gong, x., chen, x., tang, l., feng, r., dong, s., duan, j., chen, y., & zhang, r. (2021). identifying and validating subtypes within major psychiatric disorders based on frontal–posterior functional imbalance via deep learning. molecular psychiatry, 26, 2991–3002. https://doi.org/10.1038/s41380-020-00892-3 chekroud, a. m., bondar, j., delgadillo, j., doherty, g., wasil, a., fokkema, m., cohen, z., belgrave, d., derubeis, r., & iniesta, r. (2021). the promise of machine learning in predicting treatment outcomes in psychiatry. world psychiatry, 20(2), 154–170. https://doi.org/10.1002/wps.20882 colombo, m. (2021). (mis)computation in computational psychiatry. in f. calzavarini & m. viola (eds.), neural mechanisms. studies in brain and mind (pp. 427–448). springer. https://doi.org/10.1007/978-3-030-54092-0_18 cooper, r. (2004). why hacking is wrong about human kinds. british journal for the philosophy of science, 55(1), 73–85. craver, c. f. (2009). mechanisms and natural kinds. philosophical psychology, 22(5), 575–594. https://doi.org/10.1080/09515080903238930 cuthbert, b. n., & insel, t. r. (2013). toward the future of psychiatric diagnosis: the seven pillars of rdoc. bmc medicine, 11(1), 1–8. https://doi.org/10.1186/1741-7015-11-126 dattaro, l. (2021). green light for diagnostic autism app raises questions, concerns. https://www.spectrumnews.org/news/gr een-light-for-diagnostic-autism-app-raises-questions-concerns/ davies, a., veličković, p., buesing, l., blackwell, s., zheng, d., tomašev, n., tanburn, r., battaglia, p., blundell, c., & juhász, a. (2021). advancing mathematics by guiding human intuition with ai. nature, 600(7887), 70–74. https://doi.org/10.1038/s41586-021-04086-x döringer, s. (2021). “the problem-centred expert interview.” combining qualitative interviewing approaches for investigating implicit expert knowledge. international journal of social research methodology, 24(3), 265–278. https://doi.org/10.1080/13645579.2020.1766777 durstewitz, d., koppe, g., & meyer-lindenberg, a. (2019). deep neural networks in psychiatry. molecular psychiatry, 24(11), 1583–1598. https://doi.org/10.1038/s41380-019-0365-9 eitel, f., schulz, m.-a., seiler, m., walter, h., & ritter, k. (2021). promises and pitfalls of deep neural networks in neuroimaging-based psychiatric research. experimental neurology, 339, 113608. https://doi.org/10.1016/j.expneurol.2021.113608 faucher, l., & forest, d. (2021). defining mental disorder: jerome wakefield and his critics. mit press. fleck, l. (1986). some specific features of the medical way of thinking [1927]. in t. s. robert s. cohen (ed.), cognition and fact: materials on ludwik fleck (pp. 39–46). springer netherlands. https://doi.org/10.1007/978-94-009-4498-5_2 friston, k. j., stephan, k. e., montague, r., & dolan, r. j. (2014). computational psychiatry: the brain as a phantastic organ. the lancet psychiatry, 1(2), 148–158. https://doi.org/10.1016/s2215-0366(14)70275-5 given, l. m. (2015). 100 questions (and answers) about qualitative research. sage publications. hacking, i. (1991). a tradition of natural kinds. philosophical studies: an international journal for philosophy in the analytic tradition, 61(1/2), 109–126. https://doi.org/10.1007/bf00385836 hacking, i. (1996). the looping effects of human kinds. in p. sperber d. (ed.), causal cognition: a multi-disciplinary debate (pp. 351–383). oxford academic. https://doi.org/10.1093/acprof:oso/9780198524021.003.0012 hacking, i. (1999). the social construction of what? harvard university press. haslbeck, j., ryan, o., robinaugh, d. j., waldorp, l. j., & borsboom, d. (2021). modeling psychopathology: from data models to formal theories. psychological methods. https://doi.org/10.1037/met0000303 horwitz, a. v., & wakefield, j. c. (2007). the loss of sadness: how psychiatry transformed normal sorrow into depressive disorder. oxford university press. starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1017/s0140525x17002266 https://doi.org/10.1191/1478088706qp063oa https://doi.org/10.1080/2159676x.2019.1628806 https://doi.org/10.1080/19012276.2016.1270404 https://doi.org/10.1007/s40596-020-01243-8 https://doi.org/10.1016/j.bpsc.2017.11.007 https://doi.org/10.1136/bmj.f4312 https://doi.org/10.1038/s41380-020-00892-3 https://doi.org/10.1002/wps.20882 https://doi.org/10.1007/978-3-030-54092-0_18 https://doi.org/10.1080/09515080903238930 https://doi.org/10.1186/1741-7015-11-126 https://www.spectrumnews.org/news/green-light-for-diagnostic-autism-app-raises-questions-concerns/ https://www.spectrumnews.org/news/green-light-for-diagnostic-autism-app-raises-questions-concerns/ https://doi.org/10.1038/s41586-021-04086-x https://doi.org/10.1080/13645579.2020.1766777 https://doi.org/10.1038/s41380-019-0365-9 https://doi.org/10.1016/j.expneurol.2021.113608 https://doi.org/10.1007/978-94-009-4498-5_2 https://doi.org/10.1016/s2215-0366(14)70275-5 https://doi.org/10.1007/bf00385836 https://doi.org/10.1093/acprof:oso/9780198524021.003.0012 https://doi.org/10.1037/met0000303 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org georg starke, bernice s. elger, and eva de clercq 16 huys, q. j., maia, t. v., & frank, m. j. (2016). computational psychiatry as a bridge from neuroscience to clinical applications. nature neuroscience, 19(3), 404–413. https://doi.org/10.1038/nn.4238 insel, t. r., & cuthbert, b. n. (2015). brain disorders? precisely. science, 348(6234), 499–500. https://doi.org/10.1126/science.aab2358 itani, s., & rossignol, m. (2020). at the crossroads between psychiatry and machine learning: insights into paradigms and challenges for clinical applicability. frontiers in psychiatry, 11, 1029. https://doi.org/10.3389/fpsyt.2020.552262 jacobson, n. c., & bhattacharya, s. (2022). digital biomarkers of anxiety disorder symptom changes: personalized deep learning models using smartphone sensors accurately predict anxiety symptoms from ecological momentary assessments. behaviour research and therapy, 149, 104013. https://doi.org/10.1016/j.brat.2021.104013 karim, m. r., beyan, o., zappa, a., costa, i. g., rebholz-schuhmann, d., cochez, m., & decker, s. (2021). deep learningbased clustering approaches for bioinformatics. briefings in bioinformatics, 22(1), 393–415. https://doi.org/10.1093/bib/bbz170 kästner, l. (2018). integrating mechanistic explanations through epistemic perspectives. studies in history and philosophy of science part a, 68, 68–79. https://doi.org/10.1016/j.shpsa.2018.01.011 kendler, k. s. (2016). the nature of psychiatric disorders. world psychiatry, 15(1), 5–12. https://doi.org/10.1002/wps.20292 kendler, k. s., zachar, p., & craver, c. (2011). what kinds of things are psychiatric disorders? psychological medicine, 41(6), 1143–1150. https://doi.org/10.1017/s0033291710001844 koutsouleris, n., dwyer, d. b., degenhardt, f., maj, c., urquijo-castro, m. f., sanfelici, r., popovic, d., oeztuerk, o., haas, s. s., & weiske, j. (2021). multimodal machine learning workflows for prediction of psychosis in patients with clinical high-risk syndromes and recent-onset depression. jama psychiatry, 78(2), 195–209. https://doi.org/10.1001/jamapsychiatry.2020.3604 lindholm, s. k., & wickström, a. (2020). “looping effects” related to young people’s mental health: how young people transform the meaning of psychiatric concepts. global studies of childhood, 10(1), 26–38. https://doi.org/10.1177/2043610619890058 lui, j. h., marcus, d. k., & barry, c. t. (2017). evidence-based apps? a review of mental health mobile applications in a psychotherapy context. professional psychology: research and practice, 48(3), 199–210. https://doi.org/10.1037/pro0000122 martinez-martin, n., & kreitmair, k. (2018). ethical issues for direct-to-consumer digital psychotherapy apps: addressing accountability, data protection, and consent. jmir mental health, 5(2), e32. https://doi.org/10.2196/mental.9423 montague, p. r., dolan, r. j., friston, k. j., & dayan, p. (2012). computational psychiatry. trends in cognitive sciences, 16(1), 72–80. https://doi.org/10.1016/j.tics.2011.11.018 nagel, t. (1986). the view from nowhere. oxford university press. oude maatman, f. (2020). reformulating the network theory of mental disorders: folk psychology as a factor, not a fact. theory & psychology, 30(5), 703–722. https://doi.org/10.1177/0959354320921464 parascandolo, g., kilbertus, n., rojas-carulla, m., & schölkopf, b. (2018). learning independent causal mechanisms. 80, 4036–4044. https://doi.org/10.48550/arxiv.1712.00961 quaak, m., mortel, l. van de, thomas, r. m., & wingen, g. van. (2021). deep learning applications for the classification of psychiatric disorders using neuroimaging data: systematic review and meta-analysis. neuroimage: clinical, 30, 102584. https://doi.org/10.1016/j.nicl.2021.102584 reed, g. m., correia, j. m., esparza, p., saxena, s., & maj, m. (2011). the wpa-who global survey of psychiatrists’ attitudes towards mental disorders classification. world psychiatry, 10(2), 118–131. https://doi.org/10.1002/j.2051-5545.2011.tb00034.x regier, d. a., kuhl, e. a., & kupfer, d. j. (2013). the dsm‐5: classification and criteria changes. world psychiatry, 12(2), 92–98. https://doi.org/10.1002/wps.20050 robinaugh, d. j., hoekstra, r. h., toner, e. r., & borsboom, d. (2020). the network approach to psychopathology: a review of the literature 2008–2018 and an agenda for future research. psychological medicine, 50(3), 353–366. https://doi.org/10.1017/s0033291719003404 rødevand, l., bahrami, s., frei, o., lin, a., gani, o., shadrin, a., smeland, o. b., o’connell, k. s., elvsåshagen, t., & winterton, a. (2021). polygenic overlap and shared genetic loci between loneliness, severe mental disorders, and cardiovascular disease risk factors suggest shared molecular mechanisms. translational psychiatry, 11(1), 1–11. https://doi.org/10.1038/s41398-020-01142-4 salazar de pablo, g., studerus, e., vaquerizo-serrano, j., irving, j., catalan, a., oliver, d., baldwin, h., danese, a., fazel, s., & steyerberg, e. w. (2021). implementing precision psychiatry: a systematic review of individualized prediction models for clinical practice. schizophrenia bulletin, 47(2), 284–297. https://doi.org/10.1093/schbul/sbaa120 starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1038/nn.4238 https://doi.org/10.1126/science.aab2358 https://doi.org/10.3389/fpsyt.2020.552262 https://doi.org/10.1016/j.brat.2021.104013 https://doi.org/10.1093/bib/bbz170 https://doi.org/10.1016/j.shpsa.2018.01.011 https://doi.org/10.1002/wps.20292 https://doi.org/10.1017/s0033291710001844 https://doi.org/10.1001/jamapsychiatry.2020.3604 https://doi.org/10.1177/2043610619890058 https://doi.org/10.1037/pro0000122 https://doi.org/10.2196/mental.9423 https://doi.org/10.1016/j.tics.2011.11.018 https://doi.org/10.1177/0959354320921464 https://doi.org/10.48550/arxiv.1712.00961 https://doi.org/10.1016/j.nicl.2021.102584 https://doi.org/10.1002/j.2051-5545.2011.tb00034.x https://doi.org/10.1002/wps.20050 https://doi.org/10.1017/s0033291719003404 https://doi.org/10.1038/s41398-020-01142-4 https://doi.org/10.1093/schbul/sbaa120 https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org machine learning and its impact on psychiatric nosology 17 saunders, b., sim, j., kingstone, t., baker, s., waterfield, j., bartlam, b., burroughs, h., & jinks, c. (2018). saturation in qualitative research: exploring its conceptualization and operationalization. quality & quantity, 52(4), 1893–1907. https://doi.org/10.1007/s11135-017-0574-8 schulz, m.-a., chapman-rounds, m., verma, m., bzdok, d., & georgatzis, k. (2020). inferring disease subtypes from clusters in explanation space. scientific reports, 10(1), 1–6. https://doi.org/10.1038/s41598-020-68858-7 starke, g., schmidt, b., de clercq, e., & elger, b. s. (2022). explainability as fig leaf? an exploration of experts’ ethical expectations towards machine learning in psychiatry. ai and ethics. https://doi.org/10.1007/s43681-022-00177-1 stephan, k. e., & mathys, c. (2014). computational approaches to psychiatry. current opinion in neurobiology, 25, 85–92. https://doi.org/10.1016/j.conb.2013.12.007 sullivan, e. (2022). understanding from machine learning models. the british journal for the philosophy of science, 73(1), 109–133. https://doi.org/10.1093/bjps/axz035 tsou, j. y. (2007). hacking on the looping effects of psychiatric classifications: what is an interactive and indifferent kind? international studies in the philosophy of science, 21(3), 329–344. https://doi.org/10.1080/02698590701589601 van riel, r. (2016). what is constructionism in psychiatry? from social causes to psychiatric classification. frontiers in psychiatry, 7, 57. https://doi.org/10.3389/fpsyt.2016.00057 walter, h. (2013). the third wave of biological psychiatry. frontiers in psychology, 4, 582. https://doi.org/10.3389/fpsyg.20 13.00582 walter, m., alizadeh, s., jamalabadi, h., lueken, u., dannlowski, u., walter, h., olbrich, s., colic, l., kambeitz, j., koutsouleris, n., hahn, t., & dwyer, d. b. (2019). translational machine learning for psychiatric neuroimaging. progress in neuropsychopharmacology & biological psychiatry, 91, 113–121. https://doi.org/10.1016/j.pnpbp.2018.09.014 wiese, w., & friston, k. j. (2021). ai ethics in computational psychiatry: from the neuroscience of consciousness to the ethics of consciousness. behavioural brain research, 113704. https://doi.org/10.1016/j.bbr.2021.113704 winter, n. r., cearns, m., clark, s. r., leenings, r., dannlowski, u., baune, b. t., & hahn, t. (2021). from multivariate methods to an ai ecosystem. molecular psychiatry, 26, 6116–6120. https://doi.org/10.1038/s41380-021-01116-y zachar, p. (2014). a metaphysics of psychopathology. mit press. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. starke, g., elger, b. s., & de clercq, e. (2023). machine learning and its impact on psychiatric nosology. findings from a qualitative study among german and swiss experts. philosophy and the mind sciences, 4, 4. https://doi.org/10.33735/phimisci.2023.9435 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11135-017-0574-8 https://doi.org/10.1038/s41598-020-68858-7 https://doi.org/10.1007/s43681-022-00177-1 https://doi.org/10.1016/j.conb.2013.12.007 https://doi.org/10.1093/bjps/axz035 https://doi.org/10.1080/02698590701589601 https://doi.org/10.3389/fpsyt.2016.00057 https://doi.org/10.3389/fpsyg.2013.00582 https://doi.org/10.3389/fpsyg.2013.00582 https://doi.org/10.1016/j.pnpbp.2018.09.014 https://doi.org/10.1016/j.bbr.2021.113704 https://doi.org/10.1038/s41380-021-01116-y https://doi.org/10.33735/phimisci.2023.9435 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction methods results on the possibility of using ml for defining psychiatric classifications on the desirability of using ml for defining psychiatric classifications discussion conclusion perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness perspectival self-consciousness and ego-dissolution an analysis of (some) altered states of consciousness miguel ángel sebastiána (msebastian@gmail.com) abstract it is often claimed that a minimal form of self-awareness is constitutive of our conscious experience. some have considered that such a claim is plausible for our ordinary experiences but false when considered unrestrictedly on the basis of the empirical evidence from altered states. in this paper i want to reject such a reasoning. this requires, first, a proper understanding of a minimal form of self-awareness – one that makes it plausible that minimal self-awareness is part of our ordinary experiences. i will argue that it should be understood as perspectival first-person awareness (pfp-awareness): a nonconceptual identification-free self-attribution that defines the first-person perspective for our conscious experience. i will offer a detailed characterization of pfp-awareness in semantic and epistemological terms. with this tool in hand, i will review the empirical literature on altered states. i will focus on psychedelics, meditation and dreams, as they have been claimed to present the clearest cases in favor of a radical disruption of self-awareness. i will show that the rejection of the idea that minimal self-awareness is constitutive of our experience on the basis of this evidence is unfounded, for two main reasons. first, although there are good grounds to think that some forms of self-awareness that typically accompany our ordinary experiences are compromised, they do not support the claim that pfp-awareness is absent. secondly, the reports that could make us think of a radical disruption of self-awareness are most probably due to a confirmation bias – and hence we should mistrust them – derived from the expectations and metaphysical views of their subjects. keywords core de se representation ∙ dreams ∙ ego-dissolution ∙ for-me-ness ∙ meditation ∙ minimal selfconsciousness ∙ pre-reflective self-consciousness ∙ psychedelics∙ self-involving representation this article is part of a special issue on “radical disruptions of self-consciousness”, edited by thomas metzinger and raphaël millière. ainstituto de investigaciones filosóficas, universidad nacional autónoma de méxico sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-1890-8319 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 2 1 introduction in our daily life we are conscious of the objects around us, but we are also conscious of ourselves. we are conscious that we perceive those objects, that we see their colors or smell their aromas; conscious of our bodies and that we are different from the objects we perceive; and also of our plans, hopes, and fears. influenced some way or other by kant, many philosophers have claimed that self-consciousness is constitutive of our conscious experience. for example, william james (1895) suggested this when he claimed that “whatever i may be thinking of, i am always at the same time more or less aware of myself, of my personal existence” (p. 42). the plausibility of this claim will depend on the way in which such self-consciousness is unpacked; for it is not entirely obvious in what sense one is aware of one’s personal existence when one is, say, fully immersed in watching a movie or concentrated while playing chess. although we are self-conscious when we consider our plans or the kind of person we are, this form of self-consciousness is definitely not present in all our experiences. gallagher (2000) distinguishes the narrative self, corresponding roughly to the idea that we and others have of ourselves – dependent on episodic and autobiographical memory, as well as on imagination related to the planned and expected future – from the minimal self. those who defend the claim that self-consciousness is constitutive of our conscious life do not think of self-consciousness as consciousness of the narrative self but rather as involving a minimal sense of selfconsciousness: 𝑆𝐶𝑀𝐼𝑁 consciousness entails a minimal form of self-consciousness.1 in the phenomenological tradition, conscious episodes are taken to be present to the subject in an immediate first-personal way. this is unpacked in terms of a minimal form of self-consciousness which is manifest in our experience prior to reflection and introspection – pre-reflective self-consciousness. in line with these ideas, some philosophers in the analytic tradition understand the subjective aspect of our experiences in terms of self-consciousness. they have distinguished two aspects in which the phenomenal character can, at least conceptually, be decomposed: the qualitative and the subjective character (kriegel, 2009; levine, 2001). the qualitative character is what distinguishes different conscious experiences: what it is like for one to see red is different from what it is like for one to see green or to smell coffee. moreover, conscious experiences are subjective. despite the differences, there is something that all those episodes share; namely, their subjective aspect – there is something it is like for oneself. the idea is that conscious states present 1the idea that consciousness entails “self-consciousness” is ambiguous. it can mean that consciousness entails consciousness of itself (of the conscious episode itself) or consciousness of oneself (the subject of experience). elsewhere i have called them “mental-state-involving” and “self-involving” respectively (sebastián, 2012). the second reading is the one relevant to the discussion in this special issue and the one i shall assume in what follows. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 3 the world from a first-person perspective; they do not merely inform us about the way the world is but also about our relationship to it. for example, in having a visual experience one is not merely aware of one’s environment, but also that one oneself is bearing a certain relationship to it.2 the subjective character is unpacked as a minimal form of self-consciousness and intended to be a feature constitutive of all and only conscious states. this short introduction is, of course, insufficient to show that 𝑆𝐶𝑀𝐼𝑁 is true, but this is not the aim of the paper. instead, i will more modestly argue that 𝑆𝐶𝑀𝐼𝑁 cannot be ruled out by available empirical evidence, because alleged “counter-examples” to 𝑆𝐶𝑀𝐼𝑁 are in fact compatible with the claim consciousness entails a minimal form of self-consciousness. many authors have considered that a version of 𝑆𝐶𝑀𝐼𝑁 is true if we restrict its scope to our ordinary experiences. 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 ordinary consciousness entails a minimal form of selfconsciousness. but some have claimed that the unrestricted version (𝑆𝐶𝑀𝐼𝑁 ) is false for empirical reasons. in this paper i want to dispute this claim. i will assume that a minimal form of self-consciousness accompanies all our ordinary experiences (𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷) and evaluate whether the evidence from empirical research on altered states offer reasons to think that this self-consciousness is not constitutive of consciousness (𝑆𝐶𝑀𝐼𝑁 ).3 this requires first a proper understanding of the alleged minimal form of selfconsciousness, one that makes it plausible that 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 is true. section 2 2in line with the ambiguity noted in footnote 1, philosophers disagree with regard to the content of awareness, whether it entails awareness of 1) ourselves, 2) the experience itself or 3) both – for details see guillot (2016; sebastián, 2012). 1 and 3 agree that the experience concerns the subject. 1 is the view presented above and 3 is the view endorsed by those who think that the experience is represented as mine or that one oneself is having the experience. those are the views that are relevant for the discussion in this issue. i will appeal to 1 because it is the less committal and hold that one’s experience conveys, in a way to be clarified, that one oneself is in a certain state – a relational one to the environment. this view collapses into 3 if the state in which one’s experience represents oneself as being is the experience itself. it is worth remarking that it is not uncommon to start from a presentation of the subjective character in first-personal terms (e.g. zahavi & kriegel, 2016) – by appeal to a first-person perspective and employing first-person pronouns (i, me, mine), as in 1, and end up with an analysis of the phenomenon in terms of an awareness of the experience itself, as in 2. for example, kriegel (2009) explicitly agrees that the content of a human adult is de se or first-personal (p. 177). however, he denies that it is constitutively so, it constitutively entails awareness of the experience itself rather than of one oneself. the reason he provides is that the experiences of infants and other animals might fail to have such a feature (p. 178). this has always struck me as an ad hoc claim. please note that in the absence of a metasemantic theory that spells out the details of what is required for such a representation – what i will call core de se, we cannot adjudicate what happens in the case of animal or babies, as we lack access to their experiences. 3fully in line with the selfless mind manifesto (which has been incorporated into the editorial introduction, see millière & metzinger, 2020) that guides this special issue. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 4 is devoted to this task. in particular, i will argue that minimal self-consciousness should be understood in terms of perspectival first-person (core de se) awareness: pfp-awareness a non-conceptual identification-free self-attribution that characterizes the first-person perspective that consciousness offers us. i will characterize pfp-awareness in terms of its semantic and epistemological features (section 2.1); sharply distinguish it from other forms of self-consciousness (section 2.2), as well as from other notions in the vicinity which often accompany our waking experiences but not constitutively so; and explain how self-consciousness can be disrupted to different degrees while pfp-awareness remains intact (section 2.3). with this tool in hand, in section 3, i deal with some altered states of consciousness. in particular, i analyze the evidence from empirical studies of drug-induced states, dreamless sleep and meditation, as they present the clearest cases in favor of a radical dissolution of self-consciousness. 2 perspectival first-personal (pfp) awareness when we theorize about the nature and mechanisms underlying our conscious experiences, a fundamental element to be considered is their phenomenal character, what is phenomenologically manifest. this can be characterized by what the experience conveys to its subject or what the subject is aware/conscious of during the conscious episode. we can think of what is conveyed, or the content of a state, in terms of accuracy or correctness conditions. imagine that you read in the newspaper that there has been a celebration in mexico city. this news tells you that the world is a certain way, one in which there has been a celebration in mexico city; and it is accurate if the world is such a way and inaccurate if it is not. likewise, we can think of what the experience conveys to its subject as the condition under which the experience is adequate (siegel, 2010). consider an ordinary perceptual experience like the one i have when looking at my black mug of coffee. such visual experience (nonconceptually)4 conveys to me that there is a black mug. the experience is adequate if the world is such that there is a black mug. the mug and its properties constitute, in the previous example, what we can call “the primary object of experience”. it doesn’t seem controversial to claim that visual experience conveys the presence of its primary object.5 however, the primary object does not exhaust our conscious 4the use of a “that” clause to express what the experience conveys does not commit one to linguistic forms of representation, nor to the claim that the content has to be conceptual. moreover, the subject doesn’t need to posses the concepts corresponding to the terms that appear in the sentence that expresses what the experience conveys (heck, 2002), cf. toribio (2008). 5there is substantial disagreement with regard to what is the object of experience: what are the properties that enter the content of these kind of experiences. for example, on whether they are low level properties, such as maybe shape or colour (dretske, 1995; tye, 1997), or also higher-level sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 5 awareness, and few would endorse this radical form of transparency. the object of experience and the relationship our experience represents (or makes us aware of) ourselves as being in change across experiences for example, when seeing a black mug, tasting a wine or thinking that two plus two is four.6 but all these conscious episodes remain subjective. there is a first-person perspective that remains constant along, at least, all our ordinary experiences. those who think that 𝑆𝐶𝑀𝐼𝑁 is true hold that, in each of these cases, one is aware that one oneself is in a certain state, which is, in turn, responsible for the subjective aspect of experience. in this sense consciousness is claimed to entail a minimal form of self-awareness. in such cases, my conscious experience does not merely convey to me the way the environment is, but also that i relate in a certain way with such an environment; namely, that i am in a certain relational state – a visual one in this case – to the mug. a proper understanding of this claim in the presented framework requires us to spell out in what sense the experience is supposed to concern or carry information regarding its subject; that is, we need to spell out the role that the subject plays in the adequacy conditions of the state. the distinction between awareness of oneself as-object and as-subject is a classical starting point for this task.7 consider the following thoughts i (miguel sebastián, ms) might form after looking at myself wearing a sweater in a mirror: 1. that there is a sweater. 2. that ms is wearing a sweater. 3. that i see a sweater. ones, such as being an apple (block, 2014; di bona, 2017; siegel, 2010). alternatively, rather than the mug and its properties one might think that the object of perception is a mug-ish, black-ish qualitative state or sense-datum. these debates are orthogonal to the discussion in this paper. likewise, i am not interested in an analysis of the state one’s experience represents oneself as being in. the answer will plausibly depend on what we take the object of perception to be. in this example, if we think that it is the mug and its properties, then it is reasonable to assume that the experience conveys that one is in a visual relationship to it. the reasoning offered in the rest of the paper intends to be neutral on this and the reader can adapt the proposal mutatis mutandis. moreover, the claim that awareness has an object seems uncontroversial in the case of perceptual experiences, but less so in the case of other conscious experiences, such as moods – dretske (1999), for example, argues that in this case the mind is the primary object. i will ignore this complication. nothing in my argument hinges on it as will be clear in the discussion. 6unpacking awareness in representational terms does not commit us to the (popular but controversial) claim that, at the fundamental level, conscious states are representational ones. consistent with this approach, conscious states could fundamentally be, for example, states of acquaintance (williford, 2015) or states that instantiate some sort of phenomenal property and that they represent in virtue of such instantiation (chalmers, 2004). 7this distinction can be traced back at least to kant, and it influenced william james’s (1890/2007) distinction between the i and the me. it also influenced husserl’s views, and with it the idea of pre-reflective self-consciousness in the phenomenological tradition. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 6 4. that i see me. 5. that i wear a sweater. with the exception of 1, all other thoughts involve awareness of oneself. nonetheless, we would deny that 2 involves self-awareness. the content of self-awareness is expressed by means of a first-person pronoun in direct speech. now, there are two ways in which one can be self-aware: as-subject and as-object. example 3 illustrates the first one. in this case, i am aware both of myself and of a sweater the primary object. however, there is a difference between the ways i am aware of them. there is a sort of intimacy with the subject of experience that is missing with the sweater, as revealed by the fact that i can be misguided with regard to whether it is a sweater what i am seeing, but not with regard to whether i am the one seeing it. this intimacy is also missing in self-awareness as-object (in example 5, 4 exemplifies both kinds). indeed, the kind of awareness involved in self-awareness as-object is similar to the awareness i have of the mug in the visual example. as we will see in some detail, the difference is that in self-awareness as-object i identify myself with the entity i am aware of, and hence there is room for misidentification. many would doubt that the experience i have when seeing a black mug or, even more clearly, when deeply immersed in watching a film takes me as an object – and that there is any entity that all my experiences take as object, as the visual experience of the example takes the mug. if these doubts are justified and, at least, 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 is to be plausible, then minimal self-awareness ought not to be characterized in terms of self-awareness as-object. this is of course insufficient to evaluate further disputes with regard to disruption of self-awareness in altered states of consciousness. my aim in the rest of the section is to address this and characterize the minimal form of self-awareness that is suitable to make 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 plausible. i will argue (section 2.1) that we should think of minimal self-awareness as perspectival first-personal, characterized by a semantic and an epistemological feature: i) indexicality corresponding to the fact that the content has to be expressed in direct speech with a sentence that deploys the first-person pronoun – recall that this does not demand that the subject has to possess the concepts required to express the content – and ii) immunity to error through misidentification (iem) corresponding to the sort of intimacy above mentioned. perspectival awareness might seem puzzling to some. in order to mitigate perplexity, i will present a different but well acknowledged form of perspectival awareness – the spatial-centric (section 2.1.3). i then explain how perspectival first-person awareness can accommodate straightforward objections against 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 (section 2.1.4). once the intended notion of minimal selfconsciousness is clarified, i discuss how pfp-awareness relates to other notions that have been used to characterize minimal self-awareness (section 2.2), and finally how to make sense of the idea that one can be more or less self-conscious (section 2.3). sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 7 2.1 perspectivalness: indexicality and iem as individuative features 2.1.1 indexicality the kind of experience i have when looking at the black mug conveys that i am visually related to the black mug. this experience is silent with regard to identity of the entity i happen to be identical with: it does not convey that ms is visually related to the black mug, despite the fact that i am ms. and it is silent with regard to any other property that might individuate the subject: it does not convey that property f is visually related to the mug either, even if i happen to be the only one that has the property. in line with these ideas williford nicely stresses that “it is a serious confusion to identify subjective character with one’s individuality or particularity” (2015, p. 1). moreover, we think that different individuals can have experiences with the same phenomenal character. this is an intuitive folk psychological assumption that a theoretical model should better not rule out. we want to make room for the possibility that (at least in some cases) when two subjects, 𝑆1 and 𝑆2, look at the mug they have the same kind of experience. if the experience of 𝑆1 depends on representing a certain entity beyond the mug (𝑆1), then 𝑆2 cannot have this experience. 𝑆2 will rather represent another entity (𝑆2) and this will entail a difference in their awareness. nor can it be a matter of representing a generic entity (a subject). 𝑆1’s experience does not convey that a subject is seeing the mug: her experience would not be correct if another subject, say 𝑆2, were seeing the mug. the plausible claim is that all one’s experiences commonly convey something that one would express deploying an essentially indexical expression, the firstperson pronoun: “i am in such-and-such a state”. if the hypothesis that conscious experiences constitutively make us aware of ourselves is to leave the ground, then such an awareness is better understood as being first-personal or de se (castañeda, 1966; chisholm, 1957; lewis, 1979; perry, 1979).8 my awareness that i am in such-and-such a state and that ms is in such-and-such a state are different even if i am ms. this is illustrated by the fact that i can have one without the other, if, for example, i ignore the fact that i am ms. we can capture the difference between them in terms of content – their adequacy conditions – with the help of a semantic theory. contentful mental states make partitions in the space of possibilities (stalnaker, 1999). in the case of non-indexical representations such partitions are taken to be among ways the world might be. and we typically make those partitions by attributing properties to objects. my awareness that ms is in such-and-such a state distinguishes worlds in which ms is in such-and-such a state and worlds in which he is not. it attributes the property of being in such-and-such a state to ms, 8for a detailed discussion of the relationship of de se phenomena to the experience see frank (2007), sebastián (forthcoming). sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 8 and it is adequate depending on whether ms has this property in the actual world. however, possible worlds seem insufficient to capture the adequacy conditions of de se representations. lewis argues that indexical representations do not determine partitions of possible worlds, but rather of centered worlds: they are adequate depending not only on the way the world is but also on an individual. inspired by lewis (1979), we can think of first-person representations as involving a selfattribution of a property – that of being in a such-and-such a state. and this kind of representation is correct if the subject has the property she self-attributes. if self-awareness is understood in this indexical way, the tension between 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 and the idea that 𝑆1 and 𝑆2 can have experiences of the same phenomenological kind disappears. their experiences can convey alike and we can capture that in terms of the self-attribution of the same property – that of being in such-and-such a state – to their respective experiences.9 2.1.2 immunity to error through misidentification (iem) unfortunately, the notion of self-attribution is insufficient to characterize minimal self-awareness, for it does not allow us to distinguish between a first-person awareness as-subject, and a first-person awareness as-object. in the previous section i appealed to a sort of intimacy between the subject of the experience and what they are aware of. such an intimacy can be unpacked in terms of the way in which the subject cannot be misguided. in general, when one predicates f-ness of certain object a – as when one is aware that a is f – there are two ways in which one can be wrong: one might 9“content” is a technical term. i have equated content with correctness conditions as lewis does. others would maintain that the content of my representation that i am in such-and-such a state and my representation that ms is in such-and-such a state have the same content, but that they involve respectively a different kind of state, sentential meaning, guise or mode of representation; different versions of this view have been defended by kaplan (1989; perry, 1979, 1980; richard, 1983; salmon, 1986, 1989). for current purposes we can abstract from these details. my impression is, in agreement with lewis, that this is a terminological dispute (cf. feit, 2008), but readers can adapt the claims to their favourite semantics for de se representation mutatis mutandis. for example, in recanati’s theory of perspectival representation (recanati, 2007, 2012), he uses the term “content” to refer to what is explicitly represented rather than to the correctness conditions. the correctness conditions are determined by the “content” and the mode of representation (associated with frege’s notion of “force”). he argues that a feature is implicitly represented when it is part of the mode rather than the content, because the information conveyed by a representation can only concern such a feature (what in turn is intended to explain iem – see below). the state is to be evaluated for correctness with regard to the subject of experience, but the subject is not part of its content, because (in his theory) the experiential mode guarantees that the content can only be evaluated relative to the subject of experience. however, he very clearly remarks that what is part of what he calls mode can be present in experience: [t]here is absolutely no reason to consider that phenomenology supervenes on explicit content. the mode also contributes to the phenomenology, since the mode is something the subject is aware of. (2007, pp. 141–142) sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 9 be wrong about whether a is really f, and one can be right about an object being f, but misidentify such object as a (coliva, 2006). call these two kinds of errors mispredication and misidentification respectively. we can think of the self-attribution in self-awareness as-object as involving two components: i) the attribution of a certain property to a certain object o, and ii) the identification of myself with o. when i come to believe that i wear a sweater after looking at the mirror, i can be wrong because the person reflected in the mirror does not wear a pullover or because, despite the fact that person wears a pullover, i am not this person. in the sort of self-awareness involved in experience there seems to be no room for misidentification (shoemaker, 1968). we can be wrong with regard to the state we attribute to ourselves in experiences (mispredication): one might, for example, be hallucinating rather than seeing a sweater in the case described above (cf. weisberg, 2011). but it cannot be the case that we self-attribute such state because we have misidentified ourselves with someone who is in that state. the kind of selfattribution in first-person representation as-subject does not involve an identification, it does not require to take any entity as an object of awareness and identify ourselves with such an entity.10 summarizing, i have argued that if 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 is to be made plausible, then our conscious experience involves an awareness that i) does not require the possession of the corresponding concepts by the subject; ii) has to be evaluated not only with regard to the way the world is, but also with regard to the very same individual who is doing the representation; and iii) cannot be evaluated with respect to any other entity. we can call this form of non-conceptual de se awareness as subject “core de se” or “perspectival first-person (pfp) awareness.”11 due to the lack of familiarity with some semantic tools used to spell out the correctness conditions of indexical de se or first-person representation and the phenomenon of iem, some readers might find pfp-awareness a bit mysterious. i will deal with this problem next (section 2.1.3), and then explain how perspectival first-person awareness can accommodate straightforward objections against 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 (section 2.1.4). 10shoemaker further argues that this identification-free form of representation is prior to any other way in which we can attribute a property to ourselves, because identification requires that there is some property or relationship that we had already ascribed to ourselves. 11a subject attributes property f to a when it represents that a is f. s’s self-attribution of property f in the absence of an identification is not s attribution of property f to s, as illustrated by the differential semantic and epistemological features that i have used to characterize the form of self-attribution presented in the paper – see castañeda (1966) for detailed discussion of the irreducibility of the de se phenomenon. on top of that a theory of what is required from a system to make such self-attribution (a metasemantic or naturalistic theory of such a core de se representation) is required. whole books and papers are devoted to discuss what is required from a system in order to attribute a property; i.e. what it takes to have a non-indexical representation – see for example artiga (2016), dretske (1981), martínez (2013), millikan (1984), neander (2017), papineau (1984), shea (2007), skirms (2010). doing something similar and defending such a view in the case of core de se representation is something that cannot be done here if the paper is to be kept to any reasonable length. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 10 2.1.3 perspectival spatial-centric awareness i want to deflate any suspicion that perspectival awareness might raise by presenting another form of representation that the reader might find more familiar, which is also perspectival – indexical and iem: spatial-centric representation. we can be aware of locations. we can think of the middle point of the plaza del zocalo in mexico city and wish we were there. if we visit the place, we can see that at this point there is a huge flag and we can attend to this location independently of the flag. in these cases a certain location, the center of the plaza, is an object of awareness. alternatively, we can walk to the center of the plaza and look at the environment. we can see, for example, that the presidential palace is to the right of the cathedral as seen from there. as in the previous cases, a certain location is also part of what is conveyed; it is part of the correctness conditions of the state. however, it doesn’t seem adequate to think of such location as an object of awareness – at least not in the same sense as the presidential palace is in the previous example. an adequate sentence that expresses the content of our visual awareness will deploy the essential indexical expression “here”. in the case of the experience i am currently having, this would be something like – setting the subject aside for the sake of simplicity – that a black mug is located to the right of the computer from here. it does not merely convey the way the world is, but the way it is from, or relative to, a certain location. we can call this sort of awareness de hinc to contrast it with de se awareness. in the lewisian framework presented above the content is a set of centered worlds where the center is a location rather than an individual. if you were located in a very different location instead – say madrid rather than mexico – but in front of an identical visual set up, we would probably have the same kind of experience, individuated by its phenomenal character. and you would express what is conveyed by your experience using a similar expression. indeed our experiences do not reveal any further feature about the location from which the awareness is to be evaluated for correctness. this representation does not depend on identifying a particular object-location, and it is then immune to the second sort of error (iem). if my visual experience represents that there is black mug to the left of a laptop from here, one might wonder whether there is really a mug or whether it is located where one takes it to be. what cannot happen is that there is another location from which the black mug is to the left of the laptop and i misidentify it as the pole of representation. we can call this form of de hinc representation “perspectival spatial-centric”. although it is uncontroversial that most, if not all, of our visual experiences involve a perspectival spatial-centric representation, this is hardly a constitutive feature of all conscious experience since nothing spatial seems to be involved in many ordinary experiences, such as the experience of happiness or the experience one has when one considers whether madrid is the capital of spain. however, in all those cases it seems that it is conveyed that i am in a certain state. 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 requires pfp-awareness. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 11 2.1.4 pfp-awareness in ordinary experiences some have rejected that 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 is true – and 𝑆𝐶𝑀𝐼𝑁 with it. one possible reason for that is to think that self-consciousness commits us to the existence of a metaphysically unacceptable entity, the self – as is the position of buddhism for example. however, the rejection of an entity like the self is no good reason for denying 𝑆𝐶𝑀𝐼𝑁 . first, consistently with the existence of self-consciousness, many have argued that this is derived from hallucinatory representation (where there is no such a thing as the self) or illusory representations (where the self does not have the properties it appears to have).12 and secondly, and more importantly, pfp-awareness does not give us any reason to postulate a mysterious entity, because it does not reveal any detail with regard to the individuality or particularity of the subject. a theory of mental content that spells out what is required from a system to entertain this sort of representational state will then show the metaphysical commitment of the theory. in its absence there is no reason to postulate any sort of mysterious entity, in the same respect as spatial-centric awareness does not lead us to postulate the existence of any mysterious location-entity (cf. fine, (2005). once we understand perspectival awareness it is simple to explain the mistake that those who fail to appreciate self-consciousness in our ordinary experiences tend to make. for example, some (e.g. prinz, 2012) have read hume’s famous inability to find himself in introspection as claiming that we are not self-aware even in ordinary experiences. for my part, when i enter most intimately into what i call myself, i always stumble on some particular perception or other, of heat or cold, light or shade, love or hatred, pain or pleasure. i never can catch myself at any time without a perception, and never can observe any thing but the perception. (hume, 1739, p. 252) however, there is no entity to be found in introspection because pfp-awareness is not a matter of being aware of an entity, as we are aware of the primary object of experience – the mug in our visual example.13 consider again de hinc awareness – a non-controversial feature of our visual experience, which tells us how the world surrounding us is relative to a certain location. there is no such location to be found in introspection, there is no “hereness” that we can find when we introspect the content of our visual experience. analogously, in pfp-awareness there is no self to be found in introspection, but this should not lead us to conclude that there 12for review and discussion see mcclelland (2019). 13an anonymous referee has suggested an interesting alternative explanation, compatible with objectual representation – as in, for example, woźniak (2018): the inability to recognize a special and (near-) ubiquitous feature. i see two problems with this explanation. first, as we have seen, there is no common object different individuals seem to be aware of. this seems to prevent, as we have seen, the possibility that they have experiences with the same phenomenal character – there will be a difference in awareness. secondly, it doesn’t seem to be fair to hume or prinz, as they do not typically fail to appreciate ubiquitous features (mcclelland, 2015). sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 12 is no self-consciousness. when we look for the subject in introspection we are unable to find it, in the very same sense as when we look in introspection for the here we do not find a privileged location: they are not objects of awareness in the sense that the black mug in our example is. a semantic model of perspectival representation, like the one presented above, can offer a characterization of the differential role that the subject (or the location) plays. it also explains how this is fully compatible with the claim that visual experiences convey how the objects are located from here or that all experiences convey that one oneself is in a certain state. up to this point, i have offered a characterization of pfp-awareness in semantic and epistemological terms and dispelled some worries that the notion might bring. in the next subsection, i discuss how pfp-awareness relates to other forms of selfconsciousness that have been claimed to be minimal, and hence suitable to make 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 true. 2.2 pfp-awareness as minimal self-consciousness gallagher (2000) has associated the minimal form of self-consciousness with the sense of agency and ownership. the sense of agency is a form of self-awareness present in states that convey that i am the one who is causing or generating an action – the sense that, for example, i am the one who is causing the door to close or generating thoughts in my stream of consciousness. it typically accompanies our conscious life, but not constitutively so, as there seems to be no sense of agency involved in the experience one has when one contemplates the environment, and it is lost in such cases as thought insertion (ibid.). the sense of ownership is presented as the sense that i am the one who is undergoing the experience and in this sense closely related to pfp-awareness. however, the use of the term “ownership” is confusing, for it suggests a reading corresponding to a state that conveys that a certain state is mine, rather than the intended sense that what is conveyed is that i am in a certain state. this has led philosophers such as metzinger (2003) to claim that the sense of ownership is lost in thought insertion as reflected by the fact that subjects claim that the thought is not theirs. however, their experience still conveys that they are in a certain state, that of thinking (campbell, 1999; zahavi & kriegel, 2016). some authors have also thought of some sense of embodiment or bodily selfawareness as being a basic form of self-awareness. for example, thompson (2015, p. 4) considers “the embodied self-experience of being alive in the present moment, or the experience of being sentient”, and wider (2006, p. 82) thinks of “a conscious, bodily orientation toward the world” as the most basic form of self-awareness (see also blanke & metzinger, 2009). metzinger (2013) maintains that being conscious of one’s location in egocentric space is the most basic form of self-awareness. although some form of bodily self-awareness and spatial-centric location does typically accompany our daily-life experience, there doesn’t seem to be any spatial or sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 13 bodily component in the phenomenal character of the experience one has when thinking, for example, that madrid is the capital of spain. such an experience is still subjective, and plausibly conveys that one is in a certain state. one possible reaction to the characterization of minimal self-consciousness as pfp-awareness is that this should not count as full-blown self-consciousness. it concerns oneself – something about the subject is conveyed, and the subject is part of the adequacy conditions of the state –, but oneself is not an object of awareness. one might consider that full-blown self-consciousness requires that oneself is an object of representation, as the primary object is, and not what we can call “a mere pole” or “perspective” – that is, a condition beyond the state of the world for the adequacy of the state. accordingly, if i am aware that i see the black mug, or that i hear the radio, i am self-conscious only in a weak sense; whereas if i am aware that i see myself, or when i identify myself with a certain entity, as happens for example when i am aware that i am an entity with certain boundaries or identical with certain body, i am self-conscious in the strong sense. often in the debate the strong or full-blown notion of self-consciousness seems to be assumed. for example, blanke and metzinger (2009, p. 8) think that the simplest form of self-consciousness requires “a globalized form of identification with the body as a whole”;14 millière (2019, p. 5) suggests that “states of self-consciousness may broadly speaking represent the same intentional object – the self”; letheby and gerrans (2017, p. 1) assess in this respect that we represent “the self as an entity which sustains interoception, affection, cognition, and perception […]”. i do not intend to contend for the most adequate use of the term “self-consciousness”. the point that i want to make, following my reasoning in the paper up to this point, is that 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 is plausible only in so far as we understand self-consciousness in the weaker sense. moreover, this weak understanding of self-consciousness as pfp-awareness is indeed the way to capture the idea behind such a claim. as zahavi and kriegel stress: our view is not that in addition to the objects in one’s experiential field—the books, computer screen, half empty cup of coffee, and so on—there is also a self-object. rather the point is that each of these objects, when experienced, is given to one in a distinctly first-personal way. (2016, p. 38, my emphasis) 14blanke and metzinger think that this form of identification can be really minimal and hence that they are not committed to the strong reading that the isolated quote might suggest (metzinger, personal communication). be that as it may, it is useful to compare a strong form of self-consciousness with what they call “weak 1pp” (blanke & metzinger, 2009, p. 7), which is closely related to pfp-awareness. according to the authors, such a weak 1pp is “not linked with theoretically more charged notions such as ‘subject of experience’ ” but rather is “a purely geometrical feature of an egocentric model of reality” (ibid.). however, a key difference is that weak 1pp so described involves a de hinc perspective rather than de se one. certainly, for all i have said and despite my doubts, it is an open possibility that a de hinc representation could ground a de se one. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 14 i have argued that if 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷 is to be plausible, then minimal selfconsciousness has to be understood in terms of pfp-awareness. however, pfp-awareness does not seem to admit degrees. so, before moving into the discussion of the empirical evidence from altered states, we need to make sense of the idea that self-consciousness can be partially impaired. this is the purpose of the next subsection. 2.3 degrees of self-consciousness if we are to assess the possibility that self-consciousness can be disrupted but not fully so, we need to understand to what extent it makes sense to say that one can be more or less self-conscious. self-consciousness in the weak sense does not admit degrees. as i have characterized it, awareness is either pfp or it is not. this is in accordance, if 𝑆𝐶𝑀𝐼𝑁 is to be true, with the view that consciousness is not gradable (bayne, hohwy, & owen, 2016). it is not clear that the strong notion admits of degrees either. millière (2019) argues that the most plausible way in which we can make sense of degrees of (strong) self-consciousness is in terms of a scaffolding model, such that some ways in which we are aware of ourselves are more fundamental than others. he considers bodily self-awareness and spatial self-awareness as the most plausible potential candidates to occupy the fundamental level, and presents a detailed collection of experiences in which both are missing to conclude that we should abandon a scaffolding model of self-consciousness, in favor of a multidimensional one. i partially agree with this result. i think that the multidimensional model is adequate for strong selfconsciousness. this is consistent with the claim that there is a fundamental form of self-awareness: pfp-awareness. in pfp-awareness we self-attribute a property in the absence of an identification, and this form of self-awareness grounds – if shoemaker is right (see footnote 10) – any stronger sense in which we are aware of ourselves as an entity that satisfies certain features. by attributing other features to ourselves, we end up with an image of the self, an image that corresponds to a singular entity that has all of one’s experiences across time, that possesses a body and occupies a certain location in space and time. and probably other features like being immaterial and irreducible to neural processes, having free-will, initiating our conscious actions, having a history recorded in one’s biographical memory, or an essential personality consistent across times and contexts, etc.15 when we have an experience that conveys something that conflicts with the model we have of ourselves – i.e., with the properties that we attribute to ourselves – then a report of disruption of the self is likely to be produced as a result of the feeling that some feature of oneself is missing. this is important to make sense of the subject report as we are about to see. 15for a recent review of properties attributed to the self in the literature see mcclelland (2019). sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 15 3 altered states if 𝑆𝐶𝑀𝐼𝑁 is not a priori justified, but based on the phenomenological observation that there is a subjective aspect, a first-person perspective that accompanies all our ordinary experiences, then it is open to empirical falsification (even if we accept 𝑆𝐶𝑀𝐼𝑁−𝑂𝑅𝐷): is consciousness possible in the absence of this minimal form of self-consciousness? although many of us might not be able to make sense of this idea, it is intellectually enriching to keep the mind open. altered states, conscious episodes that differ some way or other from our ordinary waking experience, might illustrate what it might be like to have those experiences. and reports of these experiences have been interpreted as showing that 𝑆𝐶𝑀𝐼𝑁 is false. i think that this conclusion is not justified for two reasons. first, there is a misunderstanding of the minimal form of self-awareness involved: although there are good reasons to think that forms of self-awareness that typically accompany our ordinary experiences are compromised in altered states, there are reasons to doubt that pfp-awareness is missing. secondly, there are indeed good reasons to mistrust the subject’s report due to the confirmation bias derived from their expectations and their cosmological and metaphysical views. i will start by presenting the case of dreamless sleep, which will allow me to present the minimal form of self-attribution. then i will illustrate these two reasons in discussing the evidence from studies of meditative and drug induced states. 3.1 dreams according to some eastern traditions – the advaita vedanta and yoga – consciousness persists during dreamless sleep. to motivate this view, thompson (2015) brings back a classical philosophical debate between the advaitins and the nyayayikas. when we wake up in the morning we are not surprised by the arrival of consciousness and we have the feeling of having slept peacefully. advaitins argue that this cannot be the result of an inference and that it is rather the result of a state of consciousness lacking any particular thoughts or images. if we accept the argument we have to acknowledge a kind of experience that is radically different from waking experiences.16 but does this jeopardize 𝑆𝐶𝑀𝐼𝑁 ? a positive answer seems to be suggested by thompson, who remarks that when we sleep deeply we “do not cognize anything – there is no object to be known and no awareness of oneself as knower” (p. 5), and by windt (2015), who insists that this kind of experience is characterized by an absence of “the subject of experience, or the ‘i’.” (p. 15). however, the argument above offers no good reason to think that dreamless sleep experiences have to be characterized by an absence self-consciousness. windt et al. (2016) clarify that what these experiences lack is the sense of agency 16see also windt, nielsen, & thompson (2016) for speculative and thought-provoking discussion of empirical evidence from dream studies in favour of this sort of episode. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 16 or the sense of being a potential possessor of knowledge (metzinger, (2013), and any sense of self-other distinction. but as we have already seen none of these features are constitutive of pfp-awareness. thompson, on the other hand, proposes that the state of dreamless sleep should be characterized by a minimal form of sentience consisting of the feeling of being alive. i honestly fail to appreciate this feeling of being alive during sleep. this might well be a failure on my side and this sort of phenomenological disputes is difficult to adjudicate. however, as mcclelland (2015) stresses, one should not attribute to the opponent a kind of mistake that she is not disposed to make in contexts other than the assessment of the phenomenological claim under dispute. and i do appreciate the feeling of being alive upon awakening (cf. section 2.1.4). i think that alleged dreamless sleep experiences can be better characterized by the feeling of emptiness or nothingness, fully consistent with one’s report of peacefulness and absence of knowledge – and consistent with a feeling of duration (windt, 2015). there is an ambiguity in the characterization of this state crucial to the debate at hand. this can be understood as a state that conveys that there is emptiness or nothing (jeopardizing the truth of 𝑆𝐶𝑀𝐼𝑁 ); or as a state that conveys that i am in no state or at least in a contentless state, a state that says nothing (fully compatible with the truth of 𝑆𝐶𝑀𝐼𝑁 ). the first reading is closer to vedanta tradition which involves the absence of any form of self-awareness. one problem with this proposal is that, as there is no feeling of ignorance belonging to an “i”, an explanation of the first-person report upon awakening is still required. advaitins reply that upon awakening the ego sense is again operative and “appropriates the lingering impression or retention of not-knowing and associates this retention with itself, thereby generating the retrospective thought” (thompson, 2015, p. 10). however, a straightforward explanation of the formation of an episodic memory that explains one’s report that “i slept well and i did not know anything” is available in the second case.17 this absolute minimal form of self-ascription, where i selfattribute the property of being in a contentless state, is then consistent both with the existence of pfp-awareness and what can be characterized as undergoing an experience of nothingness or emptiness. independently of the strength of this abductive reasoning, the important thing to note for current purposes is that there is no reason for thinking of dreamless sleep experiences as experiences in which there is no self-awareness – once this is properly understood as pfp-awareness – because the alternative explanation that i have offered is available. moreover, if the proposed absolute minimal form of self-ascription can adequately characterize the experience of dreamless sleep, then it might also characterize some alleged selfless states of meditation. indeed, sankara, probably the most famous of all vedanta teachers, stresses the similarities 17my speculative proposal offers a compromise between the advaitins’ claim where there is no self-awareness and the ideas in the yoga tradition according to which deep sleep is reduced state of mind which includes the self-as-object – the feeling of presence of the entity we believe we are (ahamkara) with our habits, prejudices, desires, impulses, etc. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 17 between dreamless sleep and samadhi – the highest state in meditation –, when he remarks that there is a “natural eradication of difference in deep sleep and in samadhi” (sankara, 1980, p. 365; quoted in comans, 1993). 3.2 meditation it has been claimed that one of the mechanisms of action in meditation is related to an altered sense of self (hoelzel et al., 2011). indeed, there is evidence showing that the activity of brain networks linked to the narrative self and self-related activity is influenced – for a review of evidence see dor-ziderman, berkivich-ohana, glicksohn, & goldstein (2013). this has made of mindfulness studies an interesting field to study the possibility of conscious experience in the absence of self-consciouness. and recent results might be read as pointing in this direction. dor-ziderman et al. (2013) recorded, using magnetoencephalography (meg), the brain activity of 16 meditators and then asked the participants to – freely and in their own words – describe their experience. the experimenters were interested in three conditions associated with the narrative self, the minimal self, and a selfless experience, and gave participants instructions to guide their meditation accordingly: “try to think what characterizes you”, “try to experience what is happening to you at the present moment”, and “try to experience what is happening at the present moment, when you are not in the center.” they categorize the reports of meditators in the last condition into three groups. the most interesting for current purposes is the one that they labeled “lack of ownership”, where subjects “reported experiencing what was happening, only with the sense of agency/ownership absent” (dor-ziderman et al., 2013, p. 6). four participants, the most advanced ones, offered this kind of report, which is difficult to interpret. as we have seen, pfp-awareness is orthogonal to the sense of agency and, depending how it is understood, ownership. although the reports suggest a strong deflation of self-consciousness, there is no clear absence of pfp-awareness. in what i take to be the most significant report in the study, the subject said: “there was an experience but it had no address, it was not attached to a center or subject.” this suggests a loss of pfp-awareness. but then she clarifies: “it was not 100%, but there was no sense of an object there running the show. emptiness is the best word”, suggesting rather a disappearance of the self-as object, and fully consistent with the minimal self-attribution presented in the case of dreamless sleep. in eastern tradition samadhi, the most advanced state in meditation, is described as a feeling of union with the universe or the divine. according to hinduism, in samadhi, the atman (self, soul) frees itself of habits, prejudices, desires, impulses (ahamkara) and reveals itself as being part of, or participating in, the universe or the divine (brahman). it is in this sense that the fundamental distinction subject-object or self-other is claimed to be obliterated (comans, 1993). the feeling of union with the universe might entail a disruption of the distinction between self and other, but this distinction is not something revealed in any sense in pfpsebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 18 awareness, and hence is not a challenge to 𝑆𝐶𝑀𝐼𝑁 . the experience i have when i look at the mug (other), or when i look at my chest or at myself in a mirror (self) involves in the very same respect pfp-awareness. it is true that the latter might additionally be accompanied, at least typically, with some additional feeling that it is me who i am looking at. however this is not the kind of minimal self-awareness appropriate to make 𝑆𝐶𝑀𝐼𝑁 true.18 there is one further reason to doubt that current empirical evidence from meditation studies should lead us to think that 𝑆𝐶𝑀𝐼𝑁 is false. meditation studies have focused on buddhist practices (lutz, dunne, & davidson, 2007). in buddhist traditions there is no such a thing as the self, which is a mere illusion of our minds. buddhism rejects the foundational premise of the vedas and upanishads that “atman exists”, accepted by all major orthodox schools of hinduism – mimamsa, nyaya, samkhya, vaisesika, vedanta, and yoga. the root of human suffering derives from our identification with such an illusory entity and meditation is a way to learn this by reaching a selfless mode of experience where the identification with the self is abandoned (e.g., dalai lama, 1991). samadhi in buddhism is often described as a selfless experience. this is a problem, because we all are prone to a confirmation bias, a well established tendency to store, recall, and interpret information in a way that confirms one’s preexisting beliefs or hypothesis. are the most advanced meditators in buddhism and hinduism having different kinds of experiences or are they offering different descriptions of an otherwise similar experience? i am not calling into question that advanced meditators undergo experiences whose phenomenology is indeed different from ordinary waking experiences, nor that their experiences differ from less experienced ones – as their reports and neuroimages suggest. the claim is rather that there are good reasons to think that the information stored and recalled from their experiences as well as the interpretation of these memories as reflected in their reports is likely to be influenced by their beliefs. a control for cognitive bias in the experiments is required. considering a distinctive form of meditation might contribute to advance the discussion. the subject-object distinction is claimed to be disrupted in a more profound sense in very advanced states in non-dualistic forms of meditation, such as open-presence (josipovic, 2010, 2019; millière, carhart-harris, roseman, trautwein, & berkovich-ohana, 2018). the state of open-presence is the main meditative state that practitioners of chag-zôg style attempt to cultivate (lutz et al., 2007). the aim of the practice is to understand the nature of the conscious experience, by focusing on the invariant feature of all experiences, hence essential to consciousness. according to chag-zôg theorists, what is constitutive is a reflexive form of awareness: “as an object is being presented to an experiencing subject, reflexive awareness also 18additionally, some have claimed that minimal forms of self-awareness require the ability to distinguish self and non-self (damasio, 2000; gallagher, 2000). but this is a metasemantic or metaphysical claim. 𝑆𝐶𝑀𝐼𝑁 is a phenomenological claim, one about what the experience conveys, and this has to be carefully distinguished from the metaphysical and metasemantic claims with regard to what is required from reality or from a system in order to be able to entertain states that convey what phenomenology reveals. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 19 presents the process or occurrence of that experience”19 (ibid., p. 512). following a series of steps the object and the subject are de-emphasized, making it possible to reach the most advanced state in which there is no object nor subject of awareness, just reflexive awareness. but there are reasons to doubt that such a state is really possible. first, as lutz and colleagues stress (p. 511), the practice itself presupposes what is at stake in this debate, namely that the invariant elements of experience are selfless. only very few practitioners in each generation recognize having truly reached this level, and the most advanced ones are likely to characterize the differential phenomenology of their experience in the terms prescribed by the tradition. second, the aim of the practice is to let the experience continue even though “one de-emphasizes the particularity of the object and subject” (p. 515). in looking for invariants, the practice aims at abstracting away from the differences with regard to what is accidental about the object – the difference in the experience one has when looking at a cup and at a computer – but also what is accidental about the subject “for example, its temporal location in the narrative of personal identity or the particular emotional state that is occurring within the subjectivity” (p. 512). however, those elements, that indeed change from experience to experience, might contribute to the selfas-object, but not to pfp-awareness – which is, according to those who maintain that 𝑆𝐶𝑀𝐼𝑁 is true, an invariant element of experience. indeed, it is not clear that this element is missing, for a sustained state of open-presence is one in which one may “continue to experience phenomena without objectifying them and, ideally, without having a sense of an agentive or narrative subjectivity” (p.515). this all is consistent with there being a first-personal view that is characterized in terms of pfp-awareness. 3.3 drugs psychedelic experiences are sometimes characterized by distortions of selfconsciousness (huxley, 1954; leary, metzner, & alpert, 1964; lebedev et al., 2015), and several studies using classical psychedelic drugs (5-ht2𝐴 receptor partial agonist), such as psilocybin, mescaline, lsd (lysergic diethylamide) or dmt (dimethyltryptamine), as well as other psychoactive substances, such as ketamine, have observed a reduction of self-awareness in comparison with ordinary experiences – see millière (2017; millière et al., 2018) for a recent review. on this basis, it has been claimed that the study of psychedelic experiences offers evidence against 𝑆𝐶𝑀𝐼𝑁 (letheby & gerrans, 2017; nour, evans, nutt, & carhart-harris, 2016; and especially millière, 2017). i think that this conclusion is unjustified for the very same reason as in the case of meditation. first, there are good reasons to think that trip reports are subject to a similar cognitive bias as the one discussed in the case of meditation; secondly, a sharp distinction between pfp-awareness 19this recalls what i have called elsewhere “mental-state involving” (sebastián, 2012) developments of pre-reflective self-consciousness as in kriegel (2009; williford, 2006, 2015). sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 20 and other forms of self-awareness, together with a proper understanding of the former, offer the tools to make sense of the subjects’ reports – in a way that is also consistent with the formation of autobiographical memories on the basis of these allegedly selfless episodes – without any need to give up on 𝑆𝐶𝑀𝐼𝑁 . early research on mescaline before the popularization of psychedelics focused on perceptual hallucinatory experiences (kluever, 1928), with few mentions of absence of self-consciousness and rather mentioning bizarre body experiences, detachment from one’s body and mental states or the feeling of merging with one’s surroundings (mayer-gross & stein, 1926). the term “psychedelics” was coined by humphrey osmond in correspondence with aldous huxley to refer to a drug induced altered state of consciousness, in which “the mind” (psyche) is “revealed” (delos). huxley is indeed the most important player in the popularization of these drugs, not only through his literary work but through his intense social life promoting its use among intellectuals and artists (stevens, 1998). huxley was greatly influenced by eastern religions (huxley, 1945), and the resemblance between his descriptions of trips to samadhi should be of no surprise. although in the doors of perception, huxley considers the feeling of unity with the environment as the “final state of egolessness”, and this experience doesn’t seem to entail the absence of minimal self-awareness as pfp-awareness, later development in the dissemination of psychedelics might help to explain the view that psychedelic experiences refute 𝑆𝐶𝑀𝐼𝑁 . huxley introduced the tibetan book of the dead to timothy leary, which influenced the well known manual for lsd trips (leary et al., 1964). leary and colleagues considered the tibetan book of the dead to be a key for “those who are seeking the spiritual path of liberation” (p. 11) that huxley thought psychedelics were aimed to offer. they construed the effect of lsd finding parallels between the stages of death and rebirth in it and the stages that leary had identified during his research (gould, 2007), and defined ego-dissolution as “[…] complete transcendence – beyond words, beyond space – time, beyond self. there are no visions, no sense of self, no thoughts. there are only pure awareness and ecstatic freedom […].” despite the criticism in religious studies of the idea that a common self-less experience is a basis of buddhism (mcmahan, 2008, ch. 1), and the connection between ego-dissolution and psychedelics with tibetan buddhism (reynolds, 1989), such a connection remains popular in the public image. hence, reports of ego-dissolution by drug users might be the result of their expectations rather than a pure reflection of their phenomenology. this alternative hypothesis finds support in anthopological studies – as the difference between the reports of hindu and buddhist meditators evidences a possible cognitive bias in meditation studies. psychedelic drugs have been traditionally consumed by communities with different cosmological views outside western culture. for example, psilocybin mushrooms have been traditionally used in mesoamerica (wasson & wasson, 2011), as well as the mescaline containing peyote (la barre, 2011); ayahuasca has been used in amazonia (chaumeil, (1998) and anadenanthera seeds in south america (torres & repke, 2006), both containing sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 21 dimethyltryptamine. anthropological studies show that drug experiences in these communities are all about meeting spirits and getting power and knowledge from them, without reference, to the best of my knowledge, to ego-dissolution or mystical union with the universe.20 the influence of huxley’s work is also present in scientific studies. one of the tools most widely used to measure the features of altered states is the 5d-asc (3d-oav+2d) questionnaire (dittrich, lamparter, & maurer, 2010), consisting of 94 descriptions in which the participants have to rate how adequately they describe their experience. the descriptions are intended to measure five dimensions of the experience, two of which, oceanic boundlessness and anxious ego-dissolution, are analogous respectively to the “heaven” and “hell” aspects of huxley’s mescaline account (dittrich, 1998). importantly, although some of the descriptions entail a loss of self-consciousness, they do not indicate a loss of pfp-awareness. indeed, studerus and colleagues have shown that the descriptions in the 5d-asc are better clustered in eleven dimensions (studerus, gamma, & vollenweider, 2010, fig. 1), of which the two dimensions related to loss of self-awareness are associated with a sense of unity21 – which as we have seen huxley associated with ego-dissolution, and loss of the sense of embodiment.22 none of them jeopardizes 𝑆𝐶𝑀𝐼𝑁 . an interesting complementary set of questions to 5d-asc, intended to track ego-dissolutions in particular, has been recently validated by nour et al. (2016). unfortunately, it is also subject to significant deficits considering current purposes. the “ego-dissolution inventory” they present includes descriptions that seem to be irrelevant to evaluate the absence of pfp-awareness (“i felt a sense of union with others”, “i experienced a decrease in my sense of self-importance”, “i felt far less absorbed by my own issues and concerns”, nour et al., 2016, p. 3, table 1). their statistical validation considers a unique phenomenon and cannot distinguish the possibility of ego dissolution in two different dimensions (as-subject and asobject). in fact, there is no significant difference with regard to how effectively they track the studied condition of ego dissolution – as measured by spearman’s rank correlation coefficient (r)23 – between descriptions that concern the self-asobject (“i felt at once with the universe”, r=0.830; “i experienced a disintegration of my ‘self’ or ego”, r=0.897) and descriptions that are candidates to track the selfas-subject (“i lost all sense of ego”, r= 0.883). moreover, it is far from clear that such descriptions have to be interpreted as showing an absence of pfp-awareness. the description “i experienced a disinte20i am deeply grateful to martin fortier for his support with the review of the anthropological literature regarding drug use in mesoamerica, the amazon basin and south america. 21“everything seems to unite into oneness”, “it seemed to me that my environment and i were one”, “i experienced a touch of eternity”, “conflicts and contradictions seem to dissolve”, “i experienced past, present and future as oneness” (nour et al., 2016, p. 9). 22“it seemed to me as though i did not have a body anymore”, “i had the feeling of being outside of my body”, “i felt as though i was floating” (nour et al., 2016, p. 9). 23the spearman’s rank correlation coefficient measures the strength of a link between two sets of data. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 22 gration of my ‘self’ or ego” (ibid.) might correspond to an experience that conveys that me, the entity i am, disintegrates. but pfp-awareness cannot be a matter of being aware of a particular entity as we have seen. likewise, the description “i lost all sense of ego” (ibid.), might correspond to an experience that conveys to its subject that they themselves are in a certain state, namely one in which there is no self-awareness in the strong sense described in section 2.2. so, the report is consistent with an experience where self-awareness is disrupted but pfp-awareness remains unimpaired. furthermore, this alternative interpretation has the advantage that it can not only accommodate such reports but also the willingness to self-attribute, in a seemingly non-inferential way, the experience. millière (2017) – see also millière et al. (2018) – argues that the possibility of total disruption of self-consciousness is suggested by open descriptions of psychedelic episodes such as those offered in erowid.org. he remarks that in these narrative reports, some drug users are reluctant to use the first-person pronoun when describing their experience, suggesting that they lack a “first-personal aspect altogether” (millière, 2017, p. 14). as a paradigmatic example, he presents a description of an experience induced by the ingestion of psilocybin mushroom where the subject reports that in their experience “there existed no one, not even me, just one!” (ibid.). this seems to rather describe a typical case in psychedelic experience involving ego-dissolution in which there is a feeling of unity with one’s surroundings that is related to the disruption of ego-boundaries and hence irrelevant to pfp-awareness. it is from this disruption that the subject concludes that “yet thought continued, so would it be proper to still speak of ‘i’ even as the notion of ‘i’ seemed palpably illusory?” (ibid.). millière interprets this as a total absence of first-person perspective, whereas i am suggesting that that the report reflects the perplexity of experiencing the dissolution of the subject-object duality, as a consequence of the feeling that the entity we identify ourselves with merges with the environment into “one.” this description is indeed very close to the hindu’s description of samadhi, and as we have seen it is fully compatible with the truth of 𝑆𝐶𝑀𝐼𝑁 once the minimal form of self-consciousness is understood as pfp-awareness. further empirical data are required to decide which interpretation is the right one. moreover, in the absence of the corresponding control, we cannot decide if these reports are the result of a cognitive bias rather than a straightforward report of the phenomenology.24 24caution here is advised. my reasoning in this paper by no means entails that introspective reports cannot be trusted in general. we know that we are not infallible with regard to our phenomenology. this does not mean that introspective reports cannot be trusted – consciousness studies would hardly be possible otherwise. as kriegel (2015) argues, methodologically, conscious reports should be trusted unless we have good reasons not to do so. i have argued that there are good reasons to think that the reports of the subjects considered in the experiments are due to cognitive bias. therefore such reports should not be trusted in the absence of the corresponding control. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 23 4 conclusions many authors are sympathetic to the idea that a minimal form of selfconsciousness is constitutive of consciousness. altered states of consciousness have been presented as an empirical challenge to this view. in this paper, i have argued that such a minimal form of self-consciousness – one suitable to make it plausible that, at least, all our ordinary experiences entail self-consciousness – has to be understood as a non-conceptual identification-free de se awareness (a core de se representation that characterizes pfp-awareness). furthermore, i have argued that the conclusion that consciousness does not entail a minimal form of self-consciousness cannot be derived from current empirical research on altered states for two reasons. first, most of the reports are consistent with experiences where pfpawareness is not compromised, and more evidence is required to decide how the reports should be interpreted. secondly, there are good reasons to suspect that these reports are the result of a cognitive bias derived from the buddhist belief that there is no self in meditation, and the influence of these ideas in the public imagination of psychedelic experiences. however, i do think that altered states of consciousness open an interesting door to test the claim at stake. this requires addressing the flaws of current evidence by: 1. validating new questionnaires that are sensitive to the difference between pfp-awareness and other forms of self-awareness. one of the main challenges for empirical research on ego-dissolution, as we have seen, is that it is hard to determine what subjects are really reporting. one possibility might be to explain to the subjects what pfp-awareness is and ask them explicitly about the presence of this feature in their experience. if we accept that pfp-awareness is present in our ordinary experiences and the aim is to test whether it is a constitutive feature appealing to altered states, ordinary experiences can be used to validate the subject’s understanding or the adequacy of any description suitable to track the absence of pfp-awareness, such as “i lost all sense of ego.” for example, some subjects might tend to think that the offered description is satisfied by an immersive experience such as the one one might have while watching an engaging film. this would show that this is not an adequate description to capture pfp-awareness. experiences where strong forms of self-awareness tend to vanish, such as those involving immersion, and high demands of working memory can be used in validation moreover, pfp-awareness does not admit degrees unlike other forms of selfconsciousness. this can also be used to validate the understanding of the subjects or the suitability of the description used. we can control for the confidence of subjects in their reports. those who have a high degree of sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 24 confidence should rank the adequacy of the description only with extreme values – a three point scale might be preferred to a seven point likert scale for this purpose. alternatively, other methodologies such as open interviews with subjects who understand the difference between pfp-awareness and other forms of self-consciousness after drug use or meditation are of use. more details have to be considered to refine the adequate experimental setup – some of them beyond my expertise. however, i see no reason to think that it is not possible to determine experimentally the form of selfconsciousness whose absence the subject is reporting. 2. include controls regarding preexisting cultural beliefs and expectations to avoid cognitive bias. for this purpose, experiments with non-buddhist meditators in meditation studies, and with non-western influenced subjects in drug studies could be performed. anthropological studies and open descriptions of psychedelic episodes by non-western users are particularly interesting. there are many things to be learned from the empirical study of altered states with regard to the nature of our conscious experience. it seems that, at least for the moment, the possibility of a radical dissolution of self-consciousness is not one of them. acknowledgments for discussion and comments i am very grateful to kirareset barrera, josé luis díaz, martin fortier, rubén sebastián, to my students at the scc-philab and the participants at the 5th international workshop in cognitive sciences (uaemor). i am specially grateful to two anonymous referees and the editors of this special issue, thomas metzinger and raphaël millière, for their thoughtful comments and suggestions. financial support for this research was provided by dgapa projects ia400218 and ig400219, and the spanish ministry of science, innovation and universities via research grant pgc2018-095909-b-100. references artiga, m. (2016). teleosemantic modeling of cognitive representations. biology and philosophy, 31(4), 483–505. https://doi.org/10.1007/s10539-016-9525-3 bayne, t., hohwy, j., & owen, a. m. (2016). are there levels of consciousness? trends in cognitive science, 20(6), 405–413. https://doi.org/10.1016/j.tics.2016.03.009 blanke, o., & metzinger, t. (2009). full-body illusions and minimal phenomenal selfhood. trends in cognitive science, 13(1), 7–13. https://doi.org/10.1016/j.tics.2008.10.003 block, n. (2014). seeing-as in the light of vision science. philosophy and phenomenological research, 89(3), 560–572. https://doi.org/10.1111/phpr.12135 campbell, j. (1999). schizophrenia, the space of reasons and thinking as a motor process. the monist, 82, 609–625. https://doi.org/10.5840/monist199982426 castañeda, h.-n. (1966). “he”: a study in the logic of self-consciousness. ratio, 8, 130–157. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s10539-016-9525-3 https://doi.org/10.1016/j.tics.2016.03.009 https://doi.org/10.1016/j.tics.2008.10.003 https://doi.org/10.1111/phpr.12135 https://doi.org/10.5840/monist199982426 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 25 chalmers, d. j. (2004). the representational character of experience. in b. leiter (ed.), the future for philosophy (pp. 153–181). new york: oxford university press. chaumeil, j. p. (1998). ver, saber, poder. chamanismo de los yagua de la amazonia peruana. lima: ifea/caaap/caea-conicet. chisholm, r. (1957). perceiving: a philosophical study. cornell: cornell university press. coliva, a. (2006). error through misidentification: some varieties. the journal of philosophy, 103(8), 403–425. https://doi.org/10.5840/jphil2006103824 comans, m. (1993). the question of the importance of shamadi in modern and classical advainta vedanta. philosophy east and west, 43(1), 19–38. https://doi.org/10.2307/1399467 dalai lama, x. (1991). path to bliss. ny: snow lion. damasio, a. (2000). the feeling of what happens: body and emotion in the making of consciousness (1st ed.). boston: mariner books. di bona, e. (2017). towards a rich view of auditory experience. philosophical studies, 174, 2629–2643. https://doi.org/10.1007/s11098-016-0802-4 dittrich, a. (1998). the standardized psychometric assessment of altered states of consciousness (ascs) in humans. pharmacopsychiatry, 31(2), 80–84. https://doi.org/10.1055/s-2007-979351 dittrich, a., lamparter, d., & maurer, m. (2010). 5d-asc: questionnaire for the assessment of altered states of consciousness. a short introduction. zurich: psin plus. dor-ziderman, y., berkivich-ohana, a., glicksohn, j., & goldstein, a. (2013). mindfulness-induced selfess: a meg neurophenomenological study. frontiers in human neuroscience, 7(582). https://doi.org/10.3389/fnhum.2013.00582 dretske, f. (1981). knowledge and the flow of information. cambridge: mit press. dretske, f. (1995). naturalizing the mind. cambridge: mit press. dretske, f. (1999). the mind’s awareness of itself. philosophical studies, 95(1-2), 103–124. https://doi.org/10.1023/a:1004515508042 feit, n. (2008). belief about the self: a defense of the property theory of content. oxford university press. fine, k. (2005). tense and reality. in k. fine (ed.), modality and tense. new york: oxford university press. frank, m. (2007). non-objectal subjectivity. journal of consciousness studies, 14, 152–173. retrieved from https://www.ingentaconnect.com/content/imp/jcs/2007/00000014/f0020005/art00008 gallagher, s. (2000). philosophical conceptions of the self: implications for cognitive science. trends in cognitive sciences, 4(1), 14–21. https://doi.org/10.1016/s1364-6613(99)01417-5 gould, j. (2007). can’t buy me love: the beatles, britain, and america. new york: crown publishing group. guillot, m. (2016). i, me, mine: on a confusion concerning subjective character. review of philosophy and psychology, 8(1), 23–53. https://doi.org/10.1007/s13164-016-0313-4 heck, r. g. (2002). nonconceptual content and the “space of reasons”. philosophical review, 109, 483–523. https://doi.org/10.1215/00318108-109-4-483 hoelzel, b. k., lazar, s. w., gard, t., schuman-olivier, z., vago, d. r., & ott, u. (2011). how does mindfulness meditation work? proposing mechanisms of action from a conceptual and neural perspective. perspectives on psychological science, 6(6), 537–559. https://doi.org/10.1177/1745691611419671 hume, d. (1739). a treatise of human nature. oxford: oxford university press. huxley, a. (1945). the perennial philosophy. new york: harper; brothers. huxley, a. (1954). the doors of perception. new york: harper; brothers. james, w. (1895). the principles of psychology. henry holt. james, w. (2007). the principles of psychology. vol. 1. https://doi.org/10.1016/j.concog.2010.03.016 (original work published 1890) josipovic, z. (2010). duality and nonduality in meditation research. conciousness and cognition, 19, 1119–1121. https://doi.org/10.1016/j.concog.2010.03.016 josipovic, z. (2019). nondual awareness: consciousness-as-such as non-representational reflexivity. progress in brain research, 244, 273–298. https://doi.org/10.1016/bs.pbr.2018.10.021 kaplan, d. (1989). demonstratives. in j. almog, j. perry, & h. wettstein (eds.), themes from kaplan (pp. 481–563). new york: oxford university press. kluever, h. (1928). mescal, the divine plant and its psychological effects. london: kegan paul, trench, trubner; co. kriegel, u. (2009). subjective consciousness: a self-representational theory. new york: oxford university press. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.5840/jphil2006103824 https://doi.org/10.2307/1399467 https://doi.org/10.1007/s11098-016-0802-4 https://doi.org/10.1055/s-2007-979351 https://doi.org/10.3389/fnhum.2013.00582 https://doi.org/10.1023/a:1004515508042 https://www.ingentaconnect.com/content/imp/jcs/2007/00000014/f0020005/art00008 https://doi.org/10.1016/s1364-6613(99)01417-5 https://doi.org/10.1007/s13164-016-0313-4 https://doi.org/10.1215/00318108-109-4-483 https://doi.org/10.1177/1745691611419671 https://doi.org/10.1016/j.concog.2010.03.016 https://doi.org/10.1016/j.concog.2010.03.016 https://doi.org/10.1016/bs.pbr.2018.10.021 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org miguel ángel sebastián 26 kriegel, u. (2015). the varieties of consciousness. new york: oxford university press. la barre, w. (2011). the peyote cult. norman: university of oklahoma press. leary, t., metzner, r., & alpert, r. (1964). the psychedelic experience. a manual based on the tibetan book of the dead. new york: citadel. lebedev, a. v., loevden, m., rosenthal, g., feilding, a., nutt, d., & carhart-harris, r. l. (2015). finding the self by losing the self: neural correlates of ego-dissolution under psilocybin. human brain mapping, 36(8), 3137–3153. https://doi.org/10.1002/hbm.22833 letheby, c., & gerrans, p. (2017). self unbound: ego dissolution in psychedelic experience. neuroscience of consciousness, 2017(1), nix016. https://doi.org/10.1093/nc/nix016 levine, j. (2001). purple haze: the puzzle of consciousness. new york: oxford university press. lewis, d. (1979). attitudes de dicto and de se. philosophical review, 88(4), 513–543. https://doi.org/10.2307/2184843 lutz, a., dunne, j. d., & davidson, r. j. (2007). meditation and the neuroscience of consciousness. in p. zelazo, m. moscovitch, & e. thompson (eds.), cambridge handbook of consciousness (pp. 499–555). ny: cambridge university press. martínez, m. (2013). teleosemantics and indeterminacy. dialectica, 67(4), 427–453. https://doi.org/10.1111/1746-8361.12039 mayer-gross, w., & stein, h. (1926). über einige abänderungen der sinnestätigkeit im meskalinrausch. zeitschrift für die gesamte neurologie und psychiatrie, 101, 354–386. https://doi.org/10.1007/bf02878343 mcclelland, t. (2015). affording introspection: an alternative model of inner awareness. philosophical studies, 172(9), 2469–2492. https://doi.org/10.1007/s11098-014-0421-x mcclelland, t. (2019). against virtual selves. erkenntnis, 84(1), 21–40. https://doi.org/10.1007/s10670-017-9945-8 mcmahan, d. l. (2008). the making of buddhist modernism. oxford: oxford university press. metzinger, t. (2003). being no one: the self-model theory of subjectivity. cambridge: the mit press. metzinger, t. (2013). why are dreams interesting for philosophers? the example of minimal phenomenal selfhood, plus an agenda for future research. consciousness research, 4(746). https://doi.org/10.3389/fpsyg.2013.00746 millière, r. (2017). looking for the self: phenomenology, neurophysiology and philosophical significance of drug-induded ego dissolution. frontiers in human neuroscience, 11(245). https://doi.org/10.3389/fnhum.2017.00245 millière, r. (2019). are there degrees of self-consciousness? journal of consciousness studies, 26(3-4), 252–282. millière, r., carhart-harris, r. l., roseman, l., trautwein, f.-m., & berkovich-ohana, a. (2018). psychedelics, meditation and self-consciousness. frontiers in psychology, 9(1475), 1–29. https://doi.org/10.3389/fpsyg.2018.01475 millière, r., & metzinger, t. (2020). editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 millikan, r. g. (1984). language, thought and other biological categories. cambridge: mit press. neander, k. (2017). a mark of the mental: in defense of informational teleosemantics. cambridge: mit press. nour, n. m., evans, l., nutt, d., & carhart-harris, r. l. (2016). ego-dissolution and psychedelic: validation of the ego-dissolution inventory (edi). frontiers in human neuroscience, 10, 269. https://doi.org/10.3389/fnhum.2016.00269 papineau, d. (1984). representation and explanation. philosophy of science, 51, 550–572. retrieved from https://www.jstor.org/stable/187975 perry, j. (1979). the problem of the essential indexical. noûs, 13, 3–21. https://doi.org/10.2307/2214792 perry, j. (1980). belief and acceptance. midwest studies in philosophy, 5, 533–542. https://doi.org/10.1111/j.1475-4975.1980.tb00422.x prinz, j. (2012). the conscious brain: how attention engenders experience. new york: oxford university press. recanati, f. (2007). perspectival thought: a plea for (moderate) relativism. oxford: oxford university press. recanati, f. (2012). immunity to error through misidentification: what it is and where it comes from. in s. prosser & f. recanati (eds.), immunity to error thorugh misidentification: new essays. cambridge: cambridge university press. reynolds, j. m. (1989). self-liberation through seeing with naked awareness. new york: station hill press. richard, m. (1983). direct reference and ascriptions of belief. journal of philosophical logic, 12, 425–452. https://doi.org/10.1093/acprof:oso/9780199557950.003.0002 salmon, n. (1986). freges puzzle (vol. 75). https://doi.org/doi.org/10.1111/j.1933-1592.2007.00082.x salmon, n. (1989). illogical belief. philosophical perspectives, 3, 243–285. https://doi.org/10.2307/2214270 sankara. (1980). brahmasutra-sankarabhasyam (j. l. sastri, ed.). delhi: motilal banarsidass. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1002/hbm.22833 https://doi.org/10.1093/nc/nix016 https://doi.org/10.2307/2184843 https://doi.org/10.1111/1746-8361.12039 https://doi.org/10.1007/bf02878343 https://doi.org/10.1007/s11098-014-0421-x https://doi.org/10.1007/s10670-017-9945-8 https://doi.org/10.3389/fpsyg.2013.00746 https://doi.org/10.3389/fnhum.2017.00245 https://doi.org/10.3389/fpsyg.2018.01475 https://doi.org/10.33735/phimisci.2020.i.50 https://doi.org/10.3389/fnhum.2016.00269 https://www.jstor.org/stable/187975 https://doi.org/10.2307/2214792 https://doi.org/10.1111/j.1475-4975.1980.tb00422.x https://doi.org/10.1093/acprof:oso/9780199557950.003.0002 https://doi.org/doi.org/10.1111/j.1933-1592.2007.00082.x https://doi.org/10.2307/2214270 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org perspectival self-consciousness and ego-dissolution 27 sebastián, m. ángel. (2012). experiential awareness: do you prefer ”it” to ”me”? philosophical topics, 40(2), 155–177. https://doi.org/10.5840/philtopics201240218 sebastián, m. ángel. (forthcoming). subjective character, the ego and de se representation: phenomenological, metaphysical and representational considerations on pre-reflective self-awareness. protosociology. shea, n. (2007). consumers need information: supplementing teleosemantics with an input condition. philosophy and phenomenological research, 75(2), 404–435. https://doi.org/10.1111/j.1933-1592.2007.00082.x shoemaker, s. s. (1968). self-reference and self-awareness. the journal of philosophy, 65, 555–567. https://doi.org/10.5840/jphil201110824 siegel, s. (2010). the contents of visual experience. new york: oxford university press. skyrms, b. (2010). signals: evolution, learning, & information. oxford: oxford university press. stalnaker, r. c. (1999). context and content: essays on intentionality in speech and thought. new york: oxford university press. stevens, j. (1998). storming heaven: lsd and the american dream. new york: grove atlantics. studerus, e., gamma, a., & vollenweider, f. x. (2010). psychometric evaluation of the altered states of consciousness rating scale (oav). plos one, 5(8), e12412. https://doi.org/10.1371/journal.pone.0012412 thompson, e. (2015). dreamless sleep, the embodied mind, and consciousness. the relevance of a classical indian debate to cognitive science. in t. metzinger & j. windt (eds.), open mind. philosophy and the mind science in the 21st centrury. mit press. toribio, j. (2008). state versus content: the unfair trial of perceptual nonconceptualism. erkenntnis, 69(3), 351–361. https://doi.org/10.1007/s10670-008-9120-3 torres, c., & repke, d. (2006). anadenanthera: visionary plant of south america. abingdon: routledge. tye, m. (1997). ten problems of consciousness: a representational theory of the phenomenal mind. cambridge: the mit press. wasson, v., & wasson, g. (2011). mushrooms, russia and history (volume ii) (a complete book on mushrooms). new york: pantheon books. weisberg, j. (2011). misrepresenting consciousness. philosophical studies, 154(3), 409–433. https://doi.org/10.1007/s11098-010-9567-3 wider, k. (2006). emotion and self-consciousness. in u. kriegel & k. williford (eds.), self-representational approaches to consciouness (pp. 63–88). cambridge, ma: mit press. williford, k. (2006). the self-representational structure of consciousness. in u. kriegel & k. williford (eds.), self-representational approaches to consciousness (pp. 111–142). cambridge: the mit press. williford, k. (2015). representationalisms, subjective character and self-acquaintance. in t. metzinger & j. windt (eds.), open mind. frankfurt: mind group. windt, j. m. (2015). just in time—dreamless sleep experience as pure subjective temporality. in t. metzinger & j. windt (eds.), open mind. philosophy and the mind science in the 21st century. cambridge: mit press. windt, j. m., nielsen, t., & thompson, e. (2016). does consciousness disappear in dreamless sleep? trends in cognitive science, 20(12), 871–882. https://doi.org/10.1016/j.tics.2016.09.006 woźniak, m. (2018). ”i” and ”me”: the self in the context of consciousness. frontiers in psychiology, 9(1656). https://doi.org/10.3389/fpsyg.2018.01656 zahavi, d., & kriegel, u. (2016). for-me-ness: what it is and what it is not. in d. dahlstrom, a. elpidorou, & w. hopp (eds.), philosophy of mind and phenomenology: conceptual and empirical approaches. london: routledge. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.5840/philtopics201240218 https://doi.org/10.1111/j.1933-1592.2007.00082.x https://doi.org/10.5840/jphil201110824 https://doi.org/10.1371/journal.pone.0012412 https://doi.org/10.1007/s10670-008-9120-3 https://doi.org/10.1007/s11098-010-9567-3 https://doi.org/10.1016/j.tics.2016.09.006 https://doi.org/10.3389/fpsyg.2018.01656 https://doi.org/10.33735/phimisci.2020.i.44 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction perspectival first-personal (pfp) awareness perspectivalness: indexicality and iem as individuative features indexicality immunity to error through misidentification (iem) perspectival spatial-centric awareness pfp-awareness in ordinary experiences pfp-awareness as minimal self-consciousness degrees of self-consciousness altered states dreams meditation drugs conclusions attenuating oneself: an active inference perspective on ``selfless'' experiences attenuating oneself an active inference perspective on “selfless” experiences jakub limanowskia (j.limanowski@ucl.ac.uk) karl fristona (k.friston@ucl.ac.uk) abstract in this paper, we address reports of “selfless” experiences from the perspective of active inference and predictive processing. our argument builds upon grounding self-modelling in active inference as action planning and precision control within deep generative models – thus establishing a link between computational mechanisms and phenomenal selfhood. we propose that “selfless” experiences can be interpreted as (rare) cases in which normally congruent processes of computational and phenomenal self-modelling diverge in an otherwise conscious system. we discuss two potential mechanisms – within the bayesian mechanics of active inference – that could lead to such a divergence by attenuating the experience of selfhood: “self-flattening” via reduction in the depth of active inference and “self-attenuation” via reduction of the expected precision of self-evidence. keywords active inference ∙ self-model ∙ sensory attenuation this article is part of a special issue on “radical disruptions of self-consciousness”, edited by thomas metzinger and raphaël millière. 1 introduction a key theme of the 2018 “selfless minds” workshop1 was the discussion of reports of “selfless” experiences – conscious episodes during which, if one believes these reports, self-consciousness is partly or even completely lost (e.g., during intoxication or psychosis, cf. letheby & gerrans, 2017; millière, 2017; saks, 2007). such auniversity college london 1the manifesto that guided this workshop has been incorporated into the editorial of this special issue, see millière & metzinger (2020). limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakub limanowski and karl friston 2 reports raise an important question for the understanding of selfhood and selfconsciousness: how could one have a conscious experience – and able to report on it afterwards – in the absence of any awareness of oneself (as having the experience)? in this paper, we will approach this question from the perspective of active inference formulations of predictive processing (friston, 2010; friston et al., 2010; friston, samothrakis, & montague, 2012). active inference lends itself to describing brain and mind function, particularly when it comes to the construction of internal models of agents in their lived world (cf. hohwy, 2013; clark, 2015; wiese & metzinger, 2017). the framework has inspired much conceptual work on the nature of self-modelling and the experience of self (deane, 2020, this issue; hohwy & michael, 2017; letheby & gerrans, 2017; limanowski & blankenburg, 2013; metzinger, 2003; wiese, 2019). a key point of active inference is that predictive processing via internal models underwrites the optimal planning of actions – which rests on the notion of control; i.e., inferring the optimal course of my (physical, autonomic, or mental) action2 to minimize expected free energy. based on this formulation of planning (as inference), we will argue that some notion of “selfhood” or “self-agency” – in the sense of inference about control – is inherent in active inference. crucially, this includes the allocation of precision to sensory evidence, which corresponds to attention as a form of “mental” action (metzinger, 2017). the problem with reports of “selfless” experiences then boils down to the following question: how come people can feel like they lack selfhood, when in fact they are in control of at least some of their behaviour via their self-model? we will argue that these experiences can be interpreted as (rare) cases in which computational and phenomenal self-modelling diverge. we will consider two potential mechanisms – within the bayesian belief updating of active inference – that could lead to such a divergence by attenuating the experience of selfhood: “selfflattening” via reduction in the depth of active inference and “self-attenuation” via reduction of expected precision of self-evidence. 2 self-modelling based on predictive processing first, however, we will briefly introduce some key ideas about internal predictive (self) models in the active inference framework (see friston, 2010; hohwy & michael, 2017; limanowski & blankenburg, 2013; seth & tsakiris, 2018; wiese & metzinger, 2017, for a more exhaustive introduction). active inference sits within a larger “free-energy principle”, according to which any living system – that can be demarcated from its surroundings – will actively try to remain in a set of unsurprising states by maximizing the (marginal) likelihood of sensory samples (friston, 2010). in this scheme, free energy minimization corresponds to maximizing 2we use the term “action” here to emphasize the link between active inference in terms of action planning and conscious self-modelling. we will later show how the traditional definition of action as a subset of behaviour, characterized by conscious goal representation and a sense of agency, can be related to the depth of inference. limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org attenuating oneself: an active inference perspective on “selfless” experiences 3 bayesian model evidence, which implies a notion of “self-evidencing” (i.e., a bayesoptimal model – a free energy minimizing agent – will always try to maximize the evidence for its existence, hohwy, 2016). such “self-models” are probabilistic (predictive) mappings from causes to consequences, for example from latent or hidden states of the world to sensory observations, in which higher levels contextualize lower levels, and lower levels provide evidence for higher levels (e.g. in the form of “prediction errors”, as in predictive coding; cf. friston, rosch, parr, price, & bowman, 2017). this hierarchical scheme of recurrent message passing implies that increasingly higher-level beliefs represent increasingly abstract states of affairs at increasingly broad time scales. in such deep architectures, balancing the relative influence of prior beliefs or sensory evidence (i.e., prediction errors) on bayesian belief updating across the entire hierarchy – and between sensory evidence from different modalities – is accomplished by weighting the ascending prediction errors by their relative precision based on prior expectations under the model (adams et al., 2013a; feldman & friston, 2010; friston, 2010). precision-modulation is thus also a bayes-optimal, top-down mechanism that minimizes free-energy by optimally balancing or selecting prediction error signals for hierarchical inference – and implicit bayesian belief updating – depending on the current context. for instance, prediction errors can be afforded greater precision because they are particularly salient, or because they are particularly relevant for behaviour. a prediction error that is afforded high precision will have a relatively larger impact on inference (i.e. on the updating of the respective prior beliefs). this means that precision has to be estimated and deployed “top-down” at each level of the hierarchy. the functional role of this sort of top-down precision-modulation is equated with attention (feldman & friston, 2010; cf. edwards, adams, brown, pareés, & friston, 2012). attention is thus seen as a mechanism by which the impact of sensory evidence on belief updating can be amplified or attenuated (cf. fazekas & nanay, 2019), and in this way also accommodates formulations of selective attention, whose allocation is controlled by an interaction of top-down (cognitive) and bottom-up (sensory) factors (posner, snyder, & davidson, 1980; cf. corbetta & shulman, 2002; gilbert & li, 2013). we will later see how this can be associated with mental action and experienced selfhood. we (limanowski & blankenburg, 2013; limanowski & friston, 2018) have previously shown the close correspondence of the formal self-modelling implied by active inference with metzinger’s (2004) account of phenomenal self-modelling; i.e., the construction of a conscious mental model of the organism as a whole (including properties like agency and identity over time). one important assumption of both accounts is that the “self” is seen as a hypothesis or latent state (of being) that can be associated with a self-model. this component of a generative model arises as the (computationally) most accurate, parsimonious explanation for bottom-up multisensory information (metzinger, 2004; cf. seth, suzuki, & critchley, 2011; allen & friston, 2016; apps & tsakiris, 2014; ishida, suzuki, & grandi, 2015). from a limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakub limanowski and karl friston 4 predictive processing perspective, the hierarchical nature of the underlying computational architecture suggests a centeredness of the model on the “self” (allen & friston, 2016; limanowski & blankenburg, 2013) in that higher levels of the model will be increasingly abstract (amodal), complex, and invariant (i.e., less likely to be affected by prediction error) – and the highest-level inferred causes pertain to “myself”. we will later see how this is especially important for action planning (i.e., active inference) and, implicitly, for experienced “selfhood”. importantly, this (computational) hierarchical notion of self-modelling resonates with the spatiotemporal centeredness of experience on the phenomenal self (cf. metzinger, 2004), and with the idea of a non-conscious and bodily basis for higher forms of self-consciousness (blanke & metzinger, 2009; gallagher, 2000). such a framework can therefore be used to explain a vast variety of experimental results and even pathological bodily experience. for instance, bodily illusions are well explained as a result of bayes-optimal inference; i.e., arising from an interpretation of ambiguous sensory input under strong prior hypotheses (apps & tsakiris, 2014; brown, adams, parees, edwards, & friston, 2013; friston, 2005; limanowski, 2014). in the rubber hand illusion (botvinick & cohen, 1998), i “wrongly” adjust my perceived hand position to resolve multisensory ambiguity – but i still feel like a sane person in a normal body with just one, not two right arms (hohwy, 2013; limanowski, 2017). similarly, even when experienced self-location and first-person perspective – two major constituents of minimal phenomenal selfhood along some conceptualizations – are decoupled, a unified self is still experienced (blanke & metzinger, 2009; limanowski, 2014). in a predictive coding scheme, these observations are well explained by the fact that if prediction error can be explained away at lower levels, there is no need to adjust higher-level representations in my model. in sum, within the predictive processing framework, one can, in principle, associate certain3 computational mechanisms with the phenomenology of “being someone” – in other words, one can link computational to phenomenal (i.e., conscious mental) self-modelling. but of course, perceptual inference is only part of the selfmodelling story. we will next turn to predictive processing as formalized by active inference, which affords a different perspective on self-modelling in terms of action planning and “self-evidencing” (hohwy, 2016). specifically, we will discuss how “self-flattening” through reducing the depth of hierarchical (active) inference may play into “selfless” experiences. 3that is, a conscious (phenomenal) self-model is, of course, only a highly specific case of computational “self-modelling”. of course, there are many instances of computational “self-modelling” that are not accompanied by conscious (self) experience. in this paper, we focus on cases in which processes of phenomenal and computational self-modelling that are normally congruent diverge. limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org attenuating oneself: an active inference perspective on “selfless” experiences 5 3 “self-flattening”: the relationship between deep active inference and selfhood active inference extends perceptual inference or predictive coding by noting action offers another way to quench prediction errors; i.e., sampling sensory data in a way that confirms the model’s predictions. acting thus involves both generating a prediction of sensory input expected to result from intended movement, and “fulfilling” this prediction by executing the movement, thus effectively suppressing a prediction error signal that would otherwise emerge (adams et al., 2013a; brown et al., 2013; seth & friston, 2016). the agent must therefore also have beliefs about which course of action (or, more generally, behaviour; see below) will be optimal in a given context. hence the agent’s model must be able to entertain “counterfactually rich” representations; i.e., beliefs about several alternative potential actions and the states of affairs that these actions would bring about (friston et al., 2017; seth, 2014; cf. powers, 2005). this issue has recently been addressed by a formulation of active inference in terms of bayesian model selection – among potential courses of action and behaviour – based on their expected free energy (evaluated in the light of prior beliefs and preferences, parr & friston, 2017; friston et al., 2017). in brief, the latent variables of such models4 are hidden states and policies; hidden states generate observations and state transitions depend on a plan or action “policy” pursued by the agent. policy optimization thus entails selecting a sequence of actions, with an associated effect on state transitions and expected outcomes – and a corresponding free energy. in other words, the policy with the lowest expected surprise is most probable, i.e., the sort of policy “i am likely to pursue” (friston, 2018; friston et al., 2017). as noted above, policies are selected based on inference at multiple levels, where higher levels contextualize lower levels. thereby, as emphasized in the previous section, active inference relies on sensory evidence and on its appropriate weighting. in turn, the selected policy is one “i am likely to pursue” and will therefore specify empirical “self” priors (for action) that contextualize self-modelling – which again emphasizes the hierarchical nature of self-modelling discussed above (cf. butz, 2008; apps & tsakiris, 2014; limanowski & blankenburg, 2013; seth & tsakiris, 2018). as has been pointed out by a number of authors, this leads to the notion of “self-evidencing” inherent in active inference (hohwy, 2016; cf. friston et al., 2012; hohwy & michael, 2017; limanowski & blankenburg, 2013): an active inference agent will always try to maximize evidence for the hypothesis it entertains about itself – thus perceptual inference and inferred policies provide this kind of evidence that “i am that sort of agent”. 4the generative models in play here are usually conceptualized as discrete state space models similar to markov decision processes; but the same processes can in principle also be formulated in continuous state space, e.g. in a predictive coding scheme, since both formulations entail (hierarchical) belief updating based on sensory evidence. in discrete state spaces, many inference problems associated with planning and selecting actions become conceptually and mathematically more tractable. limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakub limanowski and karl friston 6 note that the kind of action (or behaviour in general) we are talking about here is not necessarily physical – the same principles may apply to mental processes (cf. metzinger, 2017). in particular, one can understand the active deployment of precision as a form of covert or mental action that has exactly the look and feel of attention5 (cf. metzinger, 2004). the argument goes (see limanowski & friston, 2018 for details) as follows: policy optimization necessarily entails a specification of the precision of (action-dependent) changes in hidden states that we are trying to infer. put in formal terms, policies or beliefs about action entail expectations about precision; placing confidence in the consequences of action is an inherent part of the policies from which we select our actions. this implies that beliefs about actions – in the sense of active inference as action policy optimization – cannot be subject to introspective attention because this would induce another policy (of policies) and an infinite regress. so, these “high level” beliefs about “what i am” doing are unique – and may be the computational basis of precluding an infinite regress by phenomenal self-modelling, with accompanying phenomenal transparency6 (metzinger, 2004; cf. limanowski & friston, 2018). this interpretation of active inference speaks to the concept of “attentional agency” as introduced by metzinger (2013, 2017; cf. wiese, 2019). whether we are talking about mental or physical action, the important point is that policy optimization is a special kind of inference – it is inference about which states of the world i can control; i.e., about selecting a course of my action that will minimize my expected free energy (cf. friston et al., 2012). one may now ask: is this probabilistic representation of control necessarily conscious? an answer to this question has been put forth by friston (2018) as follows: the representation of action policies – potentially, even several alternative ones, each of which specifies an expectation of how the state of the world (accessible via my sensory states) unfolds depending on my action – requires the system to embody an explicit representation of how states evolve over time. depending on how far this representation of fictive time (i.e., into the past and the future) extends, potential action policies will be temporally deeper. note that temporal depth is closely related to hierarchical (representational) depth and counterfactual richness (seth, 2014) because the deeper one goes into the future the greater the number of outcomes. one can now propose an association of temporal depth – the ability to plan and explore multiple futures – with the degree of consciousness7 it subtends: whereas non-conscious 5this can also be linked to concepts such as introspective attention, defined by metzinger (2004, p. 36) as a specific kind of introspection (introspection1). 6the concept of phenomenal transparency describes the specific case in which only the content of a mental representation, but not its construction process, is available to introspective attention – this may be the reason for why some experiences seem “real” (metzinger, 2004). 7we acknowledge that the idea that consciousness comes in unique “degrees” is not uncontroversial. however, our idea is compatible with multi-dimensional accounts of consciousness (e.g. bayne, hohwy, & owen, 2016): temporal depth would specify the degree of consciousness along a particular dimension – which leaves open whether degrees of consciousness can also be measured along other, perhaps independent dimensions. limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org attenuating oneself: an active inference perspective on “selfless” experiences 7 processes are stuck in the “here-and-now” (edelman, 2001), conscious processes operate under a “thick” model of future action and behaviour. this idea speaks to many previous definitions of consciousness as a quintessentially mnemonic process (damasio, 2012; edelman, 2003; husserl, 2008; james, 1890; seth, 2009; verschure, 2016; cf. powers, 2005; carey, 2018). note that this sort of temporal depth also grounds the agent in time (it generates a “narrative”, friston et al., 2017) – and provides an opportunity to explain the often discussed invariance of phenomenal selfhood over time (james, 1890; metzinger, 2003). in sum, under active inference, sentient systems that employ temporally thick generative models are likely conscious agents, in that at least some – deep – inference processes are associated with conscious mental states. we believe this idea fits well with the often advanced proposal that there is a basic self-consciousness in the background of any conscious experience (e.g. zahavi, 2014; damasio, 1999).8 the point we want to make here is that when consciousness arises during (deep) active inference, it will be accompanied by (minimal) consciousness of what one could call the “self-as-agent”. on this view, even in cases of altered self-perception and the absence of overt action, a system may engage in active inference, i.e., in that at least the allocation of attention is controlled (cf. metzinger, 2013; wiese, 2019). this would mean that any conscious system (including artificial ones) that can be said to engage in active inference will experience some sort of “mental” agency – generating a new transparent layer in its phenomenal self-model. but this seems to contrast with reports of very vivid phenomenology during otherwise “selfless” experiences (millière, 2017, 2020, this issue; saks, 2007). there is an interesting related case of apparent absence of attentional (or more generally, cognitive) control during (relatively) vivid conscious experience: namely, during mind wandering (metzinger, 2018; schooler, 2014). interestingly, by linking the (temporal) depth of inference to consciousness, we can, in principle, accommodate the traditional definition of action as a specific case of behaviour accompanied by a conscious goal representation and sense of agency: there are many kinds of behaviour that do not depend on deep inference (such as homoeostasis and reflexes) and are therefore not perceived as (consciously) controlled. some kinds of behaviour – i.e., actions – are based on deep inference about control and therefore have a phenomenology of agency. this distinction may also apply to different kinds of attention; i.e., whereas endogenous attention relies on deep inference and feels “deliberate”, a capture of attention by a salient stimulus feels much less “controlled”. even though mind wandering episodes may not be characterized by a loss of “selfhood”, they could in principle be linked to a reduced depth of active inference – and may therefore be described as mental behaviour rather than action (metzinger, 2017). 8whether or not this is to be interpreted as an inherent subjectivity in consciousness, i.e., a prereflective self-awareness characterized by the “first-personal mode of givenness” of all conscious experience, is beyond the scope of this paper (legrand, 2006; zahavi & parnas, 1998); but this can be traced back to sartre, husserl, and merleau-ponty if one wants to). limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakub limanowski and karl friston 8 so, could a potential explanation of “selfless” experiences be that they are related to a reduction of temporal depth of active inference – a “self-flattening” – resulting in an attenuation of phenomenal selfhood (deane, 2020, this issue)? we think this is unlikely to be the complete explanation, for the following reason. based on the (computational) argument that inference about the self – and, crucially, which states of the world it can control – is at the core of active inference, we propose that the corresponding processes of inference are also temporally deep(est). thus any “flattening” of active inference would have general effects on action and perception; i.e., it would involve a general reduction of consciousness as e.g. during certain stages of sleep, anaesthesia, or coma. moreover, one would expect that if the temporal depth of inference is indeed “flattened” in this way, these experiences should also be less accessible by memory in retrospect. whether or not this is universally true – and whether or not consciousness is in fact generally reduced during “selfless” experiences – is certainly an empirical question, so our proposal remains speculative. however, we believe that there is another (not necessarily exclusive) mechanism that could lead (or contribute) to “selfless” experiences; i.e., the control over expected precision. in the next section, we will discuss why it is important to get one’s precision expectations “right” for inference about what kind of an agent i am, and how aberrant precision control could play into the sort of self-attenuation that characterises “selfless” experiences, and thus – together with self-flattening – help to explain the apparent differences between computational and phenomenal self-modelling. 4 “self-attenuation”: the importance of expected precision as mentioned above, precision control has a fundamental role in the construction of self-representations. following active inference, this is a general role that should apply to low-level (e.g. bodily) self-representation but – as we will argue – also to higher, cognitive and conceptual levels. interestingly, the problem that the brain has to solve via precision control is often not which sensory evidence to emphasize, but which to attenuate (parr, rees, & friston, 2018). this sort of sensory attenuation is especially relevant for action – in fact, it would be impossible to initiate a movement without it. in brief, on an active inference reading, movement occurs because high-level multi-modal or amodal prior beliefs about behaviour predict proprioceptive and exteroceptive states that would ensue if the movement was performed (e.g. a particular limb trajectory). prediction error is then suppressed throughout a motor hierarchy ranging from intentions and goals to kinematics to muscle activity (kilner, friston, & frith, 2007). at the lowest level of the hierarchy, spinal reflex arcs suppress proprioceptive prediction errors by “fulfilling” the predicted movement – limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org attenuating oneself: an active inference perspective on “selfless” experiences 9 which thereby minimises exteroceptive prediction errors (e.g. the predicted visual consequences of the action). the assumption that action is driven by anticipation of its sensory effects links active inference to ideomotor accounts of action (hommel, müsseler, aschersleben, & prinz, 2001; prinz, 1997), perceptual control theory (powers, 2005), and things like the equilibrium point hypothesis for motor control (feldman, 1974). sensory attenuation is also important for the construction of a multisensory body representation – especially when sensory information from multiple modalities is conflicting. in the “rubber hand illusion” (botvinick & cohen, 1998), for example, visual information about hand position is expected to be very precise, while the (conflicting) proprioceptive information about hand position is afforded a lower precision – i.e., it is relatively down-weighted – to resolve the intersensory conflict and to maintain a coherent body representation. similar mechanisms may be in play during visuomotor adaptation, i.e., the adaptation to novel visuomotor mappings as introduced e.g. by wearing prism glasses: some experimental evidence suggests that, as in the rubber hand illusion, a temporary attenuation of irreconcilable proprioceptive information may help with this sort of adaptation to a new body representation (balslev et al., 2004; bernier, burle, vidal, hasbroucq, & blouin, 2009; limanowski & friston, 2019). this can potentially go as far as to induce an experiential “neglect” of the real (physical) body during virtual reality experiences – although this has only been shown in monkeys so far (i.e., it has been suggested that monkeys using brain-machine interfaces to control artificial limbs gradually begin to neglect their real body, carmena et al., 2003; cf. metzinger, 2007). lastly, sensory attenuation is crucial for self-other distinction. by attenuating sensory data that is self-produced, i can now emphasize externally generated – behaviourally relevant – sensory data (e.g. during finger movement, selfproduced somatosensory input from skin stretching and muscle movements is attenuated, while sensitivity to externally generated touch is enhanced, limanowski et al., 2019). thus, it is crucial to know which data to attenuate – this is a problem that any “social” brain has to solve: when interacting with conspecifics, i need to know how to balance proprioceptive and exteroceptive (e.g., visual) information to either move myself, or to be able to observe another’s movements without echopraxia (see kilner et al., 2007; friston et al., 2010; limanowski & friston, 2019 for further discussion). note that getting the precision estimates “right” often means lowering them to attenuate sensory evidence – especially when modelling oneself. in other words – and keeping in mind that the top-down allocation of precision is equated with attention – one can describe this as ignoring or “dis-attending” to certain features of oneself (clark, 2015; cf. limanowski, 2014, 2017). in the simplest case, this means lowering the precision afforded to one particular sensory modality (even purely interoceptively, seth et al., 2011; allen, levy, parr, & friston, 2019), but we will see that the same principle may hold at more complex levels of self-modelling, too. limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakub limanowski and karl friston 10 the point we want to emphasize is that the temporary attenuation of the precision of sensory “self-evidence” – which is necessary to entertain an alternative (and yet counterfactual, cf. seth, 2014) hypothesis about myself – is effectively a form of “self-attenuation”. in the case of movement, for instance, self-evidence would be constituted by proprioceptive information – conveying evidence for the fact that i am actually not moving – and attenuation of this self-evidence is necessary to enable movement, i.e., to enact a counterfactual proprioceptive hypothesis issued at higher levels of the hierarchy (adams et al., 2013a; brown et al., 2013). a tangible example of sensory attenuation is saccadic suppression, where we appear to be unable to “see” the motion induced by saccadic eye movements. a more fanciful example might be the temporary suspension of attention – which is cued by the misdirection of a magician, but remains under our (top-down precision) control – that allows us to suspend our disbelief that what we are witnessing is indeed “magic”. luckily, self-attenuation just needs to be applied transiently – e.g., at movement initiation – and thus sensory evidence can still be processed to guide inference. an interesting thought is that during such periods of “self-attenuation”, self-experience may also be altered. thomas metzinger has introduced a related idea with the concept of a “self-representational blink” occurring at the transition to mind wandering (metzinger, 2013; cf. wiese, 2019). while this idea may be difficult to test empirically, one could speculate that perhaps some inferred narrative would still maintain the self-model or self-hypothesis during these periods, just like the content of a scene does not disappear during a saccade (cf. hohwy & michael, 2017; seth & tsakiris, 2018). what if this sort of attenuation were not temporary? this would, in the long run, have grave consequences for self-experience, action, and ultimately life: of course, sensory information – self-evidence – is needed to maintain a representation of myself; i.e., the sensory evidence for me as a “self-as-agent”. so, if i continued to attenuate this evidence, my inference about myself would become quite unrealistic and i could not act properly in the world. it is likely that this would ultimately lead to the death of the agent. less severe examples of how “abnormal” precision control can affect self-experience are found in abundance in psychopathology. on this view, pathological states can be described in terms of a change in the precision of perceptual prediction errors (due to abnormal priors, cf. friston, 2005; parr et al., 2018; sterzer et al., 2018). this theme of aberrant precision control dominates many explanations of false perceptual inference in general and lack of central coherence in psychiatric syndromes in particular, such as organic psychosyndromes (collerton, perry, & mckeith, 2005), chronic pain (tabor & burr, 2019), functional motor symptoms (edwards et al., 2012), autism (lawson, rees, & friston, 2014; pellicano & burr, 2012; van de cruys et al., 2014) and schizophrenia (adams et al., 2013b; powers, mathys, & corlett, 2017). in relation to “selfless” experiences, the most interesting pathologies resulting from aberrant precision control are those characterized by misrepresentations of agency (adams et al., 2013b; brown et al., 2013; edwards et al., 2012; limanowski, limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org attenuating oneself: an active inference perspective on “selfless” experiences 11 2017). for instance, in schizophrenia, inference about the hidden causes of sensations may fail because the precision of high-level beliefs is increased to compensate for a failure to attenuate sensory prediction error during action. these overconfident beliefs generate additional, inappropriately confident, predictions about external causes – the agent is not able to infer whether it caused its sensations itself, or whether someone or something else caused them. this results in altered sensory experience and in severe cases in misattributions of agency – as e.g. demonstrated by different susceptibility of schizophrenic patients in the force-matching paradigm (brown et al., 2013). the same principles that presumably cause these – and other – hallucinations or delusions (i.e., when the perceptual system is affected, friston, 2005) can lead to functional symptoms, e.g. when precision control is abnormal in the motor system (edwards et al., 2012; parr et al., 2018). in other words, there may be a mechanistic link between dissociative syndromes (known as functional medical syndromes) and reports of selflessness (edwards et al., 2012). for example, dissociative symptoms such as self-reports of “i cannot feel my arm” are linked to aberrant central processing – as measured with electroencephalography – in empirical studies of sensory attenuation (hughes, desantis, & waszak, 2013; oestreich et al., 2015; pareés et al., 2014). likewise, in functional motor symptoms, abnormal precision seems to be assigned at intermediate levels of the motor hierarchy, which may trigger the execution of a movement – without the accompanying phenomenology of intentional movement generation (which would be associated with higher-level motor areas; edwards et al., 2012). to summarize, self-modelling in active inference – and consequently, healthy experience and behaviour – relies on the balance of sensory and prior (model) precision; both abnormally high and abnormally low precision estimates have negative consequences. as we hope to have shown with the above brief review of psychopathology, altered precision expectations (about sensory self-evidence) can affect even the highest levels of the phenomenal self-model – leading to selfother confusion and misattributions of agency. the point we want to make is that the same mechanistic explanation (in terms of sensory attenuation and attentional/precision control) can in principle be extended to account for much of the phenomenology of “selfless” experiences. these experiences could thus be interpreted as (partly) resulting from a temporary attenuation of more complex cues for the self-model (i.e., “self-evidence”), which can apparently be so strong that there is only marginal conscious perception of these cues – leading to a “false” update of the phenomenal self-model. this interpretation particularly speaks to “selfless” experiences of the sort associated with e.g. psychedelics, if one subscribes to the notion that these can be seen as “psychotomimetics” (i.e., that the associated altered state is akin to psychosis, cf. bayne & carter, 2018). such an explanation also aligns with arguments that the sort of “ego-dissolution” reported after some psychedelic experiences may be due to an impairment of those mechanisms that integrate sensory evidence into a coherent self-percept (letheby & gerrans, 2017; millière, 2017). limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakub limanowski and karl friston 12 5 conclusion and outlook we have argued that the experience of having “lost” one’s self (constituting a fundamental change to the phenomenal self-model) could arise from a combination of “self-flattening” via a loss of deep active inference and “self-attenuation” via aberrant precision expectations about sensory self-evidence (i.e., within the computational self-model realizing active inference). while the former mechanism could lead to a generally reduced temporal depth and degree of consciousness, the latter could attenuate (aspects of) the “self” from experience. however, even if that system may be “wrong” about what kind of agent it is – including experiences of “selflessness” or self-other confusion – it will still employ an optimal (computational) self-model in the sense implied by the active inference story. we therefore conclude that “selfless” experiences can be interpreted as (rare) cases in which – in an otherwise conscious system – normally congruent processes of computational and phenomenal self-modelling diverge. this divergence is an exciting area of investigation for interdisciplinary research on self-modelling and -experience. there are many fine details to the active inference story, which we have but touched upon here. for instance, whether or not – and how strongly – the temporal depth of active inference can be “flattened” by e.g. meditation or pharmacology is an exciting empirical question. for example, trained meditators (cf. berkovich-ohana, dor-ziderman, glicksohn, & goldstein, 2013) may be able to direct their attention in a way that will enhance self-attenuation in a very similar way as in pathological or psychedelic experience. in other words, they can evoke changes in neuronal precision control similar to the effects of neuromodulators. this puts precision control at the centre of the (empirical) story again. furthermore, there are cases in which sensory evidence itself can trigger a loss of transparency – i.e., a revision of beliefs about precision. such subjectively surprising changes from transparency to opacity encompass, for instance, reaching “lucidity” in a dream; i.e., becoming aware that one is dreaming (dresler et al., 2015), or certain stress situations; e.g., after accidents, when somehow everything about the situation seems “unreal” (metzinger, 2003, 2007); they can even be triggered by violations of sensorimotor expectations, such as when an afterimage is recognized as “unreal” because it does not move according to motor predictions sent to the eyes (seth, 2014). many altered states of consciousness seem to involve experiences similar to those described above, and could therefore offer an interesting tool to investigate the mechanisms grounding the “realness” of our experience empirically. a detailed investigation of altered states of consciousness under the assumption that they result from aberrant precision-weighting – and, perhaps, an associated loss of phenomenal transparency – could help understand why people may sometimes feel like they have “lost” parts of, or even their entire “self”. limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org attenuating oneself: an active inference perspective on “selfless” experiences 13 acknowledgments we thank thomas metzinger, raphaël millière, and two anonymous reviewers for their helpful comments. this work was supported by funding from the european union’s horizon 2020 research and innovation programme under the marie skłodowska-curie grant agreement no 749988 to jl. kf was funded by a wellcome trust principal research fellowship (ref: 088130/z/09/z). references adams, r. a., shipp, s., & friston, k. j. (2013a). predictions not commands: active inference in the motor system. brain structure and function, 218(3), 611–643. https://doi.org/10.1007/s00429-012-0475-5 adams, r. a., stephan, k. e., brown, h. r., frith, c. d., & friston, k. j. (2013b). the computational anatomy of psychosis. frontiers in psychiatry, 4(47). https://doi.org/10.3389/fpsyt.2013.00047 allen, m., & friston, k. j. (2016). from cognitivism to autopoiesis: towards a computational framework for the embodied mind. synthese, 195, 2459–2482. https://doi.org/10.1007/s11229-016-1288-5 allen, m., levy, a., parr, t., & friston, k. j. (2019). in the body’s eye: the computational anatomy of interoceptive inference. biorxiv, 603928. https://doi.org/10.1101/603928 apps, m. a., & tsakiris, m. (2014). the free-energy self: a predictive coding account of self-recognition. neuroscience & biobehavioral reviews, 41, 85–97. https://doi.org/10.1016/j.neubiorev.2013.01.029 balslev, d., christensen, l. o., lee, j. h., law, i., paulson, o. b., & miall, r. c. (2004). enhanced accuracy in novel mirror drawing after repetitive transcranial magnetic stimulation-induced proprioceptive deafferentation. journal of neuroscience, 24(43), 9698–9702. https://doi.org/10.1523/jneurosci.1738-04.2004 bayne, t., & carter, o. (2018). dimensions of consciousness and the psychedelic state. neuroscience of consciousness, 2018(1), niy008. https://doi.org/10.1093/nc/niy008 bayne, t., hohwy, j., & owen, a. m. (2016). are there levels of consciousness? trends in cognitive sciences, 20(6), 405–413. https://doi.org/10.1016/j.tics.2016.03.009 berkovich-ohana, a., dor-ziderman, y., glicksohn, j., & goldstein, a. (2013). alterations in the sense of time, space, and body in the mindfulness-trained brain: a neurophenomenologically-guided meg study. frontiers in psychology, 4. https://doi.org/10.3389/fpsyg.2013.00912 bernier, p. m., burle, b., vidal, f., hasbroucq, t., & blouin, j. (2009). direct evidence for cortical suppression of somatosensory afferents during visuomotor adaptation. cerebral cortex, 19(9), 2106–2113. https://doi.org/10.1093/cercor/bhn233 blanke, o., & metzinger, t. (2009). full-body illusions and minimal phenomenal selfhood. trends in cognitive sciences, 13(1), 7–13. https://doi.org/10.1016/j.tics.2008.10.003 botvinick, m., & cohen, j. (1998). rubber hands “feel” touch that eyes see. nature, 391(6669), 756. https://doi.org/10.1038/35784 brown, h., adams, r. a., parees, i., edwards, m., & friston, k. (2013). active inference, sensory attenuation and illusions. cognitive processing, 14, 411–427. https://doi.org/10.1007/s10339-013-0571-3 butz, m. v. (2008). how and why the brain lays the foundations for a conscious self. constructivist foundations, 4(1), 1–14. retrieved from https://constructivist.info/4/1/001 carey, t. a. (2018). consciousness as control and controlled perception – a perspective. annals of behavioral science, 4(2), 1–8. retrieved from http://behaviouralscience.imedpub.com/consciousness-as-control-and-controlled-perceptiona-perspective.php?aid=23059 carmena, j. m., lebedev, m. a., crist, r. e., o’doherty, j. e., santucci, d. m., dimitrov, d. f., et al. (2003). learning to control a brain-machine interface for reaching and grasping by primates. plos biology, 1(2), 193–208. https://doi.org/10.1371/journal.pbio.0000042 clark, a. (2015). surfing uncertainty: prediction, action, and the embodied mind. oxford: oxford university press. collerton, d., perry, e., & mckeith, i. (2005). why people see things that are not there: a novel perception and attention deficit model for recurrent complex visual hallucinations. behavioral and brain sciences, 28, 737–757. https://doi.org/10.1017/s0140525x05000130 corbetta, m., & shulman, g. l. (2002). control of goal-directed and stimulus-driven attention in the brain. nature reviews neuroscience, 3, 201–215. https://doi.org/10.1038/nrn755 limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s00429-012-0475-5 https://doi.org/10.3389/fpsyt.2013.00047 https://doi.org/10.1007/s11229-016-1288-5 https://doi.org/10.1101/603928 https://doi.org/10.1016/j.neubiorev.2013.01.029 https://doi.org/10.1523/jneurosci.1738-04.2004 https://doi.org/10.1093/nc/niy008 https://doi.org/10.1016/j.tics.2016.03.009 https://doi.org/10.3389/fpsyg.2013.00912 https://doi.org/10.1093/cercor/bhn233 https://doi.org/10.1016/j.tics.2008.10.003 https://doi.org/10.1038/35784 https://doi.org/10.1007/s10339-013-0571-3 https://constructivist.info/4/1/001 http://behaviouralscience.imedpub.com/consciousness-as-control-and-controlled-perception-a-perspective.php?aid=23059 http://behaviouralscience.imedpub.com/consciousness-as-control-and-controlled-perception-a-perspective.php?aid=23059 https://doi.org/10.1371/journal.pbio.0000042 https://doi.org/10.1017/s0140525x05000130 https://doi.org/10.1038/nrn755 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakub limanowski and karl friston 14 damasio, a. (1999). the feeling of what happens: body and emotion in the making of consciousness. new york: harcourt brace. damasio, a. (2012). self comes to mind: constructing the conscious brain. new york: pantheon. deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 dresler, m., wehrle, r., spoormaker, v. i., steiger, a., holsboer, f., czisch, m., & hobson, j. a. (2015). neural correlates of insight in dreaming and psychosis. sleep medicine reviews, 20, 92–99. https://doi.org/10.1016/j.smrv.2014.06.004 edelman, g. (2001). consciousness: the remembered present. annals of the new york academy of sciences, 929, 111–122. https://doi.org/10.1111/j.1749-6632.2001.tb05711.x edelman, g. m. (2003). naturalizing consciousness: a theoretical framework. proceedings of the national academy of sciences, 100(9), 5520–5524. https://doi.org/10.1073/pnas.0931349100 edwards, m. j., adams, r. a., brown, h., pareés, i., & friston, k. j. (2012). a bayesian account of “hysteria”. brain, 135(11), 3495–3512. https://doi.org/10.1093/brain/aws129 fazekas, p., & nanay, b. (2019). attention is amplification, not selection. the british journal for the philosophy of science, axy065. https://doi.org/10.1093/bjps/axy065 feldman, a. g. (1974). change in the length of the muscle as a consequence of a shift in equilibrium in the muscle-load system. biophysics, 19(3), 544–548. feldman, h., & friston, k. (2010). attention, uncertainty, and free-energy. frontiers in human neuroscience, 4(215.). https://doi.org/10.3389/fnhum.2010.00215 friston, k. (2005). a theory of cortical responses. philosophical transactions of the royal society b: biological sciences, 360(1456), 815–836. https://doi.org/10.1098/rstb.2005.1622 friston, k. (2018). am i self-conscious? (or does self-organization entail self-consciousness?). frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018.00579 friston, k. j. (2010). the free-energy principle: a unified brain theory? nature reviews neuroscience, 11(2), 127–138. retrieved from https://www.nature.com/articles/nrn2787 friston, k. j., daunizeau, j., kilner, j., & kiebel, s. j. (2010). action and behavior: a free-energy formulation. biological cybernetics, 102(3), 227–260. https://doi.org/10.1007/s00422-010-0364-z friston, k. j., rosch, r., parr, t., price, c., & bowman, h. (2017). deep temporal models and active inference. neuroscience & biobehavioral reviews, 90, 486–501. https://doi.org/10.1016/j.neubiorev.2018.04.004 friston, k., samothrakis, s., & montague, r. (2012). active inference and agency: optimal control without cost functions. biological cybernetics, 106, 523–541. https://doi.org/10.1007/s00422-012-0512-8 gallagher, s. (2000). philosophical conceptions of the self: implications for cognitive science. trends in cognitive sciences, 4(1), 14–21. https://doi.org/10.1016/s1364-6613(99)01417-5 gilbert, c. d., & li, w. (2013). top-down influences on visual processing. nature reviews neuroscience, 14, 350–363. https://doi.org/10.1038/nrn3476 hohwy, j. (2013). the predictive mind. oxford: oxford university press. hohwy, j. (2016). the self-evidencing brain. noûs, 50(2), 259–285. https://doi.org/10.1111/nous.12062 hohwy, j., & michael, j. (2017). why should any body have a self? in f. de vignemont & a. alsmith (eds.), the subject’s matter: self-consciousness and the body (pp. 363–391). cambridge, ma: mit press. hommel, b., müsseler, j., aschersleben, g., & prinz, w. (2001). codes and their vicissitudes. behavioral and brain sciences, 24(5), 910–926. https://doi.org/10.1017/s0140525x01520105 hughes, g., desantis, a., & waszak, f. (2013). mechanisms of intentional binding and sensory attenuation: the role of temporal prediction, temporal control, identity prediction, and motor prediction. psychological bulletin, 139(1), 133–151. https://doi.org/10.1037/a0028566 husserl, e. (2008). on the phenomenology of the consciousness of internal time (1893-1917) (j. b. brough, trans.). new york, ny: springer. ishida, h., suzuki, k., & grandi, l. c. (2015). predictive coding accounts of shared representations in parieto-insular networks. neuropsychologia, 70, 442–454. https://doi.org/10.1016/j.neuropsychologia.2014.10.020 james, w. (1890). the principles of psychology. new york: holt. kilner, j. m., friston, k. j., & frith, c. d. (2007). predictive coding: an account of the mirror neuron system. cognitive processing, 8, 159–166. https://doi.org/10.1007/s10339-007-0170-2 lawson, r. p., rees, g., & friston, k. j. (2014). an aberrant precision account of autism. frontiers in human neuroscience, 8(302.). https://doi.org/10.3389/fnhum.2014.00302 limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://doi.org/10.1016/j.smrv.2014.06.004 https://doi.org/10.1111/j.1749-6632.2001.tb05711.x https://doi.org/10.1073/pnas.0931349100 https://doi.org/10.1093/brain/aws129 https://doi.org/10.1093/bjps/axy065 https://doi.org/10.3389/fnhum.2010.00215 https://doi.org/10.1098/rstb.2005.1622 https://doi.org/10.3389/fpsyg.2018.00579 https://www.nature.com/articles/nrn2787 https://doi.org/10.1007/s00422-010-0364-z https://doi.org/10.1016/j.neubiorev.2018.04.004 https://doi.org/10.1007/s00422-012-0512-8 https://doi.org/10.1016/s1364-6613(99)01417-5 https://doi.org/10.1038/nrn3476 https://doi.org/10.1111/nous.12062 https://doi.org/10.1017/s0140525x01520105 https://doi.org/10.1037/a0028566 https://doi.org/10.1016/j.neuropsychologia.2014.10.020 https://doi.org/10.1007/s10339-007-0170-2 https://doi.org/10.3389/fnhum.2014.00302 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org attenuating oneself: an active inference perspective on “selfless” experiences 15 legrand, d. (2006). the bodily self: the sensori-motor roots of pre-reflective self-consciousness. phenomenology and the cognitive sciences, 5(1), 89–118. https://doi.org/10.1007/s11097-005-9015-6 letheby, c., & gerrans, p. (2017). self unbound: ego dissolution in psychedelic experience. neuroscience of consciousness, 2017(1), nix016. https://doi.org/10.1093/nc/nix016 limanowski, j. (2014). what can body ownership illusions tell us about minimal phenomenal selfhood? frontiers in human neuroscience, 8. https://doi.org/10.3389/fnhum.2014.00946 limanowski, j. (2017). (dis-)attending to the body – action and self-experience in the active inference framework. in t. metzinger & w. wiese (eds.), philosophy and predictive processing (pp. 1–13). frankfurt am main, germany: mind group. limanowski, j., & blankenburg, f. (2013). minimal self-models and the free energy principle. frontiers in human neuroscience, 7. https://doi.org/10.3389/fnhum.2013.00547 limanowski, j., & friston, k. (2018). “seeing the dark”: grounding phenomenal transparency and opacity in precision estimation for active inference. frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018.00643 limanowski, j., & friston, k. (2019). attentional modulation of vision versus proprioception during action. https://doi.org/10.1093/cercor/bhz192 limanowski, j., lopes, p., keck, j., baudisch, p., friston, k., & blankenburg, f. (2019). action-dependent processing of touch in the human parietal operculum. cerebral cortex, bhz111. https://doi.org/10.1093/cercor/bhz111 metzinger, t. (2003). phenomenal transparency and cognitive self-reference. phenomenology and the cognitive science, 2, 353–393. https://doi.org/10.1023/b:phen.0000007366.42918.eb metzinger, t. (2004). being no one: the self-model theory of subjectivity. cambridge, ma: mit press. metzinger, t. (2007). self models. scholarpedia, 2(10). https://doi.org/10.4249/scholarpedia.4174 metzinger, t. (2013). why are dreams interesting for philosophers? the example of minimal phenomenal selfhood, plus an agenda for future research. frontiers in psychology, 4. https://doi.org/10.3389/fpsyg.2013.00746 metzinger, t. (2017). the problem of mental action – predictive control without sensory sheets. in t. metzinger & w. wiese (eds.), philosophy and predictive processing. frankfurt am main: mind group. metzinger, t. (2018). why is mind wandering interesting for philosophers. in k. fox & k. christoff (eds.), the oxford handbook of spontaneous thought: mind-wandering, creativity, and dreaming (pp. 97–111). new york: oxford university press. millière, r. (2017). looking for the self: phenomenology, neurophysiology and philosophical significance of drug-induced ego dissolution. frontiers in human neuroscience, 11. https://doi.org/10.3389/fnhum.2017.00245 millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 millière, r., & metzinger, t. (2020). editorial introduction. philosophy and the mind sciences, 1(i), 1. https://doi.org/10.33735/phimisci.2020.i.50 oestreich, l. k., mifsud, n. g., ford, j. m., roach, b. j., mathalon, d. h., & whitford, t. j. (2015). subnormal sensory attenuation to self-generated speech in schizotypy: electrophysiological evidence for a “continuum of psychosis”. international journal of psychophysiology, 97(2), 131–138. https://doi.org/10.1016/j.ijpsycho.2015.05.014 pareés, i., brown, h., nuruki, a., adams, r. a., davare, m., bhatia, k. p., & edwards, m. j. (2014). loss of sensory attenuation in patients with functional (psychogenic) movement disorders. brain, 137(11), 2916–2921. https://doi.org/10.1093/brain/awu237 parr, t., & friston, k. j. (2017). working memory, attention, and salience in active inference. scientific reports, 7(1), 14678. https://doi.org/10.1038/s41598-017-15249-0 parr, t., rees, g., & friston, k. j. (2018). computational neuropsychology and bayesian inference. frontiers in human neuroscience, 12(61). https://doi.org/10.3389/fnhum.2018.00061 pellicano, e., & burr, d. (2012). when the world becomes “too real”: a bayesian explanation of autistic perception. trends in cognitive science, 16, 504–510. https://doi.org/10.1016/j.tics.2012.08.009 posner, m. i., snyder, c. r., & davidson, b. j. (1980). attention and the detection of signals. journal of experimental psychology: general, 109(2), 160–174. https://doi.org/10.1037/0096-3445.109.2.160 powers, a. r., mathys, c., & corlett, p. r. (2017). pavlovian conditioning-induced hallucinations result from overweighting of perceptual priors. science, 357, 596–600. https://doi.org/10.1126/science.aan3458 powers, w. t. (2005). behavior: the control of perception. new canaan, ct: benchmark publications. prinz, w. (1997). perception and action planning. european journal of cognitive psychology, 9(2), 129–154. https://doi.org/10.1080/713752551 saks, e. r. (2007). the center cannot hold: my journey through madness. new york, ny: hachette books. limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11097-005-9015-6 https://doi.org/10.1093/nc/nix016 https://doi.org/10.3389/fnhum.2014.00946 https://doi.org/10.3389/fnhum.2013.00547 https://doi.org/10.3389/fpsyg.2018.00643 https://doi.org/10.1093/cercor/bhz192 https://doi.org/10.1093/cercor/bhz111 https://doi.org/10.1023/b:phen.0000007366.42918.eb https://doi.org/10.4249/scholarpedia.4174 https://doi.org/10.3389/fpsyg.2013.00746 https://doi.org/10.3389/fnhum.2017.00245 https://doi.org/10.33735/phimisci.2020.i.48 https://doi.org/10.33735/phimisci.2020.i.50 https://doi.org/10.1016/j.ijpsycho.2015.05.014 https://doi.org/10.1093/brain/awu237 https://doi.org/10.1038/s41598-017-15249-0 https://doi.org/10.3389/fnhum.2018.00061 https://doi.org/10.1016/j.tics.2012.08.009 https://doi.org/10.1037/0096-3445.109.2.160 https://doi.org/10.1126/science.aan3458 https://doi.org/10.1080/713752551 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakub limanowski and karl friston 16 schooler, j. (2014). bridging the objective/subjective divide: towards a meta-perspective of science and experience. in t. metzinger & j. windt (eds.), open mind. https://doi.org/10.15502/9783958570405 seth, a. (2009). explanatory correlates of consciousness: theoretical and computational challenges. cognitive computation, 1(1), 50–63. https://doi.org/10.1007/s12559-009-9007-x seth, a. k. (2014). a predictive processing theory of sensorimotor contingencies: explaining the puzzle of perceptual presence and its absence in synesthesia. cognitive neuroscience, 5(2), 97–118. https://doi.org/10.1080/17588928.2013.877880 seth, a. k., & friston, k. j. (2016). active interoceptive inference and the emotional brain. philosophical transactions of the royal society b: biological sciences, 371(1708), 20160007. https://doi.org/10.1098/rstb.2016.0007 seth, a. k., suzuki, k., & critchley, h. d. (2011). an interoceptive predictive coding model of conscious presence. frontiers in psychology, 2. https://doi.org/10.3389/fpsyg.2011.00395 seth, a. k., & tsakiris, m. (2018). being a beast machine: the somatic basis of selfhood. trends in cognitive sciences, 22(11), 969–981. https://doi.org/10.1016/j.tics.2018.08.008 sterzer, p., adams, r. a., fletcher, p., frith, c., lawrie, s. m., muckli, l., & corlett, p. r. (2018). the predictive coding account of psychosis. biological psychiatry, 84(9), 634–643. https://doi.org/10.1016/j.biopsych.2018.05.015 tabor, a., & burr, c. (2019). bayesian learning models of pain: a call to action. current opinion in behavioral sciences, 26, 54–61. https://doi.org/10.1016/j.cobeha.2018.10.006 van de cruys, s., evers, k., van der hallen, r., van eylen, l., boets, b., de-wit, l., & wagemans, j. (2014). precise minds in uncertain worlds: predictive coding in autism. psychological review, 121, 649–675. https://doi.org/10.1037/a0037665 verschure, p. f. (2016). synthetic consciousness: the distributed adaptive control perspective. philosophical transactions of the royal society b: biological sciences, 371(1701), 20150448. https://doi.org/10.1098/rstb.2015.0448 wiese, w. (2019). explaining the enduring intuition of substantiality: the phenomenal self as an abstract ‘salience object’. journal of consciousness studies, 26(3), 64–87. retrieved from https://www.ingentaconnect.com/content/imp/jcs/2019/00000026/f0020003/art00004 wiese, w., & metzinger, t. (2017). vanilla pp for philosophers: a primer on predictive processing. in t. metzinger & w. wiese (eds.), philosophy and predictive processing. frankfurt am main: mind group. zahavi, d. (2014). self and other: exploring subjectivity, empathy, and shame. oxford: oxford university press. zahavi, d., & parnas, j. (1998). phenomenal consciousness and self-awareness: a phenomenological critique of representational theory. journal of consciousness studies, 5(5), 687–705. retrieved from https://www.ingentaconnect.com/content/imp/jcs/1998/00000005/f0020005/903 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.15502/9783958570405 https://doi.org/10.1007/s12559-009-9007-x https://doi.org/10.1080/17588928.2013.877880 https://doi.org/10.1098/rstb.2016.0007 https://doi.org/10.3389/fpsyg.2011.00395 https://doi.org/10.1016/j.tics.2018.08.008 https://doi.org/10.1016/j.biopsych.2018.05.015 https://doi.org/10.1016/j.cobeha.2018.10.006 https://doi.org/10.1037/a0037665 https://doi.org/10.1098/rstb.2015.0448 https://www.ingentaconnect.com/content/imp/jcs/2019/00000026/f0020003/art00004 https://www.ingentaconnect.com/content/imp/jcs/1998/00000005/f0020005/903 https://doi.org/10.33735/phimisci.2020.i.35 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction self-modelling based on predictive processing ``self-flattening'': the relationship between deep active inference and selfhood ``self-attenuation'': the importance of expected precision conclusion and outlook are concepts a natural kind? on concept eliminativism. are concepts a natural kind? on concept eliminativism guido löhra(loehrg@icloud.com) abstract concept eliminativists argue that we should eliminate the term ‘concept’ from our vocabulary in psychology because there is no single natural kind that is picked out by it. i argue that the most developed version of concept eliminativism by edouard machery depends on the assumption that concepts are defined as stable and context-independent bodies of information. it is this assumption that leads machery to eliminativism and it is an assumption we have reason to reject. another assumption that leads to the eliminativist conclusion and that we have reason to reject is that the type of content represented in long-term memory is the relevant property based on which we should individuate certain natural kinds in cognitive psychology. finally, i argue that certain pieces of information are functionally integrated enough to meet the conditions for being a natural kind. keywords concepts ∙ concept eliminativism ∙ concept pluralism ∙ concept hybridism ∙ contextualism ∙ natural kinds 1 introduction according to margolis and laurence (margolis & laurence, 2018, p. 1): concepts are the building blocks of thoughts. consequently, they are crucial to such psychological processes as categorization, inference, memory, learning, and decision-making. this much is relatively uncontroversial. if we take margolis and laurence’s quote as a starting point, it becomes immediately clear why the notion of concept is very dear to philosophers and cognitive scientists alike: the majority of them consider it indispensable to studying thought and language. unsurprisingly, the proposal to eliminate the term ‘concept’ from cognitive science (machery, 2005) has generated much opposition (machery, 2010; aeindhoven university of technology. löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-7028-3515 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org guido löhr 2 piccinini & scott, 2006; rice, 2016; taylor & vickers, 2017; taylor & vosgerau, 2019; vicente & martínez manrique, 2016; weiskopf, 2009). it may even be considered a serious challenge to some of philosophers’ favorite methods of analyzing or engineering concepts (blackburn, 1999; cappelen, 2018; jackson, 1998). however, machery also argues that philosophers and psychologists employ different notions of concept (machery, 2009). concepts in philosophy can be characterized by margolis and laurence’s quote above. the notion of concept common in psychology is less focused on building blocks of propositional attitudes and truth-conditional content. instead, it aims at explaining categorization and higher cognition (löhr, 2020; rey, 2010). of course, both notions are related and may both be necessary for a full account of cognition. still, if they do turn out to be different, as machery argues, then there are also two different kinds of concept eliminativism, one being more revisionary than the other. the kind of concept eliminativism defended by machery (2009) is a philosophically modest one. he argues that the term ‘concept’ in cognitive science picks out different kinds of stable bodies of information that are too heterogeneous to be considered a single natural kind. assuming that scientists intend to discover natural kinds, machery concludes that we should abandon the term ‘concept’ and instead use more fine-grained notions that do denote natural kinds. this modest kind of concept eliminativism will be discussed here. a more radical version that aims to eliminate the notion of concept as constituents of propositional attitudes will not be addressed. as far as i know, machery is skeptical of this notion as well. however, he does not explicitly argue for its elimination. 1 the question of whether we should be concept eliminativists, hybridists, or pluralists of concepts in psychology has been critically discussed in various places (piccinini & scott, 2006; rice, 2016; samuels & ferreira, 2010; vicente & martínez manrique, 2016; weiskopf, 2009). in this paper, i focus on the eliminativist argument in more detail. i am interested in the question of whether eliminativists have given us sufficient reason to accept that concepts in cognitive science are not a natural kind. to evaluate this psychological concept eliminativism, i focus on its most developed version by machery (2009). however, my argument will have consequences for pluralism and hybridism as well. 2 i reconstruct machery’s argument in the following way: (1) ‘concept’ in psychology refers to the “body of knowledge about x that is stored in long-term memory and that is used by default in the processes underlying most, if not all, higher cognitive competences 1but see machery, 2017 for a discussion of the more philosophical notion. 2note that the question of whether the term ‘concept’ in cognitive science picks out a natural kind or not (whether it should be eliminated) is different from the question of whether the term picks out a single hybrid kind or several kinds. we can argue that ‘concept’ picks out a hybrid and still recommend its elimination based on the view that this hybrid is too heterogenous to fulfill the conditions for being a natural kind. löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org are concepts a natural kind? on concept eliminativism 3 when these processes result in judgments about x.” machery (2009, p. 12) (see section 2.1) (2) the description labeled ‘concept’ picks out three fundamentally distinct types of bodies of knowledge: prototypes, exemplars, and theories. (see section 2.2) –these types of bodies of knowledge are individuated by their type of content and type of computation. prototypes store information about typical features, while exemplars store information about exemplars. theories store information about causal features. (3) natural kinds are individuated by means of sets of scientifically relevant properties. in the psychology of concepts, the relevant properties are whatever explains higher cognitive competencies, in particular, what i call ‘regularity effects’ (typicality effects, exemplar effects, and causal effects). (see section 2.3) (4) what explains regularity effects are the type of content of a body of information and the computational process that calculates similarity to this content. type of content and type of computation are therefore scientifically relevant properties. (see section 2.3) (5) if (2–4), then prototypes, exemplars, and theories are each a natural kind. (see section 2.4) (6) if (1,5), then the term ‘concept’ picks out three different natural kinds. (see section 2.4) (7) if (6), then this is confusing psychologists. therefore, we should eliminate the term ‘concept’ and replace it with more specific terminology like prototype, exemplar, and theory. (see section 2.4) i argue that we have reason to reject the first premise (section 3.1). if we reject (1), then ‘concept’ in cognitive science can in theory pick out more than bodies of information retrieved by default. i then argue that type of content of representations is not the critical property that we should be looking for when looking for a natural kind in cognitive science. instead, we should focus on the way information in long-term memory is functionally integrated, which determines how and when it is retrieved (section 3.2). prototypes, exemplars, and theories do not however differ in terms of the way their stored pieces of information are functionally integrated, at least not in the relevant sense that takes deliberation into account (section 3.3). i conclude that we have no reason to be eliminativists at this point (section 3.4). löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org guido löhr 4 2 the eliminativist argument 2.1 machery’s concept of concept in psychology to understand machery’s proposal to eliminate the notion of concept from psychology, it is crucial to first understand what exactly he takes the relevant notion of concept to be. according to machery (machery, 2009, p. 12): a concept of x is a body of knowledge about x that is stored in long-term memory and that is used by default [immediately and in a context-independent manner] in the processes underlying most, if not all, higher cognitive competences when these processes result in judgments about x. (my emphasis). note that machery, in this quote, is proposing a description of how psychologists in fact use the term ‘concept’. he does not yet propose a theory of what concepts are or of what is picked out by his description. machery argues that there is no single natural kind that is picked out by the description. instead, he argues that there are at least three distinct kinds picked out by it – prototypes, exemplars, and theories (see below). 2.2 ‘concept’ picks out three distinct types of bodies of information machery (2009) argues that the notion of concept that he takes to be central in psychology (“a body of knowledge about x that is stored in long-term memory and that is used by default in the processes underlying most, if not all, higher cognitive competences”) picks out different kinds of representations. in other words, he argues that there are three or more distinct types of bodies of information retrieved by default. machery (in chapter 4 of doing without concepts) identifies these as what psychologists have called ‘prototypes’, ‘exemplars’ and ‘theories’. following machery, they are individuated primarily in terms of their type of content. according to machery (machery, 2009, p. 84): [p]rototypes consist of knowledge about properties that objects either possess or do not possess or about properties that objects possess to some degree. the property having wings is an instance of the first type of property. the property being sweet is an instance of the second type of property: a substance can be more or less sweet. 3 3he adds: “according to some theories, prototypes merely store some knowledge about which properties are typical (or cue-valid). according to other theories, prototypes also store the degree of typicality (or of cue-validity) of the typical (or cue-valid) properties.” (p. 84), but this technicality need not concern us here. löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org are concepts a natural kind? on concept eliminativism 5 furthermore, according to machery, these bodies of information are used in cognitive processes that compute the similarity between the properties “summarized” in the prototype and the properties detected in the object in question. 4 he takes the similarity measurement in the case of prototypes to be linear. 5 according to machery (machery, 2009, p. 90): [a]n exemplar is a body of knowledge about the properties believed to be possessed by a particular member of a class. when we categorize, when we reason, and so on, we have by default in mind a set of exemplars (or, in some models, an exemplar drawn from a set of exemplars stored in long-term memory). with respect to their use in higher cognitive processes, machery (machery, 2009, p. 96) argues that exemplar-based models assume that cognitive processes involve the computation of the similarity between exemplars and other representations. (…) the second central property of exemplar-based models of cognitive processes is that the similarity measure is usually supposed to be non-linear. 6 finally, machery takes theories to be sets of beliefs retrieved by default about nomological, causal, functional, or generic properties associated with a category. in machery’s (machery, 2009, p. 101) words: thus, a theoretical concept is supposed to store some nomological, causal, functional, and/or generic knowledge about the members of its extension. for instance, a theoretical concept of dog stores some nomological, causal, functional, and/or generic knowledge about dogs. according to machery, theories are employed in cognitive processes that are different from similarity measures and more akin to causal reasoning. 7 4he writes: “their most important property is that cognitive processes are assumed to involve the computation of the similarity between prototypes and other representations. (…) the second property of these models is that the similarity computation is usually assumed to be linear.” (machery, 2009, p. 90) 5this means that the more representations two sets of representations have in common, the more similar are they judged as. 6machery (machery, 2009, p. 98) summarizes this feature of exemplar models well: “an object that is extremely similar to a specific known category member, but only moderately similar to others is more likely to be categorized as a category member than an object that is moderately similar to most known category members.” 7machery (machery, 2009, p. 106) summarizes an example from murphy & medin (1985): “if at a party, a guest jumps in the swimming pool with her clothes on, we may conclude that she is drunk. this categorization judgment does not result from matching the concept of drunken people with a representation of this guest. on the contrary, we infer that the most plausible explanation of the behavior of this guest is that she is drunk.” löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org guido löhr 6 at this point, i want to emphasize that machery does not just argue that we have intentional states with the aforementioned three kinds of content or that there may be different kinds of processes that they are used in. few people would deny, for example, that we represent certain properties as typical or essential for a category, or that we represent exemplars of a category. the crucial point is that machery argues that there is evidence that they are default bodies of knowledge. other bodies of information do not belong to the respective concept but to what machery calls ‘background knowledge’. 8 2.3 a natural kind in science is individuated in terms of scientifically relevant properties based on richard boyd’s (boyd, 1991) criteria for being a natural or scientific kind, machery (machery, 2005, p. 232) argues that bodies of information retrieved by default should only be considered a natural kind if they meet the following set of conditions: a class c of entities is a natural kind if and only if there is a large set of scientifically relevant properties such that c is the maximal class whose members tend to share these properties because of some causal mechanism. 9 what could be relevant (to the psychologist) properties that bodies of information must have in common to be considered a natural kind? generally, cognitive psychologists study bodies of information based on a number of dimensions:10 a) their content (what their constituents represent, e.g., whether they represent a feature as typical or necessary for category membership) b) the cognitive processes in which they are used (e.g., whether they are used for linear or non-linear similarity measurements) c) their functional integration (roughly, how the pieces of information are stored or connected in long-term memory) 8machery agrees that there is no reason to stop at positing just three fundamentally different kinds of bodies of information individuated in terms of what kind of information they store. he does not elaborate on what other fundamental sets of beliefs there could be, but i take it that the following two may be candidates: for example, there has been some evidence for beliefs about situations associated with a category (wiemer-hastings & xu, 2005), as well as beliefs about feelings associated with a category (kousta et al., 2011). both types of representations may form their own sets of default beliefs that function to aid categorization. another possibility could be that image-like modal representations could form its own natural kind (but this is certainly not machery’s view). 9machery (2009) says nothing about the second requirement for being a natural kind (that these bodies of knowledge possess shared properties because of some causal mechanism). 10for this list, i follow loosely machery (machery, 2009, p. 77). löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org are concepts a natural kind? on concept eliminativism 7 d) their function (what competencies they explain, e.g., categorization, linguistic understanding, analogy making, etc.) e) how they are acquired (e.g., whether they are learned or not) f) their format (e.g., whether this content is represented in an image-like perceptual or language-like amodal format) g) their neural implementation the crucial step in machery’s argument is that the dimensions (a) and (b), i.e., content and processes are the critical ones when it comes to explaining higher cognitive competencies, in particular categorization (see chapter 4 in doing without concepts). the main reason he takes this to be the case is that psychologists rely on them to explain categorization. the key idea is that we store information about the world based on previous experiences. when we see a new object, we compare the properties of this object with this set of information, i.e., the information about properties in the world we cluster together. a major part of explaining categorization, according to machery, is to explain what generates three types of what i call ‘regularity effects’. those are the effects found when studying the knowledge psychologists believe we rely on when comparing our memory to our perceptual input in classification processes. it has been argued that these regularity effects show that we do not normally seem to carve up the world by means of necessary and sufficient conditions. one kind of regularity effect is that representations of typical features of a category can have a processing advantage (‘typicality effect’). for example, animals that have many typical features of the category ‘bird’ are recognized much more quickly as birds than atypical exemplars of the same kind of animal (murphy, 2004). it is generally thought that prototype theory and its posited prototypes are especially good at explaining typicality effects. a second type of effect, what i call the ‘exemplar effect’, is roughly based on the observation that we often rely on remembered exemplars of a category (say, looney tunes’ tweety for bird) – rather than typical features – to judge category membership. exemplars were posited by exemplar theorists to explain especially such exemplar effects. third, what i call the ‘causal effect’ is based on the observation that people tend to judge category membership not just by means of its superficial features but also by more theory-like causal or essentialist properties, e.g., that birds are not mammals or that the atomic number of gold is 79 (machery, 2009; murphy, 2004; prinz, 2002). theories are posited by theory theorists to explain these causal effects. 2.4 the eliminativist conclusion the eliminativist conclusion is now easy to derive. machery argues that the term ‘concept’ in psychology picks out three types of body of information retrieved by löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org guido löhr 8 default (i.e., immediately and in a context-insensitive way): prototypes, exemplars, and theories. these bodies are essentially distinguished in terms of their different types of content and cognitive processing. machery argues that they do not differ with respect to (c) their functional organization11, (d) their function12 , and (f) their format13. machery does not say much about potential differences concerning (e) how they are acquired or (g) their neural implementation.14 why do (a) type of content and (b) type of cognitive processing matter the most? according to machery, these are the critical features of bodies of information that psychologists use to explain regularity effects. again, prototype theory was developed to explain typicality effects. exemplars are posited to explain exemplar effects and theories are posited to explain causal effects. since type of content and type of cognitive processes explain regularity effects, they are the ones that are scientifically relevant for individuating natural kinds in cognitive psychology. so, ‘concept’ picks out not one but three different natural kinds – prototypes, exemplars, and theories. finally, if ‘concept’ picks out three distinct natural kinds, this is confusing psychologists. it has motivated them to pointlessly debate what ‘concept’ properly refers to – whether concepts are prototypes, exemplars, or theories. we should instead acknowledge that they all are natural kinds and abandon the term ‘concept’ and replace it with ‘prototype’, ‘exemplar’ and ‘theory’. note that while the purpose of the remainder of this paper is to reject premises 2 and 4 of this argument, i agree with machery that the question of whether concepts are prototypes, theories, or exemplars is pointless. however, we can reach this attractive conclusion without embracing eliminativism. as we will see in the next section, while prototypes, exemplars, and theories are individuated in terms of the type of content they store, we should not do the same with the notion of concept in psychology. 3 the eliminativist argument debunked 3.1 machery’s concept of concept (against premise 1) significant weight in machery’s argument rests on the assumption that the term ‘concept’ picks out only those bodies of information that are retrieved by default. 11according to machery, the different pieces of information that make up a single default body of information are coordinated in the sense that the retrieval of one is accompanied by the retrieval of the other. 12machery (2009) takes all fundamental kinds of information to fulfill a number of different cognitive competencies, which is an essential part of his “heterogeneity hypothesis”. 13machery (2016) is very explicit that he endorses a monist amodal view of concepts. 14however, there is no reason to think that machery takes the different pieces of knowledge retrieved by default to differ fundamentally based on these last two dimensions. he also does not seem to think that they explain regularity effects and are therefore not the relevant property we should care about. löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org are concepts a natural kind? on concept eliminativism 9 what exactly are machery’s reasons for restricting his notion of concept in cognitive science to only those bodies of information that are retrieved in an immediate and context-independent manner? this question is interesting because if the term ‘concept’ is extensionally allowed to pick out more context-dependent structures, we might find a single natural kind after all. the reason why machery includes the restriction is that (i) he takes contextinsensitivity to be required to explain the observed stability or regularity of our behaviors (machery, 2009, p. 24) and (ii), more simply, because he takes this notion to be the one used by psychologists (machery, 2005, pp. 196–197). both arguments can be rejected as reasons for restricting the concept of concept in psychology in the way machery suggests. that concepts have to be stable in the sense of being context-independent is not new (fodor, 1998; rey, 2010). many philosophers of concepts assume that concepts – as compositional constituents of propositional attitudes – have to be context-independent to account for the fact that we can successfully reason and communicate. if the concept of table in the premise “all tables are made of wood” differed from the concept of table in the premise “x is a table”, we could not be able to rationally infer that x is made of wood.15 however, if, as machery says, the term ‘concept’ in cognitive science does not pick out the kind of entity that constitutes our premises or beliefs and instead denotes bodies of beliefs (not their constituents) that underly categorization, then ‘concepts’ not need to be as strictly stable as required by the logician or philosopher of language. note that this can also be agreed upon by those philosophers who argue that philosophers and psychologists refer to the same kind of thing when using the term ‘concept’ (prinz, 2002). for example, the bodies of information that underlie categorization may be highly unstable even if they still have the same “intentional content” (prinz, 2002), or “r-content” (del pinal, 2015). this is how prinz and del pinal can defend the view that the same concept may be instantiated or realized by different bodies of beliefs or belief-like bodies of information. thus, there is no conceptual reason why we could not explain regularity effects by means of context-dependent structures. hence, stability should not feature in the definition of the notion of concept in cognitive science. some have argued that contextualists can even better account for regularity effects (casasanto & lupyan, 2015; kiefer, 2018; lalumera, 2010; löhr, 2017). context-dependent structures can account for regularity effects (typicality, frequency, and exemplar effects) assuming that some pieces of knowledge in some contexts may usually be more available than others. for example, the belief that dogs bark may be reliably available in most if not all contexts. all that so-called contextualists of categorization are committed to deny is that most sets of information underlying higher cognition 15note that these strong requirements of context-independence of concepts – as constituents of truth-conditional thoughts – may not always be met by ordinary minds and speakers. we may be able to converse in ways that facilitate coordination in joint action without necessarily always having identical concept types. löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org guido löhr 10 are stable and retrieved by default come what may as a set, i.e., whether or not this body of information is largely inappropriate for the respective sentence or discourse context. second, it is not clear whether machery’s description of the use of ‘concept’ in cognitive science is accurate. many leading contemporary psychologists working on categorization like (barsalou, 1999; kiefer, 2018; pulvermüller, 2013; rosch, 2011) now prefer a more contextualist notion of concept, according to which the body of knowledge underlying higher cognitive competencies may change from situation to situation. even rosch’s not-so-recent writings (rosch, 1999, p. 69) suggest that the field has moved on to a more flexible notion, arguing that “context effects show that category prototypes and graded structure are not pre-stored as such in the mind, but rather are created anew each time ’on the fly’ from more basic features or other mental structures.” note that the reader need not accept contextualism to reject this premise. what i wanted to emphasize is that there is neither a conceptual nor an empirical reason to presuppose invariantism in this discussion. it should not be presupposed (if you agree with the reasons given above) and instead, we should leave it open to empirical investigation whether categorization for example is best explained by stable bodies of information retrieved by default or whether it is best explained by context-dependent bodies. the notion of concept in psychology should instead be defined as “bodies of beliefs or information underlying higher cognition”. 3.2 the relevant property is not content (against premise 4) the second crucial premise in machery’s argument in favor of eliminativism is that the relevant property that we should look for when searching for natural kinds in cognitive psychology is primarily the type of content of bodies of information. this is because type of content, along with type of cognitive processes, according to machery, explain what we are interested in: categorization and other higher-order cognitive abilities like reasoning or decision-making. the problem is that type of content and type of computations do not play the role of explaining our higher cognitive abilities that machery thinks they play. let’s focus again on regularity effects. believing that dogs typically bark (prototypes), or that rex is a dog (exemplars), does not explain why we recognize typical dogs faster than atypical dogs or why we recognize certain dogs via exemplars. what explains why we immediately (or after deliberation) recognize rex as a dog is that the information that barking things are usually dogs or that fido looks like rex (or that rex has a certain genetic profile) is readily available in the relevant context.16 even if we stored (somewhere in our background knowledge) the false informa16“being readily available in the relevant context” does not imply the truth of contextualism. invariantists also assume of course that the relevant information is available in the relevant context. if you are an invariantist like machery, you think that only part of the knowledge we normally think is part of a single concept is readily available in the relevant context. thus, in most contexts, löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org are concepts a natural kind? on concept eliminativism 11 tion that things that bark are cats or that rex is a cat, we would still display the same behavior if these pieces of information were suppressed in the right context. what matters to explain categorization is how our information is functionally integrated (vicente & martínez manrique, 2016) in long-term memory such that certain pieces of information become available at the right moments while others are suppressed. without the belief that dogs bark, we would probably have a hard time recognizing rex as a dog, but we should not conclude from this that the content of the information that dogs bark is what makes this piece of information special or interesting. what is important is how this content is accessed and organized in long-term memory such that it is available when we need it. consider for example machery’s (machery, 2005, p. 89) example of a prototype model of apple. this model contains several pieces of information about the diagnostic properties of apples (properties that easily set them apart from other properties, say red and round) including some information about how diagnostic this information is taken to be (largely based on typicality ratings). we can then say that one of the pieces of information stored in this prototype has the complex content that apples are typically red and round. this kind of content distinguishes prototypes from, say, exemplars, which store beliefs with a different kind of content, say, the content that the piece of fruit i ate in the morning is an apple (this might be an exemplar-like piece of information) or that apples have seeds because this is how apple trees reproduce (this would be a theory-like piece of information). intuitively, this model of our information of apples makes sense. one might think that when looking at an object, i tell you it is probably an apple because apples typically have this shape and color. this is a reliable reason for me to classify this object as an apple. it is the content of my belief that justifies my classification of the belief that apples typically look that way. however, possessing a body of information about whether apples typically or necessarily have a certain shape helps me classify things as apples only if this information is available to me in the right context. again, what explains categorizing something as an apple is that apples are red is more readily available than apples can be black and rotten in this context. and it is this difference in availability in typical apple contexts that is represented by cue validity ratings or typicality ratings and not the content of a piece of information. again, i might even have the wrong modal information about apples and still make the right rough and ready classification in typical contexts (armstrong et al., 1983). i might believe that an apple necessarily has a certain shape or that it is never yellow. if both are suppressed at the right moment, i might still display the appropriate behavior. again, what is important and what is reflected in the cue validity ratings is not the content of our bodies of information but their level of accessibility in different contexts. in other words, what explains regularity effects is how long-term memory is organized or how pieces of information are functionally the machery-style invariantist and the contextualists will make the same predictions regarding to which information is available in which context. löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org guido löhr 12 related to one another or integrated in long-term memory (in the list of dimensions above this was dimension c). this organization can be studied by employing familiar psychological experiments, such as typicality ratings. whether we represent a feature as typical or necessary is important when it comes to justifying our categorizations (e.g., when asked why we think that it is a bird). it is also important to explain why the mind is organized such that it is information about typical features of birds, rather than atypical features for example, that are more readily available. however, when it comes to explaining regularity effects, accessibility to this content is what counts. an answer to the question of why information p becomes available rather than information q likely involves the structure of the world – because birds and not dogs have wings and you asked me about birds and not dogs. that regularity effects are not explained by the type of content of the information stored, but by which bodies of information come to mind most quickly (to put it simply) may come as a surprise because psychologists often frame their theories, like prototype theory, in terms of type of content. for example, what distinguishes prototypes from exemplars is that the former represents information about typical properties and the latter information about exemplars. this is not problematic. prototype theories and exemplar theories do differ in terms of their ideas about what content is stored and what processes this stored content is used in.17 machery is right about that. but the question here is not how to individuate different kinds of bodies of information in long-term memory and not even how to individuate concepts in psychology. the question is how to individuate natural kinds in cognitive psychology. the minimal common denominator is that they should be individuated in terms of what explains typicality, exemplar, and causal effects. again, by ruling out a contextualist notion of concepts from the beginning (by definition), machery neglects the possibility that the prototypes, exemplars, and theories of birds may all have at least one thing in common besides being about the same kind: they are made most available when we come into causal contact with this kind even if this availability is context-sensitive and not default-like.18 17and it is of course interesting to ask why information about features of a category we take to be typical come to mind more quickly in some situations and why information about more theoretical features of the same category may be more available in others. the answer may simply be that features we take to be typical are, in fact, typical of the respective class of objects and that typical features are highly diagnostic and therefore useful for everyday categorizations. the most interesting fact however remains thereby unexplained: why and how does the system adapt to contexts by prioritizing certain types of contents in different situations? 18note that my proposed argumentative strategy to look for dimensions other than (a) and (b) to judge whether we should be eliminativists or not is not new. it has been suggested that a radical form of so-called “concept empiricism” or “neo empiricism” (as machery calls it), i.e., the view that all or most concepts are perceptional symbols (barsalou, 1999; prinz, 2002; pulvermüller, 2013) may, in principle, “forestall concept eliminativism” (prinz, 2002). prinz suggests that all bodies of knowledge underlying higher cognition could have something in common after all: all are perceptual symbols (dimension f). however, even though machery (2009) devotes much löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org are concepts a natural kind? on concept eliminativism 13 3.3 higher cognitive competencies and deliberation (against premise 3) higher cognition involves more than rough and ready quick categorization and concepts should explain this. what is just as important is this: explain judgments about cats, dogs, or apples when deliberating. what is just as important is explaining why we believe that a typical dog (e.g., a bulldog), an atypical exemplar of a dog (e.g., a komondor), or a burned dog with only dog dna, are all of the same kind – dogs. what is just as important is to explain why the different pieces of information are intuitively and inferentially connected and coordinated. again, what explains this is not the type of content or computational use of bodies of information. what explains this is how the information we have is integrated in memory – how accessible it is in different contexts and how it is inferentially related to other pieces of information. machery denies that different fundamental types of information in long-term memory (prototypes, exemplars, theories) are usually “coordinated” (chapter 3 of doing without concepts) based on the observation that we can hold inconsistent beliefs about the same category.19 for example, we often hold to be true that tomatoes are vegetables in one context – relying on prototypes and exemplars – and that they are fruit in another context – where theory-like information is more readily available. machery is certainly right to point out that the conceptual system is far from perfectly consistent. we often hold contradictory beliefs. however, machery tends to downplay the fact that we often cease to hold such contradictory beliefs after deliberation. again, psychologists have to explain not just fast categorization but categorization after deliberation. moreover, machery tends to downplay the important fact that there is actually a tremendous amount of coordination and consistency. i currently classify gold and water into fundamentally different categories. this could be wrong and both categories may be co-referential. however, i currently have no reason to hold this belief as gold and water present themselves in fundamentally different ways (they have very different modes of presentation). it would be irrational to merge the two categories without further evidence. thus, it seems that my beliefs about gold and water are highly coordinated space attempting to refute concept empiricism in order to support his view that ‘concept’ picks out fundamentally different kinds of representations, recall that he, too, argues for a monism of format on many occasions. in particular, machery (2016) argues that the bodies of knowledge underlying higher cognition are represented in a purely amodal format. moreover, prinz’ (2010) attempt to avoid machery’s eliminativism has, despite its perhaps somewhat misleading title, little to do with finding a common format (he focuses on organization of long-term memory as i do in this paper) and machery’s eliminativist argument is not committed to any theory of format. 19“these parts are coordinated: the parts of a given concept do not produce inconsistent outcomes, for instance, inconsistent categorization judgments.” coordination is not the same as linkage: “…when one of the parts is used, say, to categorize, we can ipso facto use the other parts of the concept for other purposes.” (machery, 2009, p. 64) löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org guido löhr 14 this familiar fregean insight must be reflected in a theory of how my knowledge about the world is organized in my mind. it justifies the assumption that the bodies of information we rely on to recognize objects as gold and water are not functionally connected in the mind (at least not in the relevant sense that explains how we categorize both) and that, unlike our different contradictory assumptions about tomatoes, they do therefore not belong to the same concept. similarly, even if we retrieve different information about tomatoes by default, this does not mean that we have two distinct concepts of tomatoes. the reason is simply that we are aware that tomatoes are not really vegetables. once we give up on the assumption that concepts have to be default structures in the mind, we can acknowledge that the beliefs that tomatoes are fruit and that they are vegetables are linked in a way that, when deliberating, we realize that one of those pieces of information is false. this does not mean that both belong to separate concepts of tomato (concept in the psychological sense). we can take the idea of deliberation as an additional factor in individuating concepts further. at this point, it is perhaps important to stress that i agree with machery’s key observation that different parts of a concept are sometimes not coordinated and, therefore, produce inconsistent outcomes. but once we lose the definitional requirement of concepts as bodies of information retrieved by default, we can think of more creative ways of individuating concepts in psychology. we might think of deliberation as a long process that could even include collective deliberation – a kind of process of conceptual engineering or revision (cappelen, 2018; machery, 2017) to increase coordination between different parts of a concept. for example, we can think of tomatoes as being (botanically speaking) fruit and also as located in the vegetable section. once we think about it a bit, we do not need to hold the belief that tomatoes are vegetables and fruit. we can just think of tomatoes as fruit, which, for historical or practical reasons, are placed in the vegetable section of supermarkets. we improve our concept of tomato by decreasing inconsistency and we do this in a way that reflects the fact that the different pieces of information are connected in a way that the pieces of information associated with gold and water are not. 3.4 why concepts might still be a natural kind (against the eliminativist conclusion) it is one thing to consider concepts hybrids of prototypes, exemplars, and theories. it is another thing to argue that they are natural kinds. to return to the question of whether concepts are a natural kind, recall machery’s (machery, 2009, p. 241) adaption of boyd’s three conditions for natural kindness. 1. there is a large set of properties that these bodies of knowledge tend to possess. löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org are concepts a natural kind? on concept eliminativism 15 2. these bodies of knowledge possess these properties because of some causal mechanism. 3. this set of properties is specific to this class of bodies of knowledge. i take it that there is currently no reason to think that these conditions are not met in the case of the concept of concept in cognitive science. a concept in cognitive science is a body of information that underlies higher cognition and that explains how we apply a category x to an object, for example. this body of information has to have at least one central property to be considered the concept of x (there might be others): it has to be functionally integrated in the right way. long-term memory is apparently functionally organized such that it prioritizes different bodies of information in different contexts, but it is also organized such that certain bodies of information that may seem disconnected under time pressure (by default), actually turn out to be inferentially related, at least when deliberating. even if they are not yet strongly related, we will most likely relate them in a consistent way when engaging in conceptual deliberation – in a way in which we would never connect our ideas about gold and water. and this simple insight, i take it, justifies that our prototypes, exemplars, and theories of x might all belong to the same concept of x and that our concept of x is a scientifically interesting kind worth positing and studying.20 at least, machery has given us no reason to reject this more conservative assumption. deliberation merely generates a context that makes all the different pieces of information that we take to belong to the same category relevant. this results in the realization that these pieces of information are connected in our minds (even if very loosely), i.e., that they are part of the same concept. thus, it is not unreasonable to hypothesize that the same mechanism that explains why some information about the same category is prioritized in one context and suppressed in another also explains why these pieces of information are also connected at least when deliberating. again, at least machery has not given us reason to think otherwise. put differently, certain bodies of information we take to be associated with the same category share a property that is at least one crucial determiner of both categorization under time pressure and after deliberation. this feature is the kind of coordination that hybrid theorists like vicente & martínez manrique (2016) have been insisting on and called “functional integration”.21 20thanks to one of the reviewers for asking me to make the following clear: i do not here claim that concepts are really a natural kind, but that machery has not given us a good reason for abandoning the idea that there might be something like a scientifically interesting kind worth studying. this is an incredibly conservative and modest claim but is enough i think to make us question whether machery’s (incredibly thought-provoking) proposal is to be adopted by the scientific community. 21michel (2022) has suggested a more detailed model of such a hybrid in a predictive processing framework. the core idea is that the different types of knowledge are not “parts” of a hybrid but arise from different ways of processing a unified structure in which prediction error minimization löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org guido löhr 16 functional integration is often suppressed under time pressure, but we know that it is there because when deliberating (when we put ourselves in the right context), we realize that a conceptual connection exists. when engaging in conceptual deliberation, this connection might be discovered as the most plausible (given our background assumptions). the mechanism that explains how the different pieces get to be connected and how they get suppressed depending on context is likely the same mechanism that explains why the connection can be intuited if it becomes relevant, e.g., when deliberating. furthermore, such connected bodies of information can be contrasted clearly with categories that we consider completely disconnected, as we saw with the difference between water and gold.22 finally, the functional integration is specific not to a sub-kind of stable bodies of information, i.e., it is not specific to prototypes, exemplars, or theories respectively. instead, it applies to the bodies of information that we ordinarily consider concepts in cognitive science. our prototypes, exemplars, and theories of tomatoes are connected while our prototypes, exemplars, and theories of water and gold are disconnected upon reflection. this means, the concept of, say tomato, is the right level of analysis as the concept contains a body of information that is connected in psychologically interesting ways that explain both fast and slow categorization. 4 conclusion the main motivation for considering whether we should be concept eliminativists or concept hybridists is that it has become apparent in recent years that no single psychological theory of concepts can explain all the data on categorization (vicente & martínez manrique, 2016, p. 72). prototype theory explains typicality effects but does poorly when it comes to explaining categorization using more theory-like representations. exemplar theory explains how we classify an atypical but familiar object but is not ideal when it comes to classifying objects based on essential features. what distinguishes these different psychological theories is that they posit representations with different kinds of content. prototype theory posits representations of typical features, exemplar theory posits representations of exemplars, and theory theory posits representations of less superficial features. is being carried out. prototypes and exemplars correspond to vertical downward processing of that structure and theories to horizontal and upward processing. details do not matter here, the point is that we have a functionally integrated hybrid, which can be considered a natural kind. it is identified and accessed via the root node of that network structure. the structure is processed in different ways, depending on the context, giving rise to the appearance of different “types” of concepts. 22what if people keep insisting that sets of knowledge are not coordinated (machery & seppälä, 2011). then one could either engage in more testing whether they are really not coordinated or functionally integrated or simply conclude that then the subject has indeed two or more concepts of, say, tomato. this does not mean that prototypes, exemplars and theories are never functionally integrated. löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org are concepts a natural kind? on concept eliminativism 17 concept eliminativism in psychology is the view that there is no reason to think that the different types of bodies of information stored in long-term memory share enough of the relevant properties to label them all ‘concept’. i argued that the most developed eliminativist theory of concepts in cognitive science is based on two assumptions we have reason to reject. we should reject the idea that concepts are by definition stable bodies of information and that they should be individuated in terms of the type of content they represent. since there are good reasons to reject these two assumptions, we have little reason (all things considered) to endorse concept eliminativism, at least until a more convincing kind has presented itself. on the contrary, i argued that we have reliable and familiar (e.g., fregean) reasons to posit hybrid concepts and to assume that they are natural kinds after all. references armstrong, s. l., gleitman, l. r., & gleitman, h. (1983). what some concepts might not be. cognition, 13(3), 263–308. https://doi.org/10.1016/0010-0277(83)90012-4 barsalou, l. w. (1999). perceptions of perceptual symbols. behavioral and brain sciences, 22(4), 637–660. https://doi.org/10 .1017/s0140525x99532147 blackburn, s. (1999). think: a compelling introduction to philosophy. oup oxford. https://doi.org/10.1093/mind/110.438.430 boyd, r. (1991). realism, anti-foundationalism and the enthusiasm for natural kinds. philosophical studies, 61(1), 127–148. https://doi.org/10.1007/bf00385837 cappelen, h. (2018). fixing language: an essay on conceptual engineering. oxford university press. https://doi.org/10.1093/ oso/9780198814719.001.0001 casasanto, d., & lupyan, g. (2015). all concepts are ad hoc concepts. in e. margolis & s. laurence (eds.), the conceptual mind: new directions in the study of concepts (pp. 543–566). mit press. https://doi.org/10.7551/mitpress/9383.003.0031 del pinal, g. (2015). dual content semantics, privative adjectives, and dynamic compositionality. semantics and pragmatics, 8, 7–1. https://doi.org/10.3765/sp.8.7 fodor, j. a. (1998). concepts: where cognitive science went wrong. oxford university press. https://doi.org/10.1093/019823 6360.001.0001 jackson, f. (1998). from metaphysics to ethics: a defence of conceptual analysis. oxford university press. https://doi.org/10 .1093/0198250614.001.0001 kiefer, m. (2018). cognitive control over unconscious cognition: flexibility and generalizability of task set influences on subsequent masked semantic priming. psychological research, 83 1556–1570. https://doi.org/10.1007/s00426-0181011-x kousta, s.-t., vigliocco, g., vinson, d. p., andrews, m., & del campo, e. (2011). the representation of abstract words: why emotion matters. journal of experimental psychology: general, 140(1), 14–34. https://doi.org/10.1037/a0021446 lalumera, e. (2010). concepts are a functional kind. behavioral and brain sciences, 33(2-3), 217–218. https://doi.org/10.101 7/s0140525x10000403 löhr, g. (2017). abstract concepts, compositionality, and the contextualism-invariantism debate. philosophical psychology, 30(6), 689–710. https://doi.org/10.1080/09515089.2017.1296941 löhr, g. (2020). concepts and categorization: do philosophers and psychologists theorize about different things? synthese, 197(5), 2171–2191. https://doi.org/10.1007/s11229-018-1798-4 machery, e. (2005). concepts are not a natural kind. philosophy of science, 72(3), 444–467. https://doi.org/10.1086/498473 machery, e. (2009). doing without concepts. oxford university press. machery, e. (2010). précis of doing without concepts. behavioral and brain sciences, 33(2-3), 295–206. https://doi.org/10.1 017/s0140525x09991531 machery, e. (2016). the amodal brain and the offloading hypothesis. psychonomic bulletin & review, 23, 1090–1095. https: //doi.org/10.3758/s13423-015-0878-4 machery, e. (2017). philosophy within its proper bounds. oxford university press. https://doi.org/10.1093/oso/978019880752 0.001.0001 löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/0010-0277(83)90012-4 https://doi.org/10.1017/s0140525x99532147 https://doi.org/10.1017/s0140525x99532147 https://doi.org/10.1093/mind/110.438.430 https://doi.org/10.1007/bf00385837 https://doi.org/10.1093/oso/9780198814719.001.0001 https://doi.org/10.1093/oso/9780198814719.001.0001 https://doi.org/10.7551/mitpress/9383.003.0031 https://doi.org/10.3765/sp.8.7 https://doi.org/10.1093/0198236360.001.0001 https://doi.org/10.1093/0198236360.001.0001 https://doi.org/10.1093/0198250614.001.0001 https://doi.org/10.1093/0198250614.001.0001 https://doi.org/10.1007/s00426-018-1011-x https://doi.org/10.1007/s00426-018-1011-x https://doi.org/10.1037/a0021446 https://doi.org/10.1017/s0140525x10000403 https://doi.org/10.1017/s0140525x10000403 https://doi.org/10.1080/09515089.2017.1296941 https://doi.org/10.1007/s11229-018-1798-4 https://doi.org/10.1086/498473 https://doi.org/10.1017/s0140525x09991531 https://doi.org/10.1017/s0140525x09991531 https://doi.org/10.3758/s13423-015-0878-4 https://doi.org/10.3758/s13423-015-0878-4 https://doi.org/10.1093/oso/9780198807520.001.0001 https://doi.org/10.1093/oso/9780198807520.001.0001 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org guido löhr 18 machery, e., & seppälä, s. (2011). against hybrid theories of concepts. anthropology and philosophy, 10, 99–126. http://dscholarship.pitt.edu/17471/ margolis, e., & laurence, s. (2018). concepts. in e. n. zalta (ed.), the stanford encyclopedia of philosophy. https://plato.stan ford.edu/archives/sum2019/entries/concepts/ michel, c. (2022). a hybrid account of concepts within the predictive processing paradigm. review of philosophy and psychology. https://doi.org/10.1007/s13164-022-00648-8 murphy, g. (2004). the big book of concepts. mit press. https://doi.org/10.1007/s13164-022-00648-8 murphy, g. l., & medin, d. l. (1985). the role of theories in conceptual coherence. psychological review, 92(2), 295. https: //doi.org/10.1037/0033-295x.92.3.289 piccinini, g., & scott, s. (2006). splitting concepts. philosophy of science, 73(4), 390–409. https://doi.org/10.1086/516806 prinz, j. (2002). furnishing the mind: concepts and their perceptual basis. cambridge university press. https://doi.org/10.7 551/mitpress/3169.001.0001 prinz, j. (2010). can concept empiricism forestall eliminativism? mind & language, 25(5), 612–621. https://doi.org/10.111 1/j.1468-0017.2010.01404.x pulvermüller, f. (2013). how neurons make meaning: brain mechanisms for embodied and abstract-symbolic semantics. trends in cognitive sciences, 17(9), 458–470. https://doi.org/10.1016/j.tics.2013.06.004 rey, g. (2010). concepts versus conceptions (again). behavioral and brain sciences, 33(2-3), 221–222. https://doi.org/10.101 7/s0140525x10000440 rice, c. (2016). concepts as pluralistic hybrids. philosophy and phenomenological research, 92(3), 597–619. https://doi.org/ 10.1111/phpr.12128 rosch, e. (1999). reclaiming concepts. journal of consciousness studies, 6(11-12), 61–77. rosch, e. h. (2011). slow lettuce: categories, concepts, fuzzy sets, and logical deduction. in r. belohlavek & g. klir (eds.), concepts and fuzzy logic (pp. 89–120). mit press. samuels, r., & ferreira, m. (2010). why don’t concepts constitute a natural kind? behavioral and brain sciences, 33(2-3), 222–223. https://doi.org/10.1017/s0140525x10000452 taylor, h., & vickers, p. (2017). conceptual fragmentation and the rise of eliminativism. european journal for philosophy of science, 7(1), 17–40. https://doi.org/10.1007/s13194-016-0136-2 taylor, s. d., & vosgerau, g. (2019). the explanatory role of concepts. erkenntnis, 86, 1045–1070. https://doi.org/10.1007/s1 0670-019-00143-0 vicente, a., & martínez manrique, f. (2016). the big concepts paper: a defence of hybridism. the british journal for the philosophy of science, 67(1), 59–88. https://doi.org/10.1093/bjps/axu022 weiskopf, d. a. (2009). the plurality of concepts. synthese, 169(1), 145–173. https://doi.org/10.1007/s11229-008-9340-8 wiemer-hastings, k., & xu, x. (2005). content differences for abstract and concrete concepts. cognitive science, 29(5), 719–736. https://doi.org/10.1207/s15516709cog0000_33 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. löhr, g. (2023). are concepts a natural kind? on concept eliminativism. philosophy and the mind sciences, 4, 16. https://doi.org/10.33735/phimisci..9632 ©the author(s). https://philosophymindscience.org issn: 2699-0369 http://d-scholarship.pitt.edu/17471/ http://d-scholarship.pitt.edu/17471/ https://plato.stanford.edu/archives/sum2019/entries/concepts/ https://plato.stanford.edu/archives/sum2019/entries/concepts/ https://doi.org/10.1007/s13164-022-00648-8 https://doi.org/10.1007/s13164-022-00648-8 https://doi.org/10.1037/0033-295x.92.3.289 https://doi.org/10.1037/0033-295x.92.3.289 https://doi.org/10.1086/516806 https://doi.org/10.7551/mitpress/3169.001.0001 https://doi.org/10.7551/mitpress/3169.001.0001 https://doi.org/10.1111/j.1468-0017.2010.01404.x https://doi.org/10.1111/j.1468-0017.2010.01404.x https://doi.org/10.1016/j.tics.2013.06.004 https://doi.org/10.1017/s0140525x10000440 https://doi.org/10.1017/s0140525x10000440 https://doi.org/10.1111/phpr.12128 https://doi.org/10.1111/phpr.12128 https://doi.org/10.1017/s0140525x10000452 https://doi.org/10.1007/s13194-016-0136-2 https://doi.org/10.1007/s10670-019-00143-0 https://doi.org/10.1007/s10670-019-00143-0 https://doi.org/10.1093/bjps/axu022 https://doi.org/10.1007/s11229-008-9340-8 https://doi.org/10.1207/s15516709cog0000_33 https://doi.org/10.33735/phimisci..9632 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction the eliminativist argument machery's concept of concept in psychology `concept' picks out three distinct types of bodies of information a natural kind in science is individuated in terms of scientifically relevant properties the eliminativist conclusion the eliminativist argument debunked machery's concept of concept (against premise 1) the relevant property is not content (against premise 4) higher cognitive competencies and deliberation (against premise 3) why concepts might still be a natural kind (against the eliminativist conclusion) conclusion naturalistic entheogenics précis of philosophy of psychedelics. naturalistic entheogenics précis of philosophy of psychedelics chris lethebya (chris.letheby@uwa.edu.au) abstract in this précis i summarise the main ideas of my book philosophy of psychedelics. the book discusses philosophical issues arising from the therapeutic use of “classic” (serotonergic) psychedelic drugs such as psilocybin and lsd. the book is organised around what i call the comforting delusion objection to psychedelic therapy: the concern that this novel and promising treatment relies essentially on the induction of non-naturalistic metaphysical beliefs, rendering it epistemically (and perhaps, therefore, ethically) objectionable. i begin the précis by summarizing material from chapters two and three of the book, which review evidence for the therapeutic efficacy of psychedelics, and the facts about their clinical use that prompt the comforting delusion objection. i then summarize materials from chapters four and five of the book, which argue that psychedelic therapy works neither by experience-independent processes of neuroplasticity, nor by inducing non-naturalistic metaphysical ideations, but by altering mental representations of the self. next, i summarise the specific, speculative account of how this might work that is developed in chapters six and seven of the book. this account is based on the predictive processing theory of brain function and the self-binding theory of self-representation. chapters eight and nine of the book argue, on the basis of this account, that psychedelic therapy can have significant epistemic and spiritual benefits that are compatible with a naturalistic worldview. i summarize this material, and then, finally, the overall conclusions about psychedelic therapy drawn in the tenth and final chapter of the book. keywords ayahuasca ∙ dmt ∙ epistemology ∙ lsd ∙ mescaline ∙ philosophy ∙ predictive processing ∙ psilocybin ∙ psychedelics ∙ self-representation ∙ spirituality this article is part of a symposium on chris letheby’s book “philosophy of psychedelics” (oup 2021), edited by chiara caporuscio and sascha benjamin fink. athe university of western australia, the university of adelaide letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-6293-7873 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 2 1 introduction in this précis i summarise the main ideas of my book philosophy of psychedelics (letheby, 2021). the book discusses philosophical issues arising from the therapeutic use of “classic” (serotonergic) psychedelic drugs such as psilocybin and lsd. the book is organised around what i call the comforting delusion objection to psychedelic therapy: the concern that this novel and promising treatment relies essentially on the induction of non-naturalistic metaphysical beliefs, rendering it epistemically (and perhaps, therefore, ethically) objectionable. in the book i develop a response to the objection which involves naturalizing the entheogenic conception of psychedelics: showing that a traditional conception of psychedelics as agents of insight and spirituality is both consistent with a naturalistic worldview and plausible in light of current scientific knowledge. besides the introduction and conclusion, the book contains eight substantive chapters, which fall naturally into pairs. in chapters two and three i review the history and phenomenology of psychedelic therapy, and evidence for its safety and efficacy. these chapters set up the problem to which the rest of the book responds, citing evidence that psychedelics’ therapeutic benefits are mediated by transcendent “mystical-type” experiences, and arguing that there is a need for a naturalistic account of psychedelic therapy that responds to this fact. i summarise this material in section 2 of the present précis. in chapters four and five of the book, i consider various theories of the mechanisms of psychedelic therapy. i argue for the following simple hypothesis: psychedelic therapy works mainly by disrupting and revising mental representations of the self. this is a genuinely psychological mechanism which involves the psychedelic experience itself—it is not a mere matter of experience-independent neurobiological changes—but it does not depend on non-naturalistic metaphysical ideations. i summarise this material in section 3 of the précis. in chapters six and seven of the book, i outline a theory of how psychedelics might cause therapeutic benefits by disrupting self-representation. this is the predictive self-binding account of psychedelic ego dissolution developed in collaboration with philip gerrans (letheby & gerrans, 2017). i summarise this material in section 4 of the précis. in chapters eight and nine of the book, i argue that, if the predictive self-binding account is on the right track, then it is plausible to regard psychedelic as reliable agents of epistemic and spiritual benefit; in other words, to embrace a naturalized entheogenic conception of these substances. the thesis of chapter eight is that the epistemic profile of psychedelic therapy, given naturalism, is fairly good, involving less risk and more benefit than one might suppose. the thesis of chapter nine is that transformative psychedelic experiences illustrate the core features of a viable naturalistic spirituality. i summarise this material in section 5 of the précis. finally, in section 6 of the précis, i recapitulate some material from the conclusion of the book. letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 3 2 setting up the problem chapter two of the book is entitled ‘on the need for a natural philosophy of psychedelics’. i begin this chapter by reviewing the history of psychedelics in science and psychiatry. the book focuses solely on the class of psychoactive substances known as “classic psychedelics” or “serotonergic hallucinogens”, the most famous examples of which are lsd, psilocybin, mescaline, and dmt. these substances, and others of their class, are believed to share a common neuropharmacological mechanism: they all alter consciousness primarily by agonism of the serotonin-2a (5-ht2a) receptor (carhart-harris, 2019). this distinguishes them from other drugs that have sometimes been called “psychedelic” on phenomenological grounds, such as the dissociative anaesthetic ketamine and the “entactogen” mdma. in the book and in this précis i reserve the term “psychedelic” for the classic, serotonergic psychedelics that are the exclusive target of my analysis. i begin chapter two by reviewing the history of psychedelics in science and psychiatry. psychedelics have long histories of ceremonial and medicinal use in various societies. they came to widespread scientific attention throughout the 1950s and 60s after albert hoffman’s 1943 discovery of the psychoactive effects of lsd. throughout the 50s and 60s these drugs were used in many different ways, including as research tools for studying psychiatric pathologies and religious experiences, and as therapeutic agents in psychiatry. the latter, somewhat surprising, use was prompted by some psychedelic subjects’ reports of transformative religious or mystical experiences leading to positive behavioural change. many clinicians and researchers believed that psychedelics had considerable therapeutic potential in various conditions, including addiction, mood disorders, and existential distress accompanying terminal illness. however, much early research suffered from methodological shortcomings. the heyday of psychedelics in science and psychiatry was cut short by the prohibition of psychedelics and the declaration of the war on drugs in the late 1960s. after this, very little human research with psychedelics took place until the 1990s (sessa, 2012). since the 1990s there has been a “renaissance” of psychedelic research in human subjects (sessa, 2012) using modern research methods and following strict safety guidelines. in the book i summarise certain core findings from this wave of research. first and foremost is the finding that psychedelics can be administered safely in moderate-to-high, profoundly consciousness-altering doses, to suitably screened and prepared volunteers under carefully controlled conditions. no evidence of serious, lasting physical or psychological harm has been reported in recent clinical trials. a second important finding is that psychedelics may indeed have therapeutic potential. several studies—many, admittedly, small and preliminary—have reported lasting psychological benefits resulting from one to three administrations of a psychedelic. these benefits encompass reductions in symptoms of anxiety, depression, and addiction, and positive personality change and increases in self-reported well-being in mentally healthy volunteers (aday et letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 4 al., 2020). these findings cannot yet be regarded as conclusive. however, there is enough evidence to take seriously the hypothesis that one to three administrations of a psychedelic can, under conducive circumstances, lead reliably to lasting psychological benefits. for the purposes of the book, i assume that this hypothesis is correct and investigate its philosophical consequences. a third finding provides an intriguing clue about the mechanisms underlying these remarkable effects. psychedelics’ effects on consciousness are highly variable, being affected not just by the dose but by the “set and setting”: the psychological state of the person taking the drug, and the environment in which they take it. under conducive conditions, as many as half to two-thirds of volunteers who receive a relatively high dose seem to have a specific type of experience: a mystical-type experience. this is defined by standard psychometric questionnaires as involving a sense of unity and transcendence of individual selfhood, as well as noetic feelings of directly encountering ultimate reality, a transcendence of time and space, a sense of sacredness, ineffability, and paradoxicality, and a deeply felt positive mood. across many studies involving different substances, doses, and populations, a pattern has emerged: the lasting psychological benefits of psychedelics seem to be mediated by the occurrence of this specific type of experience. subjects who have a mystical-type experience during the drug action tend to be the ones who experience lasting benefits, and often their psychometric ratings of the degree of mystical-type experience correlate with the degree of benefits that they experience. in sum, psychedelics seem safely and reliably to induce lasting psychological benefits by inducing transformative mystical-type experiences (james et al., 2020). it is these findings that give rise to the comforting delusion objection. the worry underlying the objection arises most clearly when we consider the best-studied therapeutic use of psychedelics: the alleviation of anxiety, depression, and existential distress accompanying terminal illness. several studies, including doubleblind, placebo-controlled trials, have found that psychedelics can occasion lasting reductions in distress in patients with a terminal diagnosis. consistent with the pattern described above, these reductions seem to be mediated by the occurrence of a mystical-type experience during the drug action (griffiths et al., 2016; ross et al., 2016). some patients speak of (re-) discovering a connection to a spiritual principle, or feeling strongly that consciousness is not intrinsically tied to the body and continues in some form after physical death. reflecting on these findings, michael pollan (2015) wondered: “is psychedelic therapy simply foisting a comforting delusion on the sick and dying?” this is the basic concern that prompts the arguments in my book. i believe the intuitive concern that pollan is expressing can be articulated as a formal argument against the therapeutic use of psychedelics (not that i think pollan would ultimately endorse this argument): letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 5 (p1) naturalism is true. (p2) if the epistemic status of psychedelic therapy is poor, then we should hesitate to recommend or prescribe it. (p3) if naturalism is true, then the epistemic status of psychedelic therapy is poor. (c) therefore, we should hesitate to recommend or prescribe psychedelic therapy. in this argument i use “naturalism” to refer to a metaphysical, rather than a methodological, doctrine: the conjunction of a generic physicalism or materialism about the mind with a denial of the existence of “paradigmatically non-natural entities such as god, angels or cartesian souls or non-natural properties” (horst, 2009, p. 225, emphasis original). there are three popular responses to the concerns embodied in this argument. one is to deny premise 1, asserting that there really is a cosmic consciousness or divine ground of being that subjects encounter during the drug effects. one exponent of this view is the late huston smith, who claimed “the basic message of the entheogens1 [is] that there is another reality that puts this one in the shade” (smith, 2003, p. 133). a second response is to deny premise 2, arguing that the epistemic profile of psychedelic therapy is relatively unimportant compared to its capacity to improve psychological well-being (flanagan & graham, 2017). a third response is to accept the conclusion and hold that we should, indeed, hesitate to recommend or prescribe psychedelic therapy due to its poor epistemic profile (lavazza, 2017). in the book i take the relatively untrodden path of arguing against premise 3. personally, i think that premises 1 and 2 of the argument are quite plausible. my concern is to show that the comforting delusion objection fails even if these two assumptions are granted—the worst-case scenario for a defender of psychedelic therapy. my strategy is to assume the truth of metaphysical naturalism and, within that constraint, show that the epistemic status of psychedelic therapy is still reasonably good: even given naturalistic assumptions, its epistemic risks are smaller, and its epistemic benefits larger, than they initially appear. in doing this i also try to show that the entheogenic conception of psychedelics as agents of insight and spirituality can be naturalized. although the sense of “naturalism” in the argument above is metaphysical, methodological naturalism also plays a role in my project. i believe that philosophy is, or ought to be, continuous with the natural sciences, and that philosophical conclusions, when possible, ought to be based upon and integrated with relevant scientific findings. as such, i propose to develop a natural philosophy of psychedelics: a trans-disciplinary synthesis of empirical findings with theoretical ideas and conceptual considerations that can answer distinctively philosophical 1another name for psychedelics; sometimes applied to any psychoactive substance when used for religious or spiritual purposes, but often treated as synonymous with “psychedelic” as i use the latter term. letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 6 questions about psychedelics. in the spirit of such a natural philosophy, i begin with a close examination of the phenomenon under investigation: the therapeutic and transformative use of psychedelics. chapter three of the book is entitled ‘the phenomenology of psychedelic therapy’. in it i provide a selective overview of experiences commonly reported by people who take psychedelics in controlled and structured settings such as clinical trials and religious rituals. in the first half of the chapter i review a variety of typical changes to perception and the sense of self. psychedelic subjects often report a distinctive set of changes to perceptual experience, especially in the visual modality. these changes can involve intensification or alteration of endogenous percepts, or novel (pseudo-) hallucinatory percepts with no apparent basis in the physical environment. the intensity of these changes can range all the way from a slight brightening and sharpening of visual experience to total immersion in other worlds, ancient civilizations, or alien realms. intensification, alteration, and novelty also occur with varying frequency in other sensory modalities. a wide variety of changes to bodily experience is reported, including alterations in the experience of body ownership, changes to internal sensations, and distortions of body image. the experiences of space and time characteristically change, with time dilation a commonly reported effect. finally, a variety of changes to the phenomenal sense of self can occur, ranging from a reduction in self-referential thought to an apparently total obliteration of all forms of self-awareness (pollan, 2018). in the second half of the chapter i review qualitative evidence concerning patients’ impressions of the therapeutic process. patients who receive psychedelics in clinical trials sometimes, but not always, describe non-naturalistic metaphysical epiphanies concerning the existence of a cosmic consciousness, spirit world, or divine reality. more consistently emphasised are experiences of psychological insight, beneficial changes to the sense of self or identity, intense and cathartic emotional experiences, and feelings of connectedness and acceptance (breeksema et al., 2020). this evidence provides initial clues that psychedelics’ therapeutic effects may not be due entirely to the induction of non-naturalistic metaphysical ideations. 3 narrowing down the mechanism chapter four of the book is entitled ‘the mechanisms of psychedelic therapy’. in this chapter i consider two diametrically opposed hypotheses about psychedelics’ therapeutic mechanisms. the first of these, the molecular neuroplasticity theory, ascribes the lasting benefits to psychedelics’ experience-independent effects on the molecular mechanisms of neuroplasticity. on this view, the remarkable changes letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 7 to consciousness surveyed in chapter three are therapeutically epiphenomenal—a mere cluster of “psychotomimetic side effects” (yang et al., 2015) 2. the second hypothesis ascribes lasting benefits to the psychedelic experience itself—indeed, to a specific component of it: the non-naturalistic metaphysical ideations that it sometimes involves. on this view, psychedelics’ therapeutic benefits stem directly from their ability to induce transcendent visions of another reality: a spirit world, divine ground of being, or joyous cosmology. this hypothesis comes in two versions. the first is the metaphysical belief theory, according to which these non-naturalistic ideations only cause psychological benefits when subjects believe that they are veridical. the second version is the metaphysical alief theory, according to which the transcendent vision of a joyous cosmology, and subsequent recollection thereof, causes psychological benefits irrespective of whether it is believed. i begin the chapter by considering the molecular neuroplasticity theory. according to this theory, psychedelics’ agonist action at 5-ht2a receptors has many distinct effects. one cluster of effects is the changes to cognitive processing and phenomenal consciousness that constitute the psychedelic experience itself; another, distinct cluster of effects includes enhancements to the molecular mechanisms of neuroplasticity. ly et al. (2018) have reported that multiple serotonergic psychedelics promote the growth of dendritic spines and increase the density of dendritic arbours in cultured mammalian neurons. this suggests that psychedelics do, indeed, increase plasticity at the level of individual neurons, independently of the many specific and variable changes to consciousness that they can induce in dependence on set and setting. while it seems likely that these low-level, experience-invariant neuroplastic effects contribute to psychedelics’ therapeutic potential, i argue that they are not the central mechanism. my grounds are simple: across multiple studies with different populations and different psychiatric conditions, the strongest predictor of lasting therapeutic benefits is not the dosage of the drug, but the occurrence of a mystical-type experience. this seems hard to account for on a pure molecular neuroplasticity theory. if the core therapeutic mechanisms were purely biological, experience-independent, and experience-invariant, then why should the most consistent predictor of good outcomes be a somewhat dose-independent variable quantifying a specific cluster of “psychotomimetic side-effects”? on the other hand, this correlation can readily be explained by the contrary hypothesis: that psychedelics durably reduce psychiatric symptoms mainly by inducing specific clusters of beneficial experiential effects—even if non-experiential effects contribute as well. i conclude that the bulk of the therapeutic benefits are caused by some aspect(s) of the psychedelic experience itself. according to the metaphysical belief and alief theories, the relevant aspect is the non-naturalistic metaphysical content of the mystical type experience: the 2yang et al. discuss ketamine, not classic psychedelics; i am citing them simply as one source of this evocative phrase. letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 8 transcendent vision of a cosmic consciousness or joyous cosmology (and, perhaps, subsequent belief in its veracity). in my view, both theories struggle to account for the qualitative evidence surveyed in chapter three. this evidence shows that some successfully treated patients do not undergo any non-naturalistic metaphysical ideations. this fact undermines the metaphysical belief and alief theories, but also creates a puzzle: given the correlation between mystical-type experiences and lasting benefits, how can successful treatment occur without non-naturalistic metaphysical ideations? the answer is simple: the psychometric construct of a mystical-type experience, as operationalized in recent psychedelic research, is not specific to experiences with non-naturalistic metaphysical contents. there are experiences which satisfy all psychometric criteria for a mystical-type experience, but do not involve any non-naturalistic ideations. the psychometric instruments cast a broad net. this fact is exemplified clearly by michael pollan’s description of a psilocybininduced experience that satisfied the psychometric criteria for a mystical-type experience, but did not tempt this avowed naturalist to believe in a cosmic consciousness or anything non-naturalistic (pollan, 2018, p. 284). all of this suggests to me that there must be some experiential factor, other than non-naturalistic ideations, that (a) correlates reasonably well with the psychometric construct of a mysticaltype experience, and (b) is responsible for the bulk of the therapeutic benefits. the conclusion of chapter four, then, is that psychedelics cause lasting benefits via some aspect of the psychedelic experience that correlates with the construct of a mystical-type experience, but is independent of non-naturalistic metaphysical ideations. in chapter five i argue that the relevant factor is changes to self-representation (i.e. to mental representations of the self). this is an independently plausible idea. changes to the sense of self3 have long been singled out as one of the most consistent, distinctive, and theoretically significant features of the psychedelic state. moreover, the sense of unity, whose corollary is a profound diminution of the ordinary sense of self, is widely regarded as the cardinal feature of the mystical-type experience, and is common to such experiences both with and without non-naturalistic contents. finally, the sorts of pathologies for which psychedelic therapy shows promise—depression, anxiety, and addiction—have all been linked to changes in self-representation (sui & gu, 2017). besides the considerations just mentioned, three lines of evidence from recent psychedelic research support the idea that psychedelic therapy works mainly by changing self-representation. the first consists in a set of studies linking good clinical outcomes to acute experiences of psychological insight. the second consists in a set of studies showing that psychedelics can increase mindfulness-related capacities. and the third consists in a set of studies linking good clinical outcomes to modulation of the default mode and salience networks. i describe these three lines of evidence, and how each of them implicates changes to the sense of self, in turn. 3here i use the terms ‘self-representation’ and ‘sense of self’ more or less interchangeably, even though, strictly speaking, the former is a cognitive and the latter a phenomenological term. letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 9 on the first count, although the most well-established predictor of lasting benefits is the mystical-type experience, a small set of studies has linked these benefits to experiences of psychological insight. one online survey study introduced a new measure called the “psychological insight questionnaire” and found that this measure was a stronger predictor of post-psychedelic reductions in depression and anxiety than the oft-used mystical experience questionnaire (davis et al., 2020). importantly, nearly every item on the piq describes some kind of change to narrative or autobiographical forms of self-awareness. the four items most commonly endorsed by participants in the survey—each endorsed by more than 95% of participants—were (davis et al., 2020, p. 41): • discovered how aspects of my life are affecting my well-being • gained a deeper understanding of previously held beliefs and/or values • awareness of information that helped me understand my life • discovered ways to see my problems with more clarity clearly the kind of “psychological insight” quantified by this questionnaire has a lot to do with gaining, or seeming to gain, new information or awareness pertaining to aspects of oneself and one’s life. in other words, this psychological insight experience centrally involves changes to the narrative self. on the second count, psychometric measures have been developed and validated to quantify the effects of mindfulness meditation training. one capacity quantified by these measures is decentering—the ability to disidentify with one’s thoughts and feelings, viewing them with a certain measure of detachment and objectivity, rather than identifying strongly with them and accepting their contents uncritically. other measures quantify the capacities to be aware of mental events, such as thoughts, feelings, and perceptions, as they happen, and to adopt a nonjudgemental or non-reactive stance towards them. several studies have found that a single psychedelic experience, without any accompanying mindfulness training, can increase these mindfulness-related capacities for anywhere from one week to three months (heuschkel & kuypers, 2020; madsen et al., 2020). in some cases, these increases have been found to correlate with therapeutic effects, such as reductions in psychiatric symptoms (gonzález et al., 2020; mian et al., 2020). and, importantly, these mindfulness-related capacities all involve changes to the sense of self—to the phenomenal experience of the subject, its boundaries, and its relations to its own mental and physical states. on the third and final count, neuroimaging findings implicate changes to self-representation in psychedelics’ therapeutic effects. neuroimaging of the psychedelic state in general has produced somewhat heterogeneous results, but when we focus on studies probing neural correlates of lasting therapeutic and transformative effects, matters become a little clearer. every relevant study that i know of has found changes to one, or both, of two key neural systems: the default letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 10 mode and salience networks. the default mode network (dmn) is so-called because it tends to be highly active in conditions of task-free, wakeful rest. its major hubs are densely connected, high-level association areas in the brain’s cortical midline whose activity typically quietens during externally-oriented or impersonal cognitive tasks. high levels of dmn activity have been observed during introspective, social, and self-referential tasks, leading to the association of this network with social cognition, mental time travel, and narrative or autobiographical forms of self-representation (davey et al., 2016). meanwhile, the salience network (sn) is named for its putative role in attributing salience (i.e. importance or behavioural relevance) to stimuli across multiple modalities and flagging these stimuli for further processing (seeley, 2019). the sn is centred on cortical midline regions implicated in a multitude of functions, including emotion, interoception (sensing the internal condition of the body), nociception (pain perception), and error detection. in contrast to the dmn, this network has been linked to more basic, pre-reflective, “minimal” or “embodied” forms of self-awareness. the consistent connection between clinical benefits and modulation of the dmn and sn is readily explained by the hypothesis that psychedelic therapy works by disintegrating and revising mental representations of the self harboured by these systems. having reviewed psychological and neuroscientific evidence concerning the mechanisms of psychedelic therapy, i conclude chapter five by outlining a metatheoretical framework for integrating these different kinds of evidence. this is the framework of neurocognitive psychiatry (gerrans, 2014) which aims at generating integrative, multi-level explanations of psychiatric disorder and treatment based on the computational theory of mind. the central idea is that the cognitive/computational or information processing level of analysis provides a necessary explanatory bridge between neurobiology and phenomenology in the study of abnormal mental states. we can explain changes to aspects of conscious experience (such as the narrative sense of self) in terms of changes to the behaviour of neural structures (such as hubs of the dmn) by theorizing the computational functions that are normally performed by those structures and how the performance of those functions is altered under the abnormal conditions in question. this means that the task of chapters 6 and 7 is to say something plausible about the computational functions of the dmn and sn, how the performance of these functions is altered by psychedelics’ pharmacological action, and how this alteration leads to the remarkable experiences reported by subjects and the symptom reductions quantified by diagnostic questionnaires. 4 unbinding the predictive self chapter six of the book is entitled ‘resetting the brain’. i begin this chapter by considering yet another theory of psychedelic therapy: the reset theory advanced by carhart-harris et al. (2017) and nichols et al. (2017). this theory starts from letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 11 the observation that the core nodes of the dmn and sn are densely connected “hub” regions which (a) are thought to facilitate communication and functional integration between multiple systems throughout the brain and (b) exhibit abnormal patterns of connectivity in pathological conditions. this observation prompts the following simple idea: perhaps psychedelics bring about therapeutic benefits by dismantling rigid and dysfunctional connectivity patterns in these pivotal connector hubs. then, “as the effect of the drug wears off, [large scale] networks can reconnect in ‘healthy’ ways, in the absence of the pathological driving force(s) that originally led to hub failure and disease” (nichols et al., 2017, p. 215). my verdict is that the reset theory, as it stands, is plausible but incomplete. on its own it does not illuminate the links between dysfunctional patterns of connectivity and psychiatric symptoms, between “hub dismantling” and mystical-type experiences, and between “healthy” network reconnection and symptom reduction. to cross this explanatory gulf, as i suggested at the end of chapter 5, we need neurocognitive theories that connect changes in the activity of neural structures to changes in the contents of conscious experience via changes to the performance of specific cognitive or information-processing functions by those neural structures. to that end, i introduce the influential predictive processing (pp) theory of brain function. pp depicts the brain as organ for prediction error minimisation, which builds models of the world and uses these to predict sensory inputs. predictively successful models are reinforced, while unpredicted inputs generate an error signal that must be cancelled by updating the model or acting in the world. importantly, the “generative models” harboured by the predictive brain have a hierarchical structure, with lower levels modelling more concrete features of the world over smaller spatiotemporal scales, and higher levels modelling increasingly large, durable, and abstract features (hohwy, 2013). the highest levels encode the brain’s most fundamental assumptions concerning self, space, time, causality, and so forth: the basic parameters of our phenomenal worlds. on one interpretation of pp, conscious experience is a kind of “controlled hallucination” generated by this modelling process. all the furniture of our experiential worlds—the tables, chairs, and people that we encounter from day to day—are the products of a thoroughly internal process of world-simulation or virtual-reality modelling. however, these mental models exhibit a feature known as phenomenal transparency: we do not experience them as models, but simply as the world itself (metzinger, 2014). it is as though we “look through” the models to their referents. as antti revonsuo (2006) puts it, our brains give us a thoroughly realistic and convincing “out-of-brain experience”. most of the time we cannot easily regard the constituents of our experiential worlds as virtual entities, or as products of a modelling process. phenomenologically speaking, they simply appear as reality, and we automatically regard them as such: transparency in action. occasionally, however, this transparency is undermined, such as in lucid dreams or stubborn perceptual illusions, and our mental representations move from transparency to phenomenal opacity: we can, unusually, experience them as representations—at least to an extent. in my view, the controlled hallucination view of consciousness letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 12 and the concepts of phenomenal transparency and opacity are indispensable for a naturalistic understanding of psychedelic experience and its transformative effects. carhart-harris and friston (2019) have proposed a pp-based explanation of psychedelic experience known as the rebus model (“relaxed beliefs under psychedelics”). according to this model, psychedelics alter consciousness by disrupting the neural substrates of the highest levels of the brain’s predictive models. the cognitive effect is to diminish the brain’s confidence in its most abstract and fundamental beliefs: to “relax” these beliefs by “decreasing their precision weighting”, in the jargon. carhart-harris and friston argue that this simple model can explain the most distinctive and important effects of psychedelics—especially at moderate-to-high doses. when it comes to therapeutic effects, the rebus model suggests that the pathologies for which psychedelics show promise are characterised by deleterious high-level beliefs, and that psychedelics, by temporarily relaxing these beliefs, afford an opportunity to revise them for the better. what kinds of beliefs are at issue? the evidence reviewed in chapter 5 suggests that self-related beliefs are the most plausible candidate. so, in chapter 7 i pursue this suggestion in more detail. the basic idea is that (a) deleterious self-related beliefs become rigidly entrenched in anxiety, depression, and addiction, and (b) psychedelics durably alleviate symptoms of these pathologies by temporarily relaxing these beliefs and thereby facilitating their beneficial revision. this is consistent with an account of psychedelics’ effects on self-representation that i developed in earlier collaborative work with philip gerrans (letheby & gerrans, 2017). chapter seven of the book is entitled ‘unbinding the self’. in it i outline the predictive self-binding theory of psychedelic ego dissolution introduced by letheby and gerrans (2017), and show how this theory can explain important facts about psychedelic therapy. the predictive self-binding theory and the rebus model share a number of key ideas. both accounts attribute core phenomenological effects of psychedelics to the disruption of predictive models implemented in highlevel cortical networks, and both accounts hold that this disruption creates an opportunity to revise these models in beneficial ways. however, the predictive self-binding theory has three distinctive, additional emphases. one is that predictive models of the self play a cognitive binding function, integrating multimodal information by attributing it to a persistent underlying entity. another is that the predictive self-model has a hierarchical structure, with lower levels implemented in cortical nodes of the sn and higher levels in nodes of the dmn. a third is that the predictive self-model functions as a centre of representational gravity, attributing salience and allocating attention in accordance with the represented goals and interests of the self. i unpack these three emphases in turn, before putting them to work in the explanation of psychedelic therapy. letheby and gerrans’ theory combines ideas from pp with the concept of selfbinding introduced by sui and humphreys (2015). in a series of studies, these researchers documented an “integrative advantage” for self-representation, findletheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 13 ing repeatedly that self-relevant information was integrated more rapidly and efficiently than self-irrelevant information, and was more difficult or costly to disintegrate later. in light of this, they posited that a key function of self-representation is to serve as an “integrative hub” for information processing, ensuring that information relevant to the self is processed efficiently across multiple modalities and tasks, by relating it to a single, core self-representation. they link this integrative function to the idea of cognitive binding—the integration of representational parts into representational wholes—resulting in the concept of self-binding. letheby and gerrans suggest that this concept can fruitfully be combined with the pp account of cognitive binding. according to pp, binding is a matter of top-down probabilistic inference (hohwy, 2013). properties are bound or integrated by models that attribute them to a single, underlying, persisting object, because such models predictively out-perform their rivals. if we apply this conception of binding to sui and humphreys’ theory of self-representation, we arrive at the idea of a representation or model that binds, or integrates, stimuli across multiple modalities by attributing these stimuli to a single, underlying, persistent entity: the self. thus, letheby and gerrans argue that predictive models achieve functionally important integration of self-related mental representations by a binding process that generates a phenomenology of cartesian selfhood: the experience as of being a unitary, indivisible subject of experience that thinks thoughts, feels feelings, authors actions, and persists through time despite changes in its properties. this basic idea of predictive self-binding promises to explain the psychedelicinduced experience of ego dissolution. if predictive self-models integrate selfrelated stimuli by representing the existence of a bounded, persistent, and unified entity, then it is no wonder that disruption to these models should lead to the phenomenal experience as of disintegrating, dying, or merging with the environment. but according to letheby and gerrans, further light can be shed on psychedelic phenomena if we posit that the predictive self-model has a hierarchical structure that maps onto distinctive neural substrates. neuroimaging studies of psychedelic ego dissolution have not found consistent results: some studies link this experience to modulation of the dmn, and others to modulation of the sn. qualitative reports also suggest that the ostensibly unitary category of “ego dissolution” is heterogeneous, encompassing such diverse phenomena as the blurring or dissolution of bodily boundaries, the loss of the sense of agency over thoughts, and perhaps even the total abolition of all forms of self-consciousness. one possible resolution of these issues is as follows: the predictive self-model, like the brain’s predictive models in general, has a hierarchical structure. lower levels model more concrete features, and higher levels more abstract. specifically, cortical nodes of the sn keep track of more immediate embodied and affective dimensions of selfhood, giving rise to the experience of the “minimal self”, while nodes of the dmn keep track of temporally extended, conceptual, and autobiographical dimensions, giving rise to the experience of the “narrative self”. psychedelics disrupt these different networks to different extents in different conditions—depending, no doubt, on set and setting—and thereby induce qualitatively distinct varieties of ego dissolution experience with distinct neural correlates. letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 14 the third and final emphasis of the predictive self-binding account is on the function of the self-model as a “centre of representational gravity”. daniel dennett (1992) famously described the self as a “centre of narrative gravity”; however, letheby and gerrans emphasise that the binding function of the predictive self-model pervades all levels of cognitive processing, not just that of explicit narrative reflection. at the perceptual level, our phenomenal worlds are egocentrically structured in relation to the body; at higher levels, they are affectively, attentionally, and narratively structured in relation to the subject’s (hypothesized) goals, interests, past, and future. the hierarchical predictive self-model implemented by the dmn and sn does not just parse the world into “me” and “notme”, into “self-relevant” and “self-irrelevant”, but also ensures that stimuli attract precious attentional and motivational resources in accordance with their putative self-relevance. the result is a phenomenal world thoroughly structured around, and filtered through, the perspective of a hypothesized subject underlying and persisting through the flow of experience. as mentioned above, this account can readily explain not only the experience of ego dissolution, but also the variability of its neural correlates and phenomenological details. however, the account also promises to explain more therapeutically relevant experiences, such as the forms of psychological insight and increases in mindfulness-related capacities canvassed in chapter five. on the first count, the story is fairly straightforward: by weakening the brain’s confidence in its fundamental hypotheses about who “i” am, psychedelics induce different forms of self-modelling. under the influence, patients can gain new perspectives on their lives, see things differently, and access information usually filtered out or ignored. they can thereby discover new, healthier self-conceptions. this sort of process is described clearly by patients in clinical trials: lisa was a latin-american female in her 50s with a family history of alcoholism, physical and emotional abuse, abandonment, and neglect. her problematic drinking began around the age of thirty and resulted in social isolation, hangovers, strong feelings of guilt and shame, and severe self-critical thoughts. […] in the second [psilocybin] session, lisa […] experienced an amplification of thought moving her into a confused and chaotic state. underneath the chaotic thinking, she identified a deep well of overwhelming sadness. she was able to eventually surrender control over her thoughts and entered into a state of peacefulness, until her thoughts quieted completely. she heard her own inner voice rupturing the quiet, whispering into her ear: “i’m going to tell you a secret. it’s the worst-kept secret in the universe because everyone knows it but you. you are a perfect creation of the universe.” at that moment she felt that everything in existence was unified and was made of love, though a part of her remained reluctant to fully believe this to be true. the voice repeatedly presented her with this reality, asking “do you believe this?” over and over until each one letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 15 of her objections had been addressed and dismissed. she examined herself and found that she finally did accept this to be true, which propelled her into a state of profound self-acceptance and wellbeing. she later said, “all there is is love, this is all that you are, this is all that matters.” following her [psilocybin] session, lisa reported that her selfcritical thoughts had dissolved and that alcohol had lost almost all of its appeal. she said that the [psilocybin] sessions had illuminated how she had been unkind to her body and had been harming herself with alcohol. she noted her ability to manage stress and found that she was making time to care for herself through socialization, relaxation, and a resumed meditation practice. she reported improved concentration, a lack of negative self-talk, decreased anxiety, and a spacious quality of mind. […] at 54 weeks, lisa reported a persisting reduction in alcohol consumption and alleviated anxiety. (bogenschutz et al., 2018, p. 5) when it comes to explaining the increases in mindfulness-related capacities, the concept of phenomenal opacity introduced in chapter 6 becomes important. as we have seen, predictive models in general exhibit the property of phenomenal transparency: they are not experienced as models or representations, but simply as reality itself. however, when the contents of self-related mental models, such as thoughts and feelings, are “unbound” from the self-model, they tend to move along the continuum from transparency to opacity. this connects with an important observation from the phenomenology of mindfulness meditation: to see thoughts and feelings as mere thoughts and feelings is ipso facto to disidentify with them (albahari 2006, pp. 63-64). unbinding and phenomenal opacity go hand-in-hand: when specific mental contents are no longer attributed to the subject of experience, they become (phenomenologically) an object of experience—a transient and fallible appearance in consciousness. by reducing the brain’s confidence in its hypotheses about who, what, and where “i” am, the boundaries of the phenomenal self can be shifted to exclude many of the thoughts, feelings, and perspectives with which we usually identify. these mental contents can then enter an open, spacious attention in the phenomenal guise of mere thoughts and feelings; no longer identified with so strongly, they evoke less defensiveness and reactivity. once again, this process of disidentifying with self-related mental contents and thereby coming to see their contingency and fallibility is described clearly by patients: two related feelings were present. one was a tremendous freedom to experience, to be i. it became very important to distinguish between “i” and “me”, the latter being an object defined by patterns and structures and responsibilities—all of which had vanished—and the former being the subject experiencing and feeling. my normal life seemed to be all me, all demands and responsibilities, a rushing burden which destroyed the pleasure and freedom of being “i”. later in the evening letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 16 the question of how to fit back into my normal life without becoming a slave of its patterns and demands became paramount. (durr, 1971, p. 79, my emphasis) for a few seconds, it was just like ‘i’m me, and there are no defining characteristics!’ […] that made me realise that i’m not a ‘smoker’. (noorani et al., 2018, p. 759) after discussing how increased mindfulness-related capacities can be explained in terms of self-unbinding, i take a brief detour and show how the predictive self-binding account and the concept of phenomenal opacity can shed light on the non-naturalistic metaphysical epiphanies that sometimes occur during psychedelic therapy (cf. metzinger, 2003). then i return to the main focus of the chapter and ask: what overall picture of psychedelic therapy falls out of the account i have developed? i argue that this account supports a two-factor theory of psychedelic therapy. the first factor is the induction of plasticity, at multiple levels: experience-independent stimulation of the mechanisms of neuroplasticity, the induction of cognitive plasticity by relaxing prior beliefs, and a phenomenal sense of freedom, an appreciation of the possibility of change, that is bound up with increased opacity of the self-model. however, the induction of plasticity alone is not sufficient for truly durable change. patients must also discover new, healthier forms of self-modelling during the psychedelic experience, and consolidate these during the subsequent period of integration; otherwise, neurocognitive business-as-usual will inexorably resume. i conclude the chapter by briefly considering philosophical issues to do with self and self-consciousness. i have argued elsewhere that psychedelic evidence supports two controversial theses: that the self does not exist (letheby & gerrans, 2017) and that there can be states of phenomenal consciousness that totally lack all forms of self-consciousness (letheby, 2020). however, both claims face considerable obstacles, and for present purposes i content myself with two weaker, but still extremely interesting, psychedelic-supported conclusions: (i) there can be states of consciousness lacking anything like the ordinary sense of self, and (ii) the kind of self that we ordinarily experience ourselves as being does not exist. theoretically and existentially, this is plenty to be getting on with. when it comes to the central thesis of the book—that the comforting delusion objection fails—the upshot of chapters four to seven is this: the epistemic risks of psychedelic therapy, given naturalism, are smaller than they initially appear. this treatment, despite first appearances, does not work centrally by inducing comforting metaphysical beliefs, but by changing how patients see themselves and relate to their own minds and lives. however, the conclusion that the comforting delusion objection fails is based not only on the claim that the epistemic risks of psychedelic therapy, given naturalism, are relatively small, but also on the claim that its epistemic benefits, given naturalism, are relatively large. arguing in detail for this latter conclusion is the task of chapter 8. and the other major thesis of the book is that the entheogenic letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 17 conception of psychedelics can be naturalized, which depends on the claim that naturalists can and should view psychedelics, not just as agents of epistemic benefit, but also as agents of spiritual benefit. arguing for this conclusion is the task of chapter 9. 5 naturalizing the entheogenic conception chapter eight of the book is entitled ‘epistemology’. i begin this chapter by rebutting some general arguments for the impossibility of psychedelic-facilitated epistemic benefit. one of these, the argument from drug use, alleges that psychedelics, as drugs, are ipso facto unable to provide epistemic benefit. this is easily refuted by pointing out the widespread practice of pharmacological cognitive enhancement, a live ethical issue in tertiary education today. a second argument, the argument from impairment, appeals to the fact that psychedelics impair the brain’s reality-representing mechanisms: they demonstrably cause perceptual misrepresentation of the external world. however, on one reading of the premise, such impairment is just one effect of psychedelics, which is consistent with some of their other effects being epistemically beneficial. on another reading, impairment is psychedelics’ only epistemically relevant effect; but on this reading, the premise is question-begging. after considering one final argument (from alterity) i conclude that “master arguments” cannot settle the question. specific proposals about putative psychedelic-facilitated epistemic benefits need to be considered individually, on their own merits. the type of knowledge most commonly discussed in epistemology, the philosophical study of knowledge, is propositional knowledge or “knowledge that”. this can be glossed, roughly, as justified true belief: holding a true belief about reality on the basis of sufficient reasons or evidence. the most obvious psychedelic candidate, from a naturalistic standpoint, is psychodynamic insight. many psychedelic subjects report epiphanies about their own previously unconscious, or otherwise unknown, mental states, such as hidden, forgotten, or repressed beliefs, memories, or desires. as thomas metzinger (2003) points out, many such subjects also show clinical improvement, and a simple explanation is that the apparent insights are genuine. however, following david jopling (2001), i note that we cannot discount the possibility of “placebo insights”: apparent insights that cause clinical benefits despite being spurious. i offer some suggestive but inconclusive reasons for thinking that such psychedelic-facilitated insights are likely to be accurate reasonably often. my conclusion is that the psychedelic experience alone never provides adequate justification for beliefs of this kind, but sober reflection in the post-session period can close the epistemic gap. probably, psychedelic subjects often gain justified true beliefs about their own mental states, through the process of (a) undergoing a psychedelic experience and (b) subsequently reflecting critically on it. a more promising candidate is ability knowledge or “knowledge how”. psychedelic subjects often report having gained some kind of new cognitive or letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 18 attentional skill, or ability, through their experiences. this dovetails with a body of evidence, surveyed in chapter 5, showing lasting increases in mindfulnessrelated capacities for decentering, non-judging, awareness, and so forth after a single psychedelic experience. interpreting a suggestion made by benny shanon (2010) in light of this evidence, i argue that psychedelic experiences very often do promote a kind of knowledge how that is essentially identical to the mindfulness-related capacities measured by psychometric questionnaires. through experiences of self-unbinding and altered patterns of attention, subjects learn how to pay attention and relate to their own mental contents in decentered, psychologically flexible ways of the kind that are systematically cultivated in mindfulness practice. i consider a sceptical objection which raises the possibility that psychedelics do not really induce the acquisition of mindfulness skills, but rather cause the typical cognitive effects of those skills via a non-experiential mechanism. i do not answer this objection conclusively, but i appeal to subjects’ testimony, and to arguments developed earlier in the book, to suggest that these psycho-epistemic benefits of the psychedelic state really are brought about by a conscious process of learning and not (solely) by low-level, non-experiential pathways. one epistemological category of obvious relevance to the psychedelic experience is knowledge by acquaintance. on an intuitive level, this is the sense of “knowledge” at issue when we claim to know someone, rather than knowing that something is the case, or knowing how to do something (pavese, 2021). the idea of knowledge by acquaintance was formulated originally by bertrand russell (1910). however, i adopt a definition given by earl conee (1994) according to which one has knowledge by acquaintance with some fact if and only if one knows that fact in the most direct way possible. using this definition, i argue that psychedelic experience often facilitates the acquisition of knowledge by acquaintance with two important facts about the subject’s mind: that it has vast, normally unrealized potential, and that the ordinary sense of self is contingent, constructed, and mutable. (both of these, in philosophical parlance, are modal facts: facts about what is possible, necessary, impossible, and so on.) on the first count, it is possible to know that one’s mind has the potential for states of mystical bliss, wonder, and awe by reading books and hearing others’ testimony. but the most direct way to know that one’s mind has this potential is for that very potential to become manifest in that very mind—a state of affairs that psychedelics often induce. on the second count, it is possible to know in similarly indirect ways that one’s ordinary sense of self is contingent and mutable. but the most direct way to know that one’s ordinary sense of self can be altered radically, or altogether absent, is for that very possibility to be actualized within one’s mind. thus, psychedelic experience often involves the acquisition of knowledge by acquaintance with important (and, indeed, therapeutically relevant) modal facts about the subject’s mind. a fourth type of epistemic benefit that i argue psychedelics often induce is the acquisition of “new knowledge of old facts”. here, the philosophical terminology is quite intuitive: someone acquires new knowledge of an old fact when (a) there letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 19 is some fact that they already know, but (b) they come to know that fact in an importantly new and different way. one important example concerns the difference between knowing in a merely intellectual fashion that one is mortal, and having the vivid, motivating, often life-changing awareness of mortality that a brush with death can induce (baillie, 2020). psychedelics seem routinely to bridge this sort of gulf. one patient, when asked whether her religious and spiritual beliefs had changed since her psilocybin session, replied: “[the psilocybin experience] brought my beliefs to life, made them real, something tangible and true – it made my beliefs more than something to think about, really something to lean on and look forward to.” (malone et al., 2018, p. 4) once again, benny shanon (2010, p. 268) has suggested something like this as a possible epistemic benefit of the ayahuasca experience (though he does not use the terminology of “new knowledge of old facts”). indeed, many subjects describe gaining new knowledge of a specific and important old fact: our radical interconnectedness with the natural world: “before [my psychedelic experience] i enjoyed nature, now i feel part of it. before i was looking at it as a thing, like tv or a painting. [but] you’re part of it, there’s no separation or distinction, you are it.” (watts et al., 2017, p. 534, emphasis original) during the psilocybin session, one cancer patient experienced feelings of “being connected to everything, i mean, everything in nature […] and it wasn’t like talking about it, which makes it an idea. it was experiential.” (swift et al., 2017, p. 497, my emphasis) of course, caution is warranted here, since psychedelics can also induce new “knowledge” (i.e. vivid and motivating phenomenal simulations) of old “facts” that are not actually facts at all. but in cases where there is a fact that a subject knows, very often the psychedelic experience can allow them to understand or grasp that fact more fully, or feel its significance more deeply, than they previously had. in my view, this constitutes a significant epistemic benefit. fifth, and finally, i argue that indirect epistemic benefits often result from therapeutic psychedelic administration. these indirect benefits are mediated by lasting psychological benefits. here i draw on lisa bortolotti’s (2015, 2020) work on epistemic benefits of suboptimal cognitive states such as delusions. bortolotti points out that certain epistemically faulty cognitions, such as so-called “motivated” or “defensive” delusions, are thought to have psychological benefits: by shielding patients from painful truths, these delusions can preserve psychosocial functioning and stave off debilitating despair. insofar as this is the case, bortolotti argues, these delusions are likely to have epistemic benefits too, for the simple reason that epistemic and psychosocial functionality are deeply intertwined in humans. we gain knowledge by exploring the world, talking to others, exposing our beliefs to letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 20 scrutiny, and so forth. thus, any cognition that preserves our propensities to explore and interact ipso facto helps to preserve epistemically beneficial capacities and behaviours. i argue that the same is true of therapeutic and transformative psychedelic experiences. people with reduced symptoms of anxiety, depression, or addiction, or with increased openness to experience or mindfulness-related capacities, are more likely to explore their world and engage with other people, and therefore more likely to gain various kinds of knowledge. thus, psychedelics’ lasting psychological benefits have flow-on epistemic benefits as well. the ultimate conclusion of chapter eight is that therapeutic psychedelic experiences are, in bortolotti’s parlance, epistemically innocent: they have real epistemic flaws, but also offer significant epistemic benefits that are typically unavailable by any other means. chapter nine of the book is entitled ‘spirituality’. the central thesis of this chapter is that psychedelic experiences reveal a viable form of naturalistic spirituality: a set of practices and experiences that (a) deserve the name “spirituality” and (b) can be pursued in good faith by those with a naturalistic worldview and a commitment to intellectual honesty (cf. metzinger, 2013). intuitively, the idea of a naturalistic spirituality might seem absurd, or even outright contradictory. the term “spirituality” would seem to connote belief in the spiritual—as opposed, metaphysically, to the natural. however, recently the term has acquired other connotations, often being used to describe an approach to matters of meaning, purpose, and transcendence that is practical, experiential, and non-dogmatic, as opposed to the (merely) doxastic and dogmatic orientation often associated with organised religion. in light of this shift, several philosophers and philosophically-inclined thinkers have wondered whether a naturalistic spirituality might be possible. in an important review article, jerome stone surveys several recent accounts of naturalistic spirituality, and extracts the following core set of ideas: we are spiritual, first, when our sense of connection is enlarged. second, we are spiritual when we aspire to greater things, when we attempt to realize our ideals. finally, we are spiritual when we ask the big questions. note that these three—connection, aspiration, and reflection on profound questions—are all forms of enlarging our selves, of breaking through the narrow walls of the ego. (stone, 2012, p. 492). in my view, the renaissance of psychedelic research provides a golden opportunity for an empirically-based, neurophilosophical inquiry into naturalised spirituality, because psychedelics reliably induce paradigmatically spiritual experiences: experiences that evoke that appellation consistently from those with otherwise divergent views about the topic. committed religious practitioners (griffiths et al., 2006) and atheistic “reluctant psychonauts” (pollan, 2018) alike find “spiritual” the most suitable word to describe their transformative psychedelic experiences. so if we want to know whether spirituality is compatible with naturalism, i suggest letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 21 that one approach is to examine closely the paradigmatically spiritual experiences that psychedelics induce, and see whether undergoing those kinds of experiences is compatible with a naturalistic outlook. this seems like a sensible and direct way to learn about the actual nature of the experiential phenomenon that is referred to by “spirituality” in its recent sense. of course, once this methodological proposal is made, answers flow naturally from the arguments of the preceding chapters. in at least some studies, subjects’ willingness to describe their experiences as “spiritually significant” correlates with psychometric ratings of mystical-type experiences (griffiths et al., 2008). this suggests that it is mystical-type experiences, in the psychometric sense, that are the paradigmatically spiritual ones, so our question becomes: are psychedelic-induced mystical-type experiences compatible with naturalism? and i have argued in earlier chapters that they are. through an analysis of qualitative evidence, i showed that there are subjects who undergo profound and transformative experiences that satisfy the definition of a mystical-type experience but do not involve any nonnaturalistic metaphysical ideations. instead, these experiences feature changes to the sense of self, profound feelings of unity and connectedness, psychological and existential insights, and deeply significant emotional shifts. not only does the account developed in the preceding chapters support the claim that spirituality can be naturalized—it also converges with the results of philosophical inquiry as summarised by stone. stone’s three pillars of naturalistic spirituality are an enlarged sense of connection, an aspiration to realize our values, and reflection on profound or ultimate questions. all three are hallmarks of psychedelic-induced mystical-type experiences, whether or not any non-naturalistic ideations are involved. moreover, stone contends that all three are “all forms of enlarging our selves, of breaking through the narrow walls of the ego”—and this, too, is consistent with the account that i have developed, which explains these hallmarks in terms of psychedelic-induced disruption to self-binding processes and consequent expansion and liberation of attention. thus, i argue that a close examination of the paradigmatically spiritual experiences induced by psychedelics, and the features in virtue of which they count as “spiritual”, supports the claim that spirituality can be naturalized. this, in turn, supports the claim that psychedelics can legitimately be regarded as agents not just of epistemic benefit, but also of spiritual benefit, from a naturalistic perspective—and, therefore, that the entheogenic conception can be naturalized. letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 22 6 conclusion in the tenth and final chapter of the book, the ‘conclusion’, i summarise the foregoing discussion, draw out some testable predictions, and suggest some directions for future research. the account i develop in the book entails several testable predictions, e.g., predictions about which variables will predict good clinical outcomes most strongly. however, i will not enumerate these predictions here; see philosophy of psychedelics section 10.1. despite its somewhat presumptuous title, philosophy of psychedelics leaves untouched many important philosophical questions about psychedelics. there are outstanding issues in the philosophy of mind, science, medicine, and psychiatry. there are a host of outstanding ethical issues. there are even many epistemological issues that i have not explored, with my focus on the possible involvement of non-naturalistic metaphysical beliefs. once again, i will not enumerate these here; see philosophy of psychedelics section 10.2. perhaps the most important, if the least precise, philosophical question about psychedelics is this: what should we make of these enigmatic substances, and how should we regard the remarkable experiences that they induce—experiences which many count among the most important of their lives? i believe that i have made some progress toward answering this question in philosophy of psychedelics. i close the book by saying what i think this progress amounts to. in brief, the comforting delusion objection fails: even given naturalism, the epistemic status of psychedelic therapy is fairly good. and, the entheogenic conception can be naturalized: even given naturalism, it is plausible to regard psychedelics as effective agents of epistemic and spiritual benefit. contra the late huston smith, the “basic message of the entheogens” is not that there is another metaphysical reality that puts this one in the shade. it is that there are other phenomenological realities, many of which put our ordinary, default mode in the shade. many of these other realities, especially the salutary ones, result largely from a reduction in the constraining influence of the predictive self-model. induced skilfully, more-or-less “selfless” forms of consciousness can show us what is possible in terms of wonder, awe, insight, acceptance, and connectedness. psychedelic therapy is not an experience-independent pharmacotherapy, nor is it primarily a matter of metaphysical conviction and consolation. rather, it is an “existential medicine”, in charles grob’s (2007) phrase, since it constitutively involves a transformative, experiential re-appraisal of basic assumptions concerning the self, the world, and the relations between the two. unconstraining cognition by unbinding the self-model; revealing the vast potential of consciousness by exposing the constructed and mutable nature of our phenomenal worlds—this is the essence of psychedelic therapy, and it is perfectly consistent with a naturalistic worldview. letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 23 acknowledgments this research was partially supported by the australian government through the australian research council’s discovery projects funding scheme (project dp190101451). the views expressed herein are those of the author and are not necessarily those of the australian government or australian research council. references aday, j. s., mitzkovitz, c. m., bloesch, e. k., davoli, c. c., & davis, a. k. (2020). long-term effects of psychedelic drugs: a systematic review. neuroscience & biobehavioral reviews, 113, 179–189. https://doi.org/10.1016/j.neubiorev.2020.03.01 7 albahari, m. (2006). analytical buddhism: the two-tiered illusion of self. palgrave. baillie, j. (2020). the recognition of nothingness. philosophical studies, 177(9), 2585–2603. https://doi.org/10.1007/s11098019-01329-6 bogenschutz, m. p., podrebarac, s. k., duane, j. h., amegadzie, s. s., malone, t. c., owens, l. t., ross, s., & mennenga, s. e. (2018). clinical interpretations of patient experience in a trial of psilocybin-assisted psychotherapy for alcohol use disorder. frontiers in pharmacology, 9. https://doi.org/10.3389/fphar.2018.00100 bortolotti, l. (2015). the epistemic innocence of motivated delusions. consciousness and cognition, 33, 490–499. https: //doi.org/10.1016/j.concog.2014.10.005 bortolotti, l. (2020). the epistemic innocence of irrational beliefs. oxford university press. breeksema, j. j., niemeijer, a. r., krediet, e., vermetten, e., & schoevers, r. a. (2020). psychedelic treatments for psychiatric disorders: a systematic review and thematic synthesis of patient experiences in qualitative studies. cns drugs, 34(9), 925–946. https://doi.org/10.1007/s40263-020-00748-y carhart-harris, r. l. (2019). how do psychedelics work? current opinion in psychiatry, 32(1), 16–21. https://doi.org/10.109 7/yco.0000000000000467 carhart-harris, r. l., & friston, k. j. (2019). rebus and the anarchic brain: toward a unified model of the brain action of psychedelics. pharmacological reviews, 71(3), 316–344. https://doi.org/10.1124/pr.118.017160 carhart-harris, r. l., roseman, l., bolstridge, m., demetriou, l., pannekoek, j. n., wall, m. b., tanner, m., kaelen, m., mcgonigle, j., murphy, k., leech, r., curran, h. v., & nutt, d. j. (2017). psilocybin for treatment-resistant depression: fmri-measured brain mechanisms. scientific reports, 7(1), 13187. https://doi.org/10.1038/s41598-017-13282-7 conee, e. (1994). phenomenal knowledge. australasian journal of philosophy, 72(2), 136–150. https://doi.org/10.1080/0004 8409412345971 davey, c. g., pujol, j., & harrison, b. j. (2016). mapping the self in the brain’s default mode network. neuroimage, 132, 390–397. https://doi.org/10.1016/j.neuroimage.2016.02.022 davis, a. k., barrett, f. s., & griffiths, r. r. (2020). psychological flexibility mediates the relations between acute psychedelic effects and subjective decreases in depression and anxiety. journal of contextual behavioral science, 15, 39–45. https: //doi.org/10.1016/j.jcbs.2019.11.004 dennett, d. c. (1992). the self as the center of narrative gravity. in f. kessel, p. cole and d. johnson (eds.), self and consciousness: multiple perspectives (pp. 275–288). psychology press. durr, r. a. (1971). poetic vision and the psychedelic experience. syracuse university press. flanagan, o., & graham, g. (2017). truth and sanity: positive illusions, spiritual delusions, and metaphysical hallucinations. in j. poland & s, tekin (eds.). extraordinary science and psychiatry: responses to the crisis in mental health research (pp. 293–313). mit press. https://doi.org/10.7551/mitpress/9780262035484.003.0013 gerrans, p. (2014). the measure of madness: philosophy of mind, cognitive neuroscience, and delusional thought. mit press. gonzález, d., cantillo, j., pérez, i., farré, m., feilding, a., obiols, j. e., & bouso, j. c. (2020). therapeutic potential of ayahuasca in grief: a prospective, observational study. psychopharmacology, 237(4), 1171–1182. https://doi.org/10.1 007/s00213-019-05446-2 griffiths, r. r., johnson, m. w., carducci, m. a., umbricht, a., richards, w. a., richards, b. d., cosimano, m. p., & klinedinst, m. a. (2016). psilocybin produces substantial and sustained decreases in depression and anxiety in patients with life-threatening cancer: a randomized double-blind trial. journal of psychopharmacology, 30(12), 1181–1197. https: //doi.org/10.1177/0269881116675513 letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neubiorev.2020.03.017 https://doi.org/10.1016/j.neubiorev.2020.03.017 https://doi.org/10.1007/s11098-019-01329-6 https://doi.org/10.1007/s11098-019-01329-6 https://doi.org/10.3389/fphar.2018.00100 https://doi.org/10.1016/j.concog.2014.10.005 https://doi.org/10.1016/j.concog.2014.10.005 https://doi.org/10.1007/s40263-020-00748-y https://doi.org/10.1097/yco.0000000000000467 https://doi.org/10.1097/yco.0000000000000467 https://doi.org/10.1124/pr.118.017160 https://doi.org/10.1038/s41598-017-13282-7 https://doi.org/10.1080/00048409412345971 https://doi.org/10.1080/00048409412345971 https://doi.org/10.1016/j.neuroimage.2016.02.022 https://doi.org/10.1016/j.jcbs.2019.11.004 https://doi.org/10.1016/j.jcbs.2019.11.004 https://doi.org/10.7551/mitpress/9780262035484.003.0013 https://doi.org/10.1007/s00213-019-05446-2 https://doi.org/10.1007/s00213-019-05446-2 https://doi.org/10.1177/0269881116675513 https://doi.org/10.1177/0269881116675513 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 24 griffiths, r. r., richards, w. a., mccann, u., & jesse, r. (2006). psilocybin can occasion mystical-type experiences having substantial and sustained personal meaning and spiritual significance. psychopharmacology, 187(3), 268–283. https: //doi.org/10.1007/s00213-006-0457-5 griffiths, r., richards, w., johnson, m., mccann, u., & jesse, r. (2008). mystical-type experiences occasioned by psilocybin mediate the attribution of personal meaning and spiritual significance 14 months later. journal of psychopharmacology, 22(6), 621–632. https://doi.org/10.1177/0269881108094300 grob, c. s. (2007). the use of psilocybin in patients with advanced cancer and existential anxiety. in m. winkelman and t. b. roberts (eds.), psychedelic medicine: new evidence for hallucinogenic substances as treatments (pp. 205–216). praeger publishers. heuschkel, k., & kuypers, k. p. c. (2020). depression, mindfulness, and psilocybin: possible complementary effects of mindfulness meditation and psilocybin in the treatment of depression. a review. frontiers in psychiatry, 11. https: //doi.org/10.3389/fpsyt.2020.00224 hohwy, j. (2013). the predictive mind. oxford university press. horst, s. (2009). naturalisms in philosophy of mind. philosophy compass, 4(1), 219–254. https://doi.org/10.1111/j.17479991.2008.00191.x james, e., robertshaw, t. l., hoskins, m., & sessa, b. (2020). psilocybin occasioned mystical-type experiences. human psychopharmacology: clinical and experimental, 35(5), e2742. https://doi.org/10.1002/hup.2742 jopling, d. a. (2001). placebo insight: the rationality of insight-oriented psychotherapy. journal of clinical psychology, 57(1), 19–36. https://doi.org/10.1002/1097-4679(200101)57:1%3c19::aid-jclp4%3e3.0.co;2-z lavazza, a. (2017). ways of being well: realistic and unrealistic well-being. in l. taddio and k. w. molin (eds.), new perspectives on realism (pp. 237–252). mimesis international. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 1–26. https://doi.org/10.33735/phimisci.2020.i.47 letheby, c. (2021). philosophy of psychedelics. oxford university press. letheby, c., & gerrans, p. (2017). self unbound: ego dissolution in psychedelic experience. neuroscience of consciousness, 2017(1). https://doi.org/10.1093/nc/nix016 ly, c., greb, a. c., cameron, l. p., wong, j. m., barragan, e. v., wilson, p. c., burbach, k. f., soltanzadeh zarandi, s., sood, a., paddy, m. r., duim, w. c., dennis, m. y., mcallister, a. k., ori-mckenney, k. m., gray, j. a., & olson, d. e. (2018). psychedelics promote structural and functional neural plasticity. cell reports, 23(11), 3170–3182. https: //doi.org/10.1016/j.celrep.2018.05.022 madsen, m. k., fisher, p. m., stenbæk, d. s., kristiansen, s., burmester, d., lehel, s., páleníček, t., kuchař, m., svarer, c., ozenne, b., & knudsen, g. m. (2020). a single psilocybin dose is associated with long-term increased mindfulness, preceded by a proportional change in neocortical 5-ht2a receptor binding. european neuropsychopharmacology, 33, 71–80. https://doi.org/10.1016/j.euroneuro.2020.02.001 malone, t. c., mennenga, s. e., guss, j., podrebarac, s. k., owens, l. t., bossis, a. p., belser, a. b., agin-liebes, g., bogenschutz, m. p., & ross, s. (2018). individual experiences in four cancer patients following psilocybin-assisted psychotherapy. frontiers in pharmacology, 9. https://doi.org/10.3389/fphar.2018.00256 metzinger, t. (2003). being no one: the self-model theory of subjectivity. mit press. metzinger, t. (2013). spirituality and intellectual honesty. self-published manuscript. https://philarchive.org/rec/metsai metzinger, t. (2014). how does the brain encode epistemic reliability? perceptual presence, phenomenal transparency, and counterfactual richness. cognitive neuroscience, 5(2), 122–124. https://doi.org/10.1080/17588928.2014.905519 mian, m. n., altman, b. r., & earleywine, m. (2020). ayahuasca’s antidepressant effects covary with behavioral activation as well as mindfulness. journal of psychoactive drugs, 52(2), 130–137. https://doi.org/10.1080/02791072.2019.1674428 nichols, d., johnson, m., & nichols, c. (2017). psychedelics as medicines: an emerging new paradigm. clinical pharmacology & therapeutics, 101(2), 209–219. https://doi.org/10.1002/cpt.557 noorani, t., garcia-romeu, a., swift, t. c., griffiths, r. r., & johnson, m. w. (2018). psychedelic therapy for smoking cessation: qualitative analysis of participant accounts. journal of psychopharmacology, 32(7), 756–769. https://doi.or g/10.1177/0269881118780612 pavese, c. (2021). knowledge how. in e. zalta (ed.) the stanford encyclopedia of philosophy. https://plato.stanford.edu/arc hives/sum2021/entries/knowledge-how/ pollan, m. (2015). the trip treatment. the new yorker. http://www.newyorker.com/magazine/2015/02/09/trip-treatment pollan, m. (2018). how to change your mind: the new science of psychedelics. penguin books. revonsuo, a. (2006). inner presence: consciousness as a biological phenomenon. mit press. ross, s., bossis, a., guss, j., agin-liebes, g., malone, t., cohen, b., mennenga, s. e., belser, a., kalliontzi, k., babb, j., su, z., corby, p., & schmidt, b. l. (2016). rapid and sustained symptom reduction following psilocybin treatment for anxiety letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s00213-006-0457-5 https://doi.org/10.1007/s00213-006-0457-5 https://doi.org/10.1177/0269881108094300 https://doi.org/10.3389/fpsyt.2020.00224 https://doi.org/10.3389/fpsyt.2020.00224 https://doi.org/10.1111/j.1747-9991.2008.00191.x https://doi.org/10.1111/j.1747-9991.2008.00191.x https://doi.org/10.1002/hup.2742 https://doi.org/10.1002/1097-4679(200101)57:1%3c19::aid-jclp4%3e3.0.co;2-z https://doi.org/10.33735/phimisci.2020.i.47 https://doi.org/10.1093/nc/nix016 https://doi.org/10.1016/j.celrep.2018.05.022 https://doi.org/10.1016/j.celrep.2018.05.022 https://doi.org/10.1016/j.euroneuro.2020.02.001 https://doi.org/10.3389/fphar.2018.00256 https://philarchive.org/rec/metsai https://doi.org/10.1080/17588928.2014.905519 https://doi.org/10.1080/02791072.2019.1674428 https://doi.org/10.1002/cpt.557 https://doi.org/10.1177/0269881118780612 https://doi.org/10.1177/0269881118780612 https://plato.stanford.edu/archives/sum2021/entries/knowledge-how/ https://plato.stanford.edu/archives/sum2021/entries/knowledge-how/ http://www.newyorker.com/magazine/2015/02/09/trip-treatment https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org naturalistic entheogenics: précis of philosophy of psychedelics 25 and depression in patients with life-threatening cancer: a randomized controlled trial. journal of psychopharmacology, 30(12), 1165–1180. https://doi.org/10.1177/0269881116675512 russell, b. (1910). knowledge by acquaintance and knowledge by description. proceedings of the aristotelian society, 11, 108–128. https://www.jstor.org/stable/4543805 seeley, w. w. (2019). the salience network: a neural system for perceiving and responding to homeostatic demands. journal of neuroscience, 39(50), 9878–9882. https://doi.org/10.1523/jneurosci.1138-17.2019 sessa, b. (2012). the psychedelic renaissance: reassessing the role of psychedelic drugs in 21st century psychiatry and society. muswell hill press. shanon, b. (2010). the epistemics of ayahuasca visions. phenomenology and the cognitive sciences, 9(2), 263–280. https://doi.org/10.1007/s11097-010-9161-3 smith, h. (2003). cleansing the doors of perception: the religious significance of entheogenic plants and chemicals. sentient publications. stone, j. a. (2012). spirituality for naturalists. zygon: journal of religion and science, 47(3), 481–500. https://doi.org/10.111 1/j.1467-9744.2012.01279.x sui, j., & gu, x. (2017). self as object: emerging trends in self research. trends in neurosciences, 40(11), 643–653. https: //doi.org/10.1016/j.tins.2017.09.002 sui, j., & humphreys, g. w. (2015). the integrative self: how self-reference integrates perception and memory. trends in cognitive sciences, 19(12), 719–728. https://doi.org/10.1016/j.tics.2015.08.015 swift, t. c., belser, a. b., agin-liebes, g., devenot, n., terrana, s., friedman, h. l., guss, j., bossis, a. p., & ross, s. (2017). cancer at the dinner table: experiences of psilocybin-assisted psychotherapy for the treatment of cancer-related distress. journal of humanistic psychology, 57(5), 488–519. https://doi.org/10.1177/0022167817715966 watts, r., day, c., krzanowski, j., nutt, d., & carhart-harris, r. (2017). patients’ accounts of increased “connectedness” and “acceptance” after psilocybin for treatment-resistant depression. journal of humanistic psychology, 57(5), 520–564. https://doi.org/10.1177/0022167817709585 yang, c., shirayama, y., zhang, j., ren, q., yao, w., ma, m., dong, c., & hashimoto, k. (2015). r-ketamine: a rapidonset and sustained antidepressant without psychotomimetic side effects. translational psychiatry, 5(9), e632–e632. https://doi.org/10.1038/tp.2015.136 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. letheby, c. (2022). naturalistic entheogenics: précis of philosophy of psychedelics. philosophy and the mind sciences, 3, 4. https://doi.org/10.33735/phimisci.2022.9627 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1177/0269881116675512 https://www.jstor.org/stable/4543805 https://doi.org/10.1523/jneurosci.1138-17.2019 https://doi.org/10.1007/s11097-010-9161-3 https://doi.org/10.1111/j.1467-9744.2012.01279.x https://doi.org/10.1111/j.1467-9744.2012.01279.x https://doi.org/10.1016/j.tins.2017.09.002 https://doi.org/10.1016/j.tins.2017.09.002 https://doi.org/10.1016/j.tics.2015.08.015 https://doi.org/10.1177/0022167817715966 https://doi.org/10.1177/0022167817709585 https://doi.org/10.1038/tp.2015.136 https://doi.org/10.33735/phimisci.2022.9627 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction setting up the problem narrowing down the mechanism unbinding the predictive self naturalizing the entheogenic conception conclusion colour variation without objective colour colour variation without objective colour derek h. browna (derek.brown.2@glasgow.ac.uk) abstract colour variation is the fact that what colour physical objects look to have depends on viewing conditions and a perceiver’s visual system. both colour relationalists and colour eliminativists regard their analyses of colour variation as central to the justification for their respective views. yet the analyses are decidedly different. colour relationalists assert that most instances of colour variation are veridical and infer from this that colours are relational properties of objects that are partly determined by perceivers. by contrast, colour eliminativists assert that colour variation is too unsystematic to ground the claim that many or most instances of colour variation are veridical. from this they infer that objects don’t have colours. i argue that the eliminativist analysis is superior. on my view, the relationalist account of veridical colour experience reduces to the assertion that objects have colour simply because they cause perceivers to have colour experiences of them. in this context, i argue, the resulting conception of veridicality is vacuous. more directly, the foundational idea of eliminativism is the opposite claim: the fact that objects cause perceivers to have colour experiences of them is on its own not sufficient to justify or ground the claim that objects have colour. the relationalist, i argue, has failed to justify anything stronger than this. in this debate we should thus side with the eliminativist: objects do not possess colour; they merely cause us to undergo colour experiences. keywords colour eliminativism ∙ colour relationalism ∙ colour variation ∙ illusion ∙ metaphysics of colour ∙ veridicality in perception 1 introduction colour variation is the fact that the colour one perceptually experiences some object1 to have is affected not only by that object, but also by environmental conditions (e.g. illumination conditions) and aspects of perceivers (e.g. types of cone aphilosophy department, centre for the study of perceptual experience, university of glasgow. 1i propose two simplifying qualifications. first, unless otherwise specified, by “objects” i mean physical, perceptible objects in the environments surrounding perceivers in our world (e.g. blueberries, cars, trees). second, objects don’t have different colours on different parts, but each is instead uniformly coloured. while there is an obvious practical tension between these claims, as many objects in our environment in fact have multiple colours, there is no logical one. since these assumptions permit me to sidestep numerous technicalities that would take me off course, i will hold them in place. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-5765-2443 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 2 cells in their eyes). thus, the colour a blueberry looks to have can vary under different illuminations and across perceivers with different colour visual systems. colour variation, and perceptual variation more generally, have always been and remain core phenomena in perceptual theory.2 most often, colour variation is used to argue against objective colour realism, the thesis that colours are objective or mind-independent properties possessed by objects.3 the familiar reasoning begins with the existence of colour variation. it proceeds by arguing that there is no principled reason to treat one experienced colour of an object as most accurate (no privilege), and thus that we should treat numerous colour variants as equally accurate (ecumenicism). the conclusion is that the best explanation of these facts requires asserting that colours are not objective.4 in this paper, i am not interested in this debate. persistent focus on the tenability of objective colour realism, while important, masks part of the significance of colour variation. to bring this into focus, i presume that the argument from colour variation against objective colour realism is successful and thus the view is (for at least this reason) false. i examine colour variation as it applies to two central remaining colour ontologies. my question is whether colour variation gives us reason to accept that colours are in some sense “out there” while nonetheless being perceiver-dependent (colour relationalism), or to slide to the seemingly more extreme view that eliminates colours from our extra-perceiver world (colour eliminativism)? is colour variation better explained by the postulate that blueberries are blue in part because they look blue to perceivers, or by the postulate that blueberries are not blue in any basic sense despite causing us to have experiences of blue? at first glance, perhaps the question is straightforward. common sense holds: the common sense conception of colour (or simply common sense colour): we routinely talk, think and behave as though colours are objective or perceiver-independent features of objects around us. 2contemporary research on colour variation demonstrates that the phenomenon is widespread. notably, numerous forms of colour variation are found among “normal” perceptions. for example, these might be due to normal variations in lighting, surrounding objects, angle of viewing, or other aspects of perceptual conditions. they could also be due to normal variations in cone absorption curves, adaptation, or other aspects of perceptual apparatuses. it is very difficult to categorize these instances of colour variation as erroneous. in addition, there are powerful arguments to resist categorizing several instances of colour variation found outside “normal” perceptions as erroneous. these include perceptual variations due to some forms of colour blindness, artificial lighting, and so on. see e.g. cohen (2009, ch. 2) for an overview. 3two important versions of objective colour realism are reflectance physicalism (e.g. bradley & tye, 2001; byrne & hilbert, 2003, 2021), and colour primitivism (e.g. allen, 2016; campbell, 1993; gert, 2017; watkins, 2005). 4russell famously employs this reasoning in the opening pages of problems of philosophy. cohen (2009, ch. 2) contains a recent, clear presentation, although his aim is not merely to reject objective colour but instead to defend colour relationalism. the latter inference is the subject of this work. kalderon (2007) is an important critique of this kind of reasoning. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 3 evidence for common sense colour includes our social and linguistic commitment to, for example, blueberries being blue (full stop). evidence also includes our experiencing colour as inhering in object surfaces, as going where the object goes and as remaining stable so long as the object is. and so on. while there is reason to question common sense colour, for dialectical purposes i presume that something in the vicinity is embedded in our colour thought, speech and behaviour.5 eliminativism asserts that objects aren’t coloured and is thus straightforwardly at odds with common sense colour. a natural worry is that any attempt the eliminativist might make to explain common sense colour will be quite revisionary, and thus that eliminativism should be regarded as a view of last resort, countenanced only once all alternatives have been overwhelmingly rejected. we find this argument not only in relationalist works (cohen, 2009, pp. 64–74) but also in eliminativist ones (boghossian & velleman, 1989). in the present context, the result is not merely an immediate victory for relationalism but a victory with significant resilience. so long as relationalism remains coherent and broadly consistent with empirical data it will always be preferable to the staunchly revisionist eliminativism. as tempting as this reasoning is, it is mistaken. crucially, the currently dominant form of relationalism, which i will call ecumenical relationalism (see §3), itself provides a revisionary explanation of common sense colour.6 thus, relationalists cannot object to eliminativism on these grounds without demonstrating that their explanation is “less revisionary” than the eliminativist one. to date no such argument has been offered, and in my judgement the two views are roughly equally revisionary when it comes to explaining common sense colour. in the interest of length i set this topic aside. my focus is instead on the eliminativist and ecumenical relationalist accounts of veridical and non-veridical perceptual experience, both of which are proposed in large part to explain colour variation. we can and should assess this core aspect of these theories independently of their explanations of common sense colour. the two views are diametrically opposed on this dimension: according to eliminativism, all colour experiences are non-veridical; according to ecumenical relationalism, virtually all colour experiences are veridical. this difference, however, rests on a common foundation. 5for recent experimental literature on folk colour beliefs see e.g. adams and hansen (2020), cohen and nichols (2010), roberts et al. (2014) and roberts and schmidtke (2019). 6i regard cohen (2009) and chirimuuta (2015) as the most important defenses of ecumenical relationalism. cohen (2009, ch. 4) recognizes that he needs to provide a revisionary explanation of common sense colour and constructs his “coarse-grained” colours to serve this purpose. coarsegrained colours are not introduced to explain colour perception or colour experience. they are instead “introduced to account for our thought and talk about color” (cohen, 2009, p. 113). they are needed precisely because the heart of ecumenical relationalism – their relational account of the colours we perceive – cannot adequately explain common sense colour. chirimuuta doesn’t provide a developed account of common sense colour. i suspect this is because her primary task is to provide an account of colour vision and colour experience. her theory of the colours we perceive is similar to cohen’s and thus equally unable to explain common sense colour on its own. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 4 both eliminativists and ecumenical relationalists accept colour variation, no privilege and ecumenicism as these are described in the opening paragraph. their difference emerges from what each infers from this foundation. eliminativists infer that colours are most likely not “out there”, and when our colour experiences routinely suggest otherwise the experiences are thereby erroneous. objects often cause us to have colour experiences, but not because the objects are coloured. by contrast, ecumenical relationalists infer that colours are “out there” in some important sense but also that they are relative to the many factors that impact colour experience. as a result, for ecumenical relationalists colour experience is virtually always veridical of the highly-relativized colours in our world. my question is: are there grounds to prefer one of these accounts of (non-)veridical perceptual experience – one of these interpretations of colour variation – over the other? here one might maintain that another argument quickly emerges in favour of ecumenical relationalism: by default, we believe that colour experiences are typically veridical; ecumenical relationalism preserves this; eliminativism does not; hence the former is preferable. i will resist this argument. first, there is plenty of value and purpose to colour experience and colour vision more broadly even if those states are non-veridical (see §§2,6,7). there is thus no theoretical requirement that colour experiences be veridical, and any theory that proposes that they are should be assessed in terms of its explanatory power against rival options. second, it follows that if we are to endorse an account of veridical and nonveridical colour experience, in this case the ecumenical relationalist account, the account must be explanatorily satisfying and more powerful than the opposing eliminativist account. i maintain that this is not achieved in this case. within ecumenical relationalism perceptual illusion is impossible (§5) – this component of their account of non-veridical experience fails. the other component pertains to hallucinations and related phenomena. these, i argue, cannot justify the ecumenical relationalist conception of veridical experience in a debate with the eliminativist (§6). this leaves me to directly compare the ecumenical relationalist conception of veridical experience with the eliminativist analysis of the relevant cases (§7). i argue that the former reduces to the assertion that colour experiences are veridical simply because they cause perceivers to have colour experiences of them. this is, pretheoretically, an exceedingly weak conception of veridicality. more directly, the foundational idea of eliminativism is the opposite claim: given the robustness of colour variation, the fact that objects cause perceivers to have colour experiences of them is on its own not sufficient to justify or ground the claim that colour experiences are veridical. eliminativists are, i argue, on safe territory here. if we are to reject eliminativism in favour of ecumenical relationalism, it must be because we can demonstrate that there is more to objects regarding colour than a power to cause perceivers to have colour experiences of them. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 5 the ecumenical relationalist has failed to do this. their account of veridical colour experience has all the advantages of theft over honest toil (russell, 2020, p. 71). we should thus side with the eliminativist over the ecumenical relationalist: objects don’t possess colour; they merely cause us to undergo colour experiences. i proceed with brief primers on eliminativism (§2) and ecumenical relationalism (§3). readers familiar with these topics can skip to §4. 2 colour eliminativism primer colour eliminativists7 maintain that there are no basic colours in our extraperceiver world. for our purposes, basic colours are properties that intrinsically have features such as hues, saturations and lightnesses.8 eliminativists are driven to this conclusion via key colour perceptual phenomena like colour variation and colour structure9. regarding colour variation and the reasoning rehearsed at the outset, the contention is that the collection of colour experiences of a given object that we are forced to regard as equally veridical is too varied and unprincipled to support the claim that any of them are veridical. thus, objects don’t have colour. before proceeding with that discussion, it is worth briefly unpacking the view. it is fair to say that eliminativists value adequate explanations of phenomena like colour variation more than, for example, offering a straightforward account of common sense colour. in this regard, theirs is a “perception first” approach to colour (as opposed, e.g., to a “language first” or “no priority” approach). they need not (and generally do not) doubt that there are blueberries, that blueberries often cause perceivers to undergo colour experiences, or that our concept blue is often meaningfully applied, and to good effect, to blueberries. what they doubt is that blueberries are blue in a basic sense, believing instead that blueberries merely have the power to cause perceivers to undergo colour experiences. insofar as we can explain what basic colours are, blueberries only have this causal role. no feature of blueberries is part of the basic nature of colour or of what constitutes basic colour. 7see, e.g. averill (2005), boghossian and velleman (1989), hardin (1988), maund (1995, 2006), plus additional works cited in this section. 8the rough idea is that colours have features such as hues, saturations and lightnesses in themselves, as opposed to by virtue of their connection to something outside colour, and because of this these features are central to the nature of colour. by contrast, one might for example hold that hues and the like are not features of colours themselves but arise from experiences of colours, and infer from this that these features are not part of the nature of colour. in my experience the relevant notion of ‘intrinsicness’ is typically left at an intuitive level, though it would be interesting to see if anything interesting arises from applying more specific concepts (e.g. focusing on ‘local’, ‘interior’, or ‘duplication-preserving’ senses of ‘intrinsicness’, or something more specific still; see marshall & weatherson 2018 for an overview of ‘intrinsicness’). 9on the argument from the structure of colour see e.g. hardin (1988) and pautz (2006). see briggs (2021) for a recent overview of colour spaces, noting that what briggs calls the “perceptual models” of human trichromatic colour are most directly relevant to the eliminativist argument from structure. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 6 within these constraints one can develop eliminativism along several dimensions. while these dimensions overlap in interesting ways, it is valuable to identify each on its own. one dimension concerns the account of basic colour. several options are available here, including: colours are properties of sense-data (russell, 1912); colours are properties of internal mental states (mcgilvray, 1994); colours are merely illusory (maund, 1995); colours are only found on eden (chalmers, 2006); colours are platonic universals (pautz, 2007); and so on. another dimension concerns how eliminativists account for common sense colour. two important strategies are dual-colour accounts brown (2006) and fictionalism (e.g. gatzia, 2010). a third dimension concerns the account of colour experience, and in particular the account of experiential error. here one might opt for a classic sensedatum model (e.g. russell, 1912), a more contemporary representationalist one (e.g. chalmers, 2006), et cetera. these dimensions, while important, are not directly relevant to my discussion. of more relevance to my discussion is how the eliminativist mitigates the charge that arises from the eliminativist commitment to widespread error in colour experience. recall, eliminativists are error-theorists about a central component of colour experience: when, in experience, colours are ascribed to objects (as they often are), there is experiential error. the charge is that this commitment seems problematic, a worry that can be expressed via several ideas. shouldn’t we avoid this outcome if possible? why would colour experience be the way it is if it was so often erroneous? is the eliminativist committed, for example, to colour experience being an evolutionary spandrel (i.e. something that was selected for, not because of its adaptive value, but by luck)? the broad eliminativist response has two parts. first, the eliminativist maintains that their account of colour experience is the best explanation of phenomena like colour variation. one way to illustrate this is to compare the eliminativist explanation with that of a rival who, in agreement with the eliminativist, accepts no privilege and ecumenicism. this of course is a central aim of what follows. second, it is important to the eliminativist that their view is compatible with colour experience, and colour vision more widely, having considerable adaptive, epistemic, social and practical value.10 colour experience often helps to correctly represent the shapes, locations and sizes of objects, and the ripeness of fruit. colour experience can be used to quickly identify and track objects and identify their material nature. colour perceptual processing assists spatial processing and, in this regard, helps us to see shapes and objects. colour vision categorizes disparate incoming wavelength information into cognitively manageable “chunks” (e.g. the blues, the greens). and, of course, colour experience makes a substantive 10worthwhile discussions of the value and purpose of colour include hardin (1992), matthen (1999), maund (2006), akins & hahn (2014) and chirimuuta (2015, esp. ch. 4). not all of these theorists advocate eliminativism, but it is straightforward that many if not all of the values or purposes of colour that they discuss are consistent with eliminativism. by contrast, hilbert (1992) offers an account of the purpose of colour vision that directly supports objective colour realism. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 7 contribution to our aesthetic lives. in short, there remain numerous values and purposes of colour experience and colour vision more generally, even if eliminativism is true. since these values and purposes arguably do not directly depend on the correct analysis of the nature of colour let us call them: theoretically neutral values of colour experience: this includes various adaptive, epistemic, social and practical uses of colour experience regardless of one’s colour theory. thus, according to the eliminativist, there are many reasons for why colour experience and colour vision are valuable, even though colour experience and colour vision don’t detect basic colours in the environment. indeed, the eliminativist can push harder and turn this defensive point into an offensive argument. the thought is that there is no need to resist an error-laden account of colour experience, given how valuable colour experience is independently of one’s views about experiential error, colour ontology, and so on. what is gained from insisting that colours are “out there”, or that colour experience is typically veridical? as you will see (§§6&7), this issue manifests itself in a fascinating way in the debate between eliminativists and ecumenical relationalists. i proceed with a brief primer of the latter. 3 colour relationalism primer colour relationalists maintain that colours are relations between perceivers, perceived objects in their environments, and the environmental conditions facilitating these perceptions. the rough thought is that indexing colours to these kinds of parameters creates space to provide a compelling analysis of colour variation. as with colour eliminativism, there are many variants of relationalism.11 on a standard approach objects in our environment have colours, but colours are perceiverdependent. blueberries are blue, but blue is a property that depends in part on perceptual experiences of blue. as mentioned in §1, it is now common for relationalists to treat most if not all instances of colour variation as veridical. thus, the class of veridical colour experiences is quite wide, including variable colour experiences due to colour blindness, strange lighting, contrast effects, those associated with unique hue perceptions, and so on. to capture this colours are defined by the highly specific relations that are sensitive to all the parameters in perceptual systems and perceptual conditions that impact veridical colour experience. thus, a given colour will be defined as a relation that includes “detailed information about the perceiver’s visual system 11i have discussed variations on relationalism elsewhere (see brown, 2015, 2017). here are some options. (1) there can be more and less reductive accounts of colour relationalism. (2) dispositionalists (cohen, 2009) typically hold that objects maintain their colours when not perceived, and interactionists (chirimuuta, 2015) deny this. (3) colours might belong to objects like blueberries that have the power to induce colour perceptual states in perceivers (most relationalists accept this), or to the perceptual interactions that occur during colour perceptions (chirimuuta, 2015). these important differences are tangential to my argument. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 8 and visual circumstance, including state of adaptation, viewing distance and angle, ambient illuminant, simultaneously seen objects, and so on” (cohen, 2009, p. 101). since this approach, which i call ecumenical relationalism, is now the dominant form of colour relationalism, it is my focus.12 let me briefly flesh it out. a natural question to ask is: are there any non-veridical colour experiences within ecumenical relationalism? there are a few, and this fact provides one important point for comparison with eliminativism. i discuss these in §§5–7. for now, simply note that the view does recognize a select few non-veridical colour experiences. since a very wide range of colour variations are veridical, a given object can correctly enter into as many different colour relations – and thus “have” as many colours – as are needed to recover this. how many colours does a given object have? while there is room to debate the matter, cohen has committed himself to: universal colour pluralism: each perceptible object in one’s environment has every colour.13 from these tenets we can isolate a schematic principle for ecumenical relationalism. it is common for relationalists to define colours by schemas that reflect the kind of dependency of colour on objects and perceptual experience that the view defends. a classic formulation is: (a) object o1 is colour c1 iff in normal conditions o1 causes normal perceivers to perceptually experience o1 as c1. since ecumenical relationalism does not define veridicality in terms of anything akin to normal perceptions or conditions, but instead includes a much wider class of cases, cohen proposes the term ‘appropriate’ with the intention that this is to be understood to include the kinds of colour variations mentioned in §1 (note 2). non-veridical colour experiences are deemed ‘inappropriate’. this yields: 12it is of course open to different relationalists to draw the lines between veridical and non-veridical colour experiences differently. i take my lead from two leading manuscripts defending this form of relationalism – cohen (2009) and chirimuuta (2015) – who agree with the characterization in the text. i will contrast their views where appropriate. one important contrast that falls outside our scope is that, whereas cohen is a representationalist about perception chirimuuta is a kind of naïve realist/relationalist, albeit one who accepts the existence of some colour illusions and hallucinations. for simplicity, i will speak of ‘veridical’ and ‘non-veridical’ colour states instead of ‘good’ and ‘bad’ ones, noting that the latter terminology is preferred by naïve realists. nothing of substance in what follows hinges on this. another important difference is that cohen is a dispositionalist (ascribing colours to objects like berries), and chirimuuta is an interactionist (ascribing colours to colour interactions between perceivers and objects). thus, for chirimuuta objects don’t “have” colours, they “enter into” the colour interactions that are colours. this difference does not affect the substance of what follows. for simplicity, i presume dispositionalism in the text and save remarks about interactionist relationalism for footnotes. 13in cohen’s words, “ordinary objects have infinitely many colours” (cohen, 2004, p. 472; see also 2009, p. 133). chirimuuta (2015) doesn’t explicitly address this issue, but her view plausibly entails universal colour pluralism. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 9 (b) object o1 is colour c1 iff in appropriate conditions o1 causes appropriate perceivers to perceptually experience o1 as c1.14 while (b) appears straightforward, it is not. first, the left-to-right conditional is problematic. according to (b), if object o1 is blue, then in appropriate conditions o1 causes appropriate perceivers to perceptually experience o1 as blue. but this is only true in a small subset of cases, and false as a general statement. an improved the left-to-right conditional is: if object o1 is blue, then there is a set (of one or more) appropriate conditions in which o1 causes the members of some set (of one or more) of appropriate perceivers to perceptually experience o1 as blue. the right-to-left conditional is not inherently problematic but nonetheless could mislead. the conditional is: if in appropriate conditions o1 causes appropriate perceivers to perceptually experience o1 as blue, then o1 is blue. this could mislead if it were taken to imply that most appropriate perceivers in most appropriate conditions will experience o1 as blue, which is generally false within ecumenical relationalism. to avoid confusion it is thus beneficial make the relevant scope explicit: if there is a set (of one or more) appropriate conditions in which o1 causes the members of some set (of one or more) of appropriate perceivers to perceptually experience o1 as blue, then o1 is blue. i therefore suggest that ecumenical relationalism is better represented by: (b*) object o1 is colour c1 iff there is a set (of one or more) appropriate conditions in which o1 causes the members of some set (of one or more) of appropriate perceivers to perceptually experience o1 as c1. this is more cleanly consistent with the wide range of veridicality conditions at the heart of ecumenical relationalism, and with universal colour pluralism. i examine its implications below. with the main tenets of ecumenical relationalism in hand15, i proceed with my argument in favour of eliminativism over relationalism. 14cohen’s specific proposal is “object x is blue iff in appropriate conditions x causes perceivers to experience blue” (cohen, 2009, p. 182). i presume (b) is a suitable generalization. i leave the reader to formulate a suitable interactionalist principle in accordance with chirimuuta’s approach. 15there are various critiques of relationalism that i will not consider. one criticism asserts that that we don’t experience colours as relations of the sort posited by relationalism (“colours don’t look like relations”) but instead experienced colours as non-relational properties of objects. if this objection holds, relationalism becomes an error-theory about an aspect of colour experience. some relationalists have responded that colours do look like dispositions or at least don’t look to be non-dispositional chirimuuta (2015). a related charge asserts that colours are non-relational not only in colour experience, but also as represented in common sense colour. if accurate, this would impact relationalism’s ability to explain common sense colour. for recent discussion see again e.g. adams and hansen (2020), cohen and nichols (2010), roberts et al. (2014) and roberts and schmidtke (2019). a third objection stems from circularity, which roughly charges that relationalism defines blue by reference to experiencing/looking blue, and vice versa, and that this is a vicious form of circularity. if successful, the circularity objection renders relationalism incoherent. for important discussions see peacocke (1984), levin (2000), and more recently maund (2012). while the eliminativist may push these kinds of criticisms in defending their view against relationalism, my focus is instead on the ecumenical relationalist conception of (non-)veridicality. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 10 4 contextualizing veridical and non-veridical colour experience for ecumenical colour relationalists, colours are mind-dependent properties of objects that we often correctly perceive in visual experience.16 i suspect that for many readers these are attractive features of the view. however, readers may also worry that it seems too easy to achieve veridical colour experiences within ecumenical relationalism. the combination of (b*), which is rather under-constrained, and universal colour pluralism, collectively seem to imply that perceptual error is hard to come by. this is a challenge that ecumenical relationalists are aware of (see below for references). my aim in the rest of this work is partly to argue that this worry is not misplaced: the ecumenical relationalist conception of veridical and non-veridical colour experience is deeply problematic. the broader, more important, aim is to argue for a positive conclusion, namely that the eliminativist view is explanatorily superior and indeed compelling. my strategy is as follows. in §5 i argue against the ecumenical relationalist conception of colour illusion. in §6 i consider the remaining kinds of non-veridical colour experience endorsed by ecumenical relationalists and argue that these are not sufficient to justify their distinction between veridical and non-veridical colour experience. this is particularly apparent when compared with the eliminativist analysis. in §7 i compare the overall eliminativist analysis with that of the ecumenical relationalist, arguing that no substance remains of the latter’s conception of veridical colour experience. the eliminativist analysis of colour variation is therefore superior to the ecumenical relationalist analysis. before delving into these details, we need some shared terminology around veridicality and non-veridicality in perception. there is no agreed theoretical system for veridicality and non-veridicality in perception, so i will make some assumptions that are broadly in accordance with an established practice. perceptual states represent the world before the subject to be some way or other (e.g. p’s perceptual state might represent that o1 is before p and that o1 is f).17 this is to say that perceptual states express contents about the perceptually available world and thus are to some degree accurate or true of that world (or fail to be such). despite our best efforts, there is nothing akin to an agreed account of perceptual representation. however, it is fruitful to think about perceptual representation in terms of the objects and properties a given representation purportedly detects or measures in the surrounding environment. i follow common practice and understand perceptual experiences as representational.18 16or, for the interactionist, properties of the mind-environment interactions that are colour perceptions. 17might perceptual states represent things outside our world? yes. for example, perhaps one hallucinates a fictional creature such as a unicorn or harry potter. these fascinating cases fall outside the purview of this discussion. 18there is a very large literature on the relation between perceptual experience and perceptual content. i presume the relation is tight enough to do the explanatory work needed here. this brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 11 what is it for a perceptual experience to be veridical? it is generally agreed that for a given perceptual experience to be veridical it must, at minimum, “match” what it represents (the matching condition), and must be “appropriately caused” by what it represents (the causal condition). if an experience represents that o1 is f, then for it to be veridical it must, at minimum, be suitably caused by o1 and o1 must in fact be f. to take a rough example, for an experience as of a sphere to be veridical it must at least be of a sphere (in which case the experience matches the target object), and must arise from looking at that sphere (in which case that sphere is appropriately causing the experience by virtue of the sphere reflecting light into one’s eyes). violating one or more of these conditions thus entails the experience is non-veridical. given that ecumenical relationalists maintain that all experiences of colour variation should be regarded as veridical, what candidates for non-veridical colour experience do they propose? they propose: hallucinations, dreams, afterimages, phosphenes, some colour experiences arising from deviant causal chains, and “lowcontrast scenarios”.19 with the exception of “low-contrast scenarios” (see §5.4), the justification offered for these types of colour experiences being non-veridical is that they are “naturally and pre-theoretically described as erroneous” (cohen, 2007, p. 349).20 thus, it is claimed, the fact that ecumenical relationalism can incorporate this ruling into its account is a virtue. to aid discussion i split the proposed kinds of non-veridical colour experience into two categories: colour illusions: this includes proposed deviant causal chain cases and “low-contrast” scenarios. “disconnected” colour experiences: this includes hallucinations, dreams, afterimages and phosphenes. let me briefly remark on this division. intuitively, a colour illusion involves experiencing something to be a colour that it isn’t, for example when one is looking at something green but experiences it as blue.21 as will be plain, the proposed dedoes not entail that experiences strongly supervene on contents or are identical with contents, though it is consistent with such theses. 19cohen (2007) emphasizes hallucinations, afterimages, and the relevant deviant causal chain case. chirimuuta emphasizes hallucinations, dreams, phosphenes (2015, pp. 155–156) and “lowcontrast scenarios” (2015, p. 181). cohen also recognizes illusions due to error in cognitive (i.e. post-perceptual) judgement. in these cases a perceiver judges her colour experience to be other than it is (e.g. she is primed to expect to experience something green, she instead experiences something as blue, but the prime prompts her to judge that she is actually experiencing the thing as green). since these are not errors of perception, and can be wholly endorsed by eliminativists, i set them aside. 20chirimuuta doesn’t offer an explicit justification for categorizing hallucinations, dreams and phosphenes as erroneous colour experiences. however, the tone of her discussion suggests that, like cohen, to her the designation is natural or intuitive. 21a more technical definition of illusion is: an experience that is being actively caused by, and is about, an instance of a property, and in which the property one experiences (or is having brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 12 viant causal chain cases and “low-contrast” scenarios are intended to be illusions in this rough sense. colour illusions, and illusions more generally, involve an experience that, though erroneous, is nonetheless actively engaging the entity it is a misperception of. in this sense illusions intuitively involve experiences that are “connected” to their intentional object, much like veridical experiences (though, obviously, with veridical experiences the “connection” is wholly a good one). for this reason, when i have cause to discuss the union of illusory and veridical experiences in §§6&7 i will refer to them as connected experiences. by contrast disconnected experiences lack an active or real-time engagement with the relevant perceptual property or object. each case listed satisfies this. dreams occur when one’s eyes are closed, and one isn’t “seeing”. phosphenes occur when one depresses the side of one’s eye and experiences a “coloured blotch” bearing no connection to what is before one (unless by luck). afterimages are colour experiences that are caused by a previously seen stimulus when that stimulus is no longer present. although they are caused by and somewhat resemble past perceptions of objects, afterimages don’t “track” those objects. instead, afterimages maintain a specific location in your visual field (e.g. above your focal point), regardless of what the worldly object does or where your gaze shifts to. hallucinations are often characterized as perceptual experiences that are internally generated and are typically of things that are in fact not there in one’s environment. for example, if you hear a voice and no one is speaking nearby you would be undergoing a hallucination. more precisely, following macpherson and batty (2016), hallucinations are perceptual states that are not causally connected to their objects in any straightforward, counterfactually supporting way. thus, although lsd can cause hallucinations, the lsd isn’t the object of the induced hallucinations and isn’t causally connected to the hallucinations in a counterfactually supporting way. where a hallucination induced by lsd happens to be a hallucination of lsd, the match is by chance and is what is often called a veridical hallucination.22,23 is the ecumenical relationalist account of non-veridical experience robust and justified, and is it superior to the eliminativist claim that all colour experience is non-veridical? it is not. to see this i first assess the ecumenical relationalist account of colour illusion. an experience “as of”) doesn’t match the property in question. this characterization is roughly in line with macpherson and batty (2016), although their conception of illusion also applies to experiences of objects. this innovation is not relevant to my aims. 22this is widely classified as a deviant causal chain case involving a hallucination and not an illusion. as we will see in the next section, these are not the kind of deviant causal chain case ecumenical relationalists put forward as illusory. i thus regard veridical hallucinations as hallucinations, and in turn as disconnected experiences (see below for details). 23two additional points. the types of disconnected cases needn’t be mutually exclusive. e.g. depending on one’s account of hallucination each of dreams, phosphenes and afterimages may or may not count as hallucinations. the disconnected terminology is neutral on these debates. finally, another important class of disconnected experiences – not discussed by ecumenical relationalists – is colour imaginings. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 13 5 colour illusion within ecumenical relationalism recall that the ecumenical relationalist account of colour illusion includes two kinds of cases: deviant causal chains and chirimuuta’s “low-contrast scenarios”.24 i argue that this account fails because their combined commitment to universal colour pluralism and to a vast class of veridical colour experiences (i.e. far beyond what’s included in normal conditions and perceivers) leaves no space to justify attributing error in the cases at issue or in connected colour experiences more generally. within ecumenical relationalism, as soon as a colour experience is connected, it is veridical. my assessment proceeds by examining: cohen’s argument from naturalness (§5.1); deviant causal chains (§5.2); synaesthesia (§5.3); and low-contrast scenarios (§5.4). i conclude that the ecumenical relationalist has not offered a viable account of colour illusion and that one is not likely forthcoming.25 5.1 argument from naturalness as mentioned, ecumenical relationalists justify their account of colour illusion (and non-veridicality more widely) by appeal to “naturalness” and “pretheoretic” appeal (cohen, 2007, p. 349; see also 2012, pp. 369–370). regardless of the naturalness of designating their proposed cases as illusory, this justification is misleading because ecumenical relationalism categorizes as veridical many colour experiences that are “naturally” and “pretheoretically” deemed non-veridical. this includes experiences involving persons with colour blindness and experiences in “funny” viewing conditions (e.g. with colour lights). it also includes experiences of many stock examples of illusions in psychological literature, for example many puzzling instances of contrast effects (e.g. the adelson illusion, kitaoka’s many examples), the watercolour illusions, neon colour spreading illusions, the mccollough effect, and so on.26 on standard accounts, these kinds of illusions are of interest to psychologists because they reveal nuances of colour visual systems by exploiting their rules and effectively “tricking” the systems. thus, these cases are not only intuitively erroneous, they are erroneous for well-motivated theoretical reasons. yet ecumenical relationalists regard them all as veridical. they cannot, therefore, justify their account of illusion on grounds of its “naturalness” 24as mentioned above (fn. 25), a third class of cases involve error due to cognitive (i.e. postperceptual) judgement. as before, i see these aside as they are tangential to this discussion. 25note that gatzia (2010) also argues that ecumenical relationalists are precluded from accommodating colour illusions. she argues that ecumenical relationalists can only accommodate error via hallucination and via cognitions about perceptual experiences (e.g. errors in beliefs about experiences). i agree with her conclusion. however, i regard it as imperative to reach the conclusion by analyzing the argument from naturalness and the following cases, none of which are discussed in gatzia’s otherwise valuable piece. 26readers are encouraged to consult sources such as the illusions index to experience the effects: https://www.illusionsindex.org/. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.illusionsindex.org/ https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 14 by cherry-picking for inclusion a few “natural” cases and excluding a host of others. the legitimacy of their account of colour illusion therefore falls squarely on its internal merit. here are the key cases. 5.2 deviant causal chains cohen’s account of colour illusion rests on his contention that colour experiences involving deviant causal chains should be deemed illusory. since perceptual experiences arising via deviant chains are generally regarded as non-veridical, he is appealing to an antecedently plausible case. his example is of a telekinetic tomato, where the tomato intervenes in any attempt to look at it and telekinetically induces a colour experience that is “opposite” from what the photon-based causal chain would have induced. for example, if one is about to gaze at the unripe tomato and experience it as green, the tomato telekinetically produces an experience of red. thus, the induced colour experience is due to an odd causal chain and is a mismatch for what would otherwise be experienced via a more “standard” visual process. since such a case is typically assessed as involving colour error, cohen argues that ecumenical relationalism can take this assessment on board. i disagree. recall from §4 that for a given perceptual experience to be veridical it must, at minimum, “match” what it represents (the matching condition), and must be “appropriately caused” by what it represents (the causal condition). if an experience represents that o1 is f, then for it to be veridical it must, at minimum, be caused by o1 and o1 must in fact be f. discussions of deviant causal chains typically presume the causal and matching conditions as minimal conditions on veridicality and proceed by examining cases that purport to violate one or both of these conditions.27 unfortunately, these discussions are notoriously tricky, with numerous odd cases and variants on those cases, applied to different theoretical proposals. i do not aim to provide a thorough analysis. instead, i will first make a general case for why ecumenical relationalists cannot import “pretheoretic” intuitions about deviant causal chain cases in support of their account of colour illusion, and then explain this in detail via cohen’s telekinetic tomato case. here is the general worry. universal colour pluralism – the thesis that every object has every colour – guarantees that the matching condition is satisfied for any colour experience of an object. this places considerable weight on the causal condition (or additional conditions) to justify the claim that a given experience is non-veridical. one way to see the broad effect of this is by recalling that a common way of characterizing deviant perceptual cases is as perceptual experiences that match their object but do so by luck or accident. often this implies that the experience could have just as easily failed to match because the match was lucky, or that, if the object were to change in relevant ways the experience would not match because the match was lucky. the difficulty is that, given universal colour pluralism, the matching condition will always be satisfied and thus it is hard to generate a substantial notion 27arstila and pihlainen (2009) is a good review of the literature on deviant causal chains. additional references are given below in both the main text and in footnotes. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 15 of luck. there is an additional concern. when trying to sort causal chains into deviant and non-deviant ones, a plausible starting place is the intuitive division between normal and abnormal perceptual conditions and perceivers, respectively. for example, one can mine cases involving normal conditions for instances of nondeviant causal chains, and mine cases involving abnormal conditions for instances of deviant chains. this strategy is unavailable to the ecumenical relationalism. since it is central to the view that there are many abnormal perceptual conditions and perceivers that yield veridical colour experiences, ecumenical relationalists must start in a different spot. the same, i submit, is likely true of various intuitive starting places for uncovering or examining cases of error due to deviant causal chains. put another way, our intuitions about when a causal chain is deviant (and thus generating perceptual error) cannot be imported into ecumenical relationalism in any straightforward manner. let me illustrate this via cohen’s example. the telekinetic tomato intervenes in any attempt to look at it and telekinetically induces a colour experience that is “opposite” from what the photon-based causal chain would have induced. for example, if the unripe tomato would have been experienced as green when looked at with one’s eyes, the tomato telekinetically induces an experience of red instead of green. let me use classic relationalism (recall (a) from §3) to illustrate a standard analysis of the case. according to classic relationalism, if a normal perceiver were to look at the tomato in normal conditions and experience it as green, then the tomato is green. given this, if the tomato telekinetically intervenes and induces an experience of red in a perceiver who looks at it, then the colour experience is erroneous. why? because the colour experience doesn’t match the tomato’s colour. in addition, we now have reason to distrust the telekinetic causal chain because in normal conditions for a normal perceiver the telekinetically-induced colour experience is erroneous. thus, both the matching and causal conditions are, all else being equal, not satisfied and the experience is cleanly erroneous. this simple analysis is unavailable to the ecumenical relationalist because their analysis begins with the premise that the tomato has every colour. by hypothesis, if our hypothetical perceiver were to look at the tomato, then the photoninduced experience would be of green and hence the telekinetically-induced experience is of red. suppose the perceiver looks at the tomato and experiences the telekinetically-induced experience of red. why is the experience of red erroneous? the ecumenical relationalist cannot assert that the experience of red fails to match the tomato’s colour, because the tomato has every colour. we therefore also don’t yet have evidence to distrust the telekinetic causal chain: all that chain has done is induce an experience that diverges from the experience that would be induced by the photon-based chain, and shifting an experience from one feature an object possesses to another it possesses isn’t evidence of deviance. thus, the simple account of why the case involves error is blocked. a different analysis might assert that telekinetic causal chains are inherently deviant. however, it is difficult to justify this claim on general grounds. to see why, recall lewis’s (1980) classic discussion of prosthetic vision. prosthetic vision brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 16 is a broad category in which perceptual experiences are induced via processes involving “prosthetics”. a standard example involves inserting a bionic eye into one’s eye socket instead of a biological eye. lewis’s influential analysis holds that prosthetic vision is on its own no barrier to veridical perception, provided experiences deemed veridical at minimum satisfy reasonable construals of the matching and causal conditions. thus, if the bionic eye were on the back of one’s head or in another part of the universe the resulting experiences could still be veridical, and (to simplify) a given experience would be veridical when it matches and doesn’t do so by luck, and would be non-veridical otherwise. lewis sees “no real need for any limits on how a prosthetic eye might work. even the least convincing cases of prosthetic vision are quite convincing enough” (lewis, 1980, p. 224). i am not sure if lewis’s permissiveness should be endorsed in general, but let me explain why its application in this context seems apt. telekinetic perception seems causally odd to us. however, presumably there are creatures in possible worlds for whom it is natural, reliable, and capable of yielding veridical experiences. in this regard telekinetic perception is not inherently defective. in cohen’s example it is humans who have telekinetically-induced colour experiences. why deem these experiences to be non-veridical when the matching condition is satisfied and the ecumenical relationalist is already committed to regarding many abnormally produced colour experiences as veridical? it is hard to give a compelling answer. consider some options. (1) consistent experiences. as described by cohen, the tomato consistently induces, via telekinesis, a colour experience that is opposite to the one that would be induced via photon-based vision. a causal mechanism that consistently produces colour experiences that match a colour of the target object has the hallmarks of veridical perception. in this case the perceiver has access to two visual causal pathways, a photon-retinal pathway and a telekinetic one, where the latter overrides the former when the subject looks at this tomato. (2) random experiences. suppose instead that the tomato induces seemingly random colour experiences. setting aside ecumenical relationalism, if an object’s colour changes randomly, then colour experiences of that object that seem random – but are in fact matching – would be veridical. within ecumenical relationalism, each of these seemingly random experiences do satisfy the matching condition. there is thus no obvious reason why any is erroneous. (3) sporadic experiences. suppose instead that the tomato only sporadically activates the telekinetic causal chain, making the causal pathway only sometimes active when viewing the tomato. lewis, for example, is decidedly unconcerned about one-off or sporadic cases28, what matters is whether an 28see e.g. lewis’s deathbed cure and loose wire cases (1980, p. 244). brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 17 experience matches and does so by luck. taking this as our lead, it is again hard to see why sporadic telekinetic chains are inherently defective, when the matching condition is persistently met. in short, the standard dialogue that arises from these kinds of cases has no traction within ecumenical relationalism. the fact that cohen’s telekinetic tomato example is standardly viewed as involving deviant chains, and thus non-veridical colour experiences, thus cannot simply be imported into the ecumenical relationalist account of colour illusion. to justify the claim that some type of causal chain is deviant, the ecumenical relationalist must provide an internal justification. none has been offered. let me make one more claim about deviant chains. there are a class of cases in the deviant causal chain literature that by design involve perceptual experiences that are intuitively hallucinatory yet match and are caused in some “indirect way” by the target object (matching hallucinations). an influential case involves a mad scientist that is directly manipulating your brain so that you have a hallucination that matches a clock on the wall that you are looking toward (strawson, 1974).29 can the ecumenical relationalist argue that this kind of case involves illusory colour experience? no. importantly, the argument in favour of categorizing these kinds of cases as erroneous rests not merely on the oddness of causal chain, but more directly on the fact that the experiences are by hypothesis hallucinatory. this is why the cases are standardly regarded not as examples of colour illusion but of matching colour hallucination. for that reason, i will return to the case when i discuss hallucinations in §6. 5.3 generalizing the worry to see how quickly the worry just expressed generalizes, briefly consider synaesthesia. there are numerous types of colour synaesthesia (see brogaard, 2021 for a current review). for simplicity consider a form of colour-grapheme synaesthesia in which persons experience specific letters to have specific colours, for example ‘a’ is experienced as blue, ‘b’ as orange, and so on. there is intuitively a weirdness to this causal chain, even though the target objects (i.e. the letters) are the consistent, external causes of the colour experiences. setting this weirdness aside for a moment, the simple explanation for why these kinds of experiences are erroneous is that, on most views, specific letters don’t have specific colours: e.g. ‘a’ is not always blue, ‘b’ is not always orange, and so on. yet this explanation is unavailable to the ecumenical relationalist given their commitment to universal colour pluralism. if the ecumenical relationalist wishes to maintain that instances of colour-grapheme synaesthesia are erroneous she must rely on the oddness of the 29a modified version holds that the scientist induces an hallucination matching that clock because you are looking toward that clock, which arguably makes the hallucination counterfactually sensitive in ways lewis would insist upon (noë, 2003). a related variant is the suite of indian cobra cases in arstila and pihlainen (2009). brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 18 causal mechanism, and this is indeed the strategy taken by cohen (2012, pp. 369– 370).30 as above, the difficulty is that her critic can argue that the neural oddity that makes these persons experience ‘a’ as blue (etc.) is, given ecumenical relationalism, not inducing colour illusions but revealing colours the letters by hypothesis have. thus, i again think it is difficult for the ecumenical relationalist to take on board the intuitive analysis of the case via the argument from “naturalness”. i suggest that a similar analysis can be put forward for a host of additional cases that are typically regarded as illusory due to “mismatched” colour experiences triggered by cognitive-neural effects. this would include cases in which memory colour influences colour experiences to depart from the intuitively correct colour (macpherson, 2012), cases in which neural deficits (e.g. aneurisms) either consistently or sporadically affect experienced colours of things, and so on. i conclude that, despite how intuitive it is to deem cohen’s telekinetic scenario as illusory, cohen has not found a workable justification for this claim. i further maintain that the point generalizes: a host of other types of colour experiences that are standardly regarded as erroneous are blocked from being justifiably categorized this way within ecumenical relationalism. cohen’s account of colour illusion thus fails. one final example comes from chirimuuta’s account. 5.4 low-contrast scenarios chirimuuta argues that “low-contrast scenarios” generate colour illusions (2015, p. 181). the underlying insight is that, with regard to stimulus information, colour visual systems are in many ways specifiable as wavelength discriminating systems, as this is what they allow sensory systems to do with incoming signals.31 scenarios with poor wavelength information in sensory signals (where “poor” is defined relative to the capacities of a given type of colour visual system) prevent a given colour visual system from doing its job. thus, colour states arising in those scenarios are non-veridical. chirimuuta offers two examples: (a) conditions of low or dim illumination, where things look grainy and grey; and (b) when one looks at a room illuminated by green light and experiences everything to be black or various shades of green. in these sorts of cases, it is argued, our colour visual systems perform sub-optimally due to environmental factors, thus we shouldn’t regard the experienced colours as “real” (2015, p. 181). this is therefore a colour-specific case that is motivated by a sound theoretical insight and arguably captures the intuitive judgement that there is colour illusion. nonetheless, this conclusion should be resisted, both on general grounds and as a justifiable claim within ecumenical relationalism. i consider these in turn. 30note that cohen (2012) is responding to allen’s charge that if the ecumenical relationalist claims that synaesthetic experiences of this sort are veridical, they would be committing themselves to a claim that “doesn’t seem especially plausible” (allen, 2012, p. 336). nothing in the text rests on allen’s charge. 31for simplicity set aside the question of whether there can in addition be “non-visual” detection systems that operate on wavelength information. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 19 setting aside ecumenical relationalism, note the technical fact that adjusting the spectral power distribution of the illuminant often can and does bring out some colours and suppress others. gardeners are quick to report that in twilight blues “pop out” and reds fade to black. this is the result of selectively diminished wavelength information (relative to fuller daylight) that over-emphasizes the short wavelength “blue” information and contains minimal long wavelength “red” information. the result is a scenario where it is easier for our colour visual systems to process the blues in part because the “blue” wavelength information isn’t getting washed out by the other wavelength information present in fuller daylight. to be sure, there is diminished colour information in such lighting conditions, and as a result our colour visual systems are able to do less with those signals, but this doesn’t entail error. similarly, in monochromatic green light one tends to experience most objects as green. one interpretation of this is as a colour illusion where you experience various non-green things to be green. a different, and in my view more illuminating, interpretation is that you experience the light to be green and, because of the dominance of the illuminant colour, the object colours are occluded from view (brown, 2014). on this latter interpretation no illusion occurs. let me also offer an internal critique that makes it hard to justify chirimuuta’s analysis within ecumenical relationalism. chirimuuta generally allows that illumination variations that impact colour experience do so veridically (e.g. one might veridically perceive something to be white under one illuminant and blue under another). why doesn’t this apply to green lit rooms? if we accept universal colour pluralism, then all objects have every colour and thus all objects are green. why not regard a green lit room as one way of “bringing out” the greenness of various things? similarly, why not regard a dimly lit room as one way of “bringing out” the greyness of various things? scenarios with wavelength information that is limited or poor relative to a given colour visual system can be regarded as scenarios that nudge such a system to select a specific subset of the world’s infinite colours, as opposed to scenarios that prevent the system from accessing any of the world’s real infinite colours. for these reasons “low-contrast scenarios” are not a good example of a colour illusion for ecumenical relationalists. 5.5 conclusion i conclude that the ecumenical relationalist attempt to carve-off a small but legitimate class of colour illusions fails. given how restricted the range of proposed cases is, and how broad the class of veridical colour experiences is within the view, it is unlikely that there are other cases to consider. therefore, within ecumenical relationalism, there are no illusions of colour experience.32 this yields a general conclusion: within ecumenical relationalism, as soon as a colour experience is connected – is caused by the object of the experience in some perceptual condition 32as before, the ecumenical relationalist, like the eliminativist, can still explain some purported examples of illusion by reference to errors in cognitive (i.e. post-perceptual) judgement. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 20 – it is veridical. the ecumenical relationalist notion of “inappropriate” conditions or experiences has no application among connected colour experiences. before assessing the significance of this (§7) i turn to disconnected experiences (§6). 6 disconnected experiences assume the above argument is correct and thus that within ecumenical relationalism all connected colour experiences – experiences of objects that are caused by those objects in some perceptual condition – are veridical. one would be right to ask whether this commitment is problematic on its own. set that question aside for a moment (see §7). the reason is because, even if this commitment is found wanting, the ecumenical relationalist might try to justify their distinction between veridical and non-veridical experience account by appeal to the remaining instances of the latter. the remaining non-veridical experiences are disconnected experiences (i.e. some combination of hallucinations, phosphenes, afterimages and dreams). perhaps examining the difference between connected and disconnected experiences can motivate an underlying distinction between veridical and nonveridical experience. intuitively, the thought is that surely colour hallucinations, phosphenes, dreams and afterimages are non-veridical and other colour experiences must be fundamentally semantically different from this (and thus veridical). my aim in this section is to neutralize this intuition, so that i can make an overall assessment in §7. the argument at issue can be formulated as follows: (1) disconnected colour experiences are non-veridical. (2) disconnected colour experiences are identifiably different from connected colour experiences. (3) the difference between disconnected and connected colour experiences justifies categorizing them differently with regard to veridicality. (4) therefore, connected colour experiences are veridical. while there are contexts where this kind of reasoning can be quite compelling, in this case it should be resisted. there are internal, pretheoretical and theoretical reasons that the ecumenical relationalist can offer in support of (1). the internal reason stems from the idea that ecumenical relationalists must regard disconnected experiences as nonveridical because, by stipulation, colours belong to objects, and the experiences at issue are “disconnected” from these objects. in effect (1) is analytic of ecumenical relationalism and thus justified to the extent that the broader view is justified. however, assuming the criticisms of ecumenical relationalism i have put forward to this point are successful, the broader view is in need of justification. thus, this brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 21 internal justification for (1) carries no weight. disconnected colour experiences are nonetheless often deemed non-veridical pretheoretically, in many if not most theories of colour perception, and by the eliminativist. i thus propose we accept (1). regarding (2), there are many differences between disconnected and connected colour experiences one might examine but our focus needs to be on differences that can help justify the veridicality of connected experiences. an obvious rough starting point is the fact that connected experiences are causally-connected to objects via perceptual conditions and disconnected ones are not. from here the ecumenical relationalist might argue, in defense of (3), that this difference in causalconnectedness is often regarded as a critical line of demarcation between perceptual veridicality and non-veridicality and should be in this case as well. i hope the following analysis, though somewhat oversimplified, will suffice to indicate why i don’t believe any way of fleshing out this generic argument is successful in this context. first, the fact that causal-connectedness is often in general regarded as a critical line of demarcation in distinctions between perceptual veridicality and nonveridicality only yields incomplete justification for (3). this is because it is widely believed (see §§4&5) that causal-connectedness is a necessary but not sufficient condition for veridicality in perception, as revealed by the vast literature on perceptual illusion. thus, even if disconnected and connected colour experiences are different in a semantically-important way, it doesn’t follow that we should generally regard connected colour experiences as veridical of colour, and, on most accounts, several connected colour experiences are non-veridical. to help illustrate the point, consider a different option. suppose objects are coloured but that all connected colour experiences are illusory, and all disconnected ones are hallucinatory. here, there are colours in the environment, but they are perceptually transcendent – we never experience them correctly. advocates of such a view would reject (3). second, the eliminativist rejects (3) and does so for independently viable reasons. the eliminativist recognizes the semantic-import of causal-connectedness in perception and uses this to help illuminate the value of colour vision and colour experience, even though (according to the eliminativist) nothing in one’s surrounding environment has colour. recall (§2) that the eliminativist is committed to: theoretically neutral values of colour experience: this includes various adaptive, epistemic, social and practical uses of colour experience regardless of one’s colour ontology. the epistemic value is of particular relevance to the current discussion. regardless of one’s colour theory, colour experience can and does, to varying degrees, afford perceptual information about objects and non-colour properties. colour experience can help discern object boundaries and shapes, track and identify objects, distinguish surface cues from illumination cues, and so on. indeed chirimuuta (2015), an ecumenical relationalist, makes this point extensively. in turn, there brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 22 are many scenarios where colour experience fails, in varying degrees, to assist with these endeavours. while all colour theories can accept these values, for the eliminativist they are of particular importance because they help relieve pressure the eliminativist otherwise faces to regard any colour experiences as veridical. the eliminativist can distinguish epistemically “good” colour experiences from “bad” ones by reference to how well an experience informs the subject about non-coloured objects and non-colour properties. what she denies is a distinction between veridical and non-veridical colour experience. to apply this to our current discussion, the eliminativist regards disconnected colour experiences as generally uninformative because they are causally disconnected from any relevant part of the environment.33 connected colour experiences categorically have more epistemic potential to be informative about non-colour objects and properties in one’s environment. but they are not infallible. connected colour experiences can succeed and fail to be informative of non-colour objects and properties in interesting ways. thus, for the eliminativist several standard examples of colour illusions are connected colour experiences in which the experienced colours poorly inform the subject about the target non-colour objects and properties. for example, the neon colour spreading illusion suggests object boundaries where there are none, the hermann grid uses colour experience to suggest objects (i.e. dots at grid intersections) that do not match any objects, and so on. a detailed discussion of these matters would be helpful, but i hope it isn’t needed to appreciate the key point. eliminativists accept (2) and propose an alternative to (3): (3*) the difference between disconnected and connected colour experiences justifies categorizing them differently with regard to their respective informativeness of non-colour objects and properties. (3*) captures a key insight that motivates (3), but when (3*) replace (3) in the argument, the inference to (4) is blocked. (3) is thus subject to general criticism, and the very credible (3*) captures a central motive for (3). (3) should therefore be rejected. it follows that the division between disconnected and connected colour experiences cannot on its own underpin the division between veridical and non-veridical colour experience. let me, at last, turn to an overall assessment. 33there are important caveats to this that fall outside our purview. for example, dreams, hallucinations and other disconnected experiences can contain useful non-colour information about things those experiences represent, and they can contain colour information about the colours one would experience were one to view what is represented. these kinds of qualifications are easily handled by eliminativists, ecumenical relationalists and various other theorists. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 23 7 overall assessment colour variation is the various colours a given object is experienced to have across the full range of perceptual conditions and the full range of perceivers. it is widespread. both eliminativists and ecumenical relationalists offer colour variation as a critical source of justification for their respective views. their interpretations of colour variation agree in several key respects. both agree that colour variation persists in a range of cases within and outside any suitable specification of “normal” conditions and perceivers and, further, that the cases are so varied and widespread that there is no principled reason to treat one experienced colour of an object as most accurate (no privilege). they also agree that because of this we should treat numerous colour variants as equally accurate (ecumenicism). finally, they agree that the best explanation of these facts requires asserting that colours are not objective or mind-independent. their disagreement concerns whether to regard the range of cases of colour variation as all veridical (ecumenical relationalism) or as all non-veridical (eliminativism). which interpretation should we prefer? to begin, both eliminativists and ecumenical relationalists accept that disconnected colour experiences (e.g. hallucinations and dreams) are non-veridical. i suggest the root justification for this claim is that these experiences are not, unless by chance, actively caused by the objects they are about, and thus a basic condition for perceptual veridicality – the causal condition – is not met. assuming my analysis in §5 is correct, ecumenical relationalists do not have a viable account of colour illusion. this entails that all colour experiences that are caused by their perceptual objects are veridical (regardless of whether the causal chain is standardly categorized as deviant). that is, all connected colour experiences are veridical within ecumenical relationalism. by contrast, eliminativists maintain that all connected colour experiences are illusory: connected experiences attribute colours to objects causing those experiences, despite those objects not (according to eliminativism) having colours. our debate between eliminativists and ecumenical relationalists thus settles on the following question: should we regard connected colour experiences as wholly veridical or as wholly illusory? put another way: do objects have colour? one might intuitively prefer ecumenical relationalism because it is, in principle, better to endorse a theory that contains veridical colour experiences to one that does not. there are several ways to unpack this intuition, none of which are convincing in this case. first, note that veridicality is not a virtue in and of itself. for example, it is intuitive to pursue theories of the imagination (or sci-fi works or imaginary numbers and so on) that are built around the non-veridicality of imaginings (or sci-fi works, etc.): things don’t become true just because we make them up. however, one might think that perception is different. one might, for example, argue that the outputs of perceptual system should by default be regarded as veridical because the purpose or adaptive function of perceptual systems is to inform us brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 24 about our environment. this argument is subject to numerous objections, many of which involve arguing that adaptive systems have to “work” (i.e. improve fitness), they do not in principle need to issue truthful verdicts about the environment. in my judgement, progress on this kind of debate requires examining the details of a given case. in the present instance, two points are relevant. first, in §§2&6 i emphasized the importance of the theoretically neutral values of colour experience. colour experience is enormously beneficial even if it doesn’t detect colours of objects. these values are thus sufficient to ground a purpose for colour vision and for colour experience. if additional adaptive value emerges from adding that colour experience is often veridical, that value needs to be specified by the ecumenical relationalist. given that no such purported adaptive value has been proposed, and that ecumenical relationalists like chirimuuta (2015, esp. ch. 4) instead emphasize the theoretically neutral values, the burden is on the ecumenical relationalist to make her case. second, when considering whether to interpret connected experiences as wholly veridical or as wholly non-veridical, the proposed conceptions of veridicality and non-veridicality are important. i believe that the conception of veridicality on offer by ecumenical relationalists is inferior to the eliminativist conception of non-veridicality. let me explain why. i begin with a general argument in favour of the eliminativist, and then outline one way it might be developed. we are assuming that within ecumenical relationalism all connected experiences are veridical. it follows that every colour experience that is caused by its object is veridical. this is decidedly close to the claim that colour experiences are veridical by virtue of being caused by their objects. a key worry is that the denial of this claim is at the core of eliminativism. all agree that objects routinely cause us to have colour experiences of them. setting aside ecumenical relationalists, arguably all agree that this alone is not sufficient to deem these colour experiences to be veridical. veridicality of perceptual states must amount to more than this. at this point, eliminativists assert that, from the details of phenomena like colour variation we learn that there is a decidedly unsystematic relationship between objects, perceptual conditions and induced colour experiences. in the absence of a systematic relationship, objects are merely causes of our various colour perceptions of them – objects have no more systematic a role in colour experience than this bare causal role. if this is the case then, the eliminativist continues, colour experiences are not veridical of objects, for veridicality must amount to more than this. let me develop this line of argument. first, one might wonder whether the appropriate response is to suggest that eliminativists and ecumenical relationalists are stuck in a verbal disagreement: eliminativists are thinking of veridical experience in a slightly “thicker” way than ecumenical relationalists. perhaps we should, the thought continues, be ecumenical when considering differing conceptions of veridical experience. i admit this is a possible line of response. however, i don’t think this response is welcome to ecumenical relationalists. they want their view to have more substance than brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 25 this. they want their claim that objects have colour to not boil down to a stipulative definition that can be reached by merely taking the eliminativist framework and swapping ‘veridical’ for ‘non-veridical’ when considering connected experiences. further, given how wide the veridicality conditions are within ecumenical relationalism – covering all connected experiences – there is a burden on the ecumenical relationalist to show that this conception of veridicality has substance. second, one can imagine a situation in which it is appropriate to endorse the veridicality of all connected experiences and use this as a test case against our own situation. thus, consider a world in which: (a) each object has one unique, intrinsic colour. (b) when a perceiver has a colour experience of an object, they experience precisely that object’s colour. that is, connected experiences always “match” their object colour. (c) the general mechanics of perception otherwise work the same as they do in our world (see below for detail). call this the world of stable colour experience (or stable world). stable world is arguably a world in which we should conclude that there are no colour illusions.34 however, this conclusion isn’t reached by stipulating that whenever an object causes a perceiver to have a colour experience of it the experience is “correct”. instead, given (a) and (c), the fact that visual systems output connected colour experiences that always match their object colours is a monumental achievement. visual systems must, subconsciously, control – perfectly – for all variations in perceptual conditions (e.g. variations in lighting, angle of viewing, etc.) and control – perfectly – for all variations perceptual systems (e.g. variations in cone absorption curves, etc.). in this world, connected experiences always match object colour, but do so via an epistemic system that not only achieves perfect results, but overcomes numerous hurdles to get there in each instance. there are no colour illusions in stable world because, for connected experiences, the matching condition is always satisfied and, crucially, satisfying it is a remarkable epistemic achievement. we understand very well what it would take for a connected colour experience to be non-veridical in this world, it just happens that visual systems are able to perfectly avoid such experiences. in stable world, all connected experiences seem veridical. suppose the eliminativist tries to deny this. she insists that all connected experiences cannot be veridical, because, if they were, it would entail that objects merely cause perceivers to have colour experiences of them, and this is insufficient to ground veridicality. this eliminativist position would be untenable precisely because this description 34that is, there are no illusions of colour experience. there could still be errors in post-perceptual judgements about colour, as stressed above, and for some this would be a non-perceptual kind of illusion. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 26 ignores the remarkable epistemic achievement required for all connected experiences to match object colours in stable world. thus, the eliminativist critique of the ecumenical relationalist conception of veridicality doesn’t strictly follow from the idea that all connected experiences are veridical within ecumenical relationalism. if our world is appropriately analogous to that of stable world, the ecumenical relationalist can resist the eliminativist challenge. i hope it is plain that the analogous situation in our world is at the other extreme from stable world. in our world, unlike in stable world, colour experiences are highly sensitive to key features of objects, various dimensions of perceptual conditions, and various dimensions of perceptual apparatuses. that is, in our world variation in colour experience is robust and in need of explanation. this is the kind of scenario in which one would expect that connected colour experiences could fail to correctly represent colour. despite this, within ecumenical relationalism all connected experiences are veridical. how, according to ecumenical relationalism, do our visual systems achieve epistemic perfection for every connected experience? the matching condition is trivially satisfied by all connected experiences because of universal colour pluralism, an ontological postulate that is the opposite of stable world, where each object has a unique colour. assuming §5 is correct, the causal condition is also trivially satisfied by all connected experiences. this is because in our world ecumenical relationalists are committed to interpreting such a wide array of colour variations as veridical, that they retain no resources to block any connected experience from being causally adequate with regards to veridicality. thus, for ecumenical relationalists, in our world all connected experiences are veridical, but this is not the remarkable epistemic achievement that it is in stable world. the achievement is instead seemingly vacuous. there is no identifiable sense in which connected colour experiences detect, track or epistemically “lock onto” colour. we aren’t even sure what would count (in our world), in principle, as a non-veridical connected colour experience within the ecumenical relationalist framework. it is at this point that the eliminativist response has traction in a way that it doesn’t within stable world. the eliminativist insists that in our world the ecumenical relationalist conception of veridicality is indeed vacuous because, within ecumenical relationalism, satisfying the matching and causal conditions requires no substantive epistemic achievement beyond an experience being connected. without a systematic relationship between objects, perceptual conditions, and induced colour experiences that can be leveraged to put substance on the proposed conception of veridicality, the eliminativist is right to insist that the ecumenical relationalist conception of veridicality should be resisted. in this case objects do seem to merely have the power to cause us to have colour experiences of them, and thus we should deem these connected experiences to all be non-veridical. in sum, the eliminativist would be wrong to insist that all connected experiences cannot be veridical. they can be. in stable world this is achieved by a miracbrown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 27 ulous type of perceptual system that perfectly controls for all potential variations in perceptual conditions and apparatuses. the eliminativist should be happy to accept this conclusion, for this scenario rests on a kind of epistemic miracle that has no bearing on our world. in our world colour experience is much messier than this. for all connected experiences to be veridical in our world a different sort of miracle is required: we must permit a notion of perceptual veridicality whose matching and causal conditions are trivially satisfied for every colour experience that is about its object. the eliminativist is right to dig in here and insist that the correct response is not to regard all such experiences as veridical, but to instead regard them all as non-veridical.35 this conclusion about the ecumenical relationalist conception of veridical experience infects the justification for their colour ontology. the reason is because on this view colours are defined in terms of the relations that constitute veridical colour experiences. recall (see §3): (b*) object o1 is colour c1 iff there is a set (of one or more) appropriate conditions in which o1 causes the members of some set (of one or more) of appropriate perceivers to perceptually experience o1 as c1. on this account veridicality is defined in terms of appropriateness such that colour experiences had by appropriate perceivers in appropriate conditions are thereby veridical and are otherwise non-veridical. since introducing (b*) in §3 i have argued that this account of veridicality is unacceptable, in part by arguing (§5) that the conception of “appropriateness” on which it is based cannot be justified. the consequence for the ontology is that what is for an object to have a colour is simply for it to have the power to induce a connected colour experience in a perceiver. that is, for an object to have a colour is for it to be able to cause a perceiver (via reflecting light into retinas or telekinesis or whatever) to have an experience of it as coloured. this, i submit, is far too weak a conception of colour. if all objects do with regards to colour is cause us to have colour experiences of objects, then objects aren’t coloured. the eliminativist view is preferable. 8 concluding thoughts colour variation is the fact that what colour a physical object looks to have depends, in quite wide-ranging and striking ways, on viewing conditions and a per35readers might be wondering about how naïve realism fits into this discussion, as naïve realists are generally resistant to the existence of perceptual illusions. for example, naïve realist kalderon (2011) has argued that there are no colour illusions. in this regard the ecumenical relationalist has an ally of sorts. however, the alliance is limited. kalderon believes in objective, mind-independent colours, in contrast to the relationalist, and kalderon’s proposed mechanisms for handling purported cases of colour illusion are correspondingly very different from those discussed to this point. in addition, in my judgement kalderon and other naïve realists have not considered a host of tough cases of colour illusion, so the naïve realist account is (at present) only of limited value (see e.g. brown, n.d., in preparation). brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 28 ceiver’s visual system. how should we interpret this phenomenon? to focus the discussion, i have assumed that there is no principled reason to treat one experienced colour of an object as most accurate (no privilege), and thus that we should treat numerous colour variants as equally accurate (ecumenicism). the question then becomes whether we should regard all variants as veridical (as the ecumenical colour relationalist contends) or as non-veridical (as the colour eliminativist contends)? i am not sure how systematic the relationship must be between colour experiences, objects and perceptual conditions to justify the veridicality of all of these variants. however, the degree of systematicity found within ecumenical relationalism falls well short of any reasonable standard. as such, the ecumenical relationalist conception of veridicality, particularly given the failure of their account of illusion (§5), is vacuous (§7). further, the account cannot be salvaged by appeal to the non-veridicality of hallucinations, dreams and the like (§6). while this can be read as a wholly critical argument, in my judgement that would be a mistake. the reason is because the pressure to jettison the ecumenical relationalist conception of veridicality is pressure to precisely embrace the colour eliminativist framework. in the face of our messy colour experiential world, the eliminativist asserts that there is no adequately systematic relationship between colour experiences, objects and perceptual conditions. it follows that, although objects routinely cause us to experience colours, that is all they do regarding colour. thus, we should embrace the conclusion that colour experiences are not veridical of objects. the eliminativist analysis of colour variation is superior to the ecumenical relationalist analysis. there are other phenomena that one might consider when assessing eliminativism and ecumenical relationalism on their own or in relation to one another. some of these were mentioned above, such as the structure of colour, the analysis of common sense colour claims (e.g. blueberries are blue), and so on. in my judgement, relative to ecumenical relationalism, eliminativism maintains a strong standing. for example, at this point it should be straightforward to readers that ecumenical relationalist colours – highly relativized features of which every object has an infinite number – are not well suited to directly explain common sense colour claims. cohen recognizes this and provides a reconstructive analysis (2009, ch. 4) that is no less revisionary than existing eliminativist proposals. there is thus an implicit recognition that ecumenical relationalism is in some ways “as extreme” as eliminativism. a thorough analysis of these other issues obviously extends outside my purview. my hope, however, is that the present analysis succeeds in elevating the credibility of colour eliminativism, and in turn motivating further development of the view. let me conclude by reasserting that the above argument specifically concerns how we should interpret colour variation on the assumption that the phenomenon provides adequate grounds to reject objective colour realism. colour variation is important on its own right, but it is absolutely central to views, like colour eliminativism and ecumenical colour relationalism, that reject objective colour brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 29 realism. this being said, eliminativism shouldn’t be rejected by objective colour realists because it ascribes error to colour experiences or because it asserts that objects aren’t coloured. it should be assessed on the value of its explanation of key phenomena like colour variation. on this measure, relative to ecumenical relationalism, eliminativism is superior. how does the eliminativist analysis of colour variation fare relative to objective colour realist ones? as readers might expect, i find that existing objective colour realist analyses of colour variation leave much to be desired. however, examining those details requires a separate work.36 references adams, z., & hansen, n. (2020). the myth of the common sense conception of color. in a. m. wikforss & t. (eds.). shifting concepts: the philosophy and psychology of conceptual variability (106-127). oxford university press. akins, k. a., & hahn, m. (2014). more than mere colouring: the role of spectral information in human vision. the british journal for the philosophy of science, 65(1), 125–171. http://www.jstor.org/stable/24562869 allen, k. (2012). colour, contextualism, and self-locating contents. croatian journal of philosophy, 12(3). https://doi.org/10 .5840/croatjphil201212324 allen, k. (2016). a naïve realist theory of colour. oxford university press. arstila, v., & pihlainen, k. (2009). the causal theory of perception revisited. erkenntnis, 70(3), 397–417. https://doi.org/10 .1007/s10670-008-9153-7 averill, e. w. (2005). toward a projectivist account of color. journal of philosophy, 102(5), 217–234. https://doi.org/10.5840/ jphil2005102525 boghossian, p. a., & velleman, j. d. (1989). colour as a secondary quality 1. mind, xcviii(389), 81–103. https://doi.org/10 .1093/mind/xcviii.389.81 bradley, p., & tye, m. (2001). of colors, kestrels, caterpillars, and leaves. the journal of philosophy, 98(9), 469. https: //doi.org/10.2307/2678495 briggs, d. (2021). colour spaces. in d. h. brown & f. macpherson (eds.), the routledge handbook of philosophy of colour ( 140–156). routledge taylor & francis group. https://doi.org/10.4324/9781351048521-12 brogaard, b. (2021). colour synaesthesia and its philosophical implications. in d. h. brown & f. macpherson (eds.), the routledge handbook of philosophy of colour (210–225). routledge taylor & francis group. https://doi.org/10.4324/97 81351048521-16 brown, d. h. (2006). on the dual referent approach to colour theory. the philosophical quarterly, 56(222), 96–113. https: //doi.org/10.1111/j.1467-9213.2005.00431.x brown, d. h. (2014). colour layering and colour constancy. philosophers, 14. https://philpapers.org/rec/brocla-2 brown, d. h. (2015). colour layering and colour relationalism. minds and machines, 25(2), 177–191. https://doi.org/10.100 7/s11023-015-9363-0 brown, d. h. (2017). colouring for and colour relationalism. analysis, 77(2), 433–449. https://doi.org/10.1093/analys/anx025 brown, d. h. & macpherson. f. (ms). the limits of objective explanations of illusions: a case study in naive realism. byrne, a., & hilbert, d. r. (2003). color realism and color science. the behavioral and brain sciences, 26(1), 3-21; discussion 22-63. https://doi.org/10.1017/s0140525x03000013 byrne, a., & hilbert, d. r. (2021). objectivist reductionism. in d. h. brown & f. macpherson (eds.), the routledge handbook of philosophy of colour. routledge taylor & francis group. https://philpapers.org/rec/byror 36for helpful discussion i am indebted to keith allen, christoph kelp, fiona macpherson, glen pettigrove, adam rieger, and to audiences at the universities of dubrovnik, glasgow, milan, and tokyo. i also benefited greatly from two anonymous referees for this journal and from the journal editor. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 http://www.jstor.org/stable/24562869 https://doi.org/10.5840/croatjphil201212324 https://doi.org/10.5840/croatjphil201212324 https://doi.org/10.1007/s10670-008-9153-7 https://doi.org/10.1007/s10670-008-9153-7 https://doi.org/10.5840/jphil2005102525 https://doi.org/10.5840/jphil2005102525 https://doi.org/10.1093/mind/xcviii.389.81 https://doi.org/10.1093/mind/xcviii.389.81 https://doi.org/10.2307/2678495 https://doi.org/10.2307/2678495 https://doi.org/10.4324/9781351048521-12 https://doi.org/10.4324/9781351048521-16 https://doi.org/10.4324/9781351048521-16 https://doi.org/10.1111/j.1467-9213.2005.00431.x https://doi.org/10.1111/j.1467-9213.2005.00431.x https://philpapers.org/rec/brocla-2 https://doi.org/10.1007/s11023-015-9363-0 https://doi.org/10.1007/s11023-015-9363-0 https://doi.org/10.1093/analys/anx025 https://doi.org/10.1017/s0140525x03000013 https://philpapers.org/rec/byror https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org derek h. brown 30 campbell, j. (1993). a simple view of colour. in john j. haldane & c. wright (eds.). reality: representation and projection (257–268). oxford university press. chalmers, d. j. (2006). perception and the fall from eden. in t. s. gendler & j. hawthorne (eds.), perceptual experience (49–125). oxford university press. chirimuuta, m. (2015). outside color: perceptual science and the puzzle of color in philosophy. the mit press. cohen, j. (2004). color properties and color ascriptions: a relationalist manifesto. the philosophical review, 113(4), 451–506. http://www.jstor.org/stable/4147999 cohen, j. (2007). a relationalist’s guide to error about color perception. noûs, 41(2), 335–353. https://doi.org/10.1111/j.14680068.2007.00650.x cohen, j. (2009). the red and the real: an essay on color ontology. oxford university press. cohen, j. (2012). redness, reality, and relationalism: reply to gert and allen. croatian journal of philosophy, 12(3). https: //philpapers.org/rec/cohrra-3 cohen, j. (2015). ecumenicism, comparability, and color, or: how to have your cake and eat it, too. minds and machines, 25(2), 149–175. https://doi.org/10.1007/s11023-014-9354-6 cohen, j., & nichols, s. (2010). colours, colour relationalism and the deliverances of introspection. analysis, 70(2), 218–228. https://doi.org/10.1093/analys/anp161 gatzia, d. e. (2010). the individual variability problem. philosophia, 38(3), 533–554. https://doi.org/10.1007/s11406-0099234-0 gert, j. (2017). primitive colors: a case study in neo-pragmatist metaphysics and philosophy of perception. oxford university press usa. https://ebookcentral.proquest.com/lib/kxp/detail.action?docid=4883863 hardin, c. l. (1988). color for philosophers: unweaving the rainbow. hackett. hardin, c. l. (1992). the virtues of illusion. philosophical studies, 68(3), 371–382. https://doi.org/10.1007/bf00694852 hilbert, d. r. (1992). what is color vision? philosophical studies, 68(3), 351–370. https://doi.org/10.1007/bf00694851 johnston, m. (1992). how to speak of the colors. philosophical studies, 68(3), 221–263. https://doi.org/10.1007/bf00694847 kalderon, m. e. (2007). color pluralism. philosophical review, 116(4), 563–601. https://doi.org/10.1215/00318108-2007-014 kalderon, m. e. (2011). color illusion. noûs, 45(4), 751–775. https://doi.org/10.1111/j.1468-0068.2010.00781.x langsam, h. (2000). why colours do look like dispositions. the philosophical quarterly, 50(198), 68–75. https://doi.org/10.1 111/1467-9213.00168 levin, j. (2000). dispositional theories of color and the claims of common sense. philosophical studies, 100(2), 151–174. https://doi.org/10.1023/a:1018660204635 lewis, d. (1980). veridical hallucination and prosthetic vision. australasian journal of philosophy, 58(3), 239–249. https: //doi.org/10.1080/00048408012341251 lewis, d. (1997). naming the colours. australasian journal of philosophy, 75(3), 325–342. https://doi.org/10.1080/00048409 712347931 macpherson, f. (2012). cognitive penetration of colour experience: rethinking the issue in light of an indirect mechanism. philosophy and phenomenological research, 84(1), 24–62. https://doi.org/10.1111/j.1933-1592.2010.00481.x macpherson, f., & batty, c. (2016). redefining illusion and hallucination in light of new cases. philosophical issues, 26(1), 263–296. https://doi.org/10.1111/phis.12086 marshall, d. & weatherson, b.n (2018). intrinsic vs. extrinsic properties. in edward n. zalta (ed). the stanford encyclopedia of philosophy (spring 2018 edition). https://plato.stanford.edu/archives/spr2018/entries/intrinsic-extrinsic/. matthen, m. (1999). the disunity of color. the philosophical review, 108(1), 47. https://doi.org/10.2307/2998260 maund, b. (1981). colour — a case for conceptual fission. australasian journal of philosophy, 59(3), 308–322. https://doi.or g/10.1080/00048408112340271 maund, b. (1995). colours: their nature and representation. https://philpapers.org/rec/mauctn-3 maund, b. (2006). the illusory theory of colours: an anti-realist theory. dialectica, 60(3), 245–268. https://doi.org/10.1111/ j.1746-8361.2006.01058.x maund, b. (2012). colour relationalism and colour irrealism/eliminativism/fictionalism. croatian journal of philosophy, 12(3). https://philpapers.org/rec/maucra-4 mcgilvray, j. a. (1994). constant colors in the head. synthese, 100(2), 197–239. https://doi.org/10.1007/bf01063810 mcginn, c. (1986). the subjective view. secondary qualities and indexical thoughts. tijdschrift voor filosofie, 48(2). https: //philpapers.org/rec/mcgtsv-7 mclaughlin, b. (2003). the place of color in nature. in rainer mausfeld & dieter heyer (eds.), colour perception: mind and the physical world. https://philpapers.org/rec/mcltpo-5 brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 http://www.jstor.org/stable/4147999 https://doi.org/10.1111/j.1468-0068.2007.00650.x https://doi.org/10.1111/j.1468-0068.2007.00650.x https://philpapers.org/rec/cohrra-3 https://philpapers.org/rec/cohrra-3 https://doi.org/10.1007/s11023-014-9354-6 https://doi.org/10.1093/analys/anp161 https://doi.org/10.1007/s11406-009-9234-0 https://doi.org/10.1007/s11406-009-9234-0 https://ebookcentral.proquest.com/lib/kxp/detail.action?docid=4883863 https://doi.org/10.1007/bf00694852 https://doi.org/10.1007/bf00694851 https://doi.org/10.1007/bf00694847 https://doi.org/10.1215/00318108-2007-014 https://doi.org/10.1111/j.1468-0068.2010.00781.x https://doi.org/10.1111/1467-9213.00168 https://doi.org/10.1111/1467-9213.00168 https://doi.org/10.1023/a:1018660204635 https://doi.org/10.1080/00048408012341251 https://doi.org/10.1080/00048408012341251 https://doi.org/10.1080/00048409712347931 https://doi.org/10.1080/00048409712347931 https://doi.org/10.1111/j.1933-1592.2010.00481.x https://doi.org/10.1111/phis.12086 https://plato.stanford.edu/archives/spr2018/entries/intrinsic-extrinsic/ https://doi.org/10.2307/2998260 https://doi.org/10.1080/00048408112340271 https://doi.org/10.1080/00048408112340271 https://philpapers.org/rec/mauctn-3 https://doi.org/10.1111/j.1746-8361.2006.01058.x https://doi.org/10.1111/j.1746-8361.2006.01058.x https://philpapers.org/rec/maucra-4 https://doi.org/10.1007/bf01063810 https://philpapers.org/rec/mcgtsv-7 https://philpapers.org/rec/mcgtsv-7 https://philpapers.org/rec/mcltpo-5 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org colour variation without objective colour 31 noë, a. (2003). causation and perception: the puzzle unraveled. analysis, 63(2), 93–100. https://doi.org/10.1093/analys/63. 2.93 pautz, a. (2006). can the physicalist explain colour structure in terms of colour experience?. australasian journal of philosophy, 84(4), 535–564. https://doi.org/10.1080/00048400601079094 pautz, a. (2007). intentionalism and perceptual presence. philosophical perspectives, 21(1), 495–541. https://doi.org/10.111 1/j.1520-8583.2007.00134.x peacocke, c. (1984). colour concepts and colour experience. synthese, 58(3), 365–381. https://doi.org/10.1007/bf00485247 roberts, p., andow, j., & schmidtke, k. (2014). colour relationalism and the real deliverances of introspection. erkenntnis, 79(5), 1173–1189. https://doi.org/10.1007/s10670-014-9600-6 roberts, p., & schmidtke, k. a. (2019). folk core beliefs about color. review of philosophy and psychology, 10(4), 849–869. https://doi.org/10.1007/s13164-019-00437-w russell, b. (1912). the problems of philosophy. barnes & noble. russell, b. (2020). introduction to mathematical philosophy. london: methuen & co ltd. strawson, p. f. (1974). causation in perception. in freedom and resentment and other essays. methuen (reprinted by routledge). watkins, m. (2005). seeing red, the metaphysics of colours without the physics. australasian journal of philosophy, 83(1), 33–52. https://doi.org/10.1080/00048400500043936 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. brown, d. h. (2022). colour variation without objective colour. philosophy and the mind sciences, 3, 17. https://doi.org/10.33735/phimisci.2022.9183 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1093/analys/63.2.93 https://doi.org/10.1093/analys/63.2.93 https://doi.org/10.1080/00048400601079094 https://doi.org/10.1111/j.1520-8583.2007.00134.x https://doi.org/10.1111/j.1520-8583.2007.00134.x https://doi.org/10.1007/bf00485247 https://doi.org/10.1007/s10670-014-9600-6 https://doi.org/10.1007/s13164-019-00437-w https://doi.org/10.1080/00048400500043936 https://doi.org/10.33735/phimisci.2022.9183 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction colour eliminativism primer colour relationalism primer contextualizing veridical and non-veridical colour experience colour illusion within ecumenical relationalism argument from naturalness deviant causal chains generalizing the worry low-contrast scenarios conclusion disconnected experiences overall assessment concluding thoughts mental disorder: an ability-based view. mental disorder: an ability-based view sanja dembića (sanja.dembic@hu-berlin.de) abstract what is it to have a mental disorder? the paper proposes an ability-based view of mental disorder. it argues that such a view is preferable to biological dysfunction views such as wakefield’s harmful dysfunction analysis and boorse’s biostatistical theory. according to the proposed view, having a mental disorder is basically a matter of having a certain type of inability (or: an ability that is not sufficiently high): the inability to respond adequately to some of one’s available reasons in some of one’s reasons-sensitive attitudes or actions, where the threshold of inability is determined by one’s being harmed. the relevant concepts of inability, reasons, and harm are sketched. the paper argues that the proposed view evades some problems of biological dysfunction views by remaining neutral on questions of causation and the evolution of the mind. furthermore, it can capture better what is distinctively “mental” about mental disorder. on the proposed view, it is the rational relations among an individual’s attitudes and actions that are “disordered” and the relevant norms in mental disorder are the norms of reasons. as further merits, the view can account for degrees of disorder, incorporate biological as well as social aspects, and elucidate the relations among disorders, symptoms, and their causes. keywords ability ∙ biological dysfunction ∙ defining mental disorder ∙ reasons ∙ mental disorder this article is part of a special issue on “models and mechanisms in philosophy of psychiatry,” edited by lena kästner and henrik walter. introduction what is it to have a mental disorder? in this paper, i propose an ability-based view of mental disorder and argue that such a view does not have certain problems that common biological ones do.1 the goal of this paper is not to give a full-fledged new account, but to sketch an alternative that is worthwhile of further exploration. i proceed as follows: first, i discuss two major biological views of mental disorder. ahumboldt-universität zu berlin. 1i use the term “biological” views of mental disorder to refer to those views which analyze the concept of mental disorder mainly in terms of biological dysfunctions. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-5824-9314 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 2 second, i delineate my own ability-based view. third, i show how the view i propose evades certain problems of the two major biological views and point out some of its further merits.2 common biological views of mental disorder refer to the concept of biological dysfunction to define “mental disorder” (boorse, 1976b, 1977, 1997, 2014; wakefield, 1992b, 1992a, 1999a, 1999b). roughly, the idea is this: an individual has a mental disorder (if and) only if they are in a condition that results from a failure of some mental mechanism to perform at least one of its biological functions. the two major competing views are wakefield’s “harmful dysfunction analysis” (hda) and boorse’s “biostatistical theory” (bst). they differ (1) with respect to their definition of “biological function” (boorse, 1976a, 2002; wakefield, 1992a, 1999a) and (2) with respect to whether they consider the presence of a biological dysfunction not only necessary, but also sufficient for having a mental disorder (in a theoretical sense of the term). wakefield (1992b) argues that the presence of harm is necessary too. boorse (1997) denies this for the presence of a mental disorder in a theoretical sense, but recognizes it for various practical senses of the term. so, the main difference between the hda and the bst lies in (1) their definitions of “biological function”. boorse offers a goal-directed system definition of “biological function”, whereas wakefield offers an evolutionary one. biological dysfunction views of mental disorder face various problems. it has been argued that the presence of a biological dysfunction (in either sense) is neither necessary nor sufficient for having a mental disorder. in this paper, i refute some replies to these objections and add two further arguments against such views. first, i argue that the concept of biological dysfunction does not track the sense of “disorder” ascribers of mental disorder primarily care about when ascribing a mental disorder, because it doesn’t account for what is specifically “mental” in “mental disorder”. second, i argue that a definition of “mental disorder” which remains neutral on the evolution of the mind and the causes of mental disorder is preferable to ones that rely on certain assumptions on either of those issues as the presented biological dysfunction views do. on the view i propose, having a mental disorder is a matter of having a certain modal property, namely an inability. roughly, the idea is this: an individual has a mental disorder if and only if they are unable (in the relevant sense) to respond adequately to some of their available reasons in some of their thinking, feeling, or acting, where the threshold of inability is determined by the individual’s being harmed. i call this the “rehability view” (rha) which is a portmanteau of “reason, harm, and ability”. i argue that the rha tracks the sense of “disorder” ascribers of mental disorder primarily care about when ascribing a mental disorder. it also remains neutral on questions of causation and the evolution of the mind. i do not discuss the harm condition, because i am interested in a practically relevant 2i present my view in contrast to the two major biological ones (and not in a stand-alone way) to avoid a quite general objection: why do we need a further view when we already have at least two very elaborated views of mental disorder? dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 3 concept of mental disorder. in this respect, both the hda and the bst recognize the relevance of harm. a note on method: my aim is to offer an “explication” (or “ameliorative analysis”) of the technical concept associated with the term “mental disorder” and not a conceptual analysis.3 roughly, the difference is this: conceptual analysis is the (descriptive) project of making explicit the (ordinary or technical) concept associated with a certain term, that is to capture all of its actual (ordinary or technical) uses. an explication is the (potentially revisionary) project of “fashioning” a concept “that will do a certain job” (millikan, 1984, p. 2). in other words, the goal of an explication is to formulate a concept that is fruitful to us for certain purposes. in order to achieve this goal it need not necessarily capture all of its actual (ordinary or technical) uses. why explication? because mental disorder is a contested subject – despite widespread agreement about many phenomena, people in the scientific community disagree about whether certain phenomena do or should fall under the concept of mental disorder – and conceptual analysis cannot give us orientation in cases of disagreement. this is because it has to reflect all of our actual uses to be extensionally adequate.4 any conceptual analysis that does not capture all of our actual uses will be at least slightly revisionary, to wit in an arbitrary way.5 by contrast, an explication is potentially revisionary in a way that is not arbitrary, because it is goal-oriented. nevertheless, even an explication of “mental disorder” aims to capture as many phenomena which are deemed mental disorders within the scientific community as possible (otherwise it wouldn’t be an explication of “our” technical concept associated with the term “mental disorder”). what are the purposes for which we, as members of a society, need the concept of mental disorder? let me mention two: (a) for the purpose of scientific classification and (b) to help us settle practical or normative questions such as who should be eligible for psychiatric or psychotherapeutic treatment and who should be excused from moral or legal responsibility. in light of these purposes, i propose the following set of adequacy conditions which any fruitful definition of “mental disorder” will have to meet: to capture the distinction between (1) mental health and disorder, (2) mental and bodily disorder, (3) mental disorder and deviance from social, legal, or moral norms, as well as to clarify (4) what about having a mental 3i discuss boorse’s and wakefield’s methods in section 1. 4someone might object that a conceptual analysis doesn’t have to capture all of our actual uses, but only the correct ones. but at least with technical terms this simply begs the question, because in cases of disagreement, people precisely disagree about which use of the technical term “mental disorder” is correct. 5there are many slightly imperfect conceptual analyses. which one should we accept? without further argument, the choice seems arbitrary. if a further argument refers to some specific purpose, the purported conceptual analysis turns out to be an explication after all. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 4 disorder it is that may justify certain normative consequences (such as eligibility for treatment or excuse from responsibility).6 in sum, i aim to provide a definition that captures not necessarily all, but as many types of phenomena which are deemed mental disorders within the scientific community as possible and that captures the aforementioned distinctions which are fruitful to us for scientific and practical purposes. i argue that the rha can capture (2) better than the hda and the bst. the paper is structured as follows. in section 1 i delineate the hda and bst and discuss some of their major problems. since the literature on both of these theories is too broad to be discussed in a single paper, i focus only on some of the problems which, as i will argue, are discussed but still unresolved. in section 2 i present the rha, my own positive conception of mental disorder. in section 3 i argue that the rha evades the aforementioned problems of the hda and bst and point out some of its further merits. section 4 concludes. 1 biological dysfunction views wakefield’s harmful dysfunction analysis (hda) of mental disorder which was originally presented in two papers (wakefield, 1992b, 1992a) can be reconstructed as follows: hda an individual s has a mental disorder if and only if s is in a condition c which results from a failure of a mental mechanism to perform at least one of its proper biological functions and c is harmful by the standards of s’s culture. wakefield specifies that “proper biological function” is to be understood in evolutionary terms, that is, in the sense of a naturally selected effect. neander (1991, p. 174) offers the following explication of “proper biological function” (pbf):7 pbf it is the/a proper [biological] function of an item (x) of an organism (o) to do that which items of x’s type did to contribute to the inclusive fitness of o’s ancestors, and which caused the genotype, of which x is the phenotypic expression, to be selected by natural selection. some clarificatory remarks. a genotype is the set of genes an individual possesses. genes can be expressed as traits by an individual. a phenotype is the set of traits an individual possesses. an individual’s phenotype is not only determined by their 6can all adequacy conditions be met by one definition? perhaps, they cannot. but i treat the view that they cannot as a last resort. in this paper, i attempt to show that they can. it remains to be seen whether this project will be successful. 7an explication of “proper biological function” was first introduced by wright (1973) and most prominently developed further by millikan (1989) and neander (1991). wakefield’s own characterization is imprecise, because it doesn’t distinguish between types and tokens. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 5 genes but also by environmental influences. natural selection operates on the phenotype – more specifically: on the variation of phenotypes. it is the phenotype that gives an individual advantages or disadvantages in the struggle for survival and reproduction. and it gives only relative advantages or disadvantages, namely relative to other variations in a population. a trait can be selected because of its effects only if having that effect counts as an adaptive variation in the population. however, what evolves is the genetic setup. evolution is the change in a genotype due to the natural selection of a phenotype. so, two of the basic assumptions of natural selection are (1) that variation in a trait is possible, and (2) that a given trait can be inherited (in the genetic sense: that it can be passed on to the next generation via transmission of genes). in light of this, a trait can possess a proper biological function only if there were variations of that trait in a population and the trait is transmitted genetically. as a consequence, the concept of proper biological function is a highly “demanding” one in the sense that one needs to know a lot about evolution and genetics to justifiably ascribe a proper biological function to a certain trait. intuitions cannot be taken at face value. by contrast, boorse defines “health” in terms of statistically normal physiological functioning and disease (pathological condition or disorder) as an impairment of it.8 boorse (2014, p. 684) formulates his biostatistical theory (bst) as follows: bst 1. the reference class is a natural class of organisms of uniform functional design; specifically, i.e. an age group of a sex of a species. 2. a normal function of an internal part or process within members of a reference class is its statistically typical contribution to survival or reproduction.9 3. health in a member of a reference class is normal functional ability: the readiness of each internal part to perform its normal functions on typical occasions with typical efficiency. 4. a disease or pathological condition is an internal state which impairs health, i.e., reduces one or more functional abilities below typical efficiency. according to boorse, the bst is an analysis of both, the concept of bodily as well as mental disorder. in mental disorders, the relevant mechanisms are mental ones.10 8boorse uses “disease” and “pathological condition” as overarching terms for all deviances from health. i’ll use “disorder” (synonymously), because it is more common in the mental health literature. 9boorse identifies “internal parts or processes” as the bearers of functions. to keep it short, i’ll use “mechanism” to denote the bearer of functions. roughly, a mechanism is a system of causally interacting parts organized such that they are “responsible” for the phenomenon-to-be-explained (see glennan, 2017; krickel, 2018). 10see boorse (1976a) for a more detailed view of how his theory applies to the realm of the mental. in short, boorse (1976a, p. 67) defends the following view: “[a] mental disturbance gets classed dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 6 on boorse’s view, the bearer of disorders are organisms. the physiology of an organism is a hierarchy of goal-directed systems. boorse (2011, p. 27) adopts a dispositional view of the goal-directedness of the relevant systems: a “physical system has the purely physical, nonintensional, property of being directed to a goal g when disposed to adjust its behavior, through some range of environmental variation, in ways needed to achieve g”. boorse contends that in physiology, the relevant highest-level goals of the organism as a whole are survival and reproduction. boorse’s conception of “normal physiological function” (npf) can be reconstructed as follows: npf a mechanism of an organism o has a normal physiological function f if and only if 1. it is statistically normal for the members of the class of organisms o to have a mechanism that causes or constitutes f and 2. f contributes causally to the individual survival or reproduction of the members of the class of organisms o. for instance, the cardiovascular system in a human has the normal physiological function to pump blood through the organism on statistically normal occasions and with at least statistically normal efficiency, because doing so contributes to the individual survival or reproduction of humans and because this is what cardiovascular systems in humans statistically normally do. let me now turn to some problems of these views. 1.1 problems of the hda the hda faces many objections. i contend that the strength of some of them depends on whether the hda is conceived of as a conceptual analysis or an explication of the concept associated with the term “mental disorder”. how should we, as readers, conceive of it? wakefield admits that he has been “sloppy” about it in some of his writings.11 in wakefield (2021, p. 282) he clarifies that he thinks of the hda as a two-step approach: “ ‘harmful dysfunction’ is a conceptual analysis prior to the evolutionary interpretation of ‘dysfunction,’ and the evolutionary interpretation of ‘function’ is an essentialist theoretical move […].” as ‘mental illness’ when some accepted explanation of it refers not to the patient’s physiology but to his feelings, beliefs, and experiences. the defining property of mental disease is mental causation.” for the purposes of this paper, it suffices to say that, according to boorse, the mental healthrelevant mechanisms are mental ones, regardless of how exactly those are to be understood, because nothing in my critical evaluation of the bst hinges on a specific account of mental mechanisms. 11for a discussion, see faucher & forest (2021, ch. 11 and ch. 12, 284). dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 7 so, on wakefield’s view, the first step is a conceptual analysis of “mental disorder” which, on his view, yields that “disorder” means “harmful dysfunction”. the second step is an explication of “function” and “dysfunction” in terms of evolutionary theory. the rationale for this explication is a reference to the best explanation: evolutionary theory provides us with the best scientific theory of the nature of functions and dysfunctions. in the end, the hda is at least partly an explication of “mental disorder”. wakefield has good reasons to conceive of the hda as an explication. if we conceived of the hda as a conceptual analysis, it would be extensionally inadequate, because – as it has been argued in the literature – the presence of a biological dysfunction in terms of pbfs is not necessary for having a mental disorder. for instance, it is conceptually possible that there be (1) mental disorders of spandrels or (2) mental disorders which are themselves adaptations. wakefield rejects these objections, but as i argue in the following, his replies are unsuccessful. (a) spandrels. according to gould & lewontin (1979), spandrels are adventitious side-effects of the development of certain functions which themselves never possessed any adaptive function. as murphy & woolfolk (2000, p. 243) point out, if there are spandrels of the mind (and failures thereof) that indicate mental disorders, but are not also failures of a pbf, then not all mental disorders are dysfunctions in the evolutionary sense. wakefield (2000, p. 254) replies that nobody ever gave an actual example of a spandrel-inspired mental disorder and that merely pointing out the possibility of such a case does not prove anything. it merely asserts what has to be shown. however, if we conceive of the hda as a conceptual analysis of “mental disorder”, then wakefield’s reply is problematic. to show that the presence of a biological dysfunction is not necessary for having a mental disorder, pointing out a hypothetical class of counterexamples is enough. it suffices to show it conceivable that there be a mental disorder without a biological dysfunction. (b) adaptations. murphy & woolfolk (2000, p. 244) claim that some mental disorders such as depression may be adaptive mechanisms (for example, it might be fitness-enhancing to conserve energy and to elicit aid from others) and selected because of that (and thus, have a biological function). so, on their view, not all mental disorders necessarily involve biological dysfunctions. wakefield (2000) replies that intuitions about mental disorders such as depression depend on the intensity of the condition and that this, in turn, correlates with our intuitions about whether they are adaptive and have a biological function. he claims that only moderate depressiveness may have been an adaptive response to a loss. but generally, such moderate cases are not considered dysfunctions or disorders. on his view, biological dysfunctions are attributed only to extreme cases where they do “not appear to be useful strategies by any stretch of the imagination” (wakefield, 2000, p. 260). so, according to wakefield, only some range of the continuum of depressiveness was selected for its beneficial effects and what falls under our concept of depressive disorder is clearly out of this range. but again, if we conceive of the hda as a conceptual analysis, then wakefield’s reply is problematic. whether a certain genetic variant is selected for or against dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 8 depends not only on the phenotype but also on which other variants exist in a population. imagine, in an analogous case, a population in which individuals have either very high levels of fear or no fear at all. even though the chances of survival and reproduction may be low for individuals with high levels of fear, they might still be higher than the chances of survival and reproduction for individuals with no fear at all, since in evolutionary terms, it is better to be “safe than sorry”. it is clearly more adaptive to be able to experience fear (in order to avoid dangerous situations) than to lack this ability completely. hence, it may be the case that extreme cases of fear (or depressiveness) were selected for their effects and possess a biological function after all. regardless of that, extreme fear is considered to be pathological. now, if we conceive of the hda as an explication, then wakefield can be revisionary about the imaginable counterexamples and argue that we should not think of spandrels and adaptations as disorders. but then the question is: why should we, as theoreticians, restrict the concept of mental disorder to biological dysfunctions in terms of pbfs to begin with? in the end, the biological dysfunction condition of the hda is not a conceptual condition, but an empirical one, and as we have seen in the explication of pbf, a highly demanding one, because it requires heritability. but the relevance of the evolutionary perspective for the concept of mental disorder is not evident. as murphy (2020) points out, medicine does not make such a restriction and, as tsou (2021, p. 44) argues, it would be “pragmatically indefensible” if we, as members of a society, stopped considering depression or post-traumatic stress disorder (ptsd) as mental disorders if it turned out that they are not caused or constituted by biological dysfunctions.12 wakefield could simply argue that medicine should make such a restriction and that we should stop considering depression or ptsd as mental disorders if it turned out that they are not caused or constituted by biological dysfunctions in terms of pbfs, because this is what our best scientific theory of the nature of functions and dysfunctions yields and our concept of mental disorder depends on the concept of dysfunction. however, as i’ll argue in section 2 and section 3, i contend that when it comes to mental disorder, the relevant concept is a different one: the inability to respond adequately to one’s available reasons. in any case, it is worth noting that the hda has two drawbacks. first, the status of the phenomena scientists consider mental disorders as mental disorders is preliminary, because it depends on our knowledge of the evolution of the mind, which so far is very limited. second, the hda restricts the causal explanations of mental disorders, because it depends on a highly demanding concept of biological 12would it really be pragmatically indefensible? what if we put them in a different category, such as “socially accepted mental problems”? this is an interesting possibility and i thank an anonymous reviewer for raising this point. tsou would have to clarify that at this point, he uses “mental disorder” in the widest sense of the term which encompasses all conditions that “depart from mental health”. this is compatible with the view that there might be subcategories, for example “mental disorders” in a more narrow sense and “socially accepted mental problems” which do not fit that narrow definition, but are nonetheless departures from mental health. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 9 dysfunction. as long as we, as scientists, don’t know enough about the evolution of the mind and the exact causes of mental disorders, a definition should stay neutral with respect to these issues to avoid unnecessary restrictions and a preliminary status of our classification. even more, defining “mental disorder” is a conceptual project that “picks out” certain phenomena as a set; finding out about the causes of mental disorder (that is of the phenomena “picked out” by the definition of mental disorder) is an empirical project that explains the occurrences of those phenomena. it is not the job of a definition to tell us about the causes of mental disorder and the evolution of the mind, but the job of the empirical sciences. a definition of “mental disorder” which remains neutral on the evolution of the mind and the causes of mental disorders would be preferable, because it would enable us to keep these projects apart and remain open with respect to new empirical findings. most importantly, i contend that the hda is missing something: it does not explicate the sense in which mental disorders are “mental” and does not relate the disorder concept to that specific concept of the mental. as a consequence, it does not track what ascribers of mental disorder primarily care about when talking about “disorder” in the psychiatric or psychotherapeutic context. to clarify this point: to evaluate whether an individual has a mental disorder, any ascriber of mental disorder should evaluate primarily the rational relations among their attitudes (and actions). consider anxiety disorder. to evaluate whether an individual has an anxiety disorder, one has to evaluate whether their fear makes sense or is reasonable in light of their own epistemic situation. do they have beliefs that give them sufficient reason to fear the situations they are, in fact, afraid of? if they do, they do not have an anxiety disorder (unless they don’t have sufficient reason for their beliefs to begin with), even if their fear is objectively inadequate. only if they don’t, they might have one. but the hda does not relate the disorder-status of mental disorders (such as anxiety disorder) to the concept of rationality, but to the concept of biological function of the relevant mental states (such as fear). however, to evaluate an individual’s reasonableness of their fear, the ascriber is not committed to any specific view about the evolution of fear. nothing hinges on whether fear is the product of evolution or an evolutionary by-product. moreover, it seems that any rational creature can be the bearer of a mental disorder, whether it is the product of an evolution or not. 1.2 problems of the bst boorse’s (2011, p. 20) bst offers a conceptual analysis of the theoretical concepts of health and disease (in my terminology: disorder) in scientific medicine. thus, its adequacy should be evaluated with respect to which conditions medicine actually considers to be healthy or disordered. the bst has been criticized extensively and boorse (1997, 2014) offers a nearly comprehensive response to his critics. i focus on some problems which, as i will argue, are still unresolved. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 10 (1) not necessary. one problem for the bst is that it does not capture universal disorders. more specifically, the bst yields that universal biological dysfunction is conceptually impossible. but this is false. conceptually, it is possible that all tokens of a certain type of entity have a biological function, but do not function properly. so, a conception of biological function that does not capture this conceptual intuition does not qualify as an adequate analysis of it. melander (1997, p. 57) gives the following example: if reticulum cancer were to become pandemic in the bovine population thereby making all or most bovine reticulums unable to break down cellulose, bovine reticulums would not typically or normally be able to break down cellulose. but contrary to the proposal, to break down cellulose would then still be a function of bovine reticulums. neander (1991, p. 182), too, argues that the bst yields an absurd consequence, because “if enough of us are stricken with disease (roughly, are dysfunctional) we cease to be diseased, which is nonsense”. if we all go blind, blindness is still a dysfunction. on her view, it is conceptually possible that a biological dysfunction affects all members of a class of organisms. spreading a disease does not make the condition less of a disease. boorse (2002, p. 95) argues that vital biological dysfunctions, if universal, would simply extinguish the species. however, even if true, this would not make it impossible for an entire species to fall ill with a deathly disorder. on boorse’s view, only less than vital universal disorders are a threat to his view, because de facto there are no vital universal disorders. to capture less than vital universal disorders, boorse (2002, p. 95) proposes that we have “to use an extended time-slice of the species” to determine which mechanisms are normal for members of a species to have. the npfs of a species are not determined only by the currently living members of that species, but also by a set of past members. so, for some mechanism to have a npf, it will have to have had it for a sufficient period of time. how long is sufficient? according to boorse (2002, p. 99), any time-slice shorter than a lifetime or two seems too short for the very idea of a species-typical functional design, since identifying many functions in maturation and reproduction requires a longitudinal view of an individual organism and its progeny. boorse points out that this is vague and that at some point vagueness is inevitable. interestingly, boorse’s proposal for less than vital universal disorders indicates that, in the end, our ascriptions of biological functions do not track statistical normality, but rather traits that have been beneficent to our ancestors. in a nutshell, it seems that identifying npfs with respect to time-slices of a certain species might be a way to identify some of the pbfs that contributed to the inclusive fitness of that species. this makes sense, since one would expect that statistics follow function and not the other way around. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 11 there is another reason for believing that the presence of a biological dysfunction in terms of npfs is not necessary for having a mental disorder. tsou (2021, p. 31) points out that certain mental disorders “might turn out to be underwritten by biological mechanisms that behave in predictable ways, but fall within the (statistically) normal range of biological functioning”.13 this echoes the spandrel objection raised against the hda: it seems that we simply do not know enough about the causes of mental disorders to evaluate whether the phenomena falling under our actual concept of mental disorder are caused by biological dysfunctions in terms of npfs. in light of this, a definition of “mental disorder” that remains neutral with respect to the causes of mental disorders seems preferable. (2) not sufficient. critics argue that the bst falsely yields that homosexuality is a mental disorder and that it must be refuted because of that (for example, cooper, 2005, p. 17 and heinz, 2014, p. 44). but this is only approximately true. the bst would yield that homosexuality is a mental disorder if, for instance, the following were true: in homosexual individuals, the mental mechanism responsible for heterosexual desire is not instantiated on statistically normal occasions c with at least the efficiency statistically normal for the members of the relevant reference class of o. causing or constituting heterosexual desire, however, is the npf of that mechanism, because it is statistically normal for members of the relevant reference class of o to have that mechanism and having heterosexual desires contributes causally to the individual survival or reproduction of the members of the relevant reference class of o. the crucial (empirical) question is whether there actually is a mental mechanism that is responsible for heterosexual desire. if there were such a mechanism, the bst would yield that homosexuality is a mental disorder in the theoretical sense. boorse himself does not consider this problematic, because (1) for empirical reasons the case is not clear and (2) even if empirical research indicated that there is such a mechanism, it would only indicate that homosexuality falls under a theoretical concept of disorder. but since normality does not entail desirability, this would have no practical significance: “we always have the right to ask, of normality, what is in it for us that we already desire.” (boorse, 1975, p. 63) what should we make of this? since boorse is interested in an analysis of the theoretical concept of disorder and homosexuality is not considered a mental disorder in clinical psychology or psychiatry, the bst is inadequate as an analysis of our actual theoretical concept of mental disorder. our actual view is not that we, as scientists, do not have sufficient empirical evidence to know whether homosexuality is a mental disorder, but that it is not; not even in a theoretical sense of the term. for people interested in an explication of the concept of mental disorder, this leaves the question open whether we should consider homosexuality a mental disorder. here the answer is “no”. being homosexual does not give its bearer a pro tanto reason for seeking psychiatric or psychotherapeutic treatment (if available). the fact that someone is homosexual is not worthy of psychiatric or 13he refers to maung (2016) and stegenga (2018, ch. 4). dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 12 psychotherapeutic concern (though the fact that they are subject to stigmatization might be). another example showing that the presence of a biological dysfunction in terms of npfs is not sufficient for having a mental disorder is diminished jealousy. according to the bst, an individual would have a mental disorder if there were a mechanism for jealousy instantiated in the individual on statistically normal occasions with less than statistically normal intensity. however, diminished jealousy is not a mental disorder and shouldn’t be considered one, because it is not worthy of psychiatric or psychotherapeutic concern. finally, a point that applies to pbf and npf views of biological dysfunction. the proper bearer of a biological function is a part, process, or mechanism of an organism. in any case, the bearer is a sub-personal entity, that is, something that is “responsible” (either causally or constitutively) for a mental phenomenon which we attribute to individuals as a whole, but not the mental phenomenon itself. mental disorders, however, are ascribed to individuals as a whole. for example, we attribute both the mental phenomenon of fear as well as an anxiety disorder to individuals as a whole, but the pbf or npf to produce fear is attributed to the fear mechanism (if there is any). this makes it possible that a dysfunctional mechanism gets compensated by some other mechanism so that it does not have any “person-level” effects. so, the presence of a biological dysfunction on a sub-personal level does not guarantee the presence of a mental disorder as a personal-level phenomenon. 2 the rehability view in this section, i present the view i want to defend. i propose the following rehability view (rha) of mental disorder14: rha an individual s has a mental disorder if and only if s is unable to respond adequately to some of their available reasons in some of their reasons-sensitive attitudes or actions in view of their mental constitution and their life circumstances, where the threshold of inability is determined by s’s being harmed. the idea is this: when we ascribe a mental disorder to an individual, we basically ascribe to them that they “cannot” (in the relevant sense) respond adequately to some of their available reasons.15 there are reasons for (or against) attitudes such as beliefs and emotions; as well as reasons for (or against) actions. to respond to 14for similar views, see edwards (1981, p. 312), gaete (2008, p. 331), graham (2010, p. 117), and nordenfelt (1987/1995). ability-based views of mental health and disorder (gaete, 2008; nordenfelt, 1987/1995) are often deemed to be too broad. the view i propose aims to amend this shortcoming by making the relevant set of inabilities more precise and by drawing on some of the most recent literature on the concepts of ability, reasons, and harm. in doing so, it also aims to make the rationality claims in edwards (1981) and graham (2010) more precise. 15if you believe that “having” certain reasons already entails that they are in some sense “available” to you, then noting that they are “available” is redundant. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 13 one’s available reasons for attitude a (or action φ) is, generally, to form attitude a (or to φ/intend to φ). to respond to one’s available reasons against attitude a (or action φ) is, generally, to omit forming attitude a (or to omit φ-ing/intending to φ). only if the individual is harmed, their condition is “clinically relevant”, that is, worthy of psychiatric or psychotherapeutic concern. only then are they in the clinically relevant sense “unable” to respond adequately to some of their available reasons. a caveat: we should not confuse an inability to φ with an inability to learn to φ. the latter is a second-order ability: the ability to acquire the ability to φ. that s does not have the ability to φ does not imply that s does not have the second-order ability to learn to φ (or: the potential to φ). i may currently lack the ability to do seventy-five pull-ups, but this does not imply that i do not have the potential to do it. in addition, individuals may not only acquire abilities, but also lose them. i used to be able to speak french pretty well, but today: pas tellement. to motivate the idea that mental disorders are intimately connected to inabilities, imagine an individual with • an anxiety disorder and saying to them “just relax!” • a major depressive disorder and saying to them “just cheer up!” • a delusional disorder and saying to them “just stop thinking that someone is following you!” • an addictive disorder and saying to them “just stop using!” not only are these responses rude and unhelpful, but they also seem to miss the crucial point. as i will argue, there is at least one relevant sense in which an individual with an anxiety disorder precisely “cannot” stop experiencing fear in certain situations; an individual with a major depressive disorder precisely “cannot” stop feeling depressed; an individual with a delusional disorder precisely “cannot” stop believing that somebody is following them; and an individual with an addictive disorder precisely “cannot” stop using drugs.16 if having the ability to φ is a necessary condition for being obliged to φ, then demanding of someone to φ when they precisely “cannot” φ in the relevant sense is, above all, inadequate. on this view, unless an individual “can” φ (in the relevant sense), we cannot demand of them to φ. the challenge for an ability-based view of mental disorder is to specify the exact sense in which an individual with a mental disorder “cannot” φ. let me describe the three core concepts of the rha – ability, reasons, and harm – in more detail. 16see dembić (2021) for a detailed account of addictive disorder in terms of inabilities. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 14 2.1 abilities according to the rha, having a mental disorder necessarily involves a certain inability, to wit in view of one’s mental constitution and one’s life circumstances, where the threshold of inability is determined by s’s being harmed. let me clarify these points.17 first, having an ability is a modal property. in stating that s has the ability to φ, we state something about what s can do, and not primarily something about what s actually does. of course, given that s φs, there seems to be a sense in which s “can” φ. in light of this, an individual’s actual behavioral pattern may sometimes play a heuristic role: given that s φs, we have reason to believe that s has the ability to φ. however, the following does not hold: s’s not φ-ing does not necessarily indicate that s is unable to φ. an individual can have the ability to φ without ever actually φing. second, we should understand locutions such as “s is unable to φ” or “s does not have the ability to φ” as follows: s does not have the ability to φ to a sufficient degree. this is because having an ability is typically not an all-or-nothing matter, but a matter of degree. i can dance better today than ten years ago (i had dance lessons), but mikhail baryshnikov is a better dancer than i am. we can compare an individual’s abilities over time or compare different individuals with respect to their abilities (at the same time or over time). theoretically, it is possible to arrange individuals along a scale of a certain ability in an ascending order. however, though abilities come in degrees, we often make categorical ascriptions: s “has” or “does not have” the ability to φ, period. and we often ask whether an individual has a certain ability, thereby expecting a yes-or-no answer. but that s does not have the ability to φ does not imply that s has the ability to φ to the degree of 0. for instance, if i were asked “can you sing?” at a casting for an opera, it would be misleading of me to answer “yes”. but, of course, this does not imply that i cannot sing at all. what i said or meant was that i cannot sing well enough to be part of the cast in an opera. this specification leads to a further question: what fixes the degree above which an ability counts as “sufficient”? in other words, what determines the threshold on the scale of a certain ability, above which ascriptions of that ability apply categorically? one possible solution is to settle this by the speaker context – the context in which the ascription is made – or more precisely: by the standards that obtain in a given speaker context (jaster, 2020, ch. 4).18 in different practical contexts, different standards obtain. which standards obtain, in turn, depends on the interests of the speakers or the purposes for which they make the ability-ascriptions in the first place. for instance, in the context of a casting for an opera the degree 17see jaster (2020, ch. 1) for an informative overview of some general features of abilities. the following summarizes the main points. 18see stalnaker (2014) on the concept of context. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 15 of singing that counts as “sufficient” will be higher than the one that obtains in the context of a karaoke bar. for which purposes will an individual’s abilities to respond adequately to their available reasons have to be good enough for them to count as not having a mental disorder? i contend that, most generally, the answer is this: to achieve a certain level of well-being. when we ascribe to an individual a mental disorder, we ascribe to them that some of their abilities to respond adequately to their available reasons is so low that they are harmed in some respect of their well-being. to spell out a substantive theory of well-being and harm that is relevant to the context of psychiatry is a task for psychiatric ethics and exceeds the scope of this paper. nevertheless, i make some remarks on the concept of harm in subsection 2.3. at this point, a critic might ask19: but what exactly is the threshold (especially since harm seems to come in degrees as well)? i contend that the threshold varies with the context and that it is not up to the theoretical philosopher to settle the exact threshold for the psychiatric context. on the view i propose, there simply is no “naturally given” threshold to which a theoretician could point. just as mental disorder, abilities come in degrees (and thus, present a “continuum” in a certain sense). any categorical distinction (ability/inability, mental health/disorder) is one that is made within that continuum by speakers for certain purposes. the threshold for psychiatric purposes (say, diagnosis and decisions concerning therapy) will have to be set by the psychiatric context. the exact threshold should not be set from the “armchair”, that is, from a philosophical and purely theoretical point of view. third, abilities are always had “in view of” some facts. this traces back to kratzer’s modal semantics. according to kratzer (1977), there is no absolute sense of “can”. rather, “s can φ” has to be understood as “s can, in view of f, φ”, where f specifies a contextually selected set of facts in view of which the ability is said to be had. so, when evaluating whether s has the ability to φ, we first have to specify the facts which are relevant to the context in which we are interested in knowing that. to illustrate this, consider a professional swimmer with a broken arm. can she swim? in some sense, she can (she is a professional swimmer), but at least in one relevant sense, she cannot (she has a broken arm, after all). we can distinguish these senses by specifying the relevant facts in view of which we evaluate whether she can swim. in view of the fact that she is a professional swimmer and abstracting away of the fact that she has a broken arm it is true to claim that she can swim. but, in view of the fact that she has a broken arm, it is also true to claim that she cannot swim. in the second sense, we do not abstract away of the fact that she has a broken arm. in the literature, this distinction is typically expressed by saying that she has the “general” but not the “specific” ability to swim (honoré, 1964 and mele, 2003). 19i thank an anonymous reviewer for raising this point. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 16 which facts are relevant when evaluating whether an individual has a mental disorder? i contend that in the psychiatric and psychotherapeutic context, ascribers of mental disorders are primarily interested in the individual’s abilities in view of their mental constitution. mental constitution is a vague concept: it comprises an individual’s relatively stable attitudes, personality traits, attribution styles, and so on. to specify them in more detail is the job of psychology and the closely related sciences. when ascribing a mental disorder, we are not interested in the individual’s abilities in view of every single one of their actual mental states at the time of diagnosis. to determine whether i have a major depressive disorder, for instance, we should not consider my brief depressiveness after reading “all quiet on the western front”. in light of that particular mental state, it might turn out that i have a depressive disorder, though i surely need not have one. in ascribing a mental disorder, we ascribe an inability in a more general sense. furthermore, some facts external to the individual are relevant too. even abilities that appear to be fully “intrinsic” are only had in view of certain extrinsic facts. for instance, when we evaluate whether an individual has the ability to hit the bull’s eye, we also hold the actual laws of nature fixed. i contend that the external circumstances that matter for mental disorder are all circumstances in the individual’s life which they cannot easily change. i call these the individual’s “life circumstances”. again, this is imprecise and will have to be specified by the relevant sciences. the rha can incorporate psychiatry-critical ideas such as that sociocultural structures, power relations, and structural barriers construct disability and mental disorder.20 recognizing that abilities relevant to mental disorder are always relative to certain life circumstances has important consequences for therapy, because, in principle, enabling an individual with a mental disorder could be achieved by other means than by “changing” their mental constitution, namely by changing their life circumstances. so, the rha is not committed to the claim that to enable an individual with a mental disorder, it is necessary to change their mental constitution. 2.2 reasons according to the rha, the abilities relevant to mental disorder are abilities to respond adequately to one’s available reasons. thus, mental disorder involves φ-ings (or aspects of φ-ings) which are sensitive to reasons. for some φ-ing (or aspect of φ-ing) to be “sensitive to reasons” means that with respect to that φ-ing (or aspect of φ-ing) it makes sense to ask the question “why?” in a certain sense (anscombe, 1957, p. 9). there are at least three different types of why-questions, only two of which reveal the “sensitivity to reasons” in the relevant sense. to illustrate them, compare the following examples: 20see foucault (1965, 1973, 1977) and tremain (2015). dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 17 1. evidential reasons: “the last train to berlin leaves before midnight.” “why should i believe that?” “that’s what the schedule says.” 2. reasons for action: “jen considers buying a gift for berislav.” “why should she do that?” “to cheer him up.” 3. mere causes: “the candle went out.” “why did that happen?” “there was a draft.” in the literature, answers to the types of questions asked in (1) and (2) are typically called “normative” or “justifying” reasons. the answer in (3) is only superficially similar to (1) and (2). it is only for attitudes and actions that a why-question as in (1) or (2) can meaningfully be asked – or in other words: that normative reasons can be meaningfully asked for and offered. mere events or processes such as a candle going out can be explained by referring to causes, but it does not make sense to ask for or to offer something like normative reasons for or against them. in light of this, it would be misleading to call the things refered to in (3) as “reasons”, even though it is an answer to a certain type of why-question. to avoid misunderstandings, it is better to call them “mere causes”. a prevalent view of normative reasons as deployed in (1) and (2) is the following: a fact (or true proposition) gives us a normative reason when it counts in favour of (or against) our responding in some way, where the response is an attitude (of some type) or an action (of some type).21 as such, normative reasons support some type of response (kiesewetter, 2017). normative reasons that support a response of some type are pro tanto reasons. if s has sufficient reason for a response (of some type), then s is justified to exhibit that response. if s has decisive reason for a response (of some type), then s should or is required to exhibit that response. actions are rationally evaluable with respect to the practical normative reasons rational creatures have for them, that is, with respect to the facts (or true propositions) that count in favour of some action φ of ours to be good or worthwhile of pursuit. we can deliberate on the pro tanto reasons we have for or against some action φ, weigh these reasons against one another, and settle the question on whether we have, all things considered, sufficient (or decisive) reason to φ. attitudes such as beliefs or emotions are rationally evaluable (1) with respect to whether they “fit together”, that is, whether they are consistent or coherent with one another and (2) with respect to the situations we are in, that is, with respect to the facts (or true propositions) that count in favour of the belief being true or the emotion being objectively adequate. on the prevalent view, a fact gives us a normative reason when it counts in favour of our responding in some way, regardless of whether we have a belief 21see, after kiesewetter (2017), scanlon (1998, p. 17), dancy (2000, p. 1), velleman (2000), gibbard (2003, pp. 188–189), finlay (2006, p. 5), thomson (2008, p. 127), raz (2009, p. 18), parfit (2011, p. 31), broome (2013, p. 54). dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 18 related to that fact. but, of course, to play a role in an individual’s reasoning or deliberation about what to do or what to believe, the normative reasons the individual has will have to be present or at least “available” to them in some sense (say, as dispositions in their memory). for instance, the fact that i have a nut allergy gives me a normative reason not to eat nuts, regardless of whether i am aware of that fact. even more, i need not even believe that i have such an allergy. but, of course, if i do not believe that i have a nut allergy, i cannot (be expected to) respond to the normative reason that the respective fact gives me. this is why the relevant abilities in mental disorder are abilities to respond adequately to one’s available reasons. having knowledge of a normative reason is sufficient for it to be available to the individual knowing it. here is a worry concerning the claim that in mental disorder, the φ-ings (or aspects of φ-ings) involved are those that are sensitive to reasons. a critic might object that this view yields some obviously false verdicts, for instance, that mood disorders are not mental disorders. why? because on a prevalent view, moods such as depressiveness are not sensitive to reasons. moods are often distinguished from emotions. whereas it is relatively uncontroversial that emotions such as sadness are sensitive to reasons, it is controversial whether moods are. emotions such as sadness can be objectively adequate or inadequate and they can “fit” to our beliefs or not. but with moods, the case appears to be different: when you wake up in the morning in a good mood, you typically do not think that you have a reason for being in such a good mood. you just happen to be in a good mood and it does not seem to be about anything. prima facie, the rha yields that mood disorders are not mental disorders. however, i contend that this line of thinking involves an implausible view of (many) moods. i contend that moods such as depressiveness are sensitive to reasons. it makes sense to ask an individual for their reasons for or against being in a particular mood. the mere fact that we often do not know why we are in a particular mood does not imply that the question does not apply. perhaps, the question does not apply to certain moods such as waking up in the morning in a good mood. but mostly, as prinz (2004) argues, moods can be objectively adequate or inadequate in light of how life is going for the affected individual quite generally.22 because of that, they are just as sensitive to reasons as emotions are. 2.3 harm according to the rha, having a mental disorder necessarily involves harm. harm is typically contrasted with well-being. typical examples of harm are: pain and 22prinz (2004, p. 185) argues that the difference between emotions and moods is not whether they represent, but what they represent: “sadness represents a particular loss, while depression represents a losing battle.” dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 19 suffering.23 an event e (or derivatively a property of, or a continuant participating in it) that causes some individual s some harm in some respect x – where x is a component of their well-being or other intrinsic good – is a “harmful event”. an individual s who is caused some harm in some respect x by some event e is “harmed”. for example, a car crash causes s to have a painful fracture which is a harm to them in some respect, because it is painful. we can say that the car crash is a “harmful event”, that s was “harmed”, and that the painful injury is a “harm”. i contend that for s to be caused some harm in some respect x is for s to be caused to be worse off in respect x than before.24 since harm comes in degrees, i understand the locution “to be harmed” as “to be harmed to a sufficient degree”. harm and well-being are not exhaustive. some events are neither harmful nor beneficent to a certain individual. for instance, that i am working on my paper today is neither harmful nor beneficent to lebron james in any respect whatsoever. also, harm needs to be distinguished from a mere deprivation of good. an individual deprived of a good is not necessarily harmed. for instance, an individual who does not win the lottery is deprived of a good, but they are not necessarily harmed by that in any respect. however, it is possible that an individual who is prevented from receiving a good is harmed in some cases, namely when it leaves them in a state which is bad to begin with. thus, an individual is not only harmed when caused to be worse off in some respect x than before, but also when being prevented from receiving a good in some respect x and therein being left in a bad state in respect x.25 some event can either be intrinsically or instrumentally good/bad for an individual. an event is intrinsically good for an individual if it has value for them in itself and it is instrumentally good if it has value for them for the sake of something else.26 for instance, dancing is intrinsically good for an individual if it has value for them in itself. having a fever may be instrumentally good for them if it has value for her for the sake of something else, say, for the sake of avoiding a difficult meeting at work. as a consequence, something may at the same time constitute an intrinsic harm for an individual and be instrumentally beneficent to them. for instance, given that vomiting is painful or unpleasant, vomiting is an intrinsic harm for an individual. however, given that it prevents them from further pain by poisoning, it is, at the same time, instrumentally beneficent to them. 23what about masochism? masochism is defined as the practice of seeking pleasure or gratification by means of pain. this does not show that pain is not intrinsically bad, but rather, that something intrinsically bad can be instrumentally good. 24this view is similar to the so-called historical view of harm (rabenberg, 2014; shiffrin, 2012). see feinberg (1986), parfit (1989), and klocksiem (2012) for a counterfactual comparative view of harm, see bradley (2012) and rabenberg (2014) for a critique. 25see rabenberg (2014, p. 19). 26this is compatible with the view that something is valuable only if and because it is valued by someone. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 20 an individual can be pro tanto or all things considered harmed by an event. for instance, a visit to a dentist is typically pro tanto harmful for a patient, because it typically involves causing them an unpleasant or painful experience. but the same visit may be harmless or even beneficent for them on balance or all things considered, because it ultimately prevents them from further unpleasant or painful experiences.27 furthermore, some type of event can be harmless or harmful in a short-term, but harmful or harmless in the long-term. smoking one cigarette probably never killed anyone, but smoking on a regular basis is clearly harmful to smokers in the long-term. it is uncontroversial that in having a mental disorder, the affected individual’s condition can be instrumentally beneficent to them. also, they need not be all things considered harmed by their condition. however, for the condition to fall under the concept of mental disorder, it needs either to constitute an intrinsic harm or to be instrumentally harmful to s in some relevant respect. an individual s can be harmed by their condition c at least in three different ways: 1. c is intrinsically bad for s in some respect x or 2. c causes s to be worse off than before in some respect x or 3. c prevents s from receiving a good in some respect x and therein leaves them in a bad state in respect x, where x is some component of s’s well-being.28 a comprehensive view of mental disorder will have to be supplemented with a substantive theory about the components of human well-being that are relevant to the psychiatric and psychotherapeutic context. (in doing so, any proponent of the view that harm is necessary to mental disorder would also have to deal with the objection that some mental disorders, such as mania, apparently come with heightened well-being.) as already stated in subsection 2.1, providing such a theory exceeds the scope of this paper. at this point, it becomes clear that the concept of mental disorder is not free from normative considerations. we, as rational creatures and members of a society, “expect” people to be able to cope with certain problems of their everyday living, in the sense that we hold them to certain standards, not simply in the sense that they will typically do so. so, whether a condition falls under the concept of mental disorder depends, in part, on our normative expectations. sometimes we criticize the normative expectations that de facto obtain in a sociocultural context. for instance, we may believe that “too many” individuals in a given sociocultural 27see bradley (2012, p. 393). sometimes the distinction is drawn between prima facie harm and harm all things considered, see klocksiem (2012). but to call it “prima facie harm” is misleading, because it suggests that something only appears to be harmful, but actually is not. however, there is a sense in which, for instance, chemotherapy is pro tanto harmful, although it can be all things considered beneficent. 28this is a version of rabenberg (2014). dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 21 context count as having a mental disorder. in such cases, we also tend to believe that our expectations obtaining in that context are too high and we do so, because we believe that we have reasons for this kind of criticism. thus, any comprehensive view of mental disorder will not only have to answer the empirical question “what normative expectations do we de facto have in a given sociocultural context x?” but also the normative one “what normative expectations should we have in a given sociocultural context x?”. 2.4 example: anxiety disorder to make the rha more tangible, let me illustrate it with an example: anxiety disorder. in light of the rha, to evaluate whether an individual has an anxiety disorder, we first have to ask: does the individual have the ability to respond adequately to their available reasons against their fear in view of their mental constitution and their life circumstances? if the answer is “yes”, their fear is non-pathological; if the answer is “no”, their fear is pathological just in case the individual is also harmed by the resulting condition. more formally, a view of anxiety disorder in terms of the rha could look roughly like this: rhaanxiety s has an anxiety disorder (if and) only if 1. s experiences fear in situations c1, c2, …, cn and 2. s is unable to not experience fear in situations c1, c2, …, cn in view of a. s’s mental constitution (including the fact that s has sufficient available reasons against experiencing fear in situations c1, c2, …, cn) and b. s’s life circumstances, where the threshold of inability is determined by s’s being harmed. according to the rha, an anxiety disorder is a mental disorder, because a certain ability to respond adequately to one’s available reasons is impaired or underdeveloped. the relevant adequacy of mental conditions is measured by the standards of reasons (and not primarily by biological, social, or moral standards). harm due to a (low) level of ability to respond adequately to one’s available reasons is what makes the condition count as a mental disorder. it is worth noting that the rha does not locate mental pathology in the veracity of the content of mental states, but in the individual’s epistemic situation, that is, in the relations among their mental states. the rha spells the individual’s “epistemic situation” out in terms of their ability to respond adequatly to their available reasons. let me clarify this point. to ascribe an anxiety disorder to s, we do not have to evaluate whether s’s fear is objectively adequate, that is, whether s is actually dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 22 in danger when in fear. experiencing objectively inadequate fear is neither necessary nor sufficient for having an anxiety disorder. it is not necessary because, conceptually, it is possible for s to have an anxiety disorder and s’s fear to be objectively adequate, namely in cases in which s is actually in danger but does not know or justifiably believe it. one could easily imagine cases in which s suddenly experiences intense fear, truly believes that they are in danger, but has no available reason whatsoever to believe that they are in danger. in such cases, s’s fear would be objectively adequate, but still pathological. why? because s’s fear is inadequate in light of their epistemic situation. experiencing objectively inadequate fear is also not sufficient for having an anxiety disorder. it is not sufficient because it is possible for s not to have an anxiety disorder even though s’s fear is objectively inadequate, namely in cases in which s justifiably believes that they are in danger. again, one could easily imagine cases in which s has every available reason to believe they are in danger even though, in fact, they are not. there can be, after all, misleading evidence. suppose, for instance, that s sees a tiger mock-up. this causes them to believe that they see a tiger and that there is a tiger. since s also believes that tigers are dangerous, s’s fear would be subjectively rational. s’s belief that there is a tiger gives them sufficient apparent reason for their fear. nevertheless, s’s fear is objectively inadequate, because, in fact, they are not in danger. in such cases, s’s fear would not be pathological. why? because s’s fear is adequate in light of their epistemic situation. 3 why adopt the rehability view? in this section, i argue that the rha does not have the aformentioned problems of the hda and the bst and that it has some further merits. 3.1 the rehability view vs. biological dysfunction views first, contra the hda, the rha has no problem with disorders of spandrels or adaptations. furthermore, contra the bst, the rha is compatible with the idea that there might be mental disorders that are caused or constituted by biological mechanisms that fall within the (statistically) normal range of biological functioning. this is because the rha is not committed to any particular causal story of mental disorder. according to the rha, having a mental disorder is a matter of having a certain inability in view of one’s mental constitution and one’s life circumstances. to have an inability is simply to exhibit a certain modal pattern. the rha is compatible with different views on how to spell out the individual’s mental constitution that underlies this pattern. an inability in view of one’s mental constitution and one’s life circumstances might be due to an impairment (a “broken” or dysfunctional biological mechanism) or due to an underdeveloped biological mechanism, but it might also be due to the fact that the individual’s life circumstances dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 23 are such that they are too great a burden to bear. to illustrate the last point, think of a person who has an addiction when at war, but who has no trouble stopping use when back home. the rha can capture the fact that it is not always required to change one’s mental constitution to get rid of a mental disorder, but that sometimes it suffices to change one’s life circumstances.29 in sum, the rha puts no restrictions on the causal explanations of mental disorders. second, according to the rha, the disorder status of the phenomena we consider mental disorders does not depend on our knowledge of the evolution of the mind, because the rha is not dependent on any evolutionary concept. because of that, the disorder status of the relevant phenomena is not (in that sense) preliminary. third, the rha can capture universal disorders. humans, in general, are able to respond adequately to their available reasons in most of their reason-sensitive attitudes and actions. (though they are also, in general, every now and then unreasonable.) even individuals with mental disorders are, in general, able to respond adequately to their available reasons in most of their reasons-sensitive attitudes or actions. it’s just in some that they are not. their inability is “local”, not global. to adopt davidson’s (1982, p. 169) words, just like irrationality, mental disorder “is a failure within the house of reason”. however, nothing in the rha hinges on how many humans are unable to respond adequately to their available reasons in some of their reason-sensitive attitudes and actions. in fact, it is compatible with the view that most or even all humans have such a local inability. hence, it can capture universal mental disorders. fourth, homosexuality. one might worry that the rha falsely yields that homosexuality is a mental disorder. a critic might argue as follows: to be homosexual is to have sexual desires only for people of one’s own sex (or gender). sexual desire is an attitude which is sensitive to reasons, because a certain question “why?” applies to it. now, consider an individual who lives in a society which prescribes the death penalty for homosexuality. it seems that the fact that there is a death penalty for homosexuality gives that individual a reason against their homosexual desire. but we shouldn’t conclude that the individual has a mental disorder if they are not able to change their sexual desire in light of that reason. the trouble with this objection is that the fact that there is a death penalty for homosexuality is no reason against one’s homosexual desire, since it is a reason “of the wrong kind”.30 i do not have a sexual desire for a certain individual because the external circumstances are favourable. considerations showing that it is good or bad for me to have a certain sexual desire in certain circumstances are like considerations showing that it is good or bad to believe something: they do not 29this does not imply that mental pathology can be constituted by life circumstances alone. according to the rha, the relevant inability is always one in view of one’s mental constitution plus life circumstances. 30see hieronymi (2005, 2013) and gertken & kiesewetter (2017) for a discussion of the wrong kind of reasons. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 24 render the belief rationally intelligible to the believer. the sexual desire is not about the external circumstances, it is about a certain individual. reasons for or against a certain sexual desire are given by the qualities of “the object” of desire: whether the sexually desired individual is worthy of sexual desire. of course, the fact that there is a death penalty for homosexuality might be a reason for an individual not to act on their homosexual desire, but it is not a reason against the homosexual desire itself. i contend that the rha correctly yields that homosexuality is not a mental disorder, because it is simply not the case that objects of homosexual desires are unworthy of such desires because they are of the same sex. nothing about the sameness of sex justifies such a verdict. this view rests on a substantive claim about values. but it seems to me correct to interpret potential conflicting views about the status of homosexuality as a mental disorder as conflicts about values (and not, as biological dysfunction views do, as conflicts which can be solved by empirical findings). fifth, the rha does not capture diminished jealousy as a mental disorder, because it is not harmful. the view yields the correct verdict, because diminished jealousy is not worthy of psychiatric or psychotherapeutic concern. sixth, the rha locates the mental disorder on the correct level of description: on the personal level. the bearers of mental disorders are individuals as a whole and not some of their parts or mechanisms. this is captured by the rha, because the bearer of abilities and harm are also individuals as a whole. seventh, the rha tracks what ascribers of mental disorder primarily care about when they talk about “disorder” in the psychiatric or psychotherapeutic context, because it ties the concept of disorder specifically to the concept of the mental which, in turn, is understood in terms of the sensitivity of attitudes or actions to (available) reasons. according to the rha, the crucial question is basically whether an individual is able to think, feel, and act reasonably in light of their epistemic situation. the relevant standards of deviance are the norms of reasons and not the “norms” of biological functions (which are “norms” in a very different sense). thus, the rha can capture what is distinctive about mental disorders. 3.2 further merits in the following, let me point out three further merits of the rha. (1) the rha can capture that mental disorders come in degrees. (2) the rha is conceptually unifying in that it leaves room for both biological and social aspects to play a role in specifying the concept of mental disorder. (3) the rha can illuminate questions about symptoms, disorders, and their causes. first, degrees. mental disorders can be more or less severe. this can be captured by the rha, because both abilities and harm come in degrees as well. consider again the example of an anxiety disorder. in light of the rha, it is easy to dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 25 determine the dimensions which are relevant to assess the severity of an anxiety disorder. the severity seems depend on: 1. the frequency, intensity, and duration of the experience of fear 2. the degree of (in)ability to not experience fear 3. the degree of harm generally, the higher (1) and (3) and the lower (2), the more severe the anxiety disorder will be. furthermore, the rha can acccount for temporary mental disorders (think of a brief psychotic episode). mental disorders can be more or less stable depending on how stable the inability is. as with the threshold, the question of how long an inability needs to last to constitute a mental disorder will have to be set by the psychiatric context. second, biological and social aspects. according to the rha, to have a mental disorder is to have a certain inability in view of: • one’s own mental constitution (relatively stable attitudes) • and one’s own life circumstances (relatively stable external circumstances) the rha doesn’t exclude the relevance of biological dysfunctions, because it is possible that the individual’s mental constitution supervenes on, is identical with, or is realized by a certain mechanism which doesn’t perform some of its biological functions. whether that is the case is an empirical question. if there is a biological dysfunction that corresponds to a certain type of mental disorder, we can specify the inability relevant to it by including the biological dysfunction in the “in view of”-part of the description. the rha doesn’t exclude the relevance of social aspects neither, because an individual’s relatively stable life circumstances include facts about their social environment as well. they include the society in which the individual lives in, the stable relationships and social roles they have, and so forth. if there is a stable set of social facts that corresponds to a certain type of mental disorder, we can specify the inability relevant to it by including those social facts in the “in view of”-part of the description. third, symptoms, disorders, and their causes. the rha can make sense of the fact that in mental disorder neither the symptoms nor the causes of the disorder need to be mental and it can elucidate the relations among symptoms, disorders, and their causes. symptoms are indicators of or evidence for a disorder. the rha contends that having a mental disorder involves being in a certain mental condition and to have a certain modal property, namely the inability to (omit) φ(-ing). what the symptoms indicate is the individual’s mental condition and some of their modal properties. in light of the rha, it is easy to see how there can be symptoms without a disorder. for example, that an individual uses a certain substance may dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 26 be a symptom of an addictive disorder. but it need not. that an individual uses a certain substance does not imply that they are unable to omit it. to evaluate whether they have an addictive disorder, we have to evaluate their corresponding modal pattern. (for that, we may need an epistemology of abilities.) when we ask about the causes of mental disorder, we ask about the causes of an individual’s mental condition and how they came to have a certain modal pattern. 4 conclusion in this paper, i proposed an ability-based view of mental disorder, according to which having a mental disorder is basically a matter of having a certain type of inability (or: an ability that is not sufficiently high): the inability to respond adequately to some of one’s available reasons in some of one’s reasons-sensitive attitudes or actions, where the threshold of inability is determined by one’s being harmed (as relevant in the psychiatric and psychotherapeutic context). i argued that the proposed view evades some of the problems of two prominent biological dysfunction views of mental disorder, the hda and the bst. most notably, i argued that the rha can account for what is specifically “mental” about mental disorders. it is the rational relations among an individual’s attitudes and actions that is “disordered” and the relevant norms in mental disorder are the norms of reasons. furthermore, i argued that the rha can account conceptually for both, social as well as biological aspects of mental disorder. i conclude that the rha is conceptually unifying while remaining highly flexible due to the fact that abilities are always had in view of certain facts and those facts can be of various types. as such, it presents a theoretical alternative that might be worthy of further exploration. acknowledgments i gratefully acknowledge funding from the deutsche forschungsgemeinschaft (dfg) within the centre for advanced studies in the humanities “human abilities”, grant number 409272951. references anscombe, g. e. m. (1957). intention. harvard university press. boorse, c. (1975). on the distinction between disease and illness. philosophy & public affairs, 5(1), 49–68. https://doi.org/10.1515/9781400853564.3 boorse, c. (1976a). what a theory of mental health should be. journal for the theory of social behaviour, 6(1), 61–84. https://doi.org/10.1111/j.1468-5914.1976.tb00359.x boorse, c. (1976b). wright on functions. philosophical review, 85(1), 70–86. https://doi.org/10.2307/2184255 boorse, c. (1977). health as a theoretical concept. philosophy of science, 44(4), 542–573. https://doi.org/10.1086/288768 boorse, c. (1997). a rebuttal on health. in j. m. humber & r. f. almeder (eds.), what is disease? (pp. 1–134). humana press. boorse, c. (2002). a rebuttal on functions. in a. ariew, r. c. cummins, & m. perlman (eds.), functions: new essays in the philosophy of psychology and biology (pp. 63–112). oxford university press. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1515/9781400853564.3 https://doi.org/10.1111/j.1468-5914.1976.tb00359.x https://doi.org/10.2307/2184255 https://doi.org/10.1086/288768 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org mental disorder: an ability-based view 27 boorse, c. (2011). concepts of health and disease. in f. gifford (ed.), philosophy of medicine (pp. 13–64). elsevier. boorse, c. (2014). a second rebuttal on health. journal of medicine and philosophy, 39(6), 683–724. https://doi.org/10.1093/ jmp/jhu035 bradley, b. (2012). doing away with harm. philosophy and phenomenological research, 85(2), 390–412. https://doi.org/10.1 111/j.1933-1592.2012.00615.x broome, j. (2013). rationality through reasoning. wile-blackwell. cooper, r. (2005). classifying madness: a philosophical examination of the diagnostic and statistical manual of mental disorders. springer. dancy, j. (2000). practical reality. oxford university press. https://doi.org/10.1111/j.1933-1592.2003.tb00298.x davidson, d. (1982). rational animals. dialectica, 36(4), 317–328. https://doi.org/10.1111/j.1746-8361.1982.tb01546.x dembić, s. (2021). defining addictive disorder abilities reconsidered. philosophers’ imprint, 21(24), 1–23. edwards, r. b. (1981). mental health as rational autonomy. journal of medicine and philosophy, 6(3), 309–322. https://doi.org/10.1093/jmp/6.3.309 faucher, l., & forest, d. (2021). defining mental disorder: jerome wakefield and his critics. the mit press. feinberg, j. (1986). wrongful life and the counterfactual element in harming. social philosophy and policy, 4(1), 145–178. https://doi.org/10.1017/s0265052500000467 finlay, s. (2006). the reasons that matter. australasian journal of philosophy, 84(1), 1–20. https://doi.org/10.1080/00048400 600571661 foucault, m. (1965). madness and civilization: a history of insanity in the age of reason. tavistock. https://doi.org/10.4324/ 9780203164693 foucault, m. (1973). the birth of the clinic: an archaeology of medical perception. pantheon books. https://doi.org/10.4324/ 9780203715109 foucault, m. (1977). discipline and punish: the birth of the prison. pantheon books. https://doi.org/10.1515/9780822390169058 gaete, a. (2008). the concept of mental disorder: a proposal. philosophy, psychiatry, and psychology, 15(4), 327–339. https://doi.org/10.1353/ppp.0.0210 gertken, j., & kiesewetter, b. (2017). the right and the wrong kind of reasons. philosophy compass, 12(5), 1–14. https://doi.org/10.1111/phc3.12412 gibbard, a. (2003). thinking how to live. harvard university press. https://doi.org/10.4159/9780674037588 glennan, s. (2017). the new mechanical philosophy. oxford university press. gould, s. j., & lewontin, r. c. (1979). the spandrels of san marco and the panglossian paradigm: a critique of the adaptionist programme. proceedings of the royal society b: biological sciences, 205(1161), 252–270. https://doi.org/10.1098/rspb.1 979.0086 graham, g. (2010). the disordered mind: an introduction to philosophy of mind and mental illness. routledge. https://doi.org/10.4324/9780203857861 heinz, a. (2014). der begriff der psychischen krankheit. suhrkamp. hieronymi, p. (2005). the wrong kind of reason. journal of philosophy, 102(9), 437–457. https://doi.org/10.5840/jphil20051 02933 hieronymi, p. (2013). the use of reasons in thought (and the use of earmarks in arguments). ethics, 124(1), 114–127. https://doi.org/10.1086/671402 honoré, a. m. (1964). can and can’t. mind, 73(292), 463–479. https://doi.org/10.1093/mind/lxxiii.292.463 jaster, r. (2020). agents’ abilities. de gruyter. https://doi.org/10.1515/9783110650464 kiesewetter, b. (2017). the normativity of rationality. oxford university press. klocksiem, j. (2012). a defense of the counterfactual comparative account of harm. american philosophical quarterly, 49(4), 285–300. https://doi.org/10.1111/papq.12031 kratzer, a. (1977). what ‚must’ and ‚can’ must and can mean. linguistics and philosophy, 1(3), 337–355. https://doi.org/10 .1007/bf00353453 krickel, b. (2018). the mechanical world the metaphysical commitments of the new mechanistic approach. springer international publishing. maung, h. h. (2016). diagnosis and causal explanation in psychiatry. studies in history and philosophy of biological and biomedical sciences, 60, 15–24. https://doi.org/10.1016/j.shpsc.2016.09.003 melander, p. (1997). analyzing functions: an essay on a fundamental notion in biology. almqvist & wiksell. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1093/jmp/jhu035 https://doi.org/10.1093/jmp/jhu035 https://doi.org/10.1111/j.1933-1592.2012.00615.x https://doi.org/10.1111/j.1933-1592.2012.00615.x https://doi.org/10.1111/j.1933-1592.2003.tb00298.x https://doi.org/10.1111/j.1746-8361.1982.tb01546.x https://doi.org/10.1093/jmp/6.3.309 https://doi.org/10.1017/s0265052500000467 https://doi.org/10.1080/00048400600571661 https://doi.org/10.1080/00048400600571661 https://doi.org/10.4324/9780203164693 https://doi.org/10.4324/9780203164693 https://doi.org/10.4324/9780203715109 https://doi.org/10.4324/9780203715109 https://doi.org/10.1515/9780822390169-058 https://doi.org/10.1515/9780822390169-058 https://doi.org/10.1353/ppp.0.0210 https://doi.org/10.1111/phc3.12412 https://doi.org/10.4159/9780674037588 https://doi.org/10.1098/rspb.1979.0086 https://doi.org/10.1098/rspb.1979.0086 https://doi.org/10.4324/9780203857861 https://doi.org/10.5840/jphil2005102933 https://doi.org/10.5840/jphil2005102933 https://doi.org/10.1086/671402 https://doi.org/10.1093/mind/lxxiii.292.463 https://doi.org/10.1515/9783110650464 https://doi.org/10.1111/papq.12031 https://doi.org/10.1007/bf00353453 https://doi.org/10.1007/bf00353453 https://doi.org/10.1016/j.shpsc.2016.09.003 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sanja dembić 28 mele, a. r. (2003). agents’ abilities. noûs, 37(3), 447–470. https://doi.org/10.1111/1468-0068.00446 millikan, r. g. (1984). language, thought and other biological categories: new foundations for realism. the mit press. millikan, r. g. (1989). in defense of proper functions. philosophy of science, 56, 288–302. https://doi.org/10.1086/289488 murphy, d. (2020). concepts of disease and health. in e. n. zalta (ed.), the stanford encyclopedia of philosophy. metaphysics research lab, stanford university. https://plato.stanford.edu/entries/health-disease/ murphy, d., & woolfolk, r. l. (2000). conceptual analysis versus scientific understanding: an assessment of wakefield’s folk psychiatry. philosophy, psychiatry, and psychology, 7(4), 271–293. neander, k. (1991). functions as selected effects: the conceptual analyst’s defense. philosophy of science, 58(2), 168–184. https://doi.org/10.1086/289610 nordenfelt, l. (1995). on the nature of health: an action-theoretic approach (2nd revised and enlarged edition). springer. https://doi.org/10.1007/978-94-011-0241-4 (original work published 1987) parfit, d. (1989). reasons and persons. oxford university press. https://doi.org/10.4324/9780429488450 parfit, d. (2011). on what matters (vol. 1). oxford university press. https://doi.org/10.1093/acprof:osobl/9780199572809.00 1.0001 prinz, j. j. (2004). gut reactions: a perceptual theory of the emotions. oxford university press. rabenberg, m. (2014). harm. ethics & social philosophy, 8(3), 1–32. https://doi.org/10.26556/jesp.v8i3.84 raz, j. (2009). reasons: explanatory and normative. in c. sandis (ed.), new essays on the explanation of action (pp. 184–202). palgrave-macmillan. https://doi.org/10.2139/ssrn.999869 scanlon, t. m. (1998). what we owe to each other. harvard university press. https://doi.org/10.1111/j.1933-1592.2003.tb00 249.x shiffrin, s. (2012). harm and its moral significance. legal theory, 18(3), 357–398. https://doi.org/10.1017/s1352325212000080 stalnaker, r. c. (2014). context. oxford university press. stegenga, j. (2018). medical nihilism. oxford university press. thomson, j. j. (2008). normativity. open court. tremain, s. (2015). foucault and the government of disability. university of michigan press. tsou, j. y. (2021). philosophy of psychiatry. cambridge university press. velleman, d. (2000). the possibility of practical reason. oxford university press. wakefield, j. c. (1992a). disorder as harmful dysfunction: a conceptual critique of dsm-iii-r’s definition of mental disorder. psychological review, 99(2), 232–247. https://doi.org/10.1037/0033-295x.99.2.232 wakefield, j. c. (1992b). the concept of mental disorder. on the boundary between biological facts and social values. american psychologist, 47(3), 373–388. https://doi.org/10.1037//0003-066x.47.3.373 wakefield, j. c. (1999a). evolutionary versus prototype analyses of the concept of disorder. journal of abnormal psychology, 108, 374–399. https://doi.org/10.1037//0021-843x.108.3.374 wakefield, j. c. (1999b). mental disorder as a black box essentialist concept. journal of abnormal psychology, 108, 465–472. https://doi.org/10.1037//0021-843x.108.3.465 wakefield, j. c. (2000). spandrels, vestigial organs, and such: reply to murphy and woolfolk’s “the harmful dysfunction analysis of mental disorder.” philosophy, psychiatry, and psychology, 7(4), 253–269. https://doi.org/10.1353/ppp.2000.0 040 wakefield, j. c. (2021). is the harmful dysfunction analysis descriptive or stipulative, and is the hda or bst the better naturalist account of dysfunction? reply to maël lemoine. in l. faucher & d. forest (eds.), defining mental disorder. jerome wakefield and his critics (pp. 367–424). the mit press. wright, l. (1973). functions. philosophical review, 82(2), 139–168. https://doi.org/10.2307/2183766 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. dembić, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4, 2. https://doi.org/10.33735/phimisci.2023.9630 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1111/1468-0068.00446 https://doi.org/10.1086/289488 https://plato.stanford.edu/entries/health-disease/ https://doi.org/10.1086/289610 https://doi.org/10.1007/978-94-011-0241-4 https://doi.org/10.4324/9780429488450 https://doi.org/10.1093/acprof:osobl/9780199572809.001.0001 https://doi.org/10.1093/acprof:osobl/9780199572809.001.0001 https://doi.org/10.26556/jesp.v8i3.84 https://doi.org/10.2139/ssrn.999869 https://doi.org/10.1111/j.1933-1592.2003.tb00249.x https://doi.org/10.1111/j.1933-1592.2003.tb00249.x https://doi.org/10.1017/s1352325212000080 https://doi.org/10.1037/0033-295x.99.2.232 https://doi.org/10.1037//0003-066x.47.3.373 https://doi.org/10.1037//0021-843x.108.3.374 https://doi.org/10.1037//0021-843x.108.3.465 https://doi.org/10.1353/ppp.2000.0040 https://doi.org/10.1353/ppp.2000.0040 https://doi.org/10.2307/2183766 https://doi.org/10.33735/phimisci.2023.9630 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction biological dysfunction views problems of the hda problems of the bst the rehability view abilities reasons harm example: anxiety disorder why adopt the rehability view? the rehability view vs. biological dysfunction views further merits conclusion dissolving the self: active inference, psychedelics, and ego-dissolution dissolving the self active inference, psychedelics, and ego-dissolution george deanea (georgejwdeane@gmail.com) abstract psychedelic drugs such as psilocybin, lsd and dmt are known to induce powerful alterations in phenomenology. perhaps of most philosophical and scientific interest is their capacity to disrupt and even “dissolve” one of the most primary features of normal experience: that of being a self. such “peak” or “mystical” experiences are of increasing interest for their potentially transformative therapeutic value. while empirical research is underway, a theoretical conception of the mechanisms underpinning these experiences remains elusive. in the following paper, psychedelic-induced ego-dissolution is accounted for within an active inference framework, as a collapse in the “temporal thickness” of an agent’s deep temporal model, as a result of lowered precision on high-level priors. the argument here is composed of three moves: first, a view of the self-model is proposed as arising within a temporally deep generative model of an embodied organism navigating an affordance landscape in the service of allostasis. next, a view of the action of psychedelics as lowering the precision of high-level priors within the generative model is unpacked in terms of a high bayesian learning rate. finally, the relaxation of high-level priors is argued to cause a “collapse” in the temporal thickness of the generative model, resulting in a collapse in the self-model and a loss of the ordinary sense of being a self. this account has implications for our understanding of ordinary self-consciousness and disruptions in self-consciousness present in psychosis, autism, depression, and dissociative disorders. the philosophical, theoretical and therapeutic implications of this account are touched upon. keywords active inference ∙ consciousness ∙ ego-dissolution ∙ predictive processing ∙ psychedelics ∙ selfmodel this article is part of a special issue on “radical disruptions of self-consciousness”, edited by thomas metzinger and raphaël millière. 1 introduction psychedelic (“mind-manifesting”) drugs are known to occasion radically altered states of consciousness, including profound changes in sensory perception, emoauniversity of edinburgh deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-2418-8282 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 2 tion, cognition, time perception, and self-consciousness (preller & vollenweider, 2016). one of the most interesting of all of these effects is the experience of egodissolution. although the experience is notoriously difficult to articulate and even considered ineffable, psychedelic researcher stanislas grof, who considers egodissolution the “main objective” of psychedelic therapy, describes it as “an ecstatic state, characterized by the loss of boundaries between the subject and the objective world, with ensuing feelings of unity with other people, nature, the entire universe, and god” (grof, 1980, p. 79).1 ego-dissolution is of considerable philosophical and theoretical value for understanding selfhood and the nature of consciousness (letheby & gerrans, 2017; millière, 2017; nour & carhart-harris, 2017).2 it is also considered to be central to the therapeutic potential of psychedelics (carhartharris & goodwin, 2017; see also letheby, 2020, limanowski & friston, 2020, sebastián, 2020, all in this special issue). despite this, very little is known about the mechanisms underpinning psychedelic-induced ego-dissolution. “predictive processing” theories of brain function (clark, 2013; friston, 2010; wiese & metzinger, 2017) have recently taken precedence in cognitive science, affording a novel theoretical framework to approach cognitive phenomena. in this paper i propose a novel account of ego-dissolution within an active inference framework. to this end, i initially furnish an account of self-modelling within active inference, where pre-reflective self-consciousness emerges in organisms as a consequence of “temporal thickness”, the need to model the consequences of potential actions over time (friston, 2018). i then give an account of the action of psychedelics within a predictive processing framework, unpacking the view that psychedelics “relax” high-level priors (carhart-harris, 2019; carhart-harris & friston, 2019) in terms of a high bayesian learning rate (hohwy, 2017; mathys et al., 2014). finally, i argue that low precision at high-levels of the inferential hierarchy results in a collapse of the temporal thickness of the generative model and the corresponding self-model, leading to the phenomenon known as ego-dissolution (see also limanowski & friston, 2020, this issue). 2 the free energy principle the free energy principle (fep) has the most ambitious explanatory scope of all “predictive processing” style theoretical frameworks (friston, 2010). it combines, subsumes and links to several brain theories, including the bayesian brain hypothesis (knill & pouget, 2004), predictive coding (mumford, 1992; rao & ballard, 1999), efficient coding (barlow, 2001) and reinforcement learning (dayan & daw, 2008). the mathematics of the theory are complex and beyond the scope of this paper (for a review see buckley, kim, mcgregor, & seth, 2017). according to the fep, 1see https://www.erowid.org for over 20,000 “trip reports” of various psychoactive drugs, including many descriptions of ego-dissolution. 2no particular theory of consciousness is assumed in this paper, but the proposal is considered consistent with predictive processing approaches to consciousness (clark, 2019; hohwy, 2012). deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 3 simply in virtue of existing, all organisms tend to minimise the entropy or dispersion of their states. this much is intuitive: the conditions that are viable for an organism are fairly narrow – deviation from homeostatic bounds, such as having a body temperature of 50 degrees centigrade, is incompatible with continued existence. organisms that fail to stay within their “species-specific window of viability” (clark, 2013 p. 13) simply cease to exist. life, on this account, resists the tendency towards disorder imposed by the second law of thermodynamics, and this principle applies at all levels – “from their gross morphology to fine details of cortical microcircuitry as well as at timescales from the neuronal to the phylogenetic” (seth & tsakiris, 2018, p. 973). organisms, then, must resist entropy, the long-term average of (information-theoretic) surprise. because this quantity is beyond direct epistemic access to an organism, according to the fep, organisms minimise a proxy variable or upper bound – dubbed (variational) free energy. free energy (under some simplifying assumptions) is equivalent to precision-weighted prediction error in predictive processing. 2.1 predictive processing on the predictive processing view, the brain has stored prior beliefs (in the form of probability distributions) about the causes of sensory inputs in the world (clark, 2013; wiese & metzinger, 2017). prior beliefs are hierarchically organised, where higher-levels encode predictions about representations at lower levels. prediction errors, arising from the discrepancy between the low-level predictions and incoming sensory signals, are passed up the hierarchy, where higher-level predictions are updated to minimise further prediction errors. perception, then, both exteroceptive and interoceptive, is the product of (approximate) bayesian inference, whereby the influence of prior beliefs and sensory evidence are weighted according to “expected precision”, e.g. confidence in the given context, to generate a posterior. inference in these schemes is thought to occur across a hierarchy of inferred causes, where higher levels encode regularities that occur at larger spatial and temporal scales (kiebel, daunizeau, & friston, 2008). in perceptual inference, sensory prediction errors can be minimised by tweaking the parameters of the generative model – that is, generating predictions to quash the influx of prediction error. prediction error can also be minimised though action by changing the incoming sensory data to fit a prediction – for instance, i can move my eyes to bring my coffee cup into view, to fulfil the prediction of a coffee cup. actions can be thought of as the fulfilment of proprioceptive (or oculomotor) predictions – an intended movement occurs as a result of predicting the proprioceptive consequences (friston et al., 2010; shadmehr, smith, & krakauer, 2010). there are detailed accounts of the neural implementation of these schemes available (bastos et al., 2012; keller & mrsic-flogel, 2018; shipp, 2016). deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 4 2.2 precision-weighting a key feature of predictive processing schemes is the contextual flexibility afforded by precision-weighting (clark, 2013; feldman & friston, 2010). precision regulates the interaction between top-down and bottom-up signals, through the synaptic gain on neuronal populations signalling prediction error, in order to approximate optimal inference over time. precision can be thought of as tracking both the reliability and relevance of the incoming sensory information, where weighting by reliability is analogous to assigning greater weight to more reliable information when updating a belief. prediction error signals with high precision (inverse variance) have greater influence in updating the top-down predictions. precision itself has to be inferred, both by the empirical variance in the sensory data itself, and according to prior expectations about precision. the optimisation of precision weighting, through updating of the precision expectations (precision-related priors), is frequently equated to attention within predictive processing (clark, 2013; feldman & friston, 2010). importantly for the current treatment, precision is thought to mediate both sensory attenuation the top-down filtering out of afferent information, and affordances, where affordances refers to the latent possibilities for action given the capabilities of the agent (cisek, 2007; friston, shiner, et al., 2012). 2.3 control-oriented inference in mandating that existence necessitates maintaining oneself within a limited repertoire of states via control-oriented predictive regulation (instrumental active inference) (seth & tsakiris, 2018), the fep aligns itself with precursors of this view, cybernetic theories that build on control, feedback and predictive modelling (e.g., the “good regulator theorem”) (conant & ashby, 1970). note that while a purely helmholtzian view of the brain might cast it in terms of inferring hidden causes in the world, casting the predictive machinery in terms of being for ensuring continued existence means that the generative model is not constrained to veridicality. rather than faithfully reconstructing the world, perception is “ultimately geared towards driving actions that preserve [the] physiological integrity of the organism. in other words, we do not perceive the world (and self) as it is, but as it is useful to do so” (seth & tsakiris, 2018, p. 975). 2.4 homeostasis homeostasis refers to the tendency of living systems to keep an “internal balance” despite changes in the surrounding environment (cannon, 1929). this has long been described in terms of control theoretic and cybernetic mechanisms, and more recently this homeostatic control is thought to involve interoceptive signals that report current physiological states (e.g., heart rate, or blood-bound glucose levels) (craig, 2002; damasio & carvalho, 2013). one way to restore bodily conditions to favourable states is to engage autonomic reflexes – for example, a hyperthermic deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 5 animal can perspire to cool down. of course, autonomic regulation alone is not sufficient to ensure continued existence – to avoid hunger or thirst the animal must engage actions, such as seeking out food and water. collectively, these actions are termed allostasis, the process via which the brain regulates the needs of the body (corcoran & hohwy, 2018; corcoran, pezzulo, & hohwy, 2019; schulkin & sterling, 2019). crucially, to stay viable on longer timescales, this action must be anticipatory – avoiding dyshomeostatic conditions before they arise (pezzulo, rigoli, & friston, 2015; sterling, 2012). 2.5 active inference the fep regards homeostasis and allostasis as the central aspects of organic life, thus the autopoietic principles at the basis of the fep act as a kind of “first prior” (allen & tsakiris, 2018). in other words, “[t]he brain is in the game of predicting the world, but only as a means to the end of embodied self-preservation” (allen & friston, 2018, p. 12). in so doing, the free energy principle collapses expected utility (instrumental value) and information gain (epistemic value) under a single quantity. on this approach, action planning is itself a form of inference, where preferences and goals are framed in terms of prior beliefs, such that these priors are fulfilled by action (botvinick & toussaint, 2012). casting value and utility purely as inferential problems may at first appear unintuitive – if an agent finds itself in consistently adverse circumstances, then such adverse circumstances should, at first pass, seem to have high probability. however, “[t]he critical step in this logic is the assumption that evolution has equipped us with the belief that low utility states are low probability, due to the fact that if our ancestors spent a lot of time in those states they would be less likely to reproduce” (gershman, 2019, p. 7). the so-called “first prior”, that of maintaining existence via homeostatic and allostatic regulatory behaviour, ensures that organisms seek to actively maintain internal and external conditions conducive to their own persistence. active inference refers to the process by which agents actively sample states of the world so as to reduce uncertainty and realise prior preferences, rendering the action selection process itself an inference problem. this arbitration occurs according to priors pertaining to expected free energy over a given course of action, or policy (friston et al., 2018; pezzulo et al., 2015). expected free energy is the free energy an agent predicts itself to average in opting to pursue a particular course of action. intuitively, some courses of action are more likely than others to lead to “expected” or desirable outcomes. a policy that has lower expected free energy is going to have a higher prior probability than a policy with higher expected free energy, because agents equipped with prior beliefs about their continued existence will pursue policies that reduce expected free energy (friston et al., 2015, 2018; kaplan & friston, 2018). crucially, agents engaging in active inference do not merely restrict themselves to the states they expect; rather they anticipate in order to minimise uncertainty about potential future outcomes (friston et al., 2015, deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 6 2018; pezzulo et al., 2016). this prospective form of control relies on the contextualization provided by higher levels in the inferential hierarchy, which anticipate the downstream consequences of actions and select policies accordingly (friston, 2010; pezzulo et al., 2015). contextualisation here depends on the relative precision at various hierarchical levels, where “precision dynamics subsume the role of arbitration” (pezzulo et al., 2015, p. 27). this approach bears similarities to other control-theoretic approaches, such as the affordance competition hypothesis (cisek, 2007; pezzulo & cisek, 2016), where an affordance is a potential for action that avails itself to an organism in its action-oriented perception of environmental features.3 on this view, perceived affordances jostle for precedence and are arbitrated on the basis of the desirability of their predicted outcomes. 3 the self in active inference this section outlines an account of how pre-reflective self-consciousness – an implicit sense of being a subject present in all experience – is structured within an active inference framework. here, the self-model is underpinned by the same inferential bayesian schemes that are increasingly being used to describe perception and action. this predictive-modelling approach to selfhood has roots in thomas metzinger’s work on conscious and unconscious self-models, and the “self-model theory of subjectivity” (blanke & metzinger, 2009; metzinger, 2003, 2009) where “[a] self-model, an inner image of the organism as a whole [is] built into the worldmodel, and this is how the consciously experienced first-person perspective develop[s]” (metzinger, 2009, p. 64). the account presented here follows the increasing focus on the embodied nature of selfhood, where “being” or “having” a body is thought to be one of the most basic aspects of the experience of being a self (allen & friston, 2018; apps & tsakiris, 2014; blanke & metzinger, 2009; limanowski & blankenburg, 2013). a growing number of researchers seek to ground selfhood and emotion in interoceptive processes, particularly in their functional relation to allostatic regulation (barrett & simmons, 2015; seth, 2015; seth & friston, 2016). a key reason for this is that interoceptive inference is apt to put greater emphasis on control over discovery (seth & friston, 2016), due to “a priori hyper-precision of visceral channels”(allen & friston, 2018, p. 7), in which interoceptive signals are assigned very high precision in virtue of communicating information about key physiological variables (seth, 2015). grounding the self-model in control-oriented active inference (seth & tsakiris, 2018) inflects perception of the affordance landscape in terms of bodily states, an idea which is nicely expressed by montague and king-casas: 3it is worth noting here that the proposal put forward in this paper is considered to be compatible with both internalist and externalist conceptions of the mind. thanks to an anonymous reviewer for suggesting this clarification. deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 7 a sated and comfortable lioness looking at two antelopes sees two unthreatening creatures against the normal backdrop of the temperate savanna. the same lioness, when hungry, sees only one thing – the most immediate prey. in another circumstance, in which the lioness may be inordinately hot, the distant, shaded tree becomes the prominent visual object in the field of view. (montague & king-casas, 2007, p. 519) this forms the basis for the view that will be unpacked in more detail in what follows, that the self-model can be understood as an “allostatic control model”, arising from the system’s sense of control of the temporally deep consequences of actions for allostasis. on this view, pre-reflective self-consciousness is underpinned by the inference about endogenous control of the sensory consequences of actions within deep goal hierarchies, where goals and preferences are framed in terms of prior beliefs, such that goals are fulfilled by actions (botvinick & toussaint, 2012; pezzulo et al., 2015). recall, action allows an organism to change the sensory input in order to conform to its generative model, as opposed to perceptual inference that involves revising model parameters to conform to the sensory input. in order to act, then, the system implicitly infers its own ability to bring about the intended sensory consequences – it is in this sense that “implicit in a model of sampling is a representation or sense of agency” (friston, 2012a, p. 173), which is closely related to what has been called the “primacy of the ‘i can’ ” (bruineberg, 2017). crucially, organisms with deep temporal models have “temporal thickness” – expectations regarding the sensory consequences of actions on multiple interlocking timescales. the following sections will unpack this conception of the self-model in terms of hierarchically deep allostatic control, starting with the notion of temporal thickness, and then moving to how motivated control hierarchies “attune” organisms to action opportunities on multiple timescales, for both proximal goals, for instance, pain motivating an organism to act so as to fulfil a “healthy body condition” prior; and distal goals, for example emotions motivating a change of circumstances pertaining to longer timescales such as moving to a different city. the discussion will then move to how deep self-models allow organisms to arbitrate between different policies and trade off outcomes on different timescales. 3.1 temporal thickness to successfully navigate the world over longer timescales, and select policies that result in survival – and not dispersion or non-existence – organisms must possess models of the future; in other words, they require deep temporal models (friston et al., 2018). the generative models that endow organisms with the capability of inferring the consequences of future actions must have the property of temporal thickness, which allows the organism to anticipate the downstream consequences of potential actions, conferring the ability to select policies or action scripts that are favourable to the organism’s continued existence (friston, 2018). the minimideane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 8 sation of surprise through active inference on the fep involves acting so as to reduce uncertainty, and to do this the system must model itself across time and counterfactuals – that is, it must model what kind of agent it is at varying degrees of temporal depth. self-modelling, then, emerges as a natural consequence of prospective action selection (friston, 2018), where the principal function of a counterfactually rich self-model is to facilitate navigation of the affordance landscape and action selection across multiple interlocking timescales – for example expectations of what an agent can do on shorter timescales inform expectations of what the agent can do over longer timescales. the functional role of having a rich self-model, then, is that it enables the organism to predict outcomes across diverse policies, and endows the organism with “what if?” capabilities (friston, 2018), which puts this picture into contact with mental time travel and offline simulation (buckner & carroll, 2007; schacter, addis, & buckner, 2008). 3.2 attuning to the world conceiving of the self-model through an active inference framework, a hierarchically deep self-model guides policy selection over various timescales in service of minimising expected free energy. in what follows, pain perception, viewed as arising through the violation of the prior of “healthy body condition” (ongaro & kaptchuk, 2019), will be used to illustrate how inferences about the self “attune” an organism to adaptive action opportunities. one key advantage that the active inference account of self-modelling has over strictly bayesian approaches is that it is goal-directed (moutoussis et al., 2014). classical models of pain perception as the consequence of physiological dysfunction are challenged by the efficacy of placebo treatments in relieving pain (anchisi & zanon, 2015), and cases in which pain is experienced without physiological disruption, as is often the case in chronic pain. instead, there is evidence to suggest that affectively charged percepts, such as pain, are best understood as resting on the same inferential mechanisms as are assumed to underpin perception and action under a predictive processing framework (büchel, geuter, sprenger, & eippert, 2014). bayesian models of pain perception (morton, el-deredy, watson, & jones, 2010) indicate that prior beliefs about the generation of painful percepts are integrated with current sensory data to infer the posterior or hidden worldly cause (the painful percept). crucially, these pain percepts incorporate the “weight” or precision of past experiences when computing the current painful percept (morton et al., 2010). on their own, however, these models of pain perception are silent on the functional role of pain as a motivator to an embodied organism (moutoussis et al., 2014). optimal inference about pain to the allostatically concerned organism is heavily dependent on the context, as anyone who has felt the pain of an injury only after danger is averted can attest to. in this way, pain perception is allostatically “tuned”: “organisms can tune their own pain perception according to both their prior beliefs and the specific biological goals they believe are attainable in that context” (moutoussis et al., 2014, p. 70). deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 9 a bayesian framework of pain perception, therefore, needs to represent the agency and aims of the organism. this is precisely what is afforded by conceiving of the self-model within an active inference framework (friston, 2012b) – as this provides the necessary context to study the self-model, across multiple hierarchical levels. like physical pain, and sharing the neural underpinnings of physical pain (eisenberger, 2012), social pain is similarly understood in inferential terms, and does not scale with “damage” per se (for instance, social rejection), as evidenced by the wide range of sensitivity people have to the same physical manipulation (eisenberger, jarcho, lieberman, & naliboff, 2006). accordingly, there is evidence to suggest that appropriately “tuning” emotional responses in social contexts allows for agents to approximate bayesian inference in policy selection given bounded cognitive capacity and rationality. for example, on a “stag hunt” game,4 agents with “prosocial” preferences can outperform agents of similar cognitive sophistication that lack social biases (yoshida, dolan, & friston, 2008). 3.3 emotion conceptualising emotions in terms of a contextualisation of bodily states has historical roots dating back to the james-lange theory of emotion (cannon, 1927) and two-factor theory of emotion (schachter & singer, 1963). lisa feldman-barrett has developed this approach specifically within the active inference framework as the “theory of constructed emotion” (barrett, 2017). according to the theory of constructed emotion, emotions are constructed in the same manner as percepts, where priors are recruited according to context to make a “best guess” at the hidden causes of (interoceptive) sensory signals. on barrett’s view, emotions arise through a context-sensitive inferential categorisation of interoceptive states. for this reason, emotions on this view are “constructions” – there are no neural or physiological signatures that reliably discriminate any emotional state (siegel et al., 2018; wager et al., 2015). rather, physiological reactions in the body occur in order to prepare it for action, and these are categorised as emotions only contextually through the predictive models recruited to explain away the incoming afferent interoceptive signals. for example, heart rate increases or decreases depending only on an anticipated action – e.g., fight or flight – and given an emotional ascription only contextually – e.g., the same bodily state could be categorised as fear in one context and anger in another. interoceptive inference is experienced as emotion in service of producing allostatic action (barrett, quigley, & hamilton, 2016; barrett & simmons, 2015; wilkinson, deane, nave, & clark, 2019). in viewing the self-model in terms of hierarchical allostatic control, interoceptive inference on the hidden causes of bodily states pertaining to longer timescales tunes perception to the world and affordances differently, such that more abstract emo4a stag hunt game is a cooperative hunting game, where in order to hunt a stag (higher-payoff), an agent must have the cooperation of a partner. an agent can catch a rabbit (lower-payoff) without cooperating. deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 10 tions might track regularities over longer time scales, informing policy selection thereon (pezzulo, 2014), and allowing for more abstract and distal outcomes to be motivationally salient. 3.4 hierarchically deep self-models viewing the self-model in terms of allostatic control renders selfhood fundamentally affective and action-oriented, such that different aspects of the self in a given context – precision on goals and preferences at different levels of the hierarchy – motivate behaviour and arbitrate between policies. on this view, the self-model inflects perception of possible actions in the world and mediates salience to facilitate the selection of policies with minimal expected free energy. critical to this picture is the notion that these various models are associated with varying degrees of temporal depth (pezzulo, rigoli, & friston, 2018). deep generative models capture increasingly distal relations between actions and outcomes within hierarchical active inference, allowing for the coordination of behaviour across different hierarchical levels, enabling goals to become prioritised relative to current context (pezzulo et al., 2018). the result is an inferential framework of hierarchically nested contextual complexity, in which lower levels track basic (and sometimes evolutionarily hard-wired) motivations or affordances, while higher levels track motivations and plans over deeper timescales. in this way, higher-level contextualization of lower sensorimotor functions optimises expected actions in terms of both long-range consequences of actions and anticipated future affordances. goals at different levels of abstraction may, of course, conflict – for instance, resolving proximal interoceptive prediction error by eating chocolate cake might conflict with the longer-term goal of sticking to a diet (pezzulo et al., 2018). alternatively, temporary deviation from homeostatic set points at lower levels may be elicited to maintain higher level set points – such as a temporary change in blood pressure and adrenaline levels to engage fight-flight behaviour, with the goal of reaching safety and maintaining physiological integrity on a longer timescale. on the view of self-modelling in terms of allostatic control described, dimensions of the self at higher-levels constrain the self at lower-levels in that the self-model “actively shapes itself over time to align with those higher level regularities” (hohwy & michael, 2017, p. 370), for example long-term goals can be decomposed into intermediate short term-goals. this section has explored how the self-model arises as a consequence of a system engaged in temporally deep active inference, as prior probabilities over particular policies depend on knowledge about what and where the system finds itself, and what actions are available to it (friston et al., 2013; moutoussis et al., 2014). through active inference, agents can use their self-model to inform their goal and policy selection in order to arrive at high probability outcomes. this could entail assigning low probability to the self occupying states that are aversive, either physically or socially – with different hierarchical levels of the self-model contributing deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 11 to different goal states. in this way, the hierarchical self-model determines salience – where “salience is literally defined by whatever has the most (or least) impact on visceral and autonomic homeostasis” (allen & tsakiris, 2018, p. 7), at increasingly deep spatiotemporal scales and levels of abstraction. 4 psychedelics one of the most striking and philosophically interesting effects of psychedelics is the radical disruptions of self-consciousness they can occasion (huxley, 2010; leary, metzner, & alpert, 1964), including apparently “selfless states” (lebedev et al., 2015; nour, evans, nutt, & carhart-harris, 2016). these states, instances of “drug-induced ego-dissolution” (died) are characterised by an experienced loss of self and/or loss of self/world boundary (millière, 2017; millière, carhart-harris, roseman, trautwein, & berkovich-ohana, 2018). died occurs most reliably under high doses of “classical” psychedelic drugs (5-ht2a receptor agonists), such as dimethyltryptamine (dmt), lysergic acid diethylamide (lsd), and psilocybin. ego-dissolution appears to be induced more reliably under psychedelics than meditation, in a dose-dependent manner, and prompted most reliably by high-doses (nour et al., 2016). recent theoretical work has explored the phenomenological and neurophysiological similarities and differences of ego-dissolution induced by drugs and meditation (millière et al., 2018; see also limanowski & friston, 2020, millière, 2020, sebastián, 2020, all in this special issue). 4.1 psychedelic therapy recent years have seen a resurgence of interest in the therapeutic potential of psychedelics. several studies have found preliminary evidence that with administration in controlled circumstances psychedelics can be both safe and therapeutic, with an emphasis on the importance of context in achieving therapeutic outcomes (carhart-harris et al., 2018). interestingly, the positive therapeutic effects seem to scale with “peak” or mystical experience in the psychedelic state (roseman, nutt, & carhart-harris, 2018). psychedelics have been shown to be effective in treating depression (lyons & carhart-harris, 2018; palhano-fontes et al., 2019), obsessivecompulsive disorder (moreno, wiegand, taitano, & delgado, 2006), end of life existential distress (griffiths et al., 2016), and have even been proposed as a potential treatment for disorders of consciousness such as the vegetative state and the minimally conscious state (scott & carhart-harris, 2019). carhart-harris interprets the therapeutic effects as a result of “relaxing” high-level beliefs, allowing for a revision of pathological beliefs that have become overly dominant and resistant to revision, coined the “tightened beliefs in response to uncertainty” (tiber) model (carhart-harris, 2019). the basic tenet is that under conditions of uncertainty the model falls back on “tightened” belief structures as a defence mechanism against intolerable stress and uncertainty. this fits with what we might expect under the fep, deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 12 as adopting shallow policies (such as addictive behaviours) may appear adaptive in the short term, rather than risking policies with greater expected free energy due to low precision or uncertainty. it has recently been proposed that psychedelics “relax” high-level priors in the generative model, allowing for the (context-dependent) revision of pathological high-level beliefs (carhart-harris & friston, 2019). both psychological insight and peak-experience in the psychedelic state appear to be predictors of long-term positive prognoses (roseman et al., 2018). 4.2 psychedelics in the predictive brain the rebus – “relaxed beliefs under psychedelics” – model of psychedelic function, offers a preliminary but promising model of psychedelic action where psychedelics, through 5-ht2a agonism, “relax” high-level priors or beliefs (carhart-harris & friston, 2019). here, the focus will be on how this mechanism may be cast under the hierarchical predictive processing framework as modulating precision-weighting. to bring this into focus, this section will review how precision-weighting sets a variable bayesian learning rate in order to highlight certain features relevant to understanding the effects of psychedelics within this framework. christoph mathys and colleagues have recently developed a mathematical tool for modelling bayesian inference modulated by expectations of volatility known as the hierarchical gaussian filter (mathys, daunizeau, friston, & stephan, 2011; mathys et al., 2014). the hierarchical gaussian filter mathys posits allows a system to optimally balance the influence of prediction errors in changing environments – in other words, to adjust its learning rate. 4.3 bayesian learning rate to recap, the predictive processing framework asserts that the brain instantiates “generative models” of the causes of incoming sensory data, iteratively updating these predictive models in light of incoming “prediction error” (clark, 2013). this predictive inference is thought to occur across a hierarchy of inferred causes, where high levels track causes and regularities operating over deeper spatial and temporal scales, and lower levels track regularities over shallower spatial and temporal scales (kiebel et al., 2008). the picture of the living or cognitive system as one which needs to optimise its own learning rate emerges out of the operationalization of bayesian inference in predictive processing, namely in terms of predictions and precision-weighted prediction errors (mathys et al., 2014). according to predictive processing the prediction is given by the prior probability (which itself comes from the previous posterior) and the prediction error is given by the difference between the prediction and the incoming sensory evidence. prediction error is weighted according to the relative precisions of the prior and the prediction error (where precision is equivalent to the inverse variance of each probability distribution). intuitively, highly precise prediction error will drag deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 13 the posterior closer to the distribution of the sensory evidence and further from the prior, and in cases of low precision weighting of the prediction error, the inference relies more on the prior. this determines the learning rate: the more certain we are that the prior hypothesis is correct, the less we should be influenced by the prediction error (the evidence), which means that the learning rate is low. conversely, the better the precision on the prediction error, the higher the learning rate; that is, the more we trust the quality of the evidence the more we should learn from it (hohwy, 2017, p. 76) in other words, the lower the learning rate, the greater the influence of top-down modulation from priors; the higher the learning rate, the greater the influence of the sensory evidence on the resulting posterior. here, precision-weighting is the key mechanism – heavily weighted prediction errors drive a higher learning rate. in order to approximate bayesian inference over time, it is essential for sensory systems to balance the learning rate appropriately. over-reliance on priors means the system will fail to learn from experience, whereas over-reliance on sensory evidence (which may be noisy) will lead the system to “overfit”. on this picture, bayesian perceptual inference that minimises prediction error on the appropriate timescale – that is, not overfitting or underfitting the model – needs to have a means of regulating the learning rate (mathys et al., 2014). this is implemented by building models of precision, or expected uncertainty, where higher-level priors track longer-term regularities that inform the relative precisions of more basic priors (hohwy, 2017). optimising the learning rate, and in-so-doing minimising prediction error over time, is a critical challenge the brain faces. this is equivalent to selecting a time frame over which to minimise prediction error. minimising prediction error over too short a timescale – overfitting – runs the risk of increasing prediction error in the long run. conversely, failing to accommodate new evidence will lead to underfitting, a failure to update predictions in light of new sensory evidence. 4.4 psychedelic action as high bayesian learning rate in line with the rebus model (carhart-harris & friston, 2019), the relaxing of high-level priors under classical (serotonergic) psychedelics5 means the system adopts a very high bayesian learning rate – that is, it is in a highly plastic state, in accordance with research showing an increase in plasticity under psychedelics (ly et al., 2018). this picture casts the perceptual effects of psychedelics – “tripping” – as rampant overfitting of the sensory data, resulting from a loss of the usual con5this account is not thought to generalise to dissociative anaesthetics, like ketamine or phencyclidine (pcp), which are thought to have quite different mechanisms of action, acting primarily at nmda glutamate receptors. comparative approaches of both the mechanisms of action of other drugs as compared to psychedelics, and the mechanisms underpinning different experiences of ego-dissolution, would be a fruitful avenue for future research. deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 14 straint exerted by higher-levels on lower-levels of the inferential hierarchy. this “rampant overfitting”, resulting from diminished influence from contextualising high-level priors tracking regularities on longer timescales, means the model fits a very short temporal scale, rapidly cycling through candidate models to account for the incoming sensory signal. it is worth highlighting a compatibility of the high bayesian learning rate approach with other accounts of the mechanism of action of psychedelics in the predictive brain. the rebus model posits the mechanism of action of psychedelics as reduced precision at high levels rather than increased precision at the sensory peripheries, as psychedelics appear to disrupt functioning via stimulation of 5-ht2a receptors on deep pyramidal neurons, thought to encode high level priors or beliefs (beliveau et al., 2017; carhart-harris & friston, 2019; jakab & goldman-rakic, 1998). in contrast, philip corlett and colleagues have suggested that psychedelics preserve normal priors and act by increasing sensory noise through enhanced ampa signalling (corlett, frith, & fletcher, 2009; corlett, honey, & fletcher, 2016). on this approach, if the relaxation of high-level priors is indeed an effect of psychedelics, it could be understood to be the result of the fact that “the persistence and strength of the sensory signal suggest that there is something to be explained” (corlett et al., 2009, p. 521). arbitrating between these two mechanistic accounts and disentangling causation – whether the relaxation of high-level priors causes the reduction in sensory gating, or reduction in sensory gating eventually lowers precision at high levels – becomes very difficult here, and it is not clear a simplistic causal account is the right approach. while identifying the mechanisms of action is a key empirical and theoretical project, one potential advantage of the high bayesian learning rate hypothesis is that it doesn’t distinguish between high precision at low levels and low precision at high levels, and as such remains agnostic over the mechanism of action while preserving the useful theoretical features of both accounts that will inform the theoretical account of ego-dissolution that follows. 4.5 evidence for the high bayesian learning rate hypothesis a high bayesian learning rate is concordant with the enhanced neural plasticity observed in individuals in a psychedelic state (barre et al., 2016; berthoux, barre, bockaert, marin, & bécamel, 2018; ly et al., 2018). while an impairment to highlevel cognition is found under psychedelics (bayne & carter, 2018), in line with the high bayesian learning rate hypothesis, low-level learning (including extinction learning) and processing appears to be unaffected or enhanced in the psychedelic state (carhart-harris & nutt, 2017; king, martin, & seymour, 1972; romano et al., 2010). further evidence for a high bayesian learning rate under psychedelics is provided by a study looking at the effect of psilocybin on kanisza triangles – perceptual objects where the brain “fills in” illusory contours using prior expectations – which found reduced filling in and a reduction in the related evoked potentials (kometer, cahn, andel, carter, & vollenweider, 2011), concordant with deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 15 the fact that a high bayesian learning rate will reduce the effect of sensory history on current perception. in binocular rivalry studies – where different images are presented to each eye simultaneously, and are typically experienced as switching from one percept to the other – reduced switch rates and increased likelihood of the percept being a fusion of the two images has been observed under psilocybin (carter et al., 2007, 2005), suggestive of less influence of priors on constraining current perception. oddball paradigms are also suggestive of a weakened influence of priors on perception under psychedelics. in a sequence of tones, an “oddball” tone (unexpected given prior experience and context) generates a “mismatch negativity”, an evoked brain response which has been interpreted in predictive coding terms as prediction error violating the expectations of the sequence (garrido, kilner, stephan, & friston, 2009). under lsd, the surprise response to oddball stimuli is blunted, suggestive of a weakened influence of prior expectations (timmermann et al., 2018). arguably, there is also phenomenological evidence for the high bayesian learning rate hypothesis. perhaps most eloquently articulated by aldous huxley: “visual impressions are greatly intensified and the eye recovers some of the perceptual innocence of childhood, when the sensum was not immediately and automatically subordinated to the concept” (huxley, 2010, p. 12). this observation lends itself to a straightforward translation into the terms of predictive processing, where “subordinated to the concept” can be understood as “constrained by higher-level priors”. more generally, psychedelic phenomenology such as dynamic distortions of spatial dimensions, where things change dramatically in size and shape can be understood as a failure of high-level priors to canalise and constrain lower level predictions. 5 psychedelic-induced ego-dissolution in active inference given this picture of the action of psychedelics within a predictive processing framework, and the characterisation of self-models in terms of allostatic control, how should states of psychedelic-induced ego dissolution be conceptualised? the proposal here is that a loss of precision on high-level priors results in a flattening of temporal depth of the affordance landscape for the organism – precisely because it is high-level priors tracking longer timescales that structure temporally deep generative models. recall, under active inference, lower and higher hierarchical levels encode regularities that unfold at faster and slower timescales respectively (kiebel et al., 2008), such as the expected consequences of action both for proximal and distal goals (pezzulo et al., 2015, 2018). adopting a high bayesian learning rate is equivalent to changing the time frame over which prediction error is minimised to fit very short timescales. as a result, the deep temporal models that typically guide action and policy selection collapse, and the faster timescales corresponddeane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 16 ing to lower levels are modelled in a much finer degree of detail (pink-hashkes, rooij, & kwisthout, 2017). on the account presented in this paper, the self-model is constructed and bolstered in relation to affordances in the environment on several interlocked timescales, where high-levels contextualise and canalise the levels below and allow for motivational orientation to action opportunities pertaining to distal outcomes. under psychedelics, the relaxation of high-level priors and the corresponding high bayesian learning rate results in a collapse in the temporal thickness of deep generative models, and a collapse in the temporal depth of the corresponding self-model, which is understood as being is bolstered according to counterfactually rich expectations of the consequences of action on multiple timescales. the collapse in temporal thickness can be understood as occurring due to a failure of sensory attenuation, occurring due to low precision at high-levels and a correspondingly high bayesian learning rate. similar stories about aberrant precision at high-levels of the hierarchy corresponding to inferences about affordances and agency have been proposed to underpin hallucinations and delusions in psychosis (adams, stephan, brown, frith, & friston, 2013; fletcher & frith, 2009; sterzer et al., 2018). distinguishing between endogenous and exogenous causes – that is, distinguishing between perceptual inputs caused by oneself and those caused by the world – is vital for an agent to be able to effectively move through action space. corollary discharges – predictions about the sensory consequences of actions – allow the system to do this by withdrawing precision from self-generated movements, and are thought to underpin experienced agency of actions (crapse & sommer, 2008; friston, 2012b). the failure to predict the consequences of movement due to a failure of sensory attenuation is thought to result in an inability to attribute agency (adams et al., 2013; brown, adams, parees, edwards, & friston, 2013); for instance, a failure of corollary discharge has been thought to cause the attribution of inner speech to an external source in voice-hearing (ford, gray, faustman, roach, & mathalon, 2007; ford & mathalon, 2005; heinks-maldonado et al., 2007). importantly, for present purposes, corollary discharge can be understood as a kind of prior (friston, 2010), and low-precision priors have been associated with the severity in psychotic symptoms and disturbances of agency in people with schizophrenia (rösler et al., 2015). a reduction of precision on highlevel priors in the psychedelic state means that the corollary discharges that would usually cancel out the expected consequences of actions fail to do so, generating an increase in prediction error at lower levels. these unexpected consequences are then attributed to external rather than internal causes, as the more prediction error is generated, the more likely an action (or thought) has exogenous rather than endogenous causes (frith, 2003). this echoes similar themes in the autism literature. in autism, the failure of sensory attenuation “leads to the hypervigilant attention to sensory detail at the expense of a hierarchically deep explanation for sensations” (picard & friston, 2014, p. 1116) leading to what has been termed a “loss of central coherence” (frith, 2003). attribution to exogenous rather than endeane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 17 dogenous causes could result in a loss of “perceptual mineness” – the background feeling that my experiences are “mine” – if, as has been argued, perceptual mineness is underpinned by anticipation of changes in perceptual inputs in relation to movements (hohwy, 2007).6 ego-dissolution is not, however, confined to a loss of agentive control over immediate action outcomes, but may be characterised by a more profound dissolution of the sense of being a self or “i” distinct from the outside world. on the view presented in this paper, pre-reflective self-consciousness arises not just through modelling control over the most immediate sensory consequences of actions, but is bolstered by inferences about endogenous control over the distal sensory consequences of allostatic action and action policies. under a high dose of a psychedelic, the temporary suspension on the gating mechanism on incoming sensory data, described in this paper in terms of a high bayesian learning rate, render both the proximal and distal sensory consequences of actions highly unpredictable, and the system ceases to have the sense of their being an agent which can (and should) be controlling sensory outcomes. several authors have emphasised the psychedelic experience is a dynamic process as opposed to a firmly designated state (masters & houston, 1966; preller & vollenweider, 2016), and different types of ego-dissolution might occur both over the course of the experience and at different dosages. for example, inferences on the boundaries of the body (blanke & metzinger, 2009) might be increasingly blurred due to a failure to attenuate the flurry of low-level prediction error. aspects of the self-model corresponding to longer timescales may break down due to a sustained failure of high levels to attenuate prediction error from low levels due to highly volatile prediction errors, consistent with the fact that bodily ego-dissolution tends to precede dissolution of narrative self (savage, 1955). this fact is also perhaps suggestive, in opposition to the high-levels posited by the rebus model, that ego-dissolution could be seen as the result of the high-levels failing to contextualise the upsurge of prediction error from across the cortex. the fact that the highest level of the self-model are “increasingly abstract, complex and invariant” (limanowski & friston, 2018, p. 5), may explain why higher levels of the self-model are going to be less perturbed by prediction error and perhaps only reliably altered at high dosages. empirical exploration of these possibilities might be a fruitful avenue for future work, in particular through bridging the neurocomputational mechanisms posited here to both the dynamics of the experience as uncovered through “microphenomenological” interviews (millière, 2017; petitmengin, 2006), and to the underlying neural correlates of the experience (timmermann et al., 2019). the account of ego-dissolution in terms of a collapse in the temporal thickness 6it is worth noting that while some comparisons can be drawn between the psychedelic state and predictive processing accounts of psychosis (sterzer et al., 2018) and autism (palmer, lawson, & hohwy, 2017), a key difference here is that in the psychedelic state the system typically begins “from a baseline of stable high-level priors (e.g., a stable ego) to which it returns as drug effects subside” (carhart-harris & friston, 2019, p. 328). deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 18 of the affordance landscape presented here should also apply to the concept of a “cognitive affordance” landscape, where the “central function of autonomous activity in the mind wandering network is to create a constant stream of affordances for cognitive agency, a continuing internal competition among possible cognitive actions” (metzinger, 2017, p. 2). metzinger argues that mental actions – such as the volitional control of endogenous attention, or retrieval of an episodic memory – have epistemic rather that pragmatic goal states. on the allostatic control model of selfhood, the self-model would be constructed and bolstered relative not only to the expectations of the control of the sensory consequences of actions, but also the consequences of mental actions, where the consequences of a mental action might be epistemic and also interoceptive (consider a case where a memory triggers an autonomic response which subsequently acts as the afferent input to an interoceptive inference underpinning a felt emotion). under psychedelics, loss of control of the expected outcomes of mental actions (as well as a loss of the pragmatic concerns usually driving which epistemic actions to take) might then also be fundamental to the experience of ego-dissolution. this idea is consistent with the fact that under psychedelics mental phenomena “take on the character of objective reality” (savage, 1955, p. 12), where the ownership of mental phenomena seems to subside and “the individual may feel like a bystander watching the mental activity of another person” (girn & christoff, 2018, p. 145). it is worth mentioning a potential implication of this view for consciousness science more broadly. the psychedelic experience and ego-dissolution are often described as an “expansion” of consciousness. friston (2018) argues that not only self-consciousness, but consciousness itself, is underpinned by temporal thickness: “consciousness is nothing more than inference about my future; namely, the selfevidencing consequences of what i could do” (friston, 2018, p. 1). states of egodissolution, understood as collapse in the temporal thickness of the generative model, suggest that while temporal thickness very much structures our normal waking experience, it is not clear that temporal thickness ought to be equated with consciousness per se (see also metzinger, 2020, this issue; sebastián, 2020, this issue). 5.1 ecstatic ego-dissolution and challenging experiences the question remains as to why the hypothesised collapse in the temporal thickness of the self-model under psychedelics can be both ecstatic and of enduring therapeutic value. to bring this into focus, it’s worth recapitulating core features of the self-model provided earlier. recall, interoceptive inference on states of the embodied self “attunes” organisms to their affordance landscape, where inferences about the state of the embodied self (e.g. hunger) prescribe certain prediction error minimising policies (e.g. finding food). inferences pertaining to allostatic consequences on longer timescales may mean higher-level imperatives trump lowerlevel drives, such as choosing to abstain from chocolate cake to stay healthy (pezdeane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 19 zulo et al., 2018). in the case of basic bodily needs, as described, these variables are controlled (seth, 2015) through action – active inference is deployed to bring the world into line with predictions, rather than adjusting predictions (via perceptual inference) to conform to the world – for instance eating when hungry (pezzulo et al., 2015). in just the same way that a hungry organism can act so as to harvest confirmatory evidence for the hypothesis “i am sated”, hypotheses relating to higher-levels of the self-model geared towards control of outcomes on longer timescales act to constrain action in the present to bring downstream outcomes closer in line with the prior expectation. overly precise priors driving action on a long timescale which are failing to be fulfilled, on this view, would be a persistent cause of suffering, due to the system consistently failing to meet (or align actions towards) the goal state (hesp, smith, allen, friston, & ramstead, 2019). under the model of psychedelic-induced ego-dissolution proposed, the high-precision highlevel priors geared towards control on multiple timescales cease to exert influence on the system due to the proposed lowering of precision of high-level priors under psychedelics. if action ordinarily arises from a process of minimising deviations between the organism’s actual (inferred) and desired trajectory (friston et al., 2010), the loss of precision on high-level priors means that, instead of driving action policies, they lose influence on the rest of the system and cease to structure pre-reflective self-consciousness to orient to action opportunities favouring their fulfilment. as these prior beliefs are relaxed, they instead become amenable to perceptual revision from the influx of (highly precise) interoceptive and exteroceptive information. the collapse in temporal depth in the psychedelic state is therefore not experienced as a loss of allostatic control, precisely for the reason that the priors pertaining to longer timescales are no longer asserting an influence on the system and constraining action (and perception) in their usual manner. this picture seems to align well with phenomenological reports of ego-dissolution: “it felt as if ‘i’ did no longer exist. there was purely my sensory perception of my environment, but sensory input was not translated into needs, feelings, or acting by ‘me’ ” (unpublished online survey data quoted in millière et al., 2018, p. 7). peak experiences under psychedelics, then, could be understood as absence of prediction errors relating to allostasis due to a flattening of the temporal depth of the affordance landscape, resulting in the feeling of “oceanic boundlessness” – a sense of immense well-being and peace. here, the “itinerant strategies” to stay within our “species-specific window of viability” (clark, 2013, p. 13), are no longer necessary as the “first prior” – the expectation or imperative for existence – is being met without conditions. following the tiber model, many psychopathologies may be due to high precision on high-level priors (carhart-harris, 2019; clark, watson, & friston, 2018). peak psychedelic experience may act as a “reset” allowing for revision of entrenched high-level beliefs that structure pre-reflective self-consciousness (and, accordingly, the affordance landscape) – opening up new domains of salience and possibility for meaningful engagement with the world, through revised and deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 20 retuned self-models. increased bottom-up information flow (particularly from the limbic system), through a high bayesian learning rate, may make entrenched high-level priors amenable to revision via perceptual inference rather than driving control via active inference. this lays the theoretical groundwork for why psychedelics may effectively treat depression: if depression is underpinned by a high precision prior of low allostatic self-efficacy (stephan et al., 2016), it follows that relaxation and revision of this prior should alleviate depressive symptoms. finally, (and speculatively), if the account of “retuning” of self-models under psychedelics presented here generalises to the bodily self (which the experiential changes in bodily selfhood would suggest) this account is suggestive of a potential role for psychedelics in the treatment of chronic pain, and for phantom limb pain – for which there has already been promising results (fanciullacci, bene, franchi, & sicuteri, 1977; ramachandran, chunharas, marcus, furnish, & lin, 2018). the primary focus so far has been on “peak” experiences, due to the growing number of papers indicating they are central to positive long-term therapeutic outcomes (roseman et al., 2018). however, while generally psychedelics are thought to be very low risk (nutt, king, & phillips, 2010), and there is evidence to suggest they are protective against mental health problems (hendricks, thorne, clark, coombs, & johnson, 2015), acute and occasionally persistent adverse psychological reactions do sometimes occur (strassman, 1984). while “complete” ego-dissolution is described as a “state of complete surrender, associated bliss, and union with all things” (carhart-harris & friston, 2019, p. 321), “incomplete” ego-dissolution – due to psychological resistance or an insufficient dose – can be characterised by intense fear, anxiety, or distress. on the account presented in this paper, this can be understood as resulting from psychological resistance, where psychological resistance here may be conceptualised as a high-precision prior on being able to control the experience, that is maintained though fear-driven endogenous attention. failure to control the experience, in violating the highly precise prior for the goal state of control, is then experienced as a loss of allostatic control, bringing with it feelings of intense fear or distress. in therapeutic contexts, encouraging users to “let go” and “surrender” to the experience (richards, 2015), could be understood in these terms, as discouraging the user from putting high (endogenous) precision on a prior for control that could result in adverse experiences when unfulfilled. these considerations highlight the essential importance of context in achieving therapeutic outcomes (carhart-harris et al., 2018). 6 conclusion psychedelics are known for their ability to profoundly alter consciousness and occasion so-called “mystical” experiences (huxley, 2010). the renaissance in psychedelic research in the past decade is beginning to shed light on the mechanisms underpinning the extraordinary states of consciousness induced by psychedelics (carhart-harris & goodwin, 2017). within psychedelic phedeane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 21 nomenology, experiences of ego-dissolution are of particular phenomenological, philosophical and therapeutic interest (letheby & gerrans, 2017; millière, 2017; nour & carhart-harris, 2017). this paper has given a preliminary account of how ego-dissolution under psychedelics can be understood in terms of predictive processing and active inference. the hypothesis here is that the action of psychedelics within the predictive processing framework is best understood as a “relaxation of high-level beliefs” (carhart-harris, 2019), and this can be unpacked in terms of a high bayesian learning rate (hohwy, 2017; mathys et al., 2014). psychedelic-induced ego-dissolution, then, results in a collapse in temporal thickness (friston, 2018) of the self-model as conceived within an active inference framework. the therapeutic effects of ego-dissolution, then, can be understood in terms of the relaxing and retuning of entrenched self-models, or a “resetting” or “opening” of the affordance landscape, allowing for the possibility of new modes of engagement with the world, oneself, and other people. acknowledgments thanks very much to mel andrews, andrew corcoran, and andy clark for invaluable critical feedback and discussion in the production of this paper. work on this article was supported by the european research council (erc) advanced grant xspect dlv-692739. references adams, r. a., stephan, k. e., brown, h. r., frith, c. d., & friston, k. (2013). the computational anatomy of psychosis. frontiers in psychiatry, 4, 47. https://doi.org/10.3389/fpsyt.2013.00047 allen, m., & friston, k. (2018). from cognitivism to autopoiesis: towards a computational framework for the embodied mind. synthese, 195(6), 2459–2482. https://doi.org/10.1007/s11229-016-1288-5 allen, m., & tsakiris, m. (2018). the body as first prior: interoceptive predictive processing and the primacy. in the interoceptive mind: from homeostasis to awareness (pp. 27–45). https://doi.org/10.1093/oso/9780198811930.003.0002 anchisi, d., & zanon, m. (2015). a bayesian perspective on sensory and cognitive integration in pain perception and placebo analgesia. plos one, 10(2), e0117270. https://doi.org/10.1371/journal.pone.0117270 apps, m. a., & tsakiris, m. (2014). the free-energy self: a predictive coding account of self-recognition. neuroscience & biobehavioral reviews, 41, 85–97. https://doi.org/10.1016/j.neubiorev.2013.01.029 barlow, h. (2001). redundancy reduction revisited. network: computation in neural systems, 12(3), 241–253. https://doi.org/10.1088/0954-898x/12/3/301 barre, a., berthoux, c., de bundel, d., valjent, e., bockaert, j., marin, p., & bécamel, c. (2016). presynaptic serotonin 2a receptors modulate thalamocortical plasticity and associative learning. proceedings of the national academy of sciences, 113(10), e1382–e1391. https://doi.org/10.1073/pnas.1525586113 barrett, l. f. (2017). the theory of constructed emotion: an active inference account of interoception and categorization. social cognitive and affective neuroscience, 12(1), 1–23. https://doi.org/10.1093/scan/nsw154 barrett, l. f., quigley, k. s., & hamilton, p. (2016). an active inference theory of allostasis and interoception in depression. philosophical transactions of the royal society b: biological sciences, 371(1708), 20160011. https://doi.org/10.1098/rstb.2016.0011 barrett, l. f., & simmons, w. k. (2015). interoceptive predictions in the brain. nature reviews neuroscience, 16(7), 419–429. https://doi.org/10.1038/nrn3950 bastos, a. m., usrey, w. m., adams, r. a., mangun, g. r., fries, p., & friston, k. (2012). canonical microcircuits for predictive coding. neuron, 76(4), 695–711. https://doi.org/10.1016/j.neuron.2012.10.038 bayne, t., & carter, o. (2018). dimensions of consciousness and the psychedelic state. neuroscience of consciousness, 2018(1), niy008. https://doi.org/10.1093/nc/niy008 deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.3389/fpsyt.2013.00047 https://doi.org/10.1007/s11229-016-1288-5 https://doi.org/10.1093/oso/9780198811930.003.0002 https://doi.org/10.1371/journal.pone.0117270 https://doi.org/10.1016/j.neubiorev.2013.01.029 https://doi.org/10.1088/0954-898x/12/3/301 https://doi.org/10.1073/pnas.1525586113 https://doi.org/10.1093/scan/nsw154 https://doi.org/10.1098/rstb.2016.0011 https://doi.org/10.1038/nrn3950 https://doi.org/10.1016/j.neuron.2012.10.038 https://doi.org/10.1093/nc/niy008 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 22 beliveau, v., ganz, m., feng, l., ozenne, b., højgaard, l., fisher, p. m., et al. (2017). a high-resolution in vivo atlas of the human brain’s serotonin system. journal of neuroscience, 37(1), 120–128. https://doi.org/10.1038/sdata.2018.63 berthoux, c., barre, a., bockaert, j., marin, p., & bécamel, c. (2018). sustained activation of postsynaptic 5-ht2a receptors gates plasticity at prefrontal cortex synapses. cerebral cortex, 29(4), 1659–1669. https://doi.org/10.1093/cercor/bhy064 blanke, o., & metzinger, t. (2009). full-body illusions and minimal phenomenal selfhood. trends in cognitive sciences, 13(1), 7–13. https://doi.org/10.1016/j.tics.2008.10.003 botvinick, m., & toussaint, m. (2012). planning as inference. trends in cognitive sciences, 16(10), 485–488. https://doi.org/10.1016/j.tics.2012.08.006 brown, h., adams, r. a., parees, i., edwards, m., & friston, k. (2013). active inference, sensory attenuation and illusions. cognitive processing, 14(4), 411–427. https://doi.org/10.1007/s10339-013-0571-3 bruineberg, j. (2017). active inference and the primacy of the ’i can’. in t. metzinger & w. wiese (eds.), philosophy and predictive processing. frankfurt am main: mind group. buckley, c. l., kim, c. s., mcgregor, s., & seth, a. k. (2017). the free energy principle for action and perception: a mathematical review. journal of mathematical psychology, 81, 55–79. https://doi.org/10.1016/j.jmp.2017.09.004 buckner, r. l., & carroll, d. c. (2007). self-projection and the brain. trends in cognitive sciences, 11(2), 49–57. https://doi.org/10.1016/j.tics.2006.11.004 büchel, c., geuter, s., sprenger, c., & eippert, f. (2014). placebo analgesia: a predictive coding perspective. neuron, 81(6), 1223–1239. https://doi.org/10.1016/j.neuron.2014.02.042 cannon, w. b. (1927). the james-lange theory of emotions: a critical examination and an alternative theory. the american journal of psychology, 39(1/4), 106–124. https://doi.org/10.2307/1415404 cannon, w. b. (1929). organization for physiological homeostasis. physiological reviews, 9(3), 399–431. https://doi.org/10.1152/physrev.1929.9.3.399 carhart-harris, r. l. (2019). how do psychedelics work? current opinion in psychiatry, 32(1), 16–21. https://doi.org/10.1097/yco.0000000000000467 carhart-harris, r. l., & friston, k. (2019). rebus and the anarchic brain: toward a unified model of the brain action of psychedelics. pharmacological reviews, 71(3), 316–344. https://doi.org/10.1124/pr.118.017160 carhart-harris, r. l., & goodwin, g. m. (2017). the therapeutic potential of psychedelic drugs: past, present, and future. neuropsychopharmacology, 42(11), 2105. https://doi.org/10.1038/npp.2017.84 carhart-harris, r. l., & nutt, d. (2017). serotonin and brain function: a tale of two receptors. journal of psychopharmacology, 31(9), 1091–1120. https://doi.org/10.1177/0269881117725915 carhart-harris, r. l., roseman, l., haijen, e., erritzoe, d., watts, r., branchi, i., & kaelen, m. (2018). psychedelics and the essential importance of context. journal of psychopharmacology, 32(7), 725–731. https://doi.org/10.1177/0269881118754710 carter, o. l., hasler, f., pettigrew, j. d., wallis, g. m., liu, g. b., & vollenweider, f. x. (2007). psilocybin links binocular rivalry switch rate to attention and subjective arousal levels in humans. psychopharmacology, 195(3), 415–424. https://doi.org/10.1007/s00213-007-0930-9 carter, o. l., pettigrew, j. d., hasler, f., wallis, g. m., liu, g. b., hell, d., & vollenweider, f. x. (2005). modulating the rate and rhythmicity of perceptual rivalry alternations with the mixed 5-ht 2a and 5-ht 1a agonist psilocybin. neuropsychopharmacology, 30(6), 1154. https://doi.org/10.1038/sj.npp.1300621 cisek, p. (2007). cortical mechanisms of action selection: the affordance competition hypothesis. philosophical transactions of the royal society b: biological sciences, 362(1485), 1585–1599. https://doi.org/10.1017/cbo9780511731525.015 clark, a. (2013). whatever next? predictive brains, situated agents, and the future of cognitive science. behavioral and brain sciences, 36(3), 181–204. https://doi.org/10.1017/s0140525x12000477 clark, a. (2019). consciousness as generative entanglement. the journal of philosophy, 116, 645–662. https://doi.org/10.5840/jphil20191161241 clark, j. e., watson, s., & friston, k. (2018). what is mood? a computational perspective. psychological medicine, 48(14), 2277–2284. https://doi.org/10.1017/s0033291718000430 conant, r. c., & ashby, r. w. (1970). every good regulator of a system must be a model of that system. international journal of systems science, 1(2), 89–97. https://doi.org/10.1080/00207727008920220 corcoran, a., & hohwy, j. (2018). allostasis, interoception, and the free energy principle: feeling our way forward. https://doi.org/10.1093/oso/9780198811930.003.0015 corcoran, a. w., pezzulo, g., & hohwy, j. (2019). from allostatic agents to counterfactual cognisers: active inference, biological regulation, and the origins of cognition. https://doi.org/10.20944/preprints201911.0083.v1 deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1038/sdata.2018.63 https://doi.org/10.1093/cercor/bhy064 https://doi.org/10.1016/j.tics.2008.10.003 https://doi.org/10.1016/j.tics.2012.08.006 https://doi.org/10.1007/s10339-013-0571-3 https://doi.org/10.1016/j.jmp.2017.09.004 https://doi.org/10.1016/j.tics.2006.11.004 https://doi.org/10.1016/j.neuron.2014.02.042 https://doi.org/10.2307/1415404 https://doi.org/10.1152/physrev.1929.9.3.399 https://doi.org/10.1097/yco.0000000000000467 https://doi.org/10.1124/pr.118.017160 https://doi.org/10.1038/npp.2017.84 https://doi.org/10.1177/0269881117725915 https://doi.org/10.1177/0269881118754710 https://doi.org/10.1007/s00213-007-0930-9 https://doi.org/10.1038/sj.npp.1300621 https://doi.org/10.1017/cbo9780511731525.015 https://doi.org/10.1017/s0140525x12000477 https://doi.org/10.5840/jphil20191161241 https://doi.org/10.1017/s0033291718000430 https://doi.org/10.1080/00207727008920220 https://doi.org/10.1093/oso/9780198811930.003.0015 https://doi.org/10.20944/preprints201911.0083.v1 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 23 corlett, p. r., frith, c. d., & fletcher, p. c. (2009). from drugs to deprivation: a bayesian framework for understanding models of psychosis. psychopharmacology, 206(4), 515–530. https://doi.org/10.1007/s00213-009-1561-0 corlett, p. r., honey, g. d., & fletcher, p. c. (2016). prediction error, ketamine and psychosis: an updated model. journal of psychopharmacology, 30(11), 1145–1155. https://doi.org/10.1177/0269881116650087 craig, a. d. (2002). how do you feel? interoception: the sense of the physiological condition of the body. nature reviews neuroscience, 3(8), 655. https://doi.org/10.1038/nrn894 crapse, t. b., & sommer, m. a. (2008). corollary discharge across the animal kingdom. nature reviews neuroscience, 9(8), 587. https://doi.org/10.1038/nrn2457 damasio, a., & carvalho, g. b. (2013). the nature of feelings: evolutionary and neurobiological origins. nature reviews neuroscience, 14(2), 143. https://doi.org/10.1038/nrn3403 dayan, p., & daw, n. d. (2008). decision theory, reinforcement learning, and the brain. cognitive, affective, & behavioral neuroscience, 8(4), 429–453. https://doi.org/10.3758/cabn.8.4.429 eisenberger, n. i. (2012). the pain of social disconnection: examining the shared neural underpinnings of physical and social pain. nature reviews neuroscience, 13(6), 421. https://doi.org/10.1038/nrn3231 eisenberger, n. i., jarcho, j. m., lieberman, m. d., & naliboff, b. d. (2006). an experimental study of shared sensitivity to physical pain and social rejection. pain, 126(1-3), 132–138. https://doi.org/10.1016/j.pain.2006.06.024 fanciullacci, m., bene, e. d., franchi, g., & sicuteri, f. (1977). phantom limb pain: sub-hallucinogenic treatment with lysergic acid diethylamide (lsd-25). headache: the journal of head and face pain, 17(3), 118–119. https://doi.org/10.1111/j.1526-4610.1977.hed1703118.x feldman, h., & friston, k. (2010). attention, uncertainty, and free-energy. frontiers in human neuroscience, 4, 215. https://doi.org/10.3389/fnhum.2010.00215 fletcher, p. c., & frith, c. d. (2009). perceiving is believing: a bayesian approach to explaining the positive symptoms of schizophrenia. nature reviews neuroscience, 10(1), 48. https://doi.org/10.1038/nrn2536 ford, j. m., gray, m., faustman, w. o., roach, b. j., & mathalon, d. h. (2007). dissecting corollary discharge dysfunction in schizophrenia. psychophysiology, 44(4), 522–529. https://doi.org/10.1111/j.1469-8986.2007.00533.x ford, j. m., & mathalon, d. h. (2005). corollary discharge dysfunction in schizophrenia: can it explain auditory hallucinations? international journal of psychophysiology, 58(2-3), 179–189. https://doi.org/10.1016/j.ijpsycho.2005.01.014 friston, k. (2010). the free-energy principle: a unified brain theory? nature reviews neuroscience, 11(2), 127. https://doi.org/10.1038/nrn2787 friston, k. (2012a). embodied inference and spatial cognition. cognitive processing, 13(1), 171–177. https://doi.org/10.1007/s10339-012-0519-z friston, k. (2012b). prediction, perception and agency. international journal of psychophysiology, 83(2), 248–252. https://doi.org/10.1016/j.ijpsycho.2011.11.014 friston, k. (2018). am i self-conscious?(or does self-organization entail self-consciousness?). frontiers in psychology, 9, 579. https://doi.org/10.3389/fpsyg.2018.00579 friston, k., daunizeau, j., kilner, j., & kiebel, s. j. (2010). action and behavior: a free-energy formulation. biological cybernetics, 102(3), 227–260. https://doi.org/10.1007/s00422-010-0364-z friston, k., rigoli, f., ognibene, d., mathys, c., fitzgerald, t., & pezzulo, g. (2015). active inference and epistemic value. cognitive neuroscience, 6(4), 187–214. https://doi.org/10.1080/17588928.2015.1020053 friston, k., rosch, r., parr, t., price, c., & bowman, h. (2018). deep temporal models and active inference. neuroscience & biobehavioral reviews, 90, 486–501. https://doi.org/10.1016/j.neubiorev.2017.04.009 friston, k., schwartenbeck, p., fitzgerald, t., moutoussis, m., behrens, t., & dolan, r. j. (2013). the anatomy of choice: active inference and agency. frontiers in human neuroscience, 7, 598. https://doi.org/10.3389/fnhum.2013.00598 friston, k., shiner, t., fitzgerald, t., galea, j. m., adams, r., brown, h., et al. (2012). dopamine, affordance and active inference. plos computational biology, 8(1), e1002327. https://doi.org/10.1371/journal.pcbi.1002327 frith, u. (2003). autism: explaining the enigma. oxford: blackwell publishing. garrido, m. i., kilner, j. m., stephan, k. e., & friston, k. (2009). the mismatch negativity: a review of underlying mechanisms. clinical neurophysiology, 120(3), 453–463. https://doi.org/10.1016/j.clinph.2008.11.029 gershman, s. j. (2019). what does the free energy principle tell us about the brain? neurons, behavior, data analysis, and theory, 4(1), 1–10. girn, m., & christoff, k. (2018). journal of consciousness studies, 25(11-12), 131–154. retrieved from https://www.ingentaconnect.com/content/imp/jcs/2018/00000025/f0020011/art00008 griffiths, r. r., johnson, m. w., carducci, m. a., umbricht, a., richards, w. a., richards, b. d., et al. (2016). psilocybin produces substantial and sustained decreases in depression and anxiety in patients with life-threatening cancer: a randomized double-blind trial. journal of psychopharmacology, 30(12), 1181–1197. https://doi.org/10.1177/0269881116675513 deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s00213-009-1561-0 https://doi.org/10.1177/0269881116650087 https://doi.org/10.1038/nrn894 https://doi.org/10.1038/nrn2457 https://doi.org/10.1038/nrn3403 https://doi.org/10.3758/cabn.8.4.429 https://doi.org/10.1038/nrn3231 https://doi.org/10.1016/j.pain.2006.06.024 https://doi.org/10.1111/j.1526-4610.1977.hed1703118.x https://doi.org/10.3389/fnhum.2010.00215 https://doi.org/10.1038/nrn2536 https://doi.org/10.1111/j.1469-8986.2007.00533.x https://doi.org/10.1016/j.ijpsycho.2005.01.014 https://doi.org/10.1038/nrn2787 https://doi.org/10.1007/s10339-012-0519-z https://doi.org/10.1016/j.ijpsycho.2011.11.014 https://doi.org/10.3389/fpsyg.2018.00579 https://doi.org/10.1007/s00422-010-0364-z https://doi.org/10.1080/17588928.2015.1020053 https://doi.org/10.1016/j.neubiorev.2017.04.009 https://doi.org/10.3389/fnhum.2013.00598 https://doi.org/10.1371/journal.pcbi.1002327 https://doi.org/10.1016/j.clinph.2008.11.029 https://www.ingentaconnect.com/content/imp/jcs/2018/00000025/f0020011/art00008 https://doi.org/10.1177/0269881116675513 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 24 grof, s. (1980). lsd psychotherapy. pomona, ca: hunter house. heinks-maldonado, t. h., mathalon, d. h., houde, j. f., gray, m., faustman, w. o., & ford, j. m. (2007). relationship of imprecise corollary discharge in schizophrenia to auditory hallucinations. archives of general psychiatry, 64(3), 286–296. https://doi.org/10.1001/archpsyc.64.3.286 hendricks, p. s., thorne, c. b., clark, c. b., coombs, d. w., & johnson, m. w. (2015). classic psychedelic use is associated with reduced psychological distress and suicidality in the united states adult population. journal of psychopharmacology, 29(3), 280–288. https://doi.org/10.1177/0269881114565653 hesp, c., smith, r., allen, m., friston, k., & ramstead, m. (2019). deeply felt affect: the emergence of valence in deep active inference. psyarxiv. https://doi.org/10.31234/osf.io/62pfd hohwy, j. (2007). the sense of self in the phenomenology of agency and perception. psyche, 13(1), 1–20. hohwy, j. (2012). attention and conscious perception in the hypothesis testing brain. frontiers in psychology, 3, 96. https://doi.org/10.3389/fpsyg.2012.00096 hohwy, j. (2017). priors in perception: top-down modulation, bayesian perceptual learning rate, and prediction error minimization. consciousness and cognition, 47, 75–85. https://doi.org/10.1016/j.concog.2016.09.004 hohwy, j., & michael, j. (2017). why should any body have a self? in f. de vignemont & a. alsmith (eds.), the subject’s matter: self-consciousness and the body (pp. 363–391). cambridge, ma: mit press. huxley, a. (2010). the doors of perception: and heaven and hell. new york, ny: random house. jakab, r. l., & goldman-rakic, p. s. (1998). proceedings of the national academy of sciences, 95(2), 735–740. https://doi.org/10.1073/pnas.95.2.735 kaplan, r., & friston, k. (2018). planning and navigation as active inference. biological cybernetics, 112(4), 323–343. https://doi.org/10.1007/s00422-018-0753-2 keller, g. b., & mrsic-flogel, t. d. (2018). predictive processing: a canonical cortical computation. neuron, 100(2), 424–435. https://doi.org/10.1016/j.neuron.2018.10.003 kiebel, s. j., daunizeau, j., & friston, k. (2008). a hierarchy of time-scales and the brain. plos computational biology, 4(11), e1000209. https://doi.org/10.1371/journal.pcbi.1000209 king, a., martin, i., & seymour, k. (1972). reversal learning facilitated by a single injection of lysergic acid diethylamide (lsd 25) in the rat. british journal of pharmacology, 45(1), 161p. https://doi.org/10.1111/j.1476-5381.1974.tb08611.x knill, d. c., & pouget, a. (2004). the bayesian brain: the role of uncertainty in neural coding and computation. trends in neurosciences, 27(12), 712–719. https://doi.org/10.1016/j.tins.2004.10.007 kometer, m., cahn, b. r., andel, d., carter, o. l., & vollenweider, f. x. (2011). the 5-ht2a/1a agonist psilocybin disrupts modal object completion associated with visual hallucinations. biological psychiatry, 69(5), 399–406. https://doi.org/10.1016/j.biopsych.2010.10.002 leary, t. f., metzner, r., & alpert, r. (1964). the psychedelic experience: a manual based on the tibetan book of the dead. lebedev, a. v., lövdén, m., rosenthal, g., feilding, a., nutt, d. j., & carhart-harris, r. l. (2015). finding the self by losing the self: neural correlates of ego-dissolution under psilocybin. human brain mapping, 36(8), 3137–3153. https://doi.org/10.1002/hbm.22833 letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 letheby, c., & gerrans, p. (2017). self unbound: ego dissolution in psychedelic experience. neuroscience of consciousness, 3(1), nix016. https://doi.org/10.1093/nc/nix016 limanowski, j., & blankenburg, f. (2013). minimal self-models and the free energy principle. frontiers in human neuroscience, 7, 547. https://doi.org/10.3389/fnhum.2013.00547 limanowski, j., & friston, k. (2018). “seeing the dark”: grounding phenomenal transparency and opacity in precision estimation for active inference. frontiers in psychology, 9, 643. https://doi.org/10.3389/fpsyg.2018.00643 limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 ly, c., greb, a. c., cameron, l. p., wong, j. m., barragan, e. v., wilson, p. c., et al. (2018). psychedelics promote structural and functional neural plasticity. cell reports, 23(11), 3170–3182. https://doi.org/10.1016/j.celrep.2018.05.022 lyons, t., & carhart-harris, r. l. (2018). more realistic forecasting of future life events after psilocybin for treatment-resistant depression. frontiers in psychology, 9, 1721. https://doi.org/10.3389/fpsyg.2018.01721 masters, r. e., & houston, j. (1966). the varieties of psychedelic experience (vol. 9289). new york, chicago, san francisco: holt, rinehart; winston new york. mathys, c., daunizeau, j., friston, k., & stephan, k. e. (2011). a bayesian foundation for individual learning under uncertainty. frontiers in human neuroscience, 5, 39. https://doi.org/10.3389/fnhum.2011.00039 deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1001/archpsyc.64.3.286 https://doi.org/10.1177/0269881114565653 https://doi.org/10.31234/osf.io/62pfd https://doi.org/10.3389/fpsyg.2012.00096 https://doi.org/10.1016/j.concog.2016.09.004 https://doi.org/10.1073/pnas.95.2.735 https://doi.org/10.1007/s00422-018-0753-2 https://doi.org/10.1016/j.neuron.2018.10.003 https://doi.org/10.1371/journal.pcbi.1000209 https://doi.org/10.1111/j.1476-5381.1974.tb08611.x https://doi.org/10.1016/j.tins.2004.10.007 https://doi.org/10.1016/j.biopsych.2010.10.002 https://doi.org/10.1002/hbm.22833 https://doi.org/10.33735/phimisci.2020.i.47 https://doi.org/10.1093/nc/nix016 https://doi.org/10.3389/fnhum.2013.00547 https://doi.org/10.3389/fpsyg.2018.00643 https://doi.org/10.33735/phimisci.2020.i.35 https://doi.org/10.1016/j.celrep.2018.05.022 https://doi.org/10.3389/fpsyg.2018.01721 https://doi.org/10.3389/fnhum.2011.00039 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 25 mathys, c., lomakina, e. i., daunizeau, j., iglesias, s., brodersen, k. h., friston, k., & stephan, k. e. (2014). uncertainty in perception and the hierarchical gaussian filter. frontiers in human neuroscience, 8, 825. https://doi.org/10.3389/fnhum.2014.00825 metzinger, t. (2003). being no one: the self-model theory of subjectivity. cambridge, ma: mit press. metzinger, t. (2009). the ego tunnel: the science of the mind and the myth of the self. new york: basic books. metzinger, t. (2017). the problem of mental action. in t. metzinger & w. wiese (eds.), philosophy and predictive processing. frankfurt am mainz: mind group. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of ”pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 millière, r. (2017). looking for the self: phenomenology, neurophysiology and philosophical significance of drug-induced ego dissolution. frontiers in human neuroscience, 11, 245. https://doi.org/10.3389/fnhum.2017.00245 millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 millière, r., carhart-harris, r. l., roseman, l., trautwein, f.-m., & berkovich-ohana, a. (2018). psychedelics, meditation, and self-consciousness. frontiers in psychology, 9, 1–29. https://doi.org/10.3389/fpsyg.2018.01475 montague, p. r., & king-casas, b. (2007). efficient statistics, common currencies and the problem of reward-harvesting. trends in cognitive sciences, 11(12), 514–519. https://doi.org/10.1016/j.tics.2007.10.002 moreno, f. a., wiegand, c. b., taitano, e. k., & delgado, p. l. (2006). safety, tolerability, and efficacy of psilocybin in 9 patients with obsessive-compulsive disorder. journal of clinical psychiatry, 67(11), 1735–1740. https://doi.org/10.4088/jcp.v67n1110 morton, d. l., el-deredy, w., watson, a., & jones, a. k. (2010). placebo analgesia as a case of a cognitive style driven by prior expectation. brain research, 1359, 137–141. https://doi.org/10.1016/j.brainres.2010.08.046 moutoussis, m., fearon, p., el-deredy, w., dolan, r. j., & friston, k. (2014). bayesian inferences about the self (and others): a review. consciousness and cognition, 25, 67–76. https://doi.org/10.1016/j.concog.2014.01.009 mumford, d. (1992). on the computational architecture of the neocortex. biological cybernetics, 66(3), 241–251. https://doi.org/10.1007/bf00198477 nour, m. m., & carhart-harris, r. l. (2017). psychedelics and the science of self-experience. the british journal of psychiatry, 210(3), 177–179. https://doi.org/10.1192/bjp.bp.116.194738 nour, m. m., evans, l., nutt, d., & carhart-harris, r. l. (2016). ego-dissolution and psychedelics: validation of the ego-dissolution inventory (edi). frontiers in human neuroscience, 10, 269. https://doi.org/10.3389/fnhum.2016.00269 nutt, d. j., king, l. a., & phillips, l. d. (2010). drug harms in the uk: a multicriteria decision analysis. the lancet, 376(9752), 1558–1565. https://doi.org/10.1016/s0140-6736(10)61462-6 ongaro, g., & kaptchuk, t. j. (2019). symptom perception, placebo effects, and the bayesian brain. pain, 160(1), 1. https://doi.org/10.1097/j.pain.0000000000001367 palhano-fontes, f., barreto, d., onias, h., andrade, k. c., novaes, m. m., pessoa, j. a., et al. (2019). rapid antidepressant effects of the psychedelic ayahuasca in treatment-resistant depression: a randomized placebo-controlled trial. psychological medicine, 49(4), 655–663. https://doi.org/10.1017/s0033291718001356 palmer, c. j., lawson, r. p., & hohwy, j. (2017). bayesian approaches to autism: towards volatility, action, and behavior. psychological bulletin, 143(5), 521. https://doi.org/10.1037/bul0000097 petitmengin, c. (2006). describing one’s subjective experience in the second person: an interview method for the science of consciousness. phenomenology and the cognitive sciences, 5(3-4), 229–269. https://doi.org/10.1007/s11097-006-9022-2 pezzulo, g. (2014). why do you fear the bogeyman? an embodied predictive coding model of perceptual inference. cognitive, affective, & behavioral neuroscience, 14(3), 902–911. https://doi.org/10.3758/s13415-013-0227-x pezzulo, g., cartoni, e., rigoli, f., pio-lopez, l., & friston, k. (2016). active inference, epistemic value, and vicarious trial and error. learning & memory, 23(7), 322–338. https://doi.org/10.1101/lm.041780.116 pezzulo, g., & cisek, p. (2016). navigating the affordance landscape: feedback control as a process model of behavior and cognition. trends in cognitive sciences, 20(6), 414–424. https://doi.org/10.1016/j.tics.2016.03.013 pezzulo, g., rigoli, f., & friston, k. (2015). active inference, homeostatic regulation and adaptive behavioural control. progress in neurobiology, 134, 17–35. https://doi.org/10.1016/j.pneurobio.2015.09.001 pezzulo, g., rigoli, f., & friston, k. (2018). hierarchical active inference: a theory of motivated control. trends in cognitive sciences, 22(4), 294–306. https://doi.org/10.1016/j.tics.2018.01.009 picard, f., & friston, k. (2014). predictions, perception, and a sense of self. neurology, 83(12), 1112–1118. https://doi.org/10.1212/wnl.0000000000000798 deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.3389/fnhum.2014.00825 https://doi.org/10.33735/phimisci.2020.i.46 https://doi.org/10.3389/fnhum.2017.00245 https://doi.org/10.33735/phimisci.2020.i.48 https://doi.org/10.3389/fpsyg.2018.01475 https://doi.org/10.1016/j.tics.2007.10.002 https://doi.org/10.4088/jcp.v67n1110 https://doi.org/10.1016/j.brainres.2010.08.046 https://doi.org/10.1016/j.concog.2014.01.009 https://doi.org/10.1007/bf00198477 https://doi.org/10.1192/bjp.bp.116.194738 https://doi.org/10.3389/fnhum.2016.00269 https://doi.org/10.1016/s0140-6736(10)61462-6 https://doi.org/10.1097/j.pain.0000000000001367 https://doi.org/10.1017/s0033291718001356 https://doi.org/10.1037/bul0000097 https://doi.org/10.1007/s11097-006-9022-2 https://doi.org/10.3758/s13415-013-0227-x https://doi.org/10.1101/lm.041780.116 https://doi.org/10.1016/j.tics.2016.03.013 https://doi.org/10.1016/j.pneurobio.2015.09.001 https://doi.org/10.1016/j.tics.2018.01.009 https://doi.org/10.1212/wnl.0000000000000798 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org george deane 26 pink-hashkes, s., rooij, i. van, & kwisthout, j. (2017). perception is in the details: a predictive coding account of the psychedelic phenomenon. proceedings of the 39th annual meeting of the cognitive science society, 2907–2912. preller, k. h., & vollenweider, f. x. (2016). phenomenology, structure, and dynamic of psychedelic states. in a. l. halberstadt, f. x. vollenweider, & d. e. nichols (eds.), behavioral neurobiology of psychedelic drugs (pp. 221–256). https://doi.org/10.1007/7854_2016_459 ramachandran, v., chunharas, c., marcus, z., furnish, t., & lin, a. (2018). relief from intractable phantom pain by combining psilocybin and mirror visual-feedback (mvf). neurocase, 24(2), 105–110. https://doi.org/10.1080/13554794.2018.1468469 rao, r. p., & ballard, d. h. (1999). predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive-field effects. nature neuroscience, 2(1), 79. https://doi.org/10.1038/4580 richards, w. a. (2015). sacred knowledge: psychedelics and religious experiences. new york: columbia university press. romano, a. g., quinn, j. l., li, l., dave, k. d., schindler, e. a., aloyo, v. j., & harvey, j. a. (2010). intrahippocampal lsd accelerates learning and desensitizes the 5-ht 2a receptor in the rabbit, romano et al. psychopharmacology, 212(3), 441–448. https://doi.org/10.1007/s00213-010-2004-7 roseman, l., nutt, d. j., & carhart-harris, r. l. (2018). quality of acute psychedelic experience predicts therapeutic efficacy of psilocybin for treatment-resistant depression. frontiers in pharmacology, 8, 974. https://doi.org/10.3389/fphar.2017.00974 rösler, l., rolfs, m., van der stigchel, s., neggers, s. f., cahn, w., kahn, r. s., & thakkar, k. n. (2015). failure to use corollary discharge to remap visual target locations is associated with psychotic symptom severity in schizophrenia. journal of neurophysiology, 114(2), 1129–1136. https://doi.org/10.1152/jn.00155.2015 savage, c. (1955). variations in ego feeling induced by d-lysergic acid diethylamide (lsd-25). psychoanalytic review, 42(1), 1–16. schachter, s., & singer, j. (1963). cognitive, social, and physiological determinants of emotional state. psychological review, 70(1), 121–122. https://doi.org/10.1037/h0038845 schacter, d. l., addis, d. r., & buckner, r. l. (2008). episodic simulation of future events: concepts, data, and applications. annals of the new york academy of sciences, 1124(1), 39–60. https://doi.org/10.1196/annals.1440.001 schulkin, j., & sterling, p. (2019). allostasis: a brain-centered, predictive mode of physiological regulation. trends in neurosciences, 42(10), 740–752. https://doi.org/10.1016/j.tins.2019.07.010 scott, g., & carhart-harris, r. l. (2019). psychedelics as a treatment for disorders of consciousness. neuroscience of consciousness, 2019(1), niz003. https://doi.org/10.1093/nc/niz003 sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 seth, a. (2015). the cybernetic bayesian brain: from interoceptive inference to sensorimotor contingencies. in t. metzinger & j. windt (eds.), open mind (pp. 1–24). frankfurt am main: mind group. seth, a. k., & friston, k. (2016). active interoceptive inference and the emotional brain. philosophical transactions of the royal society b: biological sciences, 371(1708), 20160007. https://doi.org/10.1098/rstb.2016.0007 seth, a. k., & tsakiris, m. (2018). being a beast machine: the somatic basis of selfhood. trends in cognitive sciences, 22(11), 969–981. https://doi.org/10.1016/j.tics.2018.08.008 shadmehr, r., smith, m. a., & krakauer, j. w. (2010). error correction, sensory prediction, and adaptation in motor control. annual review of neuroscience, 33, 89–108. https://doi.org/10.1146/annurev-neuro-060909-153135 shipp, s. (2016). neural elements for predictive coding. frontiers in psychology, 7, 1792. https://doi.org/10.3389/fpsyg.2016.01792 siegel, e. h., sands, m. k., van den noortgate, w., condon, p., chang, y., dy, j., et al. (2018). emotion fingerprints or emotion populations? a meta-analytic investigation of autonomic features of emotion categories. psychological bulletin, 144(4), 343. https://doi.org/10.1037/bul0000128 stephan, k. e., manjaly, z. m., mathys, c., weber, l. a., paliwal, s., gard, t., et al. (2016). allostatic self-efficacy: a metacognitive theory of dyshomeostasis-induced fatigue and depression. frontiers in human neuroscience, 10, 550. https://doi.org/10.3389/fnhum.2016.00550 sterling, p. (2012). allostasis: a model of predictive regulation. physiology & behavior, 106(1), 5–15. https://doi.org/10.1016/j.physbeh.2011.06.004 sterzer, p., adams, r. a., fletcher, p., frith, c., lawrie, s. m., muckli, l., et al. (2018). the predictive coding account of psychosis. biological psychiatry, 84(9), 634–643. https://doi.org/10.1016/j.biopsych.2018.05.015 strassman, r. j. (1984). adverse reactions to psychedelic drugs. a review of the literature. journal of nervous and mental disease, 172(10), 577–595. https://doi.org/10.1097/00005053-198410000-00001 timmermann, c., roseman, l., schartner, m., millière, r., williams, l., erritzoe, d., et al. (2019). neural correlates of the dmt experience as assessed via multivariate eeg. biorxiv, 706283. https://doi.org/10.1101/706283 deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/7854_2016_459 https://doi.org/10.1080/13554794.2018.1468469 https://doi.org/10.1038/4580 https://doi.org/10.1007/s00213-010-2004-7 https://doi.org/10.3389/fphar.2017.00974 https://doi.org/10.1152/jn.00155.2015 https://doi.org/10.1037/h0038845 https://doi.org/10.1196/annals.1440.001 https://doi.org/10.1016/j.tins.2019.07.010 https://doi.org/10.1093/nc/niz003 https://doi.org/10.33735/phimisci.2020.i.44 https://doi.org/10.1098/rstb.2016.0007 https://doi.org/10.1016/j.tics.2018.08.008 https://doi.org/10.1146/annurev-neuro-060909-153135 https://doi.org/10.3389/fpsyg.2016.01792 https://doi.org/10.1037/bul0000128 https://doi.org/10.3389/fnhum.2016.00550 https://doi.org/10.1016/j.physbeh.2011.06.004 https://doi.org/10.1016/j.biopsych.2018.05.015 https://doi.org/10.1097/00005053-198410000-00001 https://doi.org/10.1101/706283 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dissolving the self: active inference, psychedelics, and ego-dissolution 27 timmermann, c., spriggs, m. j., kaelen, m., leech, r., nutt, d. j., moran, r. j., et al. (2018). lsd modulates effective connectivity and neural adaptation mechanisms in an auditory oddball paradigm. neuropharmacology, 142, 251–262. https://doi.org/10.1016/j.neuropharm.2017.10.039 wager, t. d., kang, j., johnson, t. d., nichols, t. e., satpute, a. b., & barrett, l. f. (2015). a bayesian model of category-specific emotional brain responses. plos computational biology, 11(4), e1004066. https://doi.org/10.1371/journal.pcbi.1004066 wiese, w., & metzinger, t. (2017). vanilla pp for philosophers: a primer on predictive processing. in t. metzinger & w. wiese (eds.), philosophy and predictive processing. https://doi.org/10.15502/9783958573024 wilkinson, s., deane, g., nave, k., & clark, a. (2019). getting warmer: predictive processing and the nature of emotion. in l. candiotto (ed.), the value of emotions for knowledge (pp. 101–119). https://doi.org/10.1007/978-3-030-15667-1 yoshida, w., dolan, r. j., & friston, k. (2008). game theory of mind. plos computational biology, 4(12), e1000254. https://doi.org/10.1371/journal.pcbi.1000254 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neuropharm.2017.10.039 https://doi.org/10.1371/journal.pcbi.1004066 https://doi.org/10.15502/9783958573024 https://doi.org/10.1007/978-3-030-15667-1 https://doi.org/10.1371/journal.pcbi.1000254 https://doi.org/10.33735/phimisci.2020.i.39 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction the free energy principle predictive processing precision-weighting control-oriented inference homeostasis active inference the self in active inference temporal thickness attuning to the world emotion hierarchically deep self-models psychedelics psychedelic therapy psychedelics in the predictive brain bayesian learning rate psychedelic action as high bayesian learning rate evidence for the high bayesian learning rate hypothesis psychedelic-induced ego-dissolution in active inference ecstatic ego-dissolution and challenging experiences conclusion extended mind-wandering. extended mind-wandering jelle bruineberga,∗  (jelle.bruineberg@mq.edu.au) regina e. fabrya,∗  (regina.fabry@mq.edu.au) abstract smartphone use plays an increasingly important role in our daily lives. philosophical research that has used first wave or second wave theories of extended cognition in order to understand our engagement with digital technologies has focused on the contribution of these technologies to the completion of specific cognitive tasks (e.g., remembering, reasoning, problem-solving, navigation). however, in a considerable number of cases, everyday smartphone use is task-unrelated. in psychological research, these cases have been captured by notions such as absent-minded smartphone use (marty-dugas et al., 2018) or smartphone-related inattentiveness (liebherr et al., 2020). given the prevalence of these cases, we develop a conceptual framework that can accommodate the functional and phenomenological characteristics of task-unrelated smartphone use. to this end, we will integrate research on second wave extended cognition with mind-wandering research and introduce the concept of ‘extended mind-wandering’. elaborating the family resemblances approach to mind-wandering (seli, kane, smallwood, et al., 2018), we will argue that task-unrelated smartphone use shares many characteristics with mind-wandering. we will suggest that an empirically informed conceptual analysis of cases of extended mind-wandering can enrich current work on digitally extended cognition by specifying the influence of the attention economy on our cognitive dynamics. keywords extended mind · mind-wandering · smartphone use · attention economy 1 introduction the extended mind thesis holds that many cognitive processes and mental states can extend to entities external to the subject (clark, 2008; clark & chalmers, 1998). research on the extended mind has investigated how environmental resources make crucial functional contributions to a variety of cognitive tasks, ranging from reasoning and problem-solving to remembering and spatial navigation (menary, 2010b). adepartment of philosophy, macquarie university ∗both authors contributed equally to this paper. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-5675-3376 https://orcid.org/0000-0003-1078-1499 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 2 since the initial formulation of the extended mind thesis in 1995, the amount and complexity of smartphones and other digital technologies in our daily environment have skyrocketed. it is sometimes asserted that if not true in 1995, the extended mind thesis has certainly become true in the course of the 2010s (chalmers, 2019). indeed, smartphones seem to be exemplary mind-extenders: small, personal and portable devices that support a variety of cognitive tasks (information retrieval, communication, navigation, remembering, calculation, etc.) research on extended cognition has started to explore the philosophical implications of these recent technological developments (e.g., clowes, 2019; heersmink & sutton, 2020). for the most part, research on extended cognition has focused on how environmental resources support the completion of cognitive tasks. two biases have led to this focus. first, research on extended cognition has been mainly interested in analyses of successful cognitive episodes. for example, it has been argued that digital technologies make positive contributions to belief formation (smart, 2017) and remembering (heersmink & sutton, 2020). aagaard (2021) calls this the dogma of harmony: a tendency to capitalize on cases of cooperation between humans and technology and a deemphasizing of conflict or interference. this harmony bias has contributed to a neglect of theoretical considerations on cases of disharmonious cognition, i.e., human-technology relations that are detrimental to the manifestation of the agent’s cognitive abilities or wider concerns in particular situations (for exceptions, see gillett & heersmink, 2019; hebblewhite & gillett, 2020). second, research on extended cognition has been biased towards explorations of the completion of cognitive tasks (clark & chalmers, 1998; menary, 2010a). otto’s task is to go to moma, the question is how he is able to remember where it is located.1 the task of playing tetris mandates that the player rotates, internally or externally, the pieces and moves them to the right place. this cognitive task bias has led to a neglect of an important range of cognitive phenomena that are task-unrelated, such as mind-wandering. this is problematic, since mindwandering is estimated to account for 25-50% of waking cognition (christoff et al., 2016; killingsworth & gilbert, 2010). if task-related cognition can extend into the environment, then why should not the same hold for task-unrelated cognition?2 what would task-unrelated extended cognition look like? in this paper, we will focus on the habitual (i.e., diachronically established), unreflective or pre-reflective use of mobile devices, especially smartphones.3 most smartphone users will be fa1otto is the protagonist of clark and chalmers’ (1998) thought experiment, which helped motivate the extended mind thesis. the alzheimer’s patient otto employs his notebook to remember the location of the museum of modern art (moma) in new york. identifying the location of moma, in turn, allows him to successfully navigate to the museum. 2we will discuss the notion of ‘task-unrelated’ cognitive processes in sections 3 and 4 in detail. 3mobile devices is a rather open-ended category. mobile devices are computers (in the broad sense of the word) that are portable (as opposed to desktop-computers), are connected to the internet, and typically have an interface (janlert & stolterman, 2017). at the time of writing, the canonical example of such a mobile device is a smartphone (such as an iphone). for practical purposes, we will limit our analysis to smartphone use. however, we assume that our analysis is likely to have bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 3 miliar with phenomena like using their phone for longer than planned, finding themselves checking their phone without having decided to do so or scrolling through a social media or news feed without a particular goal (aranda & baig, 2018; marty-dugas et al., 2018). according to hiniker et al. (2016), these cases qualify as ritualistic smartphone use that is “habitual and diversionary”, as opposed to instrumental, i.e., “goal-directed and purposeful” (p. 634). paradigmatic cases of ritualistic smartphone use, as identified in hiniker’s et al. (2016) experience sampling study, include the engagement with social media and news applications.4 under the assumption that cognizers spend a significant amount of time engaging with their smartphone (kruger et al., 2017; winnick, 2016),5 we should expect that habitual and diversionary smartphone use is a ubiquitous phenomenon. given that the frequency of general smartphone use is positively correlated with habitual and diversionary smartphone use (cf. marty-dugas et al., 2018), it is reasonable to assume that task-unrelated smartphone use characterizes our cognitive lives to a significant degree. recent psychological research has started to explore these cases of habitual and diversionary smartphone use (wilmer et al., 2017). notions that have been used in the psychological literature are absent-minded smartphone use (marty-dugas et al., 2018) and smartphone-related inattentiveness (liebherr et al., 2020), which refer to smartphone use that is characterized by the absence of strong, task-related, attentional constraints. smartphone-related inattentiveness can be endogenously or exogenously generated (liebherr et al., 2020; wilmer et al., 2017). in endogenously generated cases, “the user’s own thoughts drift toward a smartphone-related activity, and thereby evince an otherwise unsolicited drive to begin interacting with the device” (wilmer et al., 2017, p. 4). in exogenously induced cases, the drift of attention is cued by the smartphone (cf. wilmer et al., 2017).6 for current purposes, we a broader scope (for example also applying to smartwatch, tablet, and laptop use). we do not take a stance on whether older (pre-digital) technologies may or may not support task-unrelated extended cognition as well. an interesting avenue for further reseach would be to what extent, if any, digital technologies would be continuous with earlier non-digital technologies with regard to their contribution to task-unrelated extended cognition. 4at the time of writing, the canonical social media apps are twitter, facebook and instagram. the dominant ways of interacting with these apps is by scrolling through a newsfeed. the items on the newsfeed are typically short texts, links, images or videos that can be liked, shared and commented on. the feed itself is dynamically updated, i.e., continuously generated as the user scrolls down a page. 5in particular, winnick’s (2016) study tracked smartphone users for 5 days and concluded that they spent, on average, 145 minutes per day on their phone across 76 sessions. upon entering a waiting area (such as a queue or a bus stop), kruger’s et al. (2017) study found, 62% of the people were observed using their smartphone. 6in the product design literature, the difference between endogenously and exogenously generated habitual use is captured by the concepts internal and external triggers (eyal, 2014). external triggers are cues in the environment that a user associates with a particular behavior (for example a notification). internal triggers are automatic associations between a thought, an emotion or a routine and the use of a particular product. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 4 will focus on endogenously generated task-unrelated smartphone use to exclude cases in which smartphone notifications (stothart et al., 2015) or the presence of a smartphone (thornton et al., 2014) may merely exogenously cue the onset of an episode of inattentive smartphone use. despite these recent research efforts, a fine-grained conceptual framework for the analysis of this kind of digital engagement is currently missing. to close this gap, we propose that such episodes of habitual, diversionary smartphone use can be conceptualized as canonical cases of extended mind-wandering. to motivate this proposal, let’s consider the following episodes: 1. robert sits in class trying to listen to a lecture. he draws out his phone and checks his social media. he catches himself scrolling through his feed, puts the phone away and returns his focus to the lecture. 2. amanda joins the queue for coffee. while progressing in the queue, she draws out her phone and checks a number of apps. when she is next in the queue, she puts her phone away and orders her coffee. these examples of habitual, diversionary smartphone use bear striking similarities to typical cases of mind-wandering (christoff et al., 2016; irving & glasser, 2020; seli, kane, smallwood, et al., 2018; smallwood & schooler, 2015). however, in contrast to the kinds of mind-wandering that have been studied in the literature (irving & glasser, 2020; smallwood & schooler, 2015), the wandering described in these examples is mediated by a smartphone and thereby qualifies, or so we will argue, as a case of extended cognition. at first glance, it might seem surprising to conceptualize these cases of habitual smartphone use as extended mind-wandering. at least initially, mind-wandering was partly defined in terms of perceptual decoupling from the environment (schooler et al., 2011; smallwood & schooler, 2006, 2015), whereas extended mental processes essentially involve sensorimotor couplings with (a specific part of) the environment. based on the more recent family resemblances approach to mind-wandering (seli, kane, smallwood, et al., 2018), we will nevertheless claim that there are important resemblances between cases of non-extended mind-wandering and extended mind-wandering. moreover, the most interesting cases of extended cognition are exactly the ones that are typically thought of as internal and in which extension is achieved through sensorimotor interaction (chalmers, 2019). in what follows, we will develop a conceptual framework that integrates research on extended cognition, mind-wandering, and habitual and diversionary smartphone use. addressing the harmony and cognitive task biases in current research on extended cognition, we will argue that cases of extended mindwandering are potentially disharmonious and require a careful and balanced normative assessment. to this end, we will first survey existing work on the extended mind (section 2) and on mind-wandering (section 3). in a second step, we will integrate these two strands of research (section 4). furthermore, we will bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 5 explore the question what difference the technological mediation could make to the phenomenological and functional characteristics of task-unrelated cognition (section 5). we will discuss the similarities and dissimilarities of extended and non-extended mind-wandering and to what extent these two forms might stand in any competition. before concluding, we will briefly situate extended mind-wandering within the normative framework of the attention economy (section 6). 2 extended mind: state of research before we are in a position to establish our conceptualization of extended mindwandering, we will first present key positions in the extended mind debate. the extended mind thesis holds that cognitive processes are not exclusively realized by processes internal to the skull. although precursors of the extended mind thesis can be found in pragmatism (james, 1890) and phenomenology (merleau-ponty, 1945), the thesis was stated in its best known form by clark and chalmers (1998). this articulation is also known as first wave extended mind. clark and chalmers’ (1998) main diagnostic tool for identifying whether a particular process is part of the mind is the parity principle: if, as we confront some task, a part of the world functions as a process which, were it done in the head, we would have no hesitation in recognizing as part of the cognitive process, then that part of the world is (so we claim) part of the cognitive process. (1998, p. 8, italics in original) in other words, the extended mind thesis mandates functional parity between an extended cognitive process and an internally realized cognitive process. both otto with his notebook and inga, another fictional character mentioned in clark and chalmers’ (1998) thought experiment who relies on her biological memory, face the same task of navigating to moma. if the only difference between relying on a notebook and relying on biological memory is that one is realized externally and the other internally, then there is no good reason to deny their functional parity. the internet provides an obvious case study for the extended mind thesis. in their original article, clark and chalmers discuss a number of criteria for functional parity (and hence for cognitive extension): reliability, trust, accessibility and past endorsement. there has been considerable debate on whether and under what circumstances these trust and glue conditions apply to the internet (clark & chalmers, 1998; halpin et al., 2010; smart, 2012, 2017). what these approaches have in common is a search for sharp criteria for including parts of the internet in the metaphysics of mind. this metaphysical approach to the extended mind does not do justice to the full scope of extended mind research. sutton (2010) identifies a second wave of extended mind research. this second wave is focused on how internal components bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 6 are complemented by external (environmental) components in giving rise to a cognitive process (menary, 2010a; for discussion, see heersmink, 2018). there is no need for external processes to substitute internal processes, but brains, bodies, and tools can form a heterogeneously assembled process in very different ways. sutton sees second wave extended mind “as more [of] an invitation to give detailed attention to these differences in specific contexts and case studies than a fixed new metaphysics of mind” (sutton, 2010, p. 206). rather than searching for precise criteria for extension based on parity, second wave extended mind operates within a multidimensional space of variation.7 operationalizing complementarity-based extended cognition requires the identification of the relevant, non-trivial reciprocal causal coupling relations between internal and external components (menary, 2006). the notion of ‘reciprocal causal coupling’ originates in early work on dynamical systems theory (beer, 2000; van gelder, 1998). it designates the causal interaction of two or more components that give rise to a cognitive process across time. however, it should be noted that the distinction between internal and external components is mostly drawn for heuristic purposes, given that “the nature of reciprocal coupling makes it difficult to study the components as separate systems because they are continuously influencing and responding to one another” (menary, 2010b, p. 4). while proponents of first wave extended mind have also relied on a notion of reciprocal causal coupling (clark, 2008; clark & chalmers, 1998), second wave theorists have argued that an analysis of reciprocal coupling relations should be supplemented by a careful consideration of the cognitive practices to which these relations contribute (menary, 2010a; sutton, 2010). to summarize, first wave theories hold that minds can extend through functional parity of biological and non-biological processes, second wave theories hold that minds are extended in virtue of complementary biological and non-biological processes. in this paper, we are interested in specifying both the similarities and dissimilarities between extended and non-extended mind-wandering. these (among others functional) differences cannot be captured by an analysis based on functional parity.8 the methodological approach of second wave extended 7in his defense of the parity principle (clark & chalmers, 1998), clark (2007) seems to be sympathetic to a complementarity-based approach. thus, he writes that the parity principle “[…] is a call for sameness of opportunity, such that bio-external elements might turn out to be parts of the machinery of cognition even if their contributions are unlike (perhaps because they are deeply complementary to) those of the biological brain” (clark, 2007, pp. 167–168, italics in original). this assumption is fully developed in complementarity-based second wave extended mind theorizing. the distinction between first and second wave accounts is a way of capturing the development of theorizing about extended cognitive processes. this development, we assume, is characterized by a shift in emphasis from parity to complementarity and corresponding methodological strategies, with clark’s (2007) treatment as a transition point in the literature on the extended mind. 8this point also holds if we understand parity not in terms of a “fine-grained similarity between inner and outer processes”, but as an expression of the “sameness of opportunity” of internal and external components to contribute to cognitive processes (clark, 2007, p. 167). this “sameness of bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 7 mind that emphasizes the complementarity of internal and external components of mental functioning is therefore better suited to support our analysis.9 the question then becomes under what circumstances mobile technologies contribute to extended mind-wandering episodes. the harmony bias identified in the introduction is visible in some of the work on second wave extended mind and the internet. for example, clowes (2019) criticizes approaches to human-technology relations that he deems intrinsically pessimistic (carr, 2011; loh & kanai, 2016). as an alternative, clowes’ proposes to build upon second wave extended mind approaches to show how technologies can give rise to new forms of cognition. clowes writes: we are rapidly building new virtual environments, props and prompts for cognition that structure a vast range of our own cognitive abilities. i believe these new cognitive props are best understood, not as the impact of autonomous technologies upon us, but rather as a vast and partly conscious construction of new embodied and embedded cognitive activities and abilities. (clowes, 2019, p. 270) the risk here is that the one-sided pessimistic view is countered by an equally onesided optimistic view, according to which new digital technologies are, in principle, conducive to cognitive abilities. whether new forms of cognition appear or disappear, are constructed or destructed should not be an a priori commitment of a theory but an open empirical question (aagaard, 2021; cecutti et al., 2021). heersmink (2016) and gillett and heersmink (2019) provide a more nuanced analysis of the cognitive effects of digital technologies. heersmink (2016) focuses on how the ubiquity of external information changes memory practices. there is very little currently available empirical work in cognitive psychology, he concludes, on which tech-pessimists base their claims that memory is deteriorating. moreover, the empirical work that is cited (i.e., sparrow et al., 2011) investigates the performance on memory tasks when facts are stored in folders on a computer in the lab. it is not obvious that findings from such an artificial setting say anything about the cognitive effects of internet use in the wild. gillett and heersmink (2019) analyse how gps-based navigation systems transform navigation and wayfinding. they conclude that gps devices “undermine the agent’s development of other skills”, “promote route knowledge more than survey knowledge” and “do not scaffold the agent to make autonomous decisions about how to solve wayfinding problems” (gillett & heersmink, 2019, p. 45). they propose to change both the gps tools themselves as well as the epistemic practices in which they are embedded. opportunity” assumption does not help us specify how external components contribute to cognitive processes and how extended and non-extended processes differ from each other. 9this means that we will not engage in any metaphysical debates about the conditions under which a smartphone (or any other environmental resource) should count as part of an agent’s cognitive system (adams & aizawa, 2001). we are interested in the epistemological question of how second wave extended mind can help us understand the complementarity of internal and external components in bringing about specific cases of mind-wandering. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 8 in these studies, the normative implications do not follow from the conceptual framework itself, but from a careful study of the phenomena. for the purposes of our paper, these studies are a step in the right direction, but still miss out on an important aspect of human cognition. they focus on the completion of cognitive tasks such as wayfinding and navigation (gillett & heersmink, 2019) and memory (heersmink, 2016). more generally, slaby characterizes this bias in the literature on the extended mind as the (implicit) endorsement of a “user/resource model” (slaby, 2016, p. 5): baseline mentality in many of the example cases under discussion is that of a fully conscious individual cognizer (“user”) who sets about pursuing a well-defined task through intentional employment of a piece of equipment or exploitation of an environmental structure (“resource”). the focus on individual cognizers who intentionally pursue a well-defined task might be one reason why, as far as we are aware, no study on extended cognition has addressed the question how the internet and digital technologies are transforming task-unrelated cognition. 3 mind-wandering: state of research mind-wandering is a relatively recent topic of study in psychology (smallwood & schooler, 2006). perhaps for this reason, there has been an active debate on what exactly qualifies as mind-wandering and what characterizes mindwandering episodes. initially, research on mind-wandering started with the assumption that the target phenomenon is characterized by task-unrelatedness and stimulus-independence (e.g., smallwood & schooler, 2015). other aspects that have been discussed in the literature include the absence of attentional guidance, a lack of intention, and a lack of meta-awareness (e.g., irving & glasser, 2020). in what follows, we will first summarize previous research on these aspects of mind-wandering. our aim is not to provide an exhaustive review of the available literature, but to identify and discuss potential characteristics of mind-wandering episodes. in a second step, we will explore the option that a family-resemblance approach to mind-wandering can help integrate research on the various aspects of mind-wandering (seli, kane, smallwood, et al., 2018). it has been suggested that a mental episode qualifies as mind-wandering if it is task-unrelated (smallwood & schooler, 2006, 2015). in early behavioral studies on mind-wandering, participants were given a primary task, for example reading for comprehension (reichle et al., 2010; schooler et al., 2004; smallwood et al., 2008). any mental activity that was self-reportedly not related to task completion was classified as a mind-wandering episode (christoff et al., 2016). this presupposes that it is possible to identify a single task at any given time – both from a bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 9 subjective first-person perspective and a scientific third-person perspective.10 this assumption is problematic for at least two reasons. first, it is not clear whether there always is a primary task (murray et al., 2020). it seems possible to mindwander in situations where no immediate cognitive task is available (murray et al., 2020), such as when waiting in line. second, it is conceivable that an agent is in the process of completing multiple tasks at the same time (metzinger, 2018), which does not necessarily amount to mind-wandering. for these reasons, irving (2016) proposes that a mental episode is task-unrelated iff it is unrelated to any of the agent’s tasks. this excludes multi-tasking, since the agent’s thoughts are unrelated to some, but not any of her tasks, but includes mind-wandering in the absence of a task (e.g., during rest). as we will discuss in section 4, this does not preclude the possibility that there are relevant distinctions between mind-wandering in the presence and absence of a task. stimulus-independence has been identified as another aspect that can help identify a mind-wandering episode (konishi & smallwood, 2016; schooler et al., 2011; smallwood & schooler, 2015). on this view, mind-wandering would occur independently from current perceptual stimulation. furthermore, it has been argued that mind-wandering episodes are characterized by perceptual decoupling (broadway et al., 2015; konishi & smallwood, 2016; schooler et al., 2011; smallwood, 2011; smallwood & schooler, 2015). according to the decoupling hypothesis, “[a]ttention is directed inwards during mind wandering; thus, representations of the external environment should be superficial” (smallwood & schooler, 2006, p. 947). the proposal that mind-wandering is stimulus-independent and therefore perceptually decoupled from the local environment has been criticized for at least two reasons. first, it has been noted that any mental episode is never sufficiently stimulus-independent, as it is always dependent upon an ongoing stream of exteroceptive and interoceptive stimulations (metzinger, 2018). second, mind-wandering “[…] can also be externally oriented towards stimuli in the current perceptual environment” (christoff et al., 2016, p. 5; see also irving & glasser, 2020; seli, kane, smallwood, et al., 2018; for empirical evidence, see mills et al., 2018). for example, upon seeing your own image during an online conversation, your thoughts might drift away from the conversation to the need of getting a haircut. the assumption that mind-wandering can be oriented both inwards as well as towards the environment is part of the folk-psychological conception of mind-wandering (irving et al., 2020, study 2). given that the notions of ‘stimulus-dependence’ and ‘stimulus-independence’ are problematic, we will talk about ‘perceptual coupling’ and ‘perceptual decoupling’ in the remainder of this paper. note that ‘coupling’ in this context is a graded notion that captures reciprocal causal relations of various 10note that this conception of mind-wandering makes a rather ambiguous appeal to the notion of ‘task’. tasks can be construed as self-imposed individual projects partially constituted by the agent’s current concerns, or as other-imposed public settings independent of an agent’s current concerns. morrison et al. (2019) find that what a subject takes to be their task often does not align with the task the experimenter takes the subject to be doing. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 10 strengths between internal and external components. just as a mental episode is never fully stimulus-independent, perceptual states and processes are never fully decoupled. for this reason, it would be more accurate to capture the episodes that are addressed by the decoupling hypothesis as ‘minimally perceptually coupled’, rather than ‘perceptually decoupled’. however, in keeping with the psychological literature, we will continue to talk about ‘perceptual decoupling’. as we have pointed out in the previous paragraphs, there are pertinent problems of characterizing mind-wandering exclusively in terms of task-unrelated and perceptually decoupled mental episodes. furthermore, this characterization does not differentiate between directed thinking and attentionally unguided thought and does not capture the temporal and attentional dynamics of mind-wandering. for these reasons, it has been suggested that mind-wandering can be conceptualized as attentionally unguided thought (irving, 2016, 2021; irving & glasser, 2020; irving & thompson, 2018). during mind-wandering episodes, irving (2016) proposes, “[…] the focus of attention drifts unguided from one topic to the next” (p. 563). mindwandering is thereby understood as the absence of attentional guidance, which is defined in the following way (irving, 2016, p. 565): an agent 𝒜 is guided to focus her attention on some information 𝑖 if and only if she has two dispositions: 1. 𝒜 is reliably disposed to focus her attention on 𝑖 and 2. if 𝒜’s attention isn’t focused on 𝑖, she notices, feels discomforted by, and is thereby disposed to correct this fact. during mind-wandering, then, a person can have the disposition to focus their attention on currently relevant information, but does not feel pulled back if her mind were to drift somewhere else. this counterfactual conceptualization of mindwandering as the absence of attentional guidance for mental processes, which is derived from theorizing in the philosophy of action (pacherie, 2008), would help specify the “unstable dynamics” of mind-wandering (irving & thompson, 2018, p. 91): it excludes cases of directed thinking, since they do involve attentional guidance (irving, 2021). this conception of mind-wandering as unguided attention is fully supported by christoff et al.’s (2016) neuroscientific framework, according to which mind-wandering is characterized by “[…] an absence of strong constraints on the contents of each state and on the transitions from one mental state to the other” (christoff et al., 2016, p. 2). the characterization of mind-wandering as a lack of attentional guidance would therefore emphasize the dynamics of mindwandering across time, rather than the ‘content’ of a mind-wandering episode relative to current task demands or available stimuli in the local environment (mills et al., 2018). the idea that all cases of mind-wandering are entirely unguided might seem controversial. when your thoughts drift from reading a book to musing about your next holidays, there is a sense in which an agent feels attracted towards these musings. to accommodate this intuition, watzl (2017) develops a view in which mindbruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 11 wandering “is guided by longer-lasting states like concerns and desires with which the subject does not identify” (p. 135). while irving (2021) sees mind-wandering as unguided, he allows for a graded notion of guidedness. for example, a brainstorming session might be very loosely guided by the theme: the agent will only feel pulled back when drifting far away from the theme of the session. this means that there probably is no sharp line to be drawn between directed thinking and mind-wandering. in addition to task-unrelatedness, perceptual decoupling, and unguided attention, philosophical and psychological research has explored the role of intention for mind-wandering. initially, a lack of intention has been identified as a defining feature of mind-wandering episodes. on this view, mind-wandering episodes are not initiated or maintained intentionally (smallwood & schooler, 2006). however, more recent empirical evidence suggests that mind-wandering can be intentional or unintentional (seli, risko, & smilek, 2016; seli, risko, smilek, & schacter, 2016).11 furthermore, several philosophers have recently argued for the possibility of intentional mind-wandering (arango-muñoz & bermúdez, 2021; irving, 2021).12 first, arango-muñoz and bermúdez (2021) suggest that “intentional mindwandering is an intentional omission of control or guidance over one’s thoughts” (p. 7735). second, irving (2021) distinguishes between the intentional initiation and the intentional maintenance of mind-wandering. in the former case, the onset of a mind-wandering episode is intentional in the sense that “[y]our choice to initiate mind-wandering is intentional, but your attention then wanders unguided” (irving, 2021, p. 637). in the latter case, a mind-wandering episode is intentionally maintained if one exerts second-order control over the first-order unguidedness of one’s thoughts: “[…] one can actively suppress the guidance of attentional focus and cultivate a wandering mode of attention” (irving, 2021, p. 638). both suggestions are consistent with the assumption that mind-wandering is characterised by unguided attention (see above). finally, researchers have explored the role of meta-awareness during mindwandering. while some have associated mind-wandering with a transient lack of meta-awareness (metzinger, 2018; smallwood & schooler, 2006), others have as11it should be noted that the psychological concept of ‘intention’ as it is frequently employed in research on mind-wandering captures the influence of cognitive control and goal-directedness on the unfolding of mental processes (see e.g., smallwood & schooler, 2006). this understanding is distinct from the concept of ‘intention’ as it is used in the philosophy of language (e.g., anscombe, 1963) and the philosophy of mind (e.g., bratman, 1984). 12engaging with the initial conception of mind-wandering as task-unrelated thought, according to which mind-wandering is unrelated to a primary task, murray and krasich (2020) have argued that task-unrelatedness and intentionality are incommensurable aspects of mental episodes. according to their puzzle of willful wandering, “an agent cannot intend to have task-unrelated thoughts, as intending to have any thought thereby makes that thought related to one’s task” on logical grounds (murray & krasich, 2020, p. 2). this puzzle can be resolved, murray and krasich (2020) argue, by revising the conception of either task-unrelatedness or intentionality. if irving’s (2016) revised conception of task-unrelated thought in combination with his characterization of mind-wandering as unguided attention is adopted (see above), the puzzle of willful wandering can be avoided. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 12 sumed that mind-wandering can become available to meta-awareness (seli, risko, & smilek, 2016; seli, risko, smilek, & schacter, 2016). a relevant strategy for accommodating cases of meta-awareness and meta-unawareness is to introduce a distinction between tuning out and zoning out (schooler et al., 2004, 2011; smallwood, 2011). in cases of tuning out, cognizers can become temporarily aware that their mind is wandering. in cases of zoning out, cognizers temporarily lack metaawareness of their current engagement in a mind-wandering episode. in the course of a given mind-wandering epsiode, cognizers can alternate between meta-aware and meta-unaware moments. this distinction between tuning out and zoning out is consistent with the view that mind-wandering is characterized by unguided attention (irving, 2016; irving & thompson, 2018). considerations on the relationship between intentionality and meta-awareness have led to the assumption that these are at least partly independent dimensions of mind-wandering (seli et al., 2017). this brief discussion of previous psychological and philosophical research raises the question how these different characteristics ascribed to mind-wandering can be accommodated. recently, it has been suggested that the development of a family-resemblances framework for research on mind-wandering could help analyze the phenomenological and functional diversity and variability of mind-wandering episodes (seli, kane, smallwood, et al., 2018). the notion of ‘family resemblances’ originates in wittgenstein’s (2009) considerations on the concepts of ‘game’ and ‘number’. just as wittgenstein (2009) assumes that the concept of ‘game’ refers to “a complicated network of similarities overlapping and crisscrossing” (2009, § 66, 36e), seli et al. (2018) suggest that the concept of ‘mind-wandering’ refers to a family of more or less similar mental episodes. on this view, ‘mind-wandering’ would be “a natural category with graded membership”, which entails that “some exemplars are more prototypical than others” (seli, kane, smallwood, et al., 2018, p. 483). furthermore, just as the concept of ‘game’ has no clear, pre-determined boundaries (wittgenstein, 2009), the concept of ‘mind-wandering’ would be depicted “as a fuzzy-boundaried and heterogeneous construct” (seli, kane, smallwood, et al., 2018, p. 485). for current purposes, we assume that the family-resemblances framework can help identify and specify the members of the mind-wandering family (seli, kane, metzinger, et al., 2018).13 episodes of mind-wandering will share at least some of the features we just outlined: task-unrelatedness, perceptual decoupling, attentional unguidedness, (lack of) intentionality and (lack of) meta-awareness. 13note, however, that christoff et al. (2018) and irving and glasser (2020) have raised concerns about the family-resemblance framework and its ability to arrive at a sufficiently specific conceptualization of ‘mind-wandering’. however, in reply to christof’s et al. (2018) criticism, according to which unguided, relatively unconstrained thought does provide “an essential, defining feature” of mind-wandering (p. 957), seli, kane, metzinger et al. (2018) point out that “[a] ‘relative lack of constraint’ is insufficiently specific to allow one to distinguish mind-wandering from other thoughts” (p. 959). entering this debate is beyond the scope of the present paper. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 13 adopting a family-resemblance approach to mind-wandering also has implications for the normative assessment of the cognitive and affective consequences or the costs and benefits of mind-wandering (mooneyham & schooler, 2013; schooler et al., 2014). according to fox and christoff (2014), research on mind-wandering has emphasized its detrimental impact on cognitive abilities: “in contrast to the more desirable pursuit of ‘rational’ thought, mw [mind-wandering] is often portrayed as undesirable – a wasteful mental diversion and potentially dangerous distraction” (p. 299). in support of this assessment of the negative consequences of mind-wandering, research has pointed out that mind-wandering impedes reading comprehension (franklin et al., 2011; schooler et al., 2004; uzzaman & joordens, 2011), performance in tests of working memory span and general intelligence (mrazek et al., 2012), knowledge acquisition in classroom and online learning environments (szpunar et al., 2013), risk-aversive driving behavior (yanko & spalek, 2014), and positively valenced emotional experiences (killingsworth & gilbert, 2010). however, it should also be noted that several beneficial effects of mind-wandering episodes have been identified, for example relief from boredom (mooneyham & schooler, 2013), as well as a positive impact on autobiographical planning (baird et al., 2011), self-insight (d’argembeau, 2018), creative incubation (baird et al., 2012), and dishabituation during learning tasks (schooler et al., 2014). if a family-resemblances approach to mind-wandering is adopted, a more nuanced normative assessment of the cognitive and affective impact of mindwandering episodes comes into view. for example, the idea that mind-wandering can provide a beneficial relief from boredom might be true for a mind-wandering episode that involves intentionally initiated, unguided thinking (cf. irving, 2021, p. 637), but might not be true for a mind-wandering episode that lacks intentionality and occurs in the presence of another task. as we will suggest in the next section, the family-resemblance framework also allows for the possibility that mental episodes that are not perceptually decoupled qualify as proper members of the ‘mind-wandering family’ and warrant, as such, a normative evaluation. 4 extended mind-wandering in this section, we will propose that certain episodes of habitual, diversionary smartphone use can be conceptualized as extended mind-wandering and qualify as proper members of the mind-wandering family. this proposal rests on the assumption that an external (environmental) resource, i.e., a smartphone displaying a social media or news feed, can be a proper component of a dynamically unfolding mind-wandering episode, thereby complementing internal components (menary, 2010a; sutton, 2010). crucially, this assumption requires a reconsideration of the view that mind-wandering is characterized by perceptual decoupling (broadway et al., 2015; konishi & smallwood, 2016; schooler et al., 2011; smallwood, 2011; smallwood & schooler, 2015). as already noted in section 2, christoff et al. (2016), irving and glasser (2020), and seli et al. (2018) have suggested that at least some bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 14 mind-wandering episodes are functionally specified by an occurrent coupling relation between a cognizer and relevant parts of the local environment. we follow these researchers by assuming that perceptually coupled mental episodes can be proper members of the mind-wandering family. this assumption helps connect research on mind-wandering and second wave extended mind: in cases of extended mind-wandering, or so we argue, internal components are complemented by an external component in virtue of a reciprocal causal coupling relation holding between them (see section 2). in the case of habitual, diversionary smartphone use, internal components influence and are influenced by the perceptual input provided by the user interface as a result of motor action. perceptual coupling, then, would be a special case of reciprocal causal coupling. at first glance, there is a discrepancy between ‘perceptual coupling’ as used in the mind-wandering literature and ‘reciprocal causal coupling’ in the extended mind literature. as a limiting case, perceptual coupling might be a case of onedirectional causal coupling (i.e. passively looking at a static object). perceptual coupling of this sort would, under standard assumptions, not qualify as sufficient for extension, but would contribute to ’embedded’ mind-wandering. the cases of habitual smartphone use under consideration, however, are characterized by active sensorimotor coupling (i.e., swiping and scrolling) between user and device, which does qualify as reciprocal causal coupling.14 episodes of habitual and diversionary smartphone use are task-unrelated just in case they are unrelated to any task. extended mind-wandering can be taskunrelated because there is no cognitive task that the user would be concurrently engaged with (e.g., during rest), or because the user is engaged in a cognitive task, but the smartphone use does not relate to that task. we propose to call the first kind of task-unrelatedness task-absent extended mind-wandering and the second kind of task-unrelatedness task-present extended mind-wandering. the two kinds of extended mind-wandering are exemplified in the examples mentioned in the introduction. robert catches himself scrolling while trying to listen to a lecture (example 1). his extended mind-wandering is task-unrelated in a situation of taskpresence, because there is a task he takes himself to be doing, but his scrolling is unrelated to that task. amanda is standing in line for coffee and habitually checks a number of apps (example 2). her extended mind-wandering is task-unrelated in a situation of task-absence, because there is no cognitive task she is attempting to complete. the distinction between task-present and task-absent extended mindwandering is important when considering the normative implications of extended mind-wandering (section 5). if we adopt irving’s (2016) conceptualization of task-unrelatedness, according to which thoughts are task-unrelated iff they are unrelated to any of the agent’s tasks, and apply this conceptualisation to cases of extended mind-wandering, we will arrive at the assumption that these cases can be intentional or unintentional. 14in what follows, the cases of perceptual coupling we consider satisfy the conditions of reciprocal causal coupling. this allows us to integrate research on mind-wandering and the extended mind. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 15 the consideration of unintentional extended mind-wandering would be consistent with marty-dugas’ et al. (2018) operationalization of ritualistic smartphone use as absent-minded behavior. this kind of mind-wandering would be characterized by a lack of intention (either at the onset or in the course of the relevant mental episode) and a temporary loss of attentional guidance. applying irving’s (2021) and arangomuñoz and bermúdez’s (2021) analysis of intentional mind-wandering to extended cases, we can describe some cases of habitual and diversionary smartphone use as the intentional omission of attentional guidance. intentionally picking up one’s smartphone to scroll aimlessly through a social media feed would be an example of the intentional initiation of an extended mind-wandering episode. an example of the intentional maintenance of an extended mind-wandering episode would be to swiftly exert cognitive meta-control to return to a first-order process of attentional unguidedness when one’s attention gets captivated by a particular online content. episodes of smartphone use that are task-unrelated and intentional or unintentional can be conceptualized as attentionally unguided. recall from section 3 that “[…] the focus of attention drifts unguided from one topic to the next” during mind-wandering episodes (irving, 2016, p. 560). this characterization would be consistent with our assumption that perceptual decoupling is not a necessary condition for the classification of a mental episode as a case of mind-wandering. in cases of extended mind-wandering characterized by unguided attention, the focus of attention shifts and drifts dynamically without any robust endogenous constraints (christoff et al., 2016). episodes of extended mind-wandering characterized by unguided attention, we suggest, can either be cases of tuning out (extended mind-wandering with metaawareness) or zoning out (extended mind-wandering without meta-awareness). just as episodes of non-extended mind-wandering can occur with or without metaawareness (irving & thompson, 2018), episodes of extended mind-wandering may or may not become temporarily available to meta-awareness. for example, while attempting to understand a lecturer’s remarks (example 1) or queuing for a cup of coffee (example 2), a person could scroll through a social media or news feed while being meta-aware (tuning out) or without being meta-aware (zoning out) of their unguided smartphone behavior. ultimately, it is an empirical question how frequent extended mind-wandering cases of zoning out and tuning out episodes are, whether there are mind-wandering episodes that are characterized by alternations between phases of zoning out and tuning out, and how these insights can help specify a theoretical assessment of extended mind-wandering characterized by unguided attention. more generally, empirical research could identify the temporal dynamics of extended mind-wandering, as well as the transitions between extended mind-wandering, non-extended mind-wandering, and task-related, attentionally guided (instrumental) smartphone use across time. we have now arrived at a specification of extended mind-wandering that allows us to categorize well-defined cases of endogenously generated, habitual smartphone use as proper members of the mind-wandering family. in contrast to bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 16 non-extended cases of mind-wandering, which have been the focus of previous psychological and philosophical research, cases of extended mind-wandering are characterized by reciprocal causal coupling. this assumption, we have shown above, is consistent with more recent work on mind-wandering. in table 1, we employ seli’s et al. (2018) terminology to analyze a number of exemplars of extended mind-wandering. c ou pl ed u ng ui de d ta sk -p re se nt in te nt io na l m et aaw ar e 1. while trying to listen to a lecture, rebecca has taken her smartphone out of her pocket absentmindedly and suddenly catches herself scrolling through her social media feed. 4 4 4 8 4 2. while trying to listen to a lecture, selma sets out to scroll through her social media feed and is aware of skimming through the most recent posts. 4 4 4 4 4 3. while trying to listen to a lecture, oliver has taken his phone out of his pocket absentmindedly and scrolls through his social media feed without noticing it. 4 4 4 8 8 4. while trying to listen to a lecture, james sets out to scroll through his social media feed and keeps on skimming through the most recent posts without noticing it. 4 4 4 4 8 5. while queuing for coffee, karen has taken her phone out of her pocket absentmindedly and suddenly catches herself scrolling through her social media feed. 4 4 8 8 4 6. while queuing for coffee, peter sets out to scroll through his social media feed and is aware of skimming through the most recent posts. 4 4 8 4 4 7. while queuing for coffee, sam has taken her phone out of her pocket absentmindedly and scrolls through her social media feed without noticing it. 4 4 8 8 8 8. while queuing for coffee, adam sets out to scroll through his social media feed and keeps on skimming through the most recent posts without noticing it. 4 4 8 4 8 table 1: overview of exemplars of extended mind-wandering. all exemplars are perceptually coupled and attentionally unguided. furthermore, all exemplars are unrelated to any task, the only difference lies in the situation-specific task-presence or task-absence. exemplars that are not characterized by intentionality are cases of unintentional extended mind-wandering. furthermore, exemplars that are not characterised by meta-awareness are cases of meta-unaware extended mindwandering. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 17 the systematization of extended mind-wandering provided in table 1 allows us to arrive at a specification of the characteristics of habitual, diversionary smartphone use (hiniker et al., 2016) and of notions such as absent-minded smartphone use (marty-dugas et al., 2018) and smartphone-related inattentiveness (liebherr et al., 2020) (see section 1). at the same time, it leads to a revision of the commonly held assumption that digital technologies such as smartphones cooperatively contribute to the completion of cognitive tasks in most (if not all) cases (see section 2). perceptual input provided by a smartphone, we have argued in this section, can contribute to episodes of extended mind-wandering. it might be objected that the cases of habitual and diversionary smartphone use that we have been considering should not be conceptualized as members of the mind-wandering family. after all, habitual and diversionary smartphone use might seem to be strikingly different from the exemplars of mind-wandering that have been discussed in the psychological and philosophical literature. in reply to this objection, we return to the family resemblances approach, according to which mind-wandering is “a fuzzy-boundaried and heterogeneous construct” that applies to a variety of cases, which show similarities and dissimilarities (seli, kane, smallwood, et al., 2018, p. 485). once we allow for the possibility that perceptual decoupling need not be a necessary characteristic of mind-wandering episodes, as suggested by christoff et al. (2016), irving & glasser (2020), and seli, kane, smallwood, et al. (2018), it becomes conceivable that particular engagements with external components, such as smartphones, non-trivially contribute to the onset and maintenance of a mind-wandering episode. furthermore, as shown above, recent empirical research strongly suggests that habitual and diversionary smartphone use can share key characteristics with seminal cases of mind-wandering (forster & lavie, 2014; hiniker et al., 2016; liebherr et al., 2020; marty-dugas et al., 2018). for these reasons, we submit that the differences between particular exemplars of extended and non-extended mind-wandering are smaller than the differences within the heterogenous class of non-extended mind-wandering. the only identifiable dissimilarity of all cases of extended mind-wandering and some, but not all cases of non-extended mind-wandering, concerns the aspect of perceptual coupling/decoupling. in order to show that this approach is misguided, a critic would need to argue why perceptual decoupling should count, after all, as the necessary condition for the identification of a mental episode as a mind-wandering episode. the point of the extended mind thesis is, of course, that sensorimotor coupling with the environment, in and of itself, does not provide grounds to drive a wedge between mental phenomena (chalmers, 2019). based on these qualifications, we will further strengthen our case for the complementarity of smartphone applications and internal components in the cases of habitual and diversionary smartphone use we have identified above. to this end, we will consider the relationship of non-extended and extended cases of mindwandering in the next section. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 18 5 comparing extended and non-extended mindwandering if the previous analysis holds, extended mind-wandering is a technologically mediated form of mind-wandering. in section 3 we have briefly reviewed the costs and benefits of non-extended forms of mind-wandering. in this section we will explore two questions. first, what kind of evidence would support the thesis that extended mind-wandering tends to replace non-extended mind-wandering? second, to what extent would extended mind-wandering share the costs and benefits of non-extended mind-wandering? to date, it is difficult to directly assess the cognitive and affective costs and benefits of extended mind-wandering. the reason is that empirical research, which directly investigates the phenomenon that we conceptualize as extended mindwandering, is currently missing. yet, in this section we will review some studies that might motivate future research on extended mind-wandering. it is clear that the overall amount of smartphone usage is going up (global mobile consumer survey, 2019), but the statistics do not differentiate between situations in which smartphones are being used (during a lecture or during alone time), whether smartphone use is attentionally guided or unguided, task-related (i.e., instrumental) or task-unrelated (i.e., habitual and diversionary), and whether a smartphone is used unintentionally or intentionally and with or without metaawareness. moreover, both the prevalence as well as the costs and benefits of extended mind-wandering may be subject to inter-individual differences (diefenbach & borrmann, 2019). this section is therefore explorative and is meant to guide future empirical research, rather than giving a definite answer to the question of the normative relationship between extended and non-extended forms of mind-wandering. we will structure our discussion by considering the following two theses, which we derive from our considerations presented above: the replacement thesis: extended mind-wandering competes for the same cognitive resources as non-extended mind-wandering and seems to partially replace non-extended mind-wandering. the functionality thesis: extended mind-wandering shares the costs of nonextended mind wandering, but does not share the benefits, especially concerning self-insight. 5.1 the replacement thesis according to the replacement thesis, extended mind-wandering would partially replace non-extended mind-wandering: as we spend more time extended mindwandering, the amount of time we spend on non-extended mind-wandering would go down. what kind of empirical evidence would be required to support this bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 19 claim? ralph et al. (2020) investigated the trade-off between what they call mediamultitasking and mind-wandering. in one phase of their behavioral experiment, participants performed a simple cognitive task (1-back),15 in the other condition, participants were able to concurrently watch a task-unrelated, optional video. during the experiment, participants were asked whether they were 1) focused on the task, 2) off-task attending to the video or 3) off-task mind-wandering. compared to the non-video condition, participants in the video condition reported being off-task more often, but reported non-extended mind-wandering significantly less. these findings suggest a trade-off between (non-extended) mind-wandering and media use, at least in the presence of a primary task. this study cannot be taken directly as evidence for the replacement thesis. the reason is that the study explored the prevalence of task-unrelated cognition in the presence versus absence of the opportunity to engage in media multitasking. however, we believe that the study design used by ralph et al. (2020) could be modified to investigate the relative frequency of extended and non-extended mindwandering. for example, in one condition, participants could be offered the opportunity to scroll through a social media feed (rather than watching a video). in the other condition, participants would be denied access to any kind of smartphone use. subjects might then be asked whether or not they were guiding their attention to the task (1-back). participants who report that their attention was temporarily unguided could then be administered further questions. for example, were they initiating or maintaining their task-unrelated mental episodes intentionally or unintentionally? were they meta-aware that they were engaging in task-unrelated cognition? a trade-off between extended and non-extended mind-wandering can also be motivated through computational studies. for example, taatgen’s et al. (2021) competition model shows that there is no principled difference between task-unrelated mind-wandering and external distraction. task-unrelated mental episodes can be triggered by the availability of mental resources and the “presence of distracting stimuli in the environment and internally in the mind” (2021, p. 86). consequently, if relevant environmental stimuli are available, extended mind-wandering might outcompete non-extended mind-wandering. could this trade-off also exist in the absence of any relevant task? while we are not aware of any empirical work that directly addresses this question, we would like to discuss two studies that might motivate future research. diefenbach and borrmann (2019) investigate the hypothesis that smartphones can act as a pacifier during periods of alone time. that is to say, smartphones can prevent us from having to stay alone with our thoughts. the authors conducted a survey in which they asked smartphone users about personality traits (capacity for solitude, need to belong, proneness to boredom), their perception of their smartphone as an attachment object, their smartphone use during alone time, and their self-reflection 15in the 1-back task, participants had to judge whether the currently presented letter matched the previously presented letter. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 20 and self-insight. the authors found that users with a higher proneness to boredom, a higher need to belong, and a lower capacity for solitude also report higher smartphone use during alone time. although this is a correlational study, it is likely that personality traits influence smartphone use and not the other way around. the authors hypothesize that smartphone users engage with their phone to steer away from negative emotions. note, however, that this might not be a successful strategy: sagioglou & greitemeyer (2014) report a consistent “affective forecasting error” in which subjects expect to feel better after using facebook, but report more negatively valenced emotional experiences the longer they use facebook. diefenbach and borrmann (2019) do not investigate the quality of smartphone use during alone time. such smartphone use could in principle be highly directed and not of the attentional unguided mind-wandering variety that we have explored in this paper. the quality of smartphone use in relationship to boredom is explored in some detail by aranda and baig (2018) in a small ethnographic study. users report: “when i’m bored, i keep going into my news app and tapping the same article over and over, hoping for a new story to read.” (participant quote 1, aranda & baig, 2018, p. 19:4) “without my phone, what would i do— just stare out the window⁈” (participant quote 2, aranda & baig, 2018, p. 19:3) “i spent 1.5 hours on [a social networking site]. i was appalled at myself. i hate when i spend time just scrolling and scrolling… it’s all mind-numbing, and i don’t benefit from any of it.” (participant quote 3, aranda & baig, 2018, p. 19:3) based on these and other reports, aranda and baig (2018) identify two behavioral cycles. the first pertains to a user’s checking habits to resist boredom, the second pertains to a shared expectation to be constantly available to others. while aranda and baig (2018) did not directly investigate the replacement of non-extended by extended mind-wandering during alone time, the cited self-reports may be interpreted as phenomenological descriptions of extended mind-wandering. specifically, participant quotes 1 and 2 might imply that non-extended mind-wandering is replaced by extended mind-wandering, where habitual and diversionary smartphone use serves as a means of overcoming boredom. these identified behavioral cycles are consistent with diefenbach and borrmann’s (2019) findings that boredom and need to belong are drivers of smartphone use. they lend support to the idea that habitual and diversionary smartphone use is affectively motivated in the absence of any task. the interpretation of the smartphone as a pacifier to prevent having to stay alone with your thoughts is backed up by research that shows that virtually any activity is preferred over having to stay with your thoughts. wilson et al. (2014) asked participants to stay alone in a room for 15 minutes and entertain themselves with their thoughts. participants had the opportunity to administer themselves an electric shock that they had previbruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 21 ously experienced and had said they would pay money to avoid. during this time, 67% of men and 25% of women self-administered a shock. the study suggests that participants find it difficult and unrewarding to stay alone with their thoughts. all in all, studies that directly investigate the trade-off between extended and non-extended mind-wandering are currently lacking. however, a number of studies do investigate phenomena closely related to extended mind-wandering. these studies provide some circumstantial evidence for a replacement hypothesis, according to which extended mind-wandering replaces non-extended mind-wandering in well-defined situational contexts. we have provided some suggestions to adapt these studies to directly investigate extended mind-wandering in future empirical research. 5.2 the functionality thesis to what extent does extended mind-wandering share the costs and benefits of non-extended mind-wandering? again, no studies have operationalized extended mind-wandering in the way developed in this paper, but a number of studies have explored absent-minded smartphone use (marty-dugas et al., 2018) or smartphonerelated inattentiveness (liebherr et al., 2020), as mentioned in the introduction. there is evidence that both forms of mind-wandering interfere with task performance in a number of domains. for example, both mind-wandering (galéra et al., 2012; yanko & spalek, 2014) and phone use (hancock et al., 2003) impair driving performance. learning in a classroom setting was found to be impaired by both digital media use (hembrooke & gay, 2003; spence et al., 2020; wood et al., 2012) and non-extended mind-wandering (risko et al., 2013; wammes et al., 2016). moreover, reading comprehension was found to be impaired by intermittent popups (liu & gu, 2020) and non-extended mind-wandering (feng et al., 2013; mcvay & kane, 2012; smallwood, 2011). the functional similarity of extended and nonextended mind-wandering has been implied by mind-wandering researchers as well. smallwood and schooler (2015, p. 510) suggest: just as the use of smartphones is valuable to society yet can cause automobile accidents (nemme & white, 2010), the rich imaginative mental life that [stimulus-independent] mind wandering affords is valuable when it is used correctly but counterproductive when it is not. the empirical question then would be whether extended mind-wandering and nonextended mind-wandering share the conditions under which they are deemed appropriate or inappropriate, productive or counter-productive. in section 3, we briefly summarized the benefits of non-extended mindwandering, such as relief from boredom (mooneyham & schooler, 2013), a positive impact on autobiographical planning (baird et al., 2011), self-insight (d’argembeau, 2018), creative incubation (baird et al., 2012), and dishabituation during learning tasks (schooler et al., 2014). in particular, d’argembeau (2018) argues that because non-extended mind-wandering tends to be about self-related bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 22 information – such as personal experiences, anticipation of plans, evaluation of one’s life situation and social relationships – it might facilitate a sense of personal identity and be conducive to long-term goals. diefenbach and borrmann (2019) correlate smartphone use during alone time and measures of self-reflection and self-insight. their study shows no significant correlation between smartphone use and engaging in self-reflection, while they do find a negative correlation between smartphone use and judgments of self-insight. one speculative interpretation of these findings is that although smartphone use might initiate self-reflection, “the external oriented type of reflection might not correspond to ‘honest’ self-reflection and thus no self-insight” (diefenbach & borrmann, 2019, p. 10). according to psychoanalyst and tech critic sherry turkle (2016), the possibility for continuous online connection prevents self-reflection, the engagement with negatively valenced emotions, and the negotiation and resolution of social conflicts in direct, real-time, face-to-face conversation with other people. already in earlier work, turkle (2008) has pointed out that the frequent use of smartphones and other digital technologies has a detrimental effect on our sense of self and our social relationships (p. 127): […] what is not being cultivated is the ability to be alone, to reflect on and contain one’s emotions. the anxiety that teens report when they are without their cell phones or their link to the internet may not speak so much to missing the easy sociability with others but of missing the self that is constituted in these relationships. following this suggestion, the personality factors diefenbach and borrmann (2019) identify as predictors for smartphone use (high need to belong, high propensity to boredom, low capacity to be alone) might themselves be shaped by the availability of smartphones (especially during adolescence). importantly then, the costs and benefits of extended mind-wandering are not limited to their impact on concurrent task performance, or to opportunities for concurrent non-extended mind-wandering. instead, we need to take a diachronic perspective to understand how extended mind-wandering habits form and selfstabilize over time and how these habits co-shape or interfere with practices of self-insight, self-regulation, and face-to-face social relationships. the interesting question here is not how one particular habit is formed (e.g., eyal, 2014), but how multiple habits are linked with one another (ramírez-vizcaya & froese, 2019) and how self-control can be developed in technological contexts (cecutti et al., 2021). to summarize: in this sub-section, we have investigated the functionality thesis, according to which extended mind-wandering would share the costs but not the benefits with non-extended mind-wandering. we discussed the possibility that there might be an overlap in the costs of both forms of mind-wandering in a number of different contexts. this possibility should be explored directly in future empirical research. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 23 6 extended mind-wandering in the attention economy in order to fully understand the ramifications of extended mind-wandering, we need to situate this phenomenon in the context of the attention economy. the notion of the ‘attention economy’ captures the economic system in which human attention is the scarce commodity (hendricks & vestergaard, 2019; williams, 2018; zuboff, 2019). the incentive for businesses operating within this economic system is to optimize a user’s engagement with their product and to thrive in the competition for their attention. a steady philosophical literature is emerging on the ethical and political implications of the attention economy (castro & pham, 2020; hanin, 2020; williams, 2018). a central assumption in the literature is that a quantitative and qualitative shift occurred in how humans relate to their world. the omnipresence of digital technologies connected to the internet makes access to information extremely cheap. information, in this literature, is anything that can consume attention. information abundance leads to attention scarcity (simon, 1971). williams (2018) conceives of information abundance in terms of a functional threshold above which persons lose control over their attentional processes (for a similar view, see gazzaley & rosen, 2016). similarly, eyal (2014) writes how “[t]he convergence of access, data, and speed is making the world a more habit-forming place” (p. 14). one of the challenges in theorizing about the cognitive (and affective) effects of the attention economy is that much of our conceptual apparatus (in philosophy and more generally) is based on the assumption of information scarcity, and might therefore be ill-equipped to deal with information abundance (cf. williams, 2018, p. 16). for williams (2018), the defining challenges of contemporary society have less to do with the management of information and more with the management of attention. given that philosophical and empirical research has only started to acknowledge that these technologies give rise to information abundance and attention scarcity, a conceptual, empirically informed framework for the investigation of the epistemological, phenomenological and functional characteristics of habitual smartphone use – and their normative implications – is still missing (williams, 2018). habitual technology use is a design feature of social media and other online resources (eyal, 2014). arguably, this makes extended mind-wandering an important manifestation of the effects of the attention economy on its users. in this paper, we have developed a proposal for the conceptualization of habitual smartphone use as a technologically mediated form of mind-wandering. future research should explore these cross-connections between the literature on the attention economy and extended mind-wandering in more detail. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 24 7 concluding remarks in this paper, we have argued that the phenomenon of habitual smartphone use can be fruitfully analyzed as a case of ‘extended mind-wandering’. integrating second wave, complementarity-based work on the extended mind with philosophical and psychological research on mind-wandering, we have proposed that well-defined cases of habitual, diversionary smartphone use can be understood as technologically mediated, extended forms of mind-wandering. these cases should count as proper members of the mind-wandering family (seli, kane, smallwood, et al., 2018). the upshot is a new conceptual framework for the theoretical and empirical investigation of habitual, task-unrelated smartphone use. this framework might be applied to other cases of habitual engagements with digital technologies in future research. this framework can help overcome the cognitive task bias and the harmony bias in research on the extended mind that we have identified in the introduction. first, by exploring task-unrelated cognition associated with habitual smartphone use, the cognitive task bias, i.e., the tendency to exclusively focus on the functional contributions of digital technologies (and other environmental resources) to the completion of cognitive tasks, can be overcome. second, our considerations of the costs and benefits of extended mind-wandering can help overcome the harmony bias identified by aagaard (2021). specifically, the functionality thesis developed in section 5 can give rise to a nuanced assessment of the cognitive (and affective) costs of our habitual engagements with smartphones. at first glance, we could have motivated our conceptualization of habitual smartphone use as extended mind-wandering by relying on the parity principle developed by first wave extended mind theorists. that is, we could have rephrased clark and chalmers (1998, p. 8) parity principle in the following way: if, as we were mind-wandering, a part of the world functions as a process which, were it done in the head, we would have no hesitation in recognizing as part of the mind-wandering episode, then that part of the world is (so we claim) part of the mind-wandering episode. it would then have been sufficient to show that cases of unmediated and mediated task-unrelated cognition are functionally on a par (clark & chalmers, 1998) or would be characterized by a “sameness of opportunity” (clark, 2007, p. 167). however, this strategy would not have allowed us to specify the criteria for classifying cases of habitual, diversionary smartphone use as members of the mind-wandering family. furthermore, this strategy would have made it necessary to engage in a metaphysical discussion about the boundaries of the (wandering) mind at the expense of an in-depth analysis of the descriptive and normative ramifications of extended mind-wandering. by adopting a second wave extended mind perspective with its commitment to the complementarity principle (menary, 2010a; sutton, 2010), we have arrived at a more nuanced description and categorization of cases of habitual smartphone use as specific members of the mind-wandering family. the assumption that smartbruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 25 phone applications can be proper components of mind-wandering episodes, and thereby complement components that are internal to the organism, has put us in the unique position to directly assess the similarities and dissimilarities of nonextended and extended cases of mind-wandering – above and beyond their (alleged) functional parity or sameness of opportunity. we hope that the conceptual framework we have developed in this paper will lead to new empirical investigations on the effects of habitual smartphone use and open up new lines of research on extended cognition, mind-wandering, and their theoretical integration. specifically, we have articulated the replacement thesis and the functionality thesis in section 5, which is based on our theoretical considerations and informed by the limited, currently available empirical evidence. they are both open to further empirical investigation, which could in turn lead to refinements and specifications. we think the conceptual tools provided in this paper might help assess the normative implications of cognitive (and affective) processes in the attention economy as suggested in section 6. hopefully, the proposed conceptual framework can contribute to timely philosophical research on the moral and political roles of digital technologies in our individual and collective lives. acknowledgments we are very grateful to max jones, antonios kaldas, markus pantsar, and john sutton for their constructive written comments on an earlier version of this paper. we would also like to thank caitrin donovan, alexander gillett, richard heersmink, richard menary, thomas metzinger, odysseus stone, graham thomas, and the participants in the workshop on “4e cognition and digital technologies” (centre for agency, values and ethics, macquarie university, june 2021) for their helpful feedback on this work. we would also like to express our deep gratitude to two anonymous reviewers for their careful and constructive criticism, to jennifer windt for her very helpful editorial comments, and to wanja wiese for his excellent editorial support. jelle bruineberg’s work on this paper was funded by a macquarie university research fellowship; regina e. fabry was financially supported by a research fellowship granted by the ruhr university bochum and an arc discovery early career research award (de210100115). references aagaard, j. (2021). 4e cognition and the dogma of harmony. philosophical psychology, 34(2), 165–181. https://doi.org/10.1 080/09515089.2020.1845640 adams, f., & aizawa, k. (2001). the bounds of cognition. philosophical psychology, 14(1), 43–64. https://doi.org/10.1080/09 515080120033571 anscombe, g. e. m. (1963). intention (2nd ed.). blackwell. aranda, j. h., & baig, s. (2018). toward” jomo”: the joy of missing out and the freedom of disconnecting. proceedings of the 20th international conference on human-computer interaction with mobile devices and services, 1–8. https://doi.org/10.1145/3229434.3229468 arango-muñoz, s., & bermúdez, j. p. (2021). intentional mind-wandering as intentional omission: the surrealist method. synthese, 199, 7727–7748. https://doi.org/10.1007/s11229-021-03135-2 baird, b., smallwood, j., mrazek, m. d., kam, j. w. y., franklin, m. s., & schooler, j. w. (2012). inspired by distraction: mind wandering facilitates creative incubation. psychological science, 23(10), 1117–1122. https://doi.org/10.1177/0956797612446024 bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1080/09515089.2020.1845640 https://doi.org/10.1080/09515089.2020.1845640 https://doi.org/10.1080/09515080120033571 https://doi.org/10.1080/09515080120033571 https://doi.org/10.1145/3229434.3229468 https://doi.org/10.1007/s11229-021-03135-2 https://doi.org/10.1177/0956797612446024 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 26 baird, b., smallwood, j., & schooler, j. w. (2011). back to the future: autobiographical planning and the functionality of mind-wandering. consciousness and cognition, 20(4), 1604–1611. https://doi.org/https://doi.org/10.1016/j.concog.2011.08.007 beer, r. d. (2000). dynamical approaches to cognitive science. trends in cognitive sciences, 4(3), 91–99. https://doi.org/10.1016/s1364-6613(99)01440-0 bratman, m. (1984). two faces of intention. the philosophical review, 93(3), 375–405. https://doi.org/10.2307/2184542 broadway, j. m., franklin, m. s., & schooler, j. w. (2015). early event-related brain potentials and hemispheric asymmetries reveal mind-wandering while reading and predict comprehension. biological psychology, 107, 31–43. https://doi.org/ 10.1016/j.biopsycho.2015.02.009 carr, n. (2011). the shallows: what the internet is doing to our brains. w. w. norton & company. castro, c., & pham, a. k. (2020). is the attention economy noxious? philosopher’s imprint, 20(17). http://hdl.handle.net/2027/spo.3521354.0020.017 cecutti, l., chemero, a., & lee, s. w. s. (2021). technology may change cognition without necessarily harming it. nature human behaviour, 1–3. https://doi.org/10.1038/s41562-021-01162-0 chalmers, d. (2019). extended cognition and extended consciousness. in m. colombo, e. irvine, & m. stapleton (eds.), andy clark and his critics (pp. 9–20). oxford university press. christoff, k., irving, z. c., fox, k. c. r., spreng, r. n., & andrews-hanna, j. r. (2016). mind-wandering as spontaneous thought: a dynamic framework. nature reviews neuroscience, 17, 718–731. https://doi.org/10.1038/nrn.2016.113 clark, a. (2008). supersizing the mind: embodiment, action, and cognitive extension. oxford university press, usa. clark, a. (2007). curing cognitive hiccups: a defense of the extended mind. the journal of philosophy, 104(4), 163–192. https://www.jstor.org/stable/20620012 clark, a., & chalmers, d. (1998). the extended mind. analysis, 58(1), 7–19. https://www.jstor.org/stable/3328150 clowes, r. w. (2019). screen reading and the creation of new cognitive ecologies. ai & society, 34(4), 705–720. https://doi.org/10.1007/s00146-017-0785-5 d’argembeau, a. (2018). mind-wandering and self-referential thought. in k. c. r. fox & k. christoff (eds.), the oxford handbook of spontaneous thought: mind wandering, creativity, and dreaming (pp. 181–191). oxford university press. diefenbach, s., & borrmann, k. (2019). the smartphone as a pacifier and its consequences: young adults’ smartphone usage in moments of solitude and correlations to self-reflection. proceedings of the 2019 chi conference on human factors in computing systems, 1–14. https://doi.org/10.1145/3290605.3300536 eyal, n. (2014). hooked: how to build habit-forming products. penguin. feng, s., d’mello, s., & graesser, a. c. (2013). mind wandering while reading easy and difficult texts. psychonomic bulletin & review, 20(3), 586–592. https://doi.org/10.3758/s13423-012-0367-y forster, s., & lavie, n. (2014). distracted by your mind? individual differences in distractibility predict mind wandering. journal of experimental psychology: learning, memory, and cognition, 40(1), 251–260. https://doi.org/10.1037/a0034108 fox, k. c. r., & christoff, k. (2014). metacognitive facilitation of spontaneous thought processes: when metacognition helps the wandering mind find its way. in s. m. fleming & c. d. frith (eds.), the cognitive neuroscience of metacognition (pp. 293–319). springer. franklin, m. s., smallwood, j., & schooler, j. w. (2011). catching the mind in flight: using behavioral indices to detect mindless reading in real time. psychonomic bulletin & review, 18(5), 992–997. https://doi.org/doi.org/10.3758/s13423-011-0109-6 galéra, c., orriols, l., m’bailara, k., laborey, m., contrand, b., ribéreau-gayon, r., masson, f., bakiri, s., gabaude, c., fort, a., maury, b., lemercier, c., cours, m., bouvard, m.-p., & lagarde, e. (2012). mind wandering and driving: responsibility case-control study. bmj, 345, e8105. https://doi.org/10.1136/bmj.e8105 gazzaley, a., & rosen, l. d. (2016). the distracted mind: ancient brains in a high-tech world. mit press. gillett, a. j., & heersmink, r. (2019). how navigation systems transform epistemic virtues: knowledge, issues and solutions. cognitive systems research, 56, 36–49. https://doi.org/https://doi.org/10.1016/j.cogsys.2019.03.004 global mobile consumer survey (2019). in deloitte nederland. https://view.deloitte.nl/tmt-gmcs-2020-theme-5.html (retrieved september 15, 2022.) halpin, h., clark, a., & wheeler, m. (2010). towards a philosophy of the web: representation, enaction, collective intelligence. in h. halpin & a. monnin (eds.), philosophical engineering: toward a philosophy of the web (pp. 21–30). john wiley & sons. hancock, p. a., lesch, m., & simmons, l. (2003). the distraction effects of phone use during a crucial driving maneuver. accident analysis & prevention, 35(4), 501–514. https://doi.org/10.1016/s0001-4575(02)00028-3 hanin, m. l. (2020). theorizing digital distraction. philosophy & technology, 1–12. https://doi.org/10.1007/s13347-02000394-8 bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.concog.2011.08.007 https://doi.org/10.1016/s1364-6613(99)01440-0 https://doi.org/10.2307/2184542 https://doi.org/10.1016/j.biopsycho.2015.02.009 https://doi.org/10.1016/j.biopsycho.2015.02.009 http://hdl.handle.net/2027/spo.3521354.0020.017 https://doi.org/10.1038/s41562-021-01162-0 https://doi.org/10.1038/nrn.2016.113 https://www.jstor.org/stable/20620012 https://www.jstor.org/stable/3328150 https://doi.org/10.1007/s00146-017-0785-5 https://doi.org/10.1145/3290605.3300536 https://doi.org/10.3758/s13423-012-0367-y https://doi.org/10.1037/a0034108 https://doi.org/doi.org/10.3758/s13423-011-0109-6 https://doi.org/10.1136/bmj.e8105 https://doi.org/10.1016/j.cogsys.2019.03.004 https://view.deloitte.nl/tmt-gmcs-2020-theme-5.html https://doi.org/10.1016/s0001-4575(02)00028-3 https://doi.org/10.1007/s13347-020-00394-8 https://doi.org/10.1007/s13347-020-00394-8 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 27 hebblewhite, w., & gillett, a. j. (2020). every step you take, we’ll be watching you: nudging and the ramifications of gps technology. ai & society, 1–13. https://doi.org/10.1007/s00146-020-01098-5 heersmink, r. (2016). the internet, cognitive enhancement, and the values of cognition. minds and machines, 26(4), 389–407. https://doi.org/10.1007/s11023-016-9404-3 heersmink, r. (2018). the narrative self, distributed memory, and evocative objects. philosophical studies, 175(8), 1829–1849. https://doi.org/10.1007/s11098-017-0935-0 heersmink, r., & sutton, j. (2020). cognition and the web: extended, transactive, or scaffolded? erkenntnis, 85(1), 139–164. https://doi.org/10.1007/s10670-018-0022-8 hembrooke, h., & gay, g. (2003). the laptop and the lecture: the effects of multitasking in learning environments. journal of computing in higher education, 15(1), 46–64. https://doi.org/10.1007/bf02940852 hendricks, v. f., & vestergaard, m. (2019). reality lost: markets of attention, misinformation and manipulation. springer international publishing. https://doi.org/10.1007/978-3-030-00813-0 hiniker, a., patel, s. n., kohno, t., & kientz, j. a. (2016). why would you do that? predicting the uses and gratifications behind smartphone-usage behaviors. proceedings of the 2016 acm international joint conference on pervasive and ubiquitous computing, 634–645. https://doi.org/10.1145/2971648.2971762 irving, z. c. (2016). mind-wandering is unguided attention: accounting for the “purposeful” wanderer. philosophical studies, 173(2), 547–571. https://doi.org/https://doi.org/10.1007/s11098-015-0506-1 irving, z. c. (2021). drifting and directed minds: the significance of mind-wandering for mental agency. the journal of philosophy, 118(11), 614–644. https://doi.org/10.5840/jphil20211181141 irving, z. c., & glasser, a. (2020). mind-wandering: a philosophical guide. philosophy compass, 15(1), e12644. https://doi.org/https://doi.org/10.1111/phc3.12644 irving, z. c., glasser, a., gopnik, a., pinter, v., & sripada, c. (2020). what does “mind-wandering” mean to the folk? an empirical investigation. cognitive science, 44(10), e12908. https://doi.org/10.1111/cogs.12908 irving, z. c., & thompson, e. (2018). the philosophy of mind-wandering. in k. c. r. fox & k. christoff (eds.), the oxford handbook of spontaneous thought: mind-wandering, creativity, and dreaming. oxford university press. james, w. (1890). the principles of psychology. henry holt. janlert, l.-e., & stolterman, e. (2017). things that keep us busy: the elements of interaction. mit press. killingsworth, m. a., & gilbert, d. t. (2010). a wandering mind is an unhappy mind. science, 330(6006), 932. https://doi.org/10.1126/science.1192439 konishi, m., & smallwood, j. (2016). shadowing the wandering mind: how understanding the mind‐wandering state can inform our appreciation of conscious experience. wiley interdisciplinary reviews: cognitive science, 7(4), 233–246. https://doi.org/https://doi.org/10.1002/wcs.1392 kruger, d. j., duan, a., juhasz, d., phaneuf, c. v., sreenivasa, v., saunders, c. m., heyblom, a. m., sonnega, p. a., day, m. l., & misevich, s. l. (2017). cell phone use latency in a midwestern usa university population. journal of technology in behavioral science, 2(1), 56–59. https://doi.org/10.1007/s41347-017-0012-8 liebherr, m., schubert, p., antons, s., montag, c., & brand, m. (2020). smartphones and attention, curse or blessing? a review on the effects of smartphone usage on attention, inhibition, and working memory. computers in human behavior reports, 1, 100005. https://doi.org/https://doi.org/10.1016/j.chbr.2020.100005 liu, y., & gu, x. (2020). media multitasking, attention, and comprehension: a deep investigation into fragmented reading. educational technology research and development, 68(1), 67–87. https://doi.org/10.1007/s11423-019-09667-2 loh, k. k., & kanai, r. (2016). how has the internet reshaped human cognition? the neuroscientist: a review journal bringing neurobiology, neurology and psychiatry, 22(5), 506–520. https://doi.org/10.1177/1073858415595005 marty-dugas, j., ralph, b. c. w., oakman, j. m., & smilek, d. (2018). the relation between smartphone use and everyday inattention. psychology of consciousness: theory, research, and practice, 5(1), 46–62. https://doi.org/10.1037/cns0000131 mcvay, j. c., & kane, m. j. (2012). why does working memory capacity predict variation in reading comprehension? on the influence of mind wandering and executive attention. journal of experimental psychology: general, 141(2), 302–320. https://doi.org/10.1037/a0025250 menary, r. (2006). attacking the bounds of cognition. philosophical psychology, 19(3), 329–344. https://doi.org/10.1080/09 515080600690557 menary, r. (2010a). cognitive integration and the extended mind. in r. menary (ed.), the extended mind (pp. 227–243). mit press. menary, r. (ed.). (2010b). the extended mind. mit press. merleau-ponty, m. (1945). phenomenology of perception, trans. by colin smith. routledge. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s00146-020-01098-5 https://doi.org/10.1007/s11023-016-9404-3 https://doi.org/10.1007/s11098-017-0935-0 https://doi.org/10.1007/s10670-018-0022-8 https://doi.org/10.1007/bf02940852 https://doi.org/10.1007/978-3-030-00813-0 https://doi.org/10.1145/2971648.2971762 https://doi.org/10.1007/s11098-015-0506-1 https://doi.org/10.5840/jphil20211181141 https://doi.org/10.1111/phc3.12644 https://doi.org/10.1111/cogs.12908 https://doi.org/10.1126/science.1192439 https://doi.org/10.1002/wcs.1392 https://doi.org/10.1007/s41347-017-0012-8 https://doi.org/10.1016/j.chbr.2020.100005 https://doi.org/10.1007/s11423-019-09667-2 https://doi.org/10.1177/1073858415595005 https://doi.org/10.1037/cns0000131 https://doi.org/10.1037/a0025250 https://doi.org/10.1080/09515080600690557 https://doi.org/10.1080/09515080600690557 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 28 metzinger, t. (2018). why is mind wandering interesting for philosophers? in k. c. r. fox & k. christoff (eds.), the oxford handbook of spontaneous thought: mind-wandering, creativity, dreaming and clinical conditions (pp. 97–111). oxford university press. mills, c., raffaelli, q., irving, z. c., stan, d., & christoff, k. (2018). is an off-task mind a freely-moving mind? examining the relationship between different dimensions of thought. consciousness and cognition, 58, 20–33. https://doi.org/10.1016/j.concog.2017.10.003 mooneyham, b. w., & schooler, j. w. (2013). the costs and benefits of mind-wandering: a review. canadian journal of experimental psychology/revue canadienne de psychologie expérimentale, 67(1), 11–18. https://doi.org/10.1037/a0031569 morrison, h., mcbriar, s., powell, h., proudfoot, j., stanley, s., fitzgerald, d., & callard, f. (2019). what is a psychological task? the operational pliability of “task” in psychological laboratory experimentation. engaging science, technology, and society, 5(00), 61–85. https://doi.org/10.17351/ests2019.274 mrazek, m. d., smallwood, j., franklin, m. s., chin, j. m., baird, b., & schooler, j. w. (2012). the role of mind-wandering in measurements of general aptitude. journal of experimental psychology: general, 141(4), 788–798. https://doi.org/10.1 037/a0027968 murray, s., & krasich, k. (2020). can the mind wander intentionally? mind & language, 1–12. https://doi.org/10.1111/mila .12332 murray, s., krasich, k., schooler, j. w., & seli, p. (2020). what’s in a task? complications in the study of the task-unrelatedthought variety of mind wandering. perspectives on psychological science, 15(3), 572–588. https://doi.org/10.1177/1745 691619897966 nemme, h. e., & white, k. m. (2010). texting while driving: psychosocial influences on young people’s texting intentions and behaviour. accident analysis & prevention, 42(4), 1257–1265. https://doi.org/10.1016/j.aap.2010.01.019 pacherie, e. (2008). the phenomenology of action: a conceptual framework. cognition, 107(1), 179–217. https://doi.org/10 .1016/j.cognition.2007.09.003 ralph, b. c. w., smith, a. c., seli, p., & smilek, d. (2020). yearning for distraction: evidence for a trade-off between media multitasking and mind wandering. canadian journal of experimental psychology/revue canadienne de psychologie expérimentale, 74(1), 56–72. https://doi.org/10.1037/cep0000186 ramírez-vizcaya, s., & froese, t. (2019). the enactive approach to habits: new concepts for the cognitive science of bad habits and addiction. frontiers in psychology, 10. https://doi.org/10.3389/fpsyg.2019.00301 reichle, e. d., reineberg, a. e., & schooler, j. w. (2010). eye movements during mindless reading. psychological science, 21(9), 1300–1310. https://doi.org/10.1177/0956797610378686 risko, e. f., buchanan, d., medimorec, s., & kingstone, a. (2013). everyday attention: mind wandering and computer use during lectures. computers & education, 68, 275–283. https://doi.org/10.1016/j.compedu.2013.05.001 sagioglou, c., & greitemeyer, t. (2014). facebook’s emotional consequences: why facebook causes a decrease in mood and why people still use it. computers in human behavior, 35, 359–363. https://doi.org/10.1016/j.chb.2014.03.003 schooler, j. w., mrazek, m. d., franklin, m. s., baird, b., mooneyham, b. w., zedelius, c., & broadway, j. m. (2014). the middle way: finding the balance between mindfulness and mind-wandering. the psychology of learning and motivation, 60, 1–33. https://doi.org/10.1016/b978-0-12-800090-8.00001-9 schooler, j. w., reichle, e. d., & halpern, d. v. (2004). zoning out while reading. in d. t. levin (ed.), thinking and seeing: visual metacognition in adults and children (pp. 203–226). mit press. schooler, j. w., smallwood, j., christoff, k., handy, t. c., reichle, e. d., & sayette, m. a. (2011). meta-awareness, perceptual decoupling and the wandering mind. trends in cognitive sciences. https://doi.org/10.1016/j.tics.2011.05.006 seli, p., kane, m. j., metzinger, t., smallwood, j., schacter, d. l., maillet, d., schooler, j. w., & smilek, d. (2018). the family-resemblances framework for mind-wandering remains well clad. trends in cognitive sciences, 22(11), 959–961. https://doi.org/10.1016/j.tics.2018.07.007 seli, p., kane, m. j., smallwood, j., schacter, d. l., maillet, d., schooler, j. w., & smilek, d. (2018). mind-wandering as a natural kind: a family-resemblances view. trends in cognitive sciences, 22(6), 479–490. https://doi.org/10.1016/j.tics .2018.03.010 seli, p., ralph, b. c. w., risko, e. f., schooler, j. w., schacter, d. l., & smilek, d. (2017). intentionality and meta-awareness of mind wandering: are they one and the same, or distinct dimensions? psychonomic bulletin & review, 24(6), 1808–1818. https://doi.org/10.3758/s13423-017-1249-0 seli, p., risko, e. f., & smilek, d. (2016). on the necessity of distinguishing between unintentional and intentional mind wandering. psychological science, 27(5), 685–691. https://doi.org/10.1177/0956797616634068 seli, p., risko, e. f., smilek, d., & schacter, d. l. (2016). mind-wandering with and without intention. trends in cognitive sciences, 20(8), 605–617. https://doi.org/10.1016/j.tics.2016.05.010 bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.concog.2017.10.003 https://doi.org/10.1037/a0031569 https://doi.org/10.17351/ests2019.274 https://doi.org/10.1037/a0027968 https://doi.org/10.1037/a0027968 https://doi.org/10.1111/mila.12332 https://doi.org/10.1111/mila.12332 https://doi.org/10.1177/1745691619897966 https://doi.org/10.1177/1745691619897966 https://doi.org/10.1016/j.aap.2010.01.019 https://doi.org/10.1016/j.cognition.2007.09.003 https://doi.org/10.1016/j.cognition.2007.09.003 https://doi.org/10.1037/cep0000186 https://doi.org/10.3389/fpsyg.2019.00301 https://doi.org/10.1177/0956797610378686 https://doi.org/10.1016/j.compedu.2013.05.001 https://doi.org/10.1016/j.chb.2014.03.003 https://doi.org/10.1016/b978-0-12-800090-8.00001-9 https://doi.org/10.1016/j.tics.2011.05.006 https://doi.org/10.1016/j.tics.2018.07.007 https://doi.org/10.1016/j.tics.2018.03.010 https://doi.org/10.1016/j.tics.2018.03.010 https://doi.org/10.3758/s13423-017-1249-0 https://doi.org/10.1177/0956797616634068 https://doi.org/10.1016/j.tics.2016.05.010 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org extended mind-wandering 29 simon, h. a. (1971). designing organizations for an information-rich world. in computers, communication, and the public interest. johns hopkins university press. slaby, j. (2016). mind invasion: situated affectivity and the corporate life hack. frontiers in psychology, 7. https://doi.org/10 .3389/fpsyg.2016.00266 smallwood, j. (2011). mind‐wandering while reading: attentional decoupling, mindless reading and the cascade model of inattention. language and linguistics compass, 5(2), 63–77. https://doi.org/10.1111/j.1749-818x.2010.00263.x smallwood, j., mcspadden, m., & schooler, j. w. (2008). when attention matters: the curious incident of the wandering mind. memory & cognition, 36(6), 1144–1150. https://doi.org/10.3758/mc.36.6.1144 smallwood, j., & schooler, j. w. (2006). the restless mind. psychological bulletin, 132(6), 946–958. https://doi.org/10.1037/ 0033-2909.132.6.946 smallwood, j., & schooler, j. w. (2015). the science of mind wandering: empirically navigating the stream of consciousness. annual review of psychology, 66, 487–518. https://doi.org/10.1146/annurev-psych-010814-015331 smart, p. (2012). the web-extended mind. metaphilosophy, 43(4), 446–463. https://doi.org/10.1111/j.1467-9973.2012.01756.x smart, p. (2017). extended cognition and the internet: a review of current issues and controversies. philosophy & technology, 30(3), 357–390. https://doi.org/10.1007/s13347-016-0250-2 sparrow, b., liu, j., & wegner, d. m. (2011). google effects on memory: cognitive consequences of having information at our fingertips. science, 333(6043), 776–778. https://doi.org/10.1126/science.1207745 spence, a., beasley, k., gravenkemper, h., hoefler, a., ngo, a., ortiz, d., & campisi, j. (2020). social media use while listening to new material negatively affects short-term memory in college students. physiology & behavior, 227, 113172. https://doi.org/10.1016/j.physbeh.2020.113172 stothart, c., mitchum, a., & yehnert, c. (2015). the attentional cost of receiving a cell phone notification. journal of experimental psychology: human perception and performance, 41(4), 893–897. https://doi.org/10.1037/xhp0000100 sutton, j. (2010). exograms and interdisciplinarity: history, the extended mind, and the civilizing process. in r. menary (ed.), the extended mind (pp. 189–225). mit press. szpunar, k., moulton, s., & schacter, d. (2013). mind wandering and education: from the classroom to online learning. frontiers in psychology, 4, 495. https://doi.org/10.3389/fpsyg.2013.00495 taatgen, n. a., vugt, m. k. van, daamen, j., katidioti, i., huijser, s., & borst, j. p. (2021). the resource-availability model of distraction and mind-wandering. cognitive systems research, 68, 84–104. https://doi.org/10.1016/j.cogsys.2021.03.001 thornton, b., faires, a., robbins, m., & rollins, e. (2014). the mere presence of a cell phone may be distracting: implications for attention and task performance. social psychology, 45(6), 479–488. https://doi.org/10.1027/1864-9335/a000216 turkle, s. (2008). always-on/always-on-you: the tethered self. in j. katz (ed.), handbook of mobile communication studies (pp. 121–137). the mit press. turkle, s. (2016). reclaiming conversation: the power of talk in a digital age. penguin books. uzzaman, s., & joordens, s. (2011). the eyes know what you are thinking: eye movements as an objective measure of mind wandering. consciousness and cognition, 20(4), 1882–1886. https://doi.org/10.1016/j.concog.2011.09.010 van gelder, t. (1998). the dynamical hypothesis in cognitive science. behavioral and brain sciences, 21(5), 615–628. https: //doi.org/10.1017/s0140525x98001733 wammes, j. d., seli, p., cheyne, j. a., boucher, p. o., & smilek, d. (2016). mind wandering during lectures ii: relation to academic performance. scholarship of teaching and learning in psychology, 2(1), 33–48. https://doi.org/10.1037/stl000 0055 watzl, s. (2017). structuring mind: the nature of attention and how it shapes consciousness. oxford university press. williams, j. (2018). stand out of our light: freedom and resistance in the attention economy. cambridge university press. wilmer, h. h., sherman, l. e., & chein, j. m. (2017). smartphones and cognition: a review of research exploring the links between mobile technology habits and cognitive functioning. frontiers in psychology, 8, 605. https://doi.org/10.3389/ fpsyg.2017.00605 wilson, t. d., reinhard, d. a., westgate, e. c., gilbert, d. t., ellerbeck, n., hahn, c., brown, c. l., & shaked, a. (2014). just think: the challenges of the disengaged mind. science, 345(6192), 75–77. https://doi.org/10.1126/science.1250830 winnick, m. (2016). putting a finger on our phone obsession. https://blog.dscout.com/mobile-touches wittgenstein, l. (2009). philosophische untersuchungen philosophical investigations: the german text, with an english translation by g. e. m. anscombe, p. m. s. hacker and joachim schulte (p. m. s. hacker & j. schulte, eds.; 4th ed.). wiley-blackwell. wood, e., zivcakova, l., gentile, p., archer, k., de pasquale, d., & nosko, a. (2012). examining the impact of off-task multi-tasking with technology on real-time classroom learning. computers & education, 58(1), 365–374. https://doi.org/10.1016/j.compedu.2011.08.029 bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.3389/fpsyg.2016.00266 https://doi.org/10.3389/fpsyg.2016.00266 https://doi.org/10.1111/j.1749-818x.2010.00263.x https://doi.org/10.3758/mc.36.6.1144 https://doi.org/10.1037/0033-2909.132.6.946 https://doi.org/10.1037/0033-2909.132.6.946 https://doi.org/10.1146/annurev-psych-010814-015331 https://doi.org/10.1111/j.1467-9973.2012.01756.x https://doi.org/10.1007/s13347-016-0250-2 https://doi.org/10.1126/science.1207745 https://doi.org/10.1016/j.physbeh.2020.113172 https://doi.org/10.1037/xhp0000100 https://doi.org/10.3389/fpsyg.2013.00495 https://doi.org/10.1016/j.cogsys.2021.03.001 https://doi.org/10.1027/1864-9335/a000216 https://doi.org/10.1016/j.concog.2011.09.010 https://doi.org/10.1017/s0140525x98001733 https://doi.org/10.1017/s0140525x98001733 https://doi.org/10.1037/stl0000055 https://doi.org/10.1037/stl0000055 https://doi.org/10.3389/fpsyg.2017.00605 https://doi.org/10.3389/fpsyg.2017.00605 https://doi.org/10.1126/science.1250830 https://blog.dscout.com/mobile-touches https://doi.org/10.1016/j.compedu.2011.08.029 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jelle bruineberg and regina e. fabry 30 yanko, m. r., & spalek, t. m. (2014). driving with the wandering mind: the effect that mind-wandering has on driving performance. human factors, 56(2), 260–269. https://doi.org/10.1177/0018720813495280 zuboff, s. (2019). the age of surveillance capitalism: the fight for a human future at the new frontier of power. profile books. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. bruineberg, j., & fabry, r. e. (2022). extended mind-wandering. philosophy and the mind sciences, 3, 15. https://doi.org/10.33735/phimisci.2022.9190 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1177/0018720813495280 https://doi.org/10.33735/phimisci.2022.9190 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction extended mind: state of research mind-wandering: state of research extended mind-wandering comparing extended and non-extended mind-wandering the replacement thesis the functionality thesis extended mind-wandering in the attention economy concluding remarks emotions as emergent properties. emotions as emergent properties elena walsha(elenawalsh@gmail.com) abstract in recent decades, affective scientists have begun using concepts and tools from dynamical systems theory (dst) to characterise emotional processes. this article considers how the concept of emergence might be used to develop this approach. emotions are explicated as ‘emergent products’ that diachronically constrain the operations of their parts in virtue of feedback loops (a classical feature of nonlinear dynamical systems). the explication is shown to be broadly consistent with what is sometimes called ‘pattern’ emergence. casting emotions as emergent patterns is shown to shed light on a major conceptual and empirical challenge emotion theorists have faced over the past century: identifying and measuring the presence of emotional episodes (hollenstein & lanteigne, 2014; lindquist et al., 2012), dubbed here the ‘boundary problem’ (colombetti, 2014). in particular, the explication suggests seeing the emotional ‘signatures’ thought to accompany emotional episodes as fragile and context-bound: likely to hold under a relevant class of interventions (woodward, 2007), but not without exception beyond that class. this in turn may suggest a need to significantly revise the statistical methods currently used to measure the presence of emotional ‘signatures’. the casting of emotions as emergent patterns also functions as a further case study supporting the value of ‘pattern emergence’ (winning & bechtel, 2019) as a powerful vehicle for characterising the objects of investigation — compound and context-sensitive — ubiquitous in the biological sciences. keywords constraints ∙ dynamical systems theory ∙ emergence ∙ emotion 1 introduction over the past two decades, affective scientists and emotion theorists from different traditions have begun to describe emotions as ‘emergent’ properties. the cognitive appraisal theorist scherer (2009a, 2009b) has characterised emotions as emergent properties of dynamical systems, and giovanna colombetti (2014, p. 78) has described emotions as ‘emergent organismic patterns’. the psychological constructionist lisa feldman barrett (2006b, 2009, 2014) has identified emotions with emergent phenomena constructed from more primitive psychological ‘building blocks’, while james a. coan (2010; 2015) has outlined an ‘emergent variable’ model of emotion that uses statistical methods to define emergentist theories of emotion. auniversity of wollongong. walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-9313-0968 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org elena walsh 2 the concept of emergence itself has a long and controversial history, first coming to prominence in the hands of the so-called british emergentists — j. s. mill, c. d. broad, and samuel alexander — in the late nineteenth and early-twentieth centuries. following a period of relative neglect during the mid-twentieth century, discussions of emergence have once again become a locus of philosophical attention, motivated in part by increasing deployment of the concept across the sciences. it has been argued that deployment of the concept offers explanatory gains across various scientific disciplines (mitchell, 2012), while others note that an overly liberal use of the concept threatens trivialising the notion (sartenaer, 2016). stated simply, emergence refers to a part-whole relation such that the whole is both determined by its parts, but has some novel features with respect to them. though characterised variously in the philosophical literature (e.g., chalmers, 2008; guay & sartenaer, 2016; humphreys, 2016; winning & bechtel, 2019), it is generally agreed that the following two conditions must be met for a higher-level property of some basal-level process or structure to count as an ‘emergent’ (e) product of its base (b). these conditions demand that e be in some sense determined by b (the ‘determination condition’) whilst at the same novel with respect to b (the ‘novelty condition’). 1 in the emotion literature, emergence talk appears within the dynamical systems approach to emotion (coan, 2010; colombetti, 2014; lewis, 2005; lewis & granic, 2000; meuleman, 2015; scherer, 2009a; walsh, 2021). this approach applies concepts and formal modeling tools from dynamical systems theory (dst) — a branch of mathematics used to describe the behaviour of complex systems that evolve over time — to the explanation of emotional processes. colombetti (2014, p. 58) notes that dynamical approaches in affective science have ‘an undeniably discursive and speculative character’ and, as yet, only advance ‘quite general hypotheses about [the] emergent character [of emotions]’. missing from this literature is an explanation of whether and how emotional phenomena meet the two conditions for emergence that form the focal point for discussing emergence in the philosophical literature. i seek to address this gap here. the article begins by introducing the dynamical approach to emotion (section 2), and then considers how the two conditions for emergence (determination and novelty) might apply to it (section 3). i here adopt a ‘pattern emergence’ approach (humphreys, 2016; winning & bechtel, 2019), upon which the emergent phenomenon of interest is a new pattern appearing in the life of a complex system, rather than a new entity or property. i then propose developing the pattern emergence approach to shed light on one of the major conceptual and empirical challenges emotion theorists have faced over the past century: identifying and measuring the presence of emotional episodes (hollenstein 1here i follow the terminology of guay & sartenaer (2016), but these two conditions fairly represent, i think, generally agreed-upon criteria for emergence. a brief background of approaches to emergence and their relationship to these two conditions is provided in section 2. walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org emotions as emergent properties 3 & lanteigne, 2014; lindquist et al., 2012). following colombetti (2014), i dub this issue the ‘boundary problem’ and discuss it in section 4. i aim to show that a suitably developed conception of emotions as emergent properties can explain the historical difficulty the field of emotion science has faced in resolving the boundary problem. to better understand the problem, consider that emotions appear to occur over discrete periods of time, in opposition to, say, moods, or periods of relative emotional neutrality. 2 how do we know if an emotion is present or not present? the answer generally accepted in emotion theory is that an emotion is present when it’s components co-occur for some temporal period. these components include appraisal, action tendencies, physiological changes, facial and vocal expression, and subjective feeling (hutto et al., 2018; scherer & moors, 2019). for example, an emotional episode of fear might be characterised as a pattern of coordinated changes comprising a threat appraisal, fearful facial affect, elevated sympathetic arousal, and an urge to flee a threatening situation (hollenstein & lanteigne, 2014). the boundary problem arises because the component processes that comprise an emotional episode are involved in ongoing homeostatic processes that aren’t limited to emotion. consider fear. craniofacial muscles produce fearful faces, but are also used for biting, forming sounds, and squinting. sympathetic nervous system arousal occurs in fear, but is also involved in modulating breathing rate, hormone flow, and intestinal motility. in other words, emotional and non-emotional episodes are made up of the same ‘stuf’. but if the building blocks of emotion just are the building blocks of the human organism, and emotional episodes are like waves in a sea of ongoing physiological and psychological fluctuations, how do we tell the two apart? 2the focus in this paper is on emotional episodes that develop across ‘real’ time. by this i mean emotional episodes that typically develop over seconds or milliseconds, but that may last longer. two clarifications can be made here. first, real-time explanation contrasts with developmental or ontogenetic explanation, which in our context refers to the development of an emotional temperament or disposition (such as a tendency to be angry, or fearful, or anxious, etc.) that is set during development and persists throughout the life-span (see greenwood, 2015; sarto-jackson, 2022, for discussion). secondly, exactly how long emotional episodes last is a matter of controversy in the literature. verduyn et al. (verduyn et al., 2015, p. 330) describe the ‘traditional belief’ that emotions are ‘bursts of brief duration, lasting for seconds or a couple of minutes at most’. in contrast, it has also been proposed that emotional episodes might last for several minutes, hours, or even longer (frijda et al., 1991; see frijda, 1994; lewis, 2005; sander et al., 2005; verduyn et al., 2015). finally, it has been proposed that actions motivated by an emotion may occur days after the original emotional episode (scarantino, 2014). the ‘dynamical approach’ (section 2) focuses, in large part, on ‘transient synchronisation’ of emotion component parts thought to occur across shorter periods of time (e.g., milliseconds to minutes). having said that, dynamical approaches also describe the possibility of emotional episodes that last longer (see colombetti, 2014, for discussion). much of the existing data on component synchronisation focuses (for practical reasons) on identifying synchronisation across shorter temporal periods rather than longer ones. there is no in-principle reason that we should not attempt to measure whether component synchronisation occurs across longer periods. walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org elena walsh 4 a possible solution is to differentiate at least some types of emotion (e.g., anger, fear) 3 from ongoing homeostatic processes on the basis of distinctive patterns or ‘signatures’ 4 formed for the duration of the episode. various criteria have been offered to identify such patterns (loaiza, 2021; scarantino, 2015), and how we conceptualise these criteria matters when assessing the relevant bodies of empirical data. the data have been interpreted variously: as providing some support for the existence of emotional signatures, weak support, or no support at all (barrett, 2006a; hollenstein & lanteigne, 2014; lench et al., 2011; lindquist et al., 2012; scarantino, 2015). this is because whether or not the evidence supports the existence of emotional signatures depends on how a pattern or signature is itself defined (loaiza, 2021). having outlined the boundary problem, i propose explicating the claim that emotions are emergent properties with reference to the notion of ‘constraints’ (section 4). the term has its roots in classical physics but has been used more recently to explain mechanisms in biological systems (bechtel, 2017; hooker, 2013; pattee, 1971; winning & bechtel, 2018). constraints limit the freedom of options available to a system as it evolves, just as friction or an inclined plane limit or constrain the speed of a moving object (hooker, 2013). i argue that emotional episodes can be differentiated from ongoing psychophysiological homeostasis during periods of time in which emotion parts appear to constrain each other. i then suggest that if emotions operate via mechanisms of constraint, we need to consider new approaches to measuring emotional synchronisation (section 5), ones that align with a construal of emotions as emergent properties. it is not my purpose to outline exactly what these methods ought to be. instead, i offer a few tentative suggestions and explain why they make sense in context of the explication of emotions as emergent properties provided here. in particular, i explain that understanding emotions with reference to constraints demands seeing tokens of a given emotion type as sharing a family resemblance, 5 rather than being identical across contexts (e.g., not all instances of anger will have an identical ‘signature’ across persons and situations). this is because the causal factors (discussed in section 5) that will continuously influence such a system will prevent any emotional patterns produced from conforming strictly to an idealised type. it is hoped that this approach to construing emotions as emergent properties may inspire new approaches to measuring the presence of emotional ‘signatures’, while also functioning as a new case study supporting the value of ‘pattern emergence’ as a powerful 3theorists generally attempt to find evidence for emotional signatures by targeting ‘basic emotions’. associated with the work of paul ekman, the ‘basic’ emotions of anger, fear, disgust, joy, surprise, and sadness are thought by some emotion researchers to be a product of a phylogenetic history and to appear universally in the human species (see ekman, 1980). 4various terms are used to describe these patterns, e.g., ‘synchronisation’ (scherer, 2009a, 2009b), ‘concordance’ (bulteel et al., 2014), and ‘soft assembly’ (colombetti, 2014). for discussion of the claim that ‘almost all’ models of emotion presuppose the idea of an emotional signature characterised by transient component coordination, see hollenstein & lanteigne (2014). 5in the sense of frequently co-occurring properties, as described for instance by boyd (1991). walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org emotions as emergent properties 5 vehicle for characterising the objects of investigation — compound and contextsensitive — ubiquitous in the biological sciences. 2 dynamical patterns the proposal that emotions are emergent phenomena is by no means common or well-accepted in emotion theory. rather, it features in the work of a small body of theorists with an overarching interest in the temporal dynamics of emotional episodes as they unfold over time. those who adopt a dynamical approach to characterising emotion sit broadly within what have been called enactivist (scarantino & sousa, 2021) and ‘network’ (moors, 2022) approaches to emotion. these approaches do not neatly oppose approaches that see emotions as feelings, as judgments, as perceptions, or as motivational states (see scarantino & sousa, 2021, for a typology and discussion of these positions). rather, the difference lies in explanatory focus: those adopting a dynamical approach are less interested in identifying the core feature or features of emotional states, but rather in using concepts and tools drawn from dynamical systems theory to help describe how the component parts that make up emotional episodes change across time, with the ultimate goal of improving our knowledge of the mechanisms that produce, sustain, and dampen emotional episodes. a much-discussed feature of dynamical systems is that they can give rise to emergent properties. in the context of emotion, the dynamical system is hypothesised to be identical with most subsystems of the human organism, or the human organism itself (see meuleman, 2015, for discussion), and emergence is defined specifically in relation to the formation of emotional patterns (that is, patterns that form over time). colombetti (2014, p. 77), for instance, describes emotional episodes as ‘dynamical’ or ‘coherent’ patterns that form against a background of what she calls ‘primordial affectivity’ or ‘sense-making activity’ (colombetti, 2014, pp. 1–24). the notion of primordial affectivity has some overlap with the idea of ‘core affect’ described by psychological constructionists (e.g., russell, 2003) as the ‘bare’ or ‘raw’ feeling of bodily arousal, prior to conceptualising, verbalising or labeling. however, colombetti’s (2014, p. 19) notion of primordial affectivity, drawn from the enactivist and phenomenological traditions, is broader, denoting a very primitive sort of teleological directedness, i.e., the ability that even very simple living systems have to ‘be sensitive to what matters to them’. an example of this in a simple system is the movement of a motile bacteria toward a sugar gradient. a crucial feature of this approach, then, is that emotional episodes do not form against an otherwise non-affective background. instead, emotional episodes are ‘coherent patterns of an organism that is always already affective in virtue of its being a sense-making system’ (colombetti, 2014, p. 77). for colombetti, emotional episodes have distinctive configurations not because of the preformationist assumptions espoused by basic emotions theorists (i.e., ekman et al., 1972; ekman, 1973, 1980) but rather in virtue of a history of mutual influences between evowalsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org elena walsh 6 lutionary and developmental factors that shape the phenotypic trajectory of the organism. on this ‘softly assembled’ approach (see especially colombetti, 2014, pp. 69–74) all emotional episodes, including those associated with basic emotions, are open and flexible structures sensitive to context. the identification of emotion with dynamical patterns also appears in component process models of emotion. for instance, in the component process model (cpm) of scherer (1984, 2009b, 2009a), emotion is a process of coordination or synchronisation of emotion components, driven by appraisal from a relatively uncoordinated or loosely coupled state. here, the concept of synchronisation is used to separate business-as-usual system operations from distinctively emotional patterns. it is only when a minimum threshold of synchronisation is achieved amongst components that an emotional episode begins. the emotional episode is identical with the ongoing interactions of the emotion components over time until synchronisation falls below the minimum required threshold. emotion is described as emergent here (scherer, 2009b, 2009a) in the sense that the emotional episode forms a pattern that emerges over time. that this pattern appears is clear if we imagine an emotion appearing and disappearing, in a multitude of infinitely varied formations, on a state space that tracks how the interaction of components forms attractor basins. each attractor is a visual representation of the emergent pattern formed by each emotional episode. the emergent phenomenon here simply is the pattern, identical with the emotion itself. though the accounts of colombetti and scherer are distinctive, 6 both share the broad idea that emotion component interactivity produces patterns of ‘synchronisation’ or ‘soft assembly’, and that these patterns differentiate emotional from non-emotional episodes. 3 emergent dynamical patterns though colombetti and scherer both identify emotional patterns with ‘emergent’ emotional episodes, neither attempts to show whether and how these patterns meet the two-fold criteria for emergence introduced above: determination: e is determined by b. novelty: e is novel with respect to b. i now turn to this task. these criteria can be cashed out in various ways, corresponding to different types of emergence, and it is not my intention to provide a comprehensive survey here (existing taxonomical approaches are discussed in guay & sartenaer, 2016; humphreys, 2016; sartenaer, 2016). instead, i focus on what has been called ‘pattern emergence’ (humphreys, 2016; winning & bechtel, 6both colombetti and scherer hypothesise interactivity among emotion components, but the cpm insists on a fixed sequence of appraisal checks playing the driving role in emotion component changes. the approaches also differ with respect to the degree to which they endorse an enactivist approach to cognition. for discussion, see colombetti (2014, pp. 49–52) and moors et al. (2013). walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org emotions as emergent properties 7 2019), an approach that aligns emergent phenomena with the formation of patterns produced by the dynamical activity of complex systems, and is thus suitably paired with the case of emotional patterns understood as putatively emergent phenomena. note firstly that, on a pattern emergence approach, the determination condition is met because the emergent whole e and its base b are in a constitutional partwhole relationship. across the period of time an emotional episode is present, the parts change along with the whole (for discussion of this conception of diachronic constitution, see craver & bechtel, 2007). for example, during an episode of anger, there is typically a sharp increase, followed by a gradual decrease, in heart rate and nervous system arousal, concomitant with rumination on the theme of blameworthiness. this is the sense in which the emotion as a whole is constituted by its parts. now, it is true that these components are present when an emotional episode is not (heart rate can increase even if we are not feeling emotion, for instance, and rumination can occur without anger). the difference for our purposes is that identifying the emotion requires recourse to a different form of description: we cannot identify the presence of an emotional episode simply by describing the current state of the various parts making up the emotion, or even by providing values for variables associated with each of these parts. this is because, on a dynamical approach, we are looking for a relationship of temporary synchronisation that holds among the parts when represented as variables in a time-dependent statistical model. as such, the pattern that constitutes the emotion can only be described using an alternative level of description. 7 in practice, this alternative level of description will include terms referring to statistical measures of synchronisation (such as a moving principal component analysis or pca, see hollenstein & lanteigne, 2014), as well as terms that describe the strength of interactivity between the component parts (feedback loops and constraints are examples of such terms, and are discussed in detail below). we can now turn to the novelty condition. some varieties of emergence focus on the generation of ontological novelty, such as new properties, laws, or causal powers that arise with e (these can arise either synchronically, diachronically, or both, see guay & sartenaer, 2016). other approaches focus on epistemological novelty, where e is a novel representation of the relevant system that is unpredictable, unexplainable, or impossible to describe when armed only with b-level information (this sort of novelty can also occur either synchronically or diachronically). 8 novelty on a pattern emergence approach is sometimes cast as epistemological and diachronic (humphreys, 2008, 2016). here, the focus is on an emergent pattern e whose formation is impossible to predict at time t2 when armed only with information about b at time t1. however, it has also been argued that the unpredictable, 7theorists describing pattern emergence make reference to ‘an alternative description’ (pattee, 1971, pp. 85–86) or a ‘higher-level language’ (winning & bechtel, 2019, p. 141) through which emergent patterns are identified. 8i am grateful to an anonymous reviewer for raising this point. walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org elena walsh 8 emergent pattern also produces a form of ontological novelty. a proposed example of the latter is the emergence of ‘goal-directed behaviour’ in complex and adaptive biological systems (winning & bechtel, 2019). the possibility that emotion may be ontologically emergent in this sense is an interesting one, but i focus below on epistemological novelty cast as unpredictability, since this will, as we shall see in section 6, prove useful for thinking about the boundary problem (introduced in section 4). we will be in a better position to characterise the sort of unpredictability we are faced with in the case of emotion by considering a simple example. imagine a student who feels fear at t2 when called on to speak in class. component changes associated with fear occur: the student adopts the facial expression associated with fear (wide eyes, raised eyebrows) and concomitant physiological changes occur (e.g., raised heart rate, respiration, and blood flow, see hyde et al., 2019). the prediction of what emotion the student will feel at t2 cannot be made going merely on measures of basal-level components (the student’s facial expression at time t1, heart rate at time t1, and so on). since knowledge of b-level markers at t1 do not enable knowledge of e-level markers at t2, the criterion for unpredictability is met in a rather minimal sense. however, this notion of unpredictability might be thought too weak to classify emergent phenomena (for discussion see sartenaer, 2016). how does it differ from, say, the difficulty of predicting if a car will speed up at t2 when armed only with knowledge of the state of the engine at t1? without knowing that the driver has pressed on the accelerator pedal at t1, we cannot predict e (acceleration) at t2. and we presumably would not want to say that acceleration is an emergent property of an engine, lest we risk trivialising the notion of emergence. 9 there is, nonetheless, an important point of contrast between the case of acceleration as a putative emergent property, and the case of emotion. the function of a car engine (e.g., what elicits acceleration, how much acceleration is produced when pressing on the pedal with a certain degree of force) is determined when the engine is designed. its mature function is not a product of an adaptive developmental history unique to each car engine. but the case of emotion most likely is. consider firstly the fact that human emoters respond very differently to identical environmental input. this is so both with regard to the initiation of an emotional episode, and its regulation (how it develops, and what leads it to either continue or terminate). consider initiation first: being cut off in traffic on the freeway may prompt mild annoyance in one person, rage in a second person, and may inspire no emotional change in a third (imagine a third person who notices being cut off but continues enjoying the view of the water over the bridge while listening to the radio). a gentle dig might amuse one person, make a second person feel angry, and make a third person feel sad. now consider regulation: out of two people who respond with anger to being cut off in traffic, or the mildly insensitive dig, one might quickly return to a state of relative emotional neutrality, while another 9i am grateful to an anonymous reviewer for raising this point. walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org emotions as emergent properties 9 might remain angry for minutes (or hours, or perhaps even days). this variation in emotional elicitation and expression has been dubbed the ‘singularity problem’ (morag, 2016). it is likely that this variation is due to developmental learning. consider, for instance, the tendency to respond to environmental input with anger. influential work by dodge & somberg (1987) suggested that repeated experiences of abuse, such as aggression and/or violence lead children to develop a ‘hostile attributional bias’ (hab) that filters subsequent experience. dodge claimed that those who make hostile attributions will, when exposed to a frustrating social stimulus, such as being hit in the back with a ball, tend to interpret the stimulus as an aggressive cue and thus respond aggressively. dodge & somberg (1987) called this a sort of habitual ‘cue distortion’ (also see anderson & graham, 2007), one that manifests as dispositional anger that ignites or persists even when outer circumstances do not seem to warrant it. development of hab might explain cases of variation in emoting in which individuals respond with anger to a neutral stimulus. one benefit of a dynamical approach to emotion is that it offers concepts and modelling techniques that can be used to represent how emotional dispositions are formed over developmental time (especially in childhood and adolescence) and come to influence how occurrent emotional episodes are triggered and expressed in a given individual (for discussion, see lewis, 2005; lewis & liu, 2011; walsh, 2021). and if it is indeed the case that the triggers and dynamical trajectories associated with emotional episodes are influenced by an individual’s life history,10 e at t2 will prove unpredictable even when we are armed with knowledge of b at t1 and knowledge of an observable trigger (such as a teacher calling on a student to speak, or a person being unexpectedly hit on the back with a ball). we could certainly make educated guesses in these cases, but we will be much less certain of our predictions than in the case where we try to predict that a car will accelerate with knowledge of an observable trigger (pressing on the accelerator pedal). the difference is that emotions are flexible by design: unlike blinking when an object is suddenly moved close to the eye, emotions do not unfold in the same way for all people when faced with similar triggers. this is a well-documented feature of emergence in complex adaptive systems (humphreys, 2016; winning & bechtel, 2019), and it is in this stronger sense that e is unpredictable in the emotion case. 4 the boundary problem so far, i have described emergent emotions as patterns produced by a system at t2 that are unpredictable when armed only with knowledge of the system set-up and any relevant environmental triggers at t1. this definition presupposes that we can clearly distinguish a pattern from a non-pattern. but how do we tell the difference 10the evidence supporting this view is compelling (for a recent summary and history, see sartojackson, 2022). walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org elena walsh 10 between ‘synchronised’ emotion components and ongoing homeostatic processes? this is, as was explained in section 1, the boundary problem. emotion theorists who have tried to measure synchronisation have sought emotional ‘signatures’ for given emotion categories (anger, fear, disgust, etc.). these ‘signatures’ refer to covariational patterns among one or more emotion components that identify tokens of a particular emotion type. whilst there have been efforts to detect emotional signatures, the consensus view is that the empirical evidence provides either weak or no support lench et al. (2011). however, these studies are limited in at least two significant ways. first, many only measure covariance between two components, with few measuring covariances among three or more components of emotion (barrett, 2006a; meuleman, 2015). second, the measured covariances are assumed to be static (meaning covariation is not indexed to time, but is instead presented as a summary state at a single moment in time, see meuleman, 2015). having said this, recent work on emotional synchronisation has begun to identify time-indexed covariational patterns among emotion components (hollenstein & lanteigne, 2014). some of this work uses multivariate analysis of variance (manova) across designated time slices and uses various psychological and physiological measures as proxies for changes in emotion components. such work is in its infancy but has found significant evidence showing feedback between multiple emotion components (for discussion and examples of recent empirical work, see bulteel et al., 2014; hollenstein & lanteigne, 2014). in addition, methods for collecting and dynamically analysing multi-componential, time-indexed data are developing in sophistication, and dynamical analyses of various emotional phenomena are being used with increasing frequency (see kuppens & verduyn, 2017). unfortunately, however, there are not yet studies using specifically nonlinear, timeindexed methods to specifically assess multi-component emotional synchronisation (see meuleman, 2015). in section 4, i explore the idea that constraints (feedback loops between emotion components) constrain or reduce the output options available to the system as a whole during the period of time for which an emotional episode is present. feedback loops are, as we shall see, temporal phenomena. it is also well-recognised that capturing their dynamics requires the use of nonlinear methods (i return to this point below, see also lucky et al., 2020; scherer, 2009a). if emotions are inherently temporal phenomena that operate via feedback loops, we should expect that static and linear statistical methods will be unable to reveal their presence. i now discuss how feedback loops are involved in emotion production, and then show how the notion of emotions as emergent patterns provides the conceptual backdrop that motivates and justifies the application of nonlinear dynamical analysis to the search for emotional signatures. walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org emotions as emergent properties 11 5 constraints in classical physics, laws governing motion characterise the way a system might evolve, and constraints limit the degrees of freedom available to the system as it evolves, just as friction or an inclined plane limit or constrain the speed of a moving object (hooker, 2013). a simple example of a physical constraint is a bowl that constrains the movement of a marble. the marble can move in any direction, but only as far as the edges of the bowl. another example is a skeleton, described by hooker (2013, p. 761) as a ‘disabling constraint, for example limiting the movements of limbs (cf. an octopus)’. an octopus limb has more ‘degrees of freedom’ with respect to forming curves than a human arm, for instance, since the latter can only bend rigidly at the elbow. a number of authors have applied the notion of constraint to feedback loops in biological systems (bechtel, 2017; hooker, 2013; pattee, 1971; winning & bechtel, 2018). here, feedback loops produce a constraining effect by amplifying or dampening the output of a mechanism over time. a feedback loop is a system where outputs at time t-1 are routed back as system inputs at t. an ‘output’ refers to any variable that can be measured, and that is part of the phenomenon being investigated. when a variable is described as an ‘output’ rather than a variable, this denotes the value that a variable (such as heart rate) is taking at a particular moment in time t of interest to the observer of the system (rather than indicating an output to system activity considered as separate from the phenomenon itself). a positive feedback loop features outputs routed back as inputs that tend to amplify the strength of the incoming signal, or the level of perturbation of the system. (a simple example is a bank account with compound interest. deposits into the account increase the total balance, and compound interest amplifies this effect). a negative feedback loop produces the opposite effect, dampening or weakening system perturbation. below, i rehearse bechtel’s (2017) example of a feedback loop operative in the context of circadian biology, to illustrate the relationship between feedback loops and output options available to a system. this is followed by an example of just one feedback loop (between arousal and appraisal) that is likely involved in the case of emotion, and that also affects system outputs. consider how the notion of constraint can be used to characterise feedback control in circadian biology (bechtel, 2017; hardin et al., 1990). in particular, consider the relationship between oscillatory phase and the lower-level parts of the mechanism (the genes, proteins, etc.). it is now known that entrainment to light comes when a light signal from the retina to the suprachiasmatic nucleus functions to increase transcription of the per gene (bechtel, 2017). the function of light on transcription varies depending on the time of day. during the day, light has no effect on transcription because per transcription is already at its maximum. during early evening, light exposure delays the phase of the oscillation. this is because, early in the evening, per concentration diminishes due to an internal feedback mechanism. late at night however (e.g., approaching dawn) light exposure advances the phase. this is because during this time per concentration is beginning to increase due to walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org elena walsh 12 endogenous oscillation. as such, light received during this phase will speed up transcription, resulting in daytime levels of the per protein being reached earlier, and an advance in the phase of the oscillator. the important point here is that circadian rhythm can be construed as a mechanism whose output varies even when the type of input (i.e., light) remains constant. sometimes the one input type will increase per transcription, and other times it won’t. what this variation in output depends on are feedback mechanisms internal to the mechanism itself, such as feedback about the concentrations of the per protein in the nucleus: when per levels are low, synthesis of new per proceeds, but as per accumulates, it inhibits further synthesis. the product is a system that responds differentially to the same input. the effect on per transcription during, for example, the state in which per concentrations in the nucleus are high is determined by conditions generated within the mechanism, not from outside it. this system-dependent differential output is produced by constraint. this is so in the sense that the mechanism operates with reduced degrees of freedom, meaning that the total possible output options of the mechanism is reduced. for example, feedback about the concentrations of the per protein in the nucleus can reduce output options such that, if per transcription is increasing, the system will tend toward more quickly reaching daytime-typical concentrations of per. no longer is there the option for concentrations of per to increase or decrease. in this case, output options are reduced to those consistent with an increase in per concentration. below, i suggest that we can apply the notion of constraints to emotion. in particular, feedback loops are thought to be involved in emotion (see hollenstein, 2015), and i suggest that we see them functioning to reduce the output options available to a system, where here the ‘system’ is an individual reacting to their environment. to make the point, i consider a very simple case of one feedback loop – between arousal and appraisal – that is likely involved in the emotion of anger. i first briefly summarise the evidence base relating these components in the case of anger, and then provide an example to illustrate why the notion of constraint is useful. to start with, there is evidence that heightened positive functional connectivity with the thalamus during angry rumination reflects a reciprocal relationship between rumination and arousal (denson, 2013). the thalamus is a key structure in the brain that acts as a relay station, receiving sensory information from various parts of the body and relaying it to the relevant brain regions for further processing. in the context of angry rumination, the heightened positive functional connectivity with the thalamus indicates an increased level of communication between the thalamus and other brain regions involved in executive functions, such as the prefrontal cortex. this suggests that the executive functions associated with the prefrontal cortex (e.g., planning revenge), co-occur with an increased sense of arousal. there is further support for the idea that there is reciprocal influence between arousal and appraisal. consider firstly how arousal influences appraisal. there is growing support for the view that arousal motivates and influences reasoning walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org emotions as emergent properties 13 processes. for instance, haidt (2001, 2013) showed that a person’s ‘gut reaction’ to a vignette appeared to drive reasoning processes in a biased way: toward a reasoned assessment of the situation described in the vignette as conforming to the initial feeling the vignette provoked (e.g., get angry first, justify your anger later). more recently, it has been suggested that the appraisals we develop in response to environmental input (and, consequently, the way in which we ruminate about such appraisals) may be influenced by our interoceptive awareness of physiological changes in the body (miller & clark, 2018; wilkinson et al., 2019). now let’s consider how appraisal influences arousal. it is known that the perception of physiological activity, otherwise known as interoceptive awareness, can heighten the intensity of emotional arousal (dunn et al., 2010). consistent with the role of the insula in interoceptive awareness, denson et al. (2009) found, for instance, that right anterior insula activation is positively correlated with selfreported angry state rumination. thalamus activation is also important due to its role in emotion processing, emotion experience, and emotional control (marchand, 2010). together, increased activation in the right anterior insula and thalamus during rumination may be associated with a heightened experience of emotional arousal (see also denson, 2013). we can now consider an example (represented in figure 1 below). imagine a high-school art student working on a painting in class. at time t1, she is in a relatively unemotional state. her nervous system is unaroused, and she is daydreaming vaguely about various topics (what she will buy for lunch from the canteen, whether she will get an a in this class, etc.). at t2, her desk partner accidentally elbows her as he switches paintbrushes. the first student immediately experiences physiological arousal prompted by the unexpected pain in her arm (e.g., increased cardiac output and sympathetic nervous system arousal, see sinha et al., 1992). she then looks at her classmate’s facial expression, which is ambiguous, somewhere between shock and surprise. due to her aroused state, however, she interprets this expression as hostile, developing an appraisal consistent with the thematic content of anger, namely that a ‘demeaning offense against me or mine’ has occurred (lazarus, 1968; silvia & warburton, 2006). at t3, the student’s appraisal that her desk partner intended to hurt her now feeds back to stimulate further physiological arousal (her heart rate increases further, and her cheeks begin to redden in the familiar flush of anger). at the same time, her increased arousal stimulates and sustains a developing appraisal of her classmate as a demeaning bully deserving of blame or retaliation (at this point, perhaps she scowls at him or demands an apology). at t4, interoceptive awareness of her state of sympathetic arousal further constrains the student’s evaluation of the event (her classmate’s putatively genuine apology is met with resistance, and she gets up and abruptly leaves the classroom, ruminating on her classmate’s vicious attempt to ruin her art work).11 11the student’s response may strike the reader as inflexible, or at least as non-proportional, relative to the ambiguous nature of the initial provocation (i thank an anonymous reviewer for raising this point). in my view, it is such cases of inflexible or non-proportional emoting that best illustrate instances in which internal processes (such as the feedback loop described) are playing a walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org elena walsh 14 figure 1: positive feedback between arousal and appraisal producing constraint during an episode of anger. grey circles represent appraisals made across successive moments of time, white circles represent physiological arousal. circle size represents the degree to which the system is constrained, with a smaller size indicating greater constraint. arrows represent positive feedback between arousal and appraisal. at t1, arousal and appraisal are relatively unsynchronised. as synchronisation occurs (t2 to t4), the system is increasingly constrained. in this example, a positive feedback loop between arousal and appraisal operates to generate and sustain an instance of anger. the feedback loop acts as a system constraint in the following sense: at t2, physiological arousal results in the student developing an initial appraisal consistent with anger (e.g., that of a demeaning offence). whereas at t1, the student was thinking idly about various topics, at t2 she only has a single point of focus, namely the idea that her classmate intended to hurt her. this appraisal then constrains her state of arousal (her nervous system continues to move toward heightened sympathetic arousal, consistent with the development of anger). finally, at t4, heightened arousal and an angry appraisal remain tightly coupled, as the student abruptly leaves her desk so she can ruminate on her classmate’s apparent wrongdoing. figure 1 above provides an illustration. circles represent appraisal (grey) and physiological arousal (white). the circle size represents the flexibility or degrees of freedom available for each variable during successive moments in time (with a smaller size indicating a reduction in degrees of freedom). arrows represent positive feedback between arousal and appraisal. decreasing circle size shows a decrease in flexibility as a result of reciprocal feedback between arousal and appraisal. though the episode of anger is initiated by an environmental stimulus, it more significant role in determining how the emotional episode unfolds than the relevant environmental triggers. this might provide one way to understand individuals with inflexible emoting styles (walsh, 2021). nonetheless, the empirical evidence cited suggests that the feedback loop described will still present itself in more flexible and moderate presentations of anger. walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org emotions as emergent properties 15 is developed and sustained by the student’s continuing rumination and concomitant arousal. this process results in a reduction in the range of values that may be taken by measures of arousal and appraisal during the episode of anger. this result manifests in the student’s sharp focus on the blameworthiness of their classmate (instead of her thinking about the situation in a more flexible and open-ended way), and in her heightened sympathetic nervous system arousal (instead of experiencing a variable heart rate, for instance, the student’s heart rate instead increases steadily as she continues to ruminate on the blameworthiness of her classmate). 6 flexible boundaries i now want to suggest that if emotions operate via mechanisms of constraint, we need to consider new statistical approaches to measuring emotional synchronisation. it is not my purpose here to outline exactly what these methods ought to be. instead, i offer a few tentative suggestions and explain why they make sense in context of a pattern emergence approach to emotion. finally, i explain how the novelty condition is met on this new approach. first of all, note that in our simplified case involving anger, we saw that the output options available with respect to arousal and appraisal are reduced to those consistent with anger (e.g., increased cardiac output, rumination consistent with the theme of other-blame) and that it is the presence of feedback between arousal and appraisal that helps to develop and sustain the emotional episode. i characterised this feedback as a constraint on the system that left it operating with reduced degrees of freedom. despite this, constraints do not completely fix or determine how a dynamical episode of anger will develop and sustain (or resolve) itself across time: constraints reduce the degrees of freedom available to the system when an emotional episode is occurring, rather than fixing them completely. let me describe two other influences (separate to the constraints) that influence the dynamical pattern constituting the emotion. firstly, ongoing changes in the material and social environment will modulate the development of the emotional episode, as will internal homeostatic processes that do not feature as emotion components in a given statistical analysis. both processes can be seen as ‘control parameters’ – factors that necessarily affect the state of the system, but are not necessarily in turn affected by the system. an example of an environmental change is a social interaction providing negative feedback that dampens the intensity of the emotional state. for instance, perhaps at t5, a reassuring hug and a soothing word from a friend in the hallway interrupts the student’s narrow focus on blaming her classmate, and she begins thinking of alternative interpretations of her classmate’s behaviour. at the same time, her heart rate begins to decrease (see figure 2). here, external input into the system acts as negative feedback, dampening the student’s emotional response. (without positive feedback, emotional episodes could not reach a finite zenith, and without negative feedback, they would never resolve into neutral or less intense states.) an walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org elena walsh 16 figure 2: control parameters during an episode of anger. dashed grey lines show external control parameters modulating the degree to which the system is constrained. control parameter a modulates appraisal at t2 (the student observes the classmate’s facial expression and interprets it as hostile). from t2 to t4, feedback between arousal and appraisal constrain the system towards a stable state of anger (as per figure 1). control parameter b influences appraisal at t5. example of an internal homeostatic process influencing the system is an existing case of hypertension that exacerbates the student’s interoceptive awareness of her increased cardiac output (see yigla et al., 2006) associated with a surge of anger. importantly, these ongoing changes, represented as control parameters, may only be visible when emotions are modelled as nonlinear, time-dependent functions. a linear function is such that any change in an independent variable x (e.g., x could be an emotional stimulus, such as a scene from a scary movie) will always produce a corresponding change in a dependent variable y (e.g., y could be a physiological or expressive change indicative of fear). in this way, a linear function is predictable. but the control parameters just described are, in their essence, unpredictable. therefore, we should not expect that traditional linear statistical analysis (of the sort typically used to measure emotional synchronisation) will reveal the presence of emergent emotional patterns. the second issue is that these ongoing environmental changes may not only influence how an emotional episode develops and sustains itself across time, but whether or not a given covariational pattern ‘counts’ as a token of a given emotion type. consider again the case of the hypertensive individual. the threshold for physiological markers of the presence of anger could be expected to differ in experimental settings featuring hypertensive patients versus those featuring nonhypertensive patients (since hypertension increases cardiac output, which is frewalsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org emotions as emergent properties 17 figure 3: ‘bow-tie’ shape produced by a cellular automaton. the cells of the cellular automaton are black or white. after many rule-based computational steps are performed, the ‘bow-tie’ pattern is displayed (for discussion, see humphreys, 2008). quently used as a proxy indicator for anger-related arousal). furthermore, it may be that measures of covariance that are statistically significant may be an intraindividual phenomenon, such that the more we try to generalise across persons, the more fragile the generalisations drawn become. therefore, we might be asking for too much if we seek a single, context-independent measure. instead, we might need to seek measures that fall within a range of indicators for statistical significance, with each range relativised to a specific context of inquiry. i close with a few words about how the notion of pattern emergence may encourage us to be more flexible with respect to the search for emotional signatures. as was noted earlier, the distinctive feature of novelty on a pattern emergence approach is that the emergent pattern e produced at t2 is unpredictable when armed only with knowledge about the basal-state b at t1. we are now in a position to further clarify the notion of unpredictability at play here. firstly note that there is no need to assume that the emerging pattern at t2 has some sort of essential property, or even that it conforms precisely to some or other category type or definition (see humphreys, 2016 for discussion). it is sufficient that the pattern be unpredictable and historically related to the basal-level configuration at t1. to make the point clearer, consider the ‘bow-tie’ shape produced by a cellular automaton following a set of rules that determine how its states evolve over time (figure 3 below). the condition for the ‘bow-tie’ being an emergent pattern is that it is produced by the evolution of the cellular automaton over time. if the pattern is produced by, say, creating a matching stamp and stamping it on to a piece of paper, it does not count as an emergent property (for discussion see humphreys, 2008). if emotions are novel in this historical, system-dependent sense, we have an additional reason not to demand that tokens of emotion conform to an ideal or essentialised type. this is that emergent patterns are principally defined by their walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org elena walsh 18 unpredictable emergence in the life of a given system. there need not be an additional requirement that the pattern produced conform to a strict type identity. nonetheless, this does not mean that it is impossible to identify broad commonalities among all emotional episodes of, say, anger, or fear, or disgust as they begin to emerge. i have argued that the notion of constraint can explain why emotions appear to occur over discrete periods of time, in opposition to, say, moods, or periods of relative non-emotionality (e.g., that a sudden flash of anger, say, seems to have a discernable time course, concomitant with flushed cheeks, a raised voice, and an increased heart rate). the explanation is that the dynamical system (identifiable with the human organism itself) is constrained when an emotional episode is present such that it operates with reduced degrees of freedom. an emotional episode will produce a pattern whose exact configuration is unpredictable in any specific case, but which, when compared across persons and situations, will exhibit broad configurative similarities that can be described with approximate covariance measures. the degree to which such a patterned identity emerges will reflect the degree to which the system is constrained. 7 conclusion this article explored what it might mean to characterise emotions as emergent properties, building on existing proposals to construe emotions as dynamical patterns whose emergence is unpredictable in the life of a given system (e.g., a given emoter). it proposed aligning the view that emotions are emergent properties with a ‘pattern’ emergence approach in which novelty is cast as epistemic unpredictability. it also argued that the emotion mechanism as a whole functions to reduce the degrees of freedom of its various components, via the operation of constraints (realised via the feedback loops characteristic of nonlinear dynamical systems). it might be hypothesised that the operation of these constraints, and the patterns to which they correspond, themselves appear at the ‘edge of chaos’12 — at the boundary between the random behaviour of a complex system and a self-organised state (humphreys, 2016; packard, 1988). it has been recognised that one of the distinctive features of the nature of explanation in the special sciences is seeking generalisations that hold under a relevant class of interventions, rather than exceptionless laws (for discussion see woodward, 2000, 2007). a characterisation of emergence as constraint suggests that previous attempts to identify emotional signatures may have been in search of a chimera — a precise, lawlike and contextually-invariant signature for any given emotion type. construing emotions as emergent properties in the manner proposed here offers an alternative to pre-emptively abandoning the task of seeking 12the phrase was coined by physicist norman packard (1988), but it refers to phenomena that arise out of what appears to be random or unorganised behaviour in a complex system. the ‘edge’ is the temporal point at which self-organised behaviour can be said to emerge. walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org emotions as emergent properties 19 emotional signatures for at least some emotion categories. instead, emotional signatures might be hypothesised to have measures that fall within specified ranges, that feature inter-individual variation, that are context-sensitive, and that are identifiable with patterns only visible under the lens of dynamical systems analyses. such measures would be ‘invariant’ in the sense described by woodward (2000): they would hold under a relevant class of interventions, even if not exceptionlessly beyond that domain. seeking such measures yields an approach to empirical investigation consistent with the nature of generalisations distinctive of the special sciences — limited and more fragile, but not non-existent. acknowledgments some of the ideas in this paper were presented to members of the centre for technomoral futures at the edinburgh futures institute, university of edinburgh. these ideas have evolved from research that began whilst working within the theory and method in biosciences lab at the charles perkins centre, university of sydney. i received valuable feedback from both groups. i also received instrumental feedback from laura kotevska at the university of sydney, and from two anonymous reviewers who provided assistance in refining the central argument of the paper. i am thankful for the generous support for this research provided by the university of wollongong’s aegis renew grant scheme (20222023), and for the support of my colleagues at the school of liberal arts, university of wollongong. references anderson, k. b., & graham, l. m. (2007). hostile attribution bias. in r. f. baumesiter & k. d. vohs (eds.), encyclopedia of social psychology (pp. 445–446). sage publications. barrett, l. f. (2006a). are emotions natural kinds? perspectives on psychological science, 1(28), 28–58. https://doi.org/10.111 1/j.1745-6916.2006.00003.x barrett, l. f. (2006b). solving the emotion paradox: categorization and the experience of emotion. personality and social psychology review, 10(1), 20–46. https://doi.org/10.1207/s15327957pspr1001_2 barrett, l. f. (2009). the future of psychology: connecting mind to brain. perspectives on psychological science, 4(4), 326–339. https://doi.org/10.1111/j.1745-6924.2009.01134.x barrett, l. f. (2014). the conceptual act model: a précis. emotion review, 6(4), 292–297. https://doi.org/10.1177/17540739 14534479 bechtel, w. (2017). explicating top-down causation using networks and dynamics. philosophy of science, 84(2), 253–274. https://doi.org/10.1086/690718 boyd, r. (1991). realism, anti-foundationalism and the enthusiasm for natural kinds. philosophical studies, 61(1-2), 127–148. https://doi.org/10.1007/bf00385837 bulteel, k. et al. (2014). decon: a tool to detect emotional concordance in multivariate time series data of emotional responding. biological psychology, 98, 29–42. https://doi.org/10.1016/j.biopsycho.2013.10.011 chalmers, d. j. (2008). strong and weak emergence. in p. clayton & p. davies (eds.), the re-emergence of emergence (pp. 244–256). oxford university press. coan, j. a. (2010). emergent ghosts of the emotion machine. emotion review, 2(3), 274–285. https://doi.org/10.1177/175407 3910361978 coan, j. a., & gonzalez, m. z. (2015). emotions as emergent variables. in l. f. barrett & j. a. russell (eds.), the psychological construction of emotion (pp. 209–225). the guilford press. colombetti, g. (2014). the feeling body: affective science meets the enactive mind. mit press. craver, c., & bechtel, w. (2007). top-down causation without top-down causes. biology & philosophy, 22, 547–563. https://doi.org/10.1007/s10539-006-9028-8 walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1111/j.1745-6916.2006.00003.x https://doi.org/10.1111/j.1745-6916.2006.00003.x https://doi.org/10.1207/s15327957pspr1001_2 https://doi.org/10.1111/j.1745-6924.2009.01134.x https://doi.org/10.1177/1754073914534479 https://doi.org/10.1177/1754073914534479 https://doi.org/10.1086/690718 https://doi.org/10.1007/bf00385837 https://doi.org/10.1016/j.biopsycho.2013.10.011 https://doi.org/10.1177/1754073910361978 https://doi.org/10.1177/1754073910361978 https://doi.org/10.1007/s10539-006-9028-8 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org elena walsh 20 denson, t. f. (2013). the multiple systems model of angry rumination. personality and social psychology review: an official journal of the society for personality and social psychology, 17(2), 103–123. https://doi.org/10.1177/1088868312467086 denson, t. f., pedersen, w. c., ronquillo, j., & nandy, a. s. (2009). the angry brain: neural correlates of anger, angry rumination, and aggressive personality. journal of cognitive neuroscience, 21(4), 734–744. https://doi.org/10.1162/jo cn.2009.21051 dodge, k. a., & somberg, d. r. (1987). hostile attributional biases among aggressive boys are exacerbated under conditions of threats to the self. child development, 213–224. https://doi.org/10.2307/1130303 dunn, b. d., galton, h. c., morgan, r., evans, d., oliver, c., meyer, m., cusack, r., lawrence, a. d., & dalgleish, t. (2010). listening to your heart: how interoception shapes emotion experience and intuitive decision making. psychological science, 21(12), 1835–1844. https://doi.org/10.1177/0956797610389191 ekman, p. (1973). darwin and facial expression: a century of research in review. academic press. ekman, p. (1980). biological and cultural contributions to body and facial movement in the expression of emotions. in a. o. rorty (ed.), explaining emotions (pp. 73–102). university of california press. ekman, p., friesen, w. v., & ellsworth, p. (1972). emotion in the human face: guidelines for research and an integration of findings. pergamon press. frijda, n. h. (1994). varieties of affect: emotions and episodes, moods, and sentiments. in p. ekman & r. j. davidson (eds.), the nature of emotion. fundamental questions (pp. 59–67). oxford university press. frijda, n. h., mesquita, b., sonnemans, j., & goozen, s. van. (1991). the duration of affective phenomena or emotions, sentiments and passions. in k. t. strongman (ed.), international review of studies on emotion (pp. 187–225). wiley. greenwood, j. (2015). becoming human: the ontogenesis, metaphysics, and expression of human emotionality. mit press. guay, a., & sartenaer, o. (2016). a new look at emergence. or when after is different. european journal for philosophy of science, 6, 297–322. https://doi.org/10.1007/s13194-016-0140-6 haidt, j. (2001). the emotional dog and its rational tail: a social intuitionist approach to moral judgment. psychological review, 108(4), 814. https://doi.org/10.1037/0033-295x.108.4.814 haidt, j. (2013). the righteous mind: why good people are divided by politics and religion. vintage. hardin, p. e., hall, j. c., & rosbash, m. (1990). feedback of the drosophila period gene product on circadian cycling of its messenger rna levels. nature, 343(6258), 536–540. https://doi.org/10.1038/343536a0 hollenstein, t. (2015). this time, it’s real: affective flexibility, time scales, feedback loops, and the regulation of emotion. emotion review, 7(4), 308–315. https://doi.org/10.1177/1754073915590621 hollenstein, t., & lanteigne, d. (2014). models and methods of emotional concordance. biological psychology, 98, 1–5. https://doi.org/10.1016/j.biopsycho.2013.12.012 hooker, c. (2013). on the import of constraints in complex dynamical systems. foundations of science, 18(4), 757–780. https://doi.org/10.1007/s10699-012-9304-9 humphreys, p. (2008). synchronic and diachronic emergence. minds & machines, 18, 431–442. https://doi.org/10.1007/s110 23-008-9125-3 humphreys, p. (2016). emergence: a philosophical account. oxford university press. hutto, d. d., robertson, i., & kirchhoff, m. (2018). a new, better bet: rescuing and revising basic emotion theory. frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018.01217 hyde, j., ryan, k. m., & waters, a. m. (2019). psychophysiological markers of fear and anxiety. curr psychiatry rep, 21, 56. https://doi.org/10.1007/s11920-019-1036-x kuppens, p., & verduyn, p. (2017). emotion dynamics. current opinion in psychology, 17, 22–26. https://doi.org/10.1016/j. copsyc.2017.06.004 lazarus, r. s. (1968). emotions and adaptation: conceptual and empirical relations. nebraska symposium on motivation, 16, 175–266. lench, h. c., flores, s. a., & bench, s. w. (2011). discrete emotions predict changes in cognition, judgment, experience, behavior, and physiology: a meta-analysis of experimental emotion elicitations. psychological bulletin, 137, 834–855. https://doi.org/10.1037/a0024244 lewis, m. d. (2005). bridging emotion theory and neurobiology through dynamic systems modeling. behavioral and brain sciences, 28(2), 169–194. https://doi.org/10.1017/s0140525x0500004x lewis, m. d., & granic, i. (2000). emotion, development, and self-organization: dynamic systems approaches to emotional development. cambridge university press. lewis, m. d., & liu, z.-x. (2011). three time scales of neural self-organization underlying basic and nonbasic emotions. emotion review, 3(4), 416–423. https://doi.org/10.1177/1754073911410748 walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1177/1088868312467086 https://doi.org/10.1162/jocn.2009.21051 https://doi.org/10.1162/jocn.2009.21051 https://doi.org/10.2307/1130303 https://doi.org/10.1177/0956797610389191 https://doi.org/10.1007/s13194-016-0140-6 https://doi.org/10.1037/0033-295x.108.4.814 https://doi.org/10.1038/343536a0 https://doi.org/10.1177/1754073915590621 https://doi.org/10.1016/j.biopsycho.2013.12.012 https://doi.org/10.1007/s10699-012-9304-9 https://doi.org/10.1007/s11023-008-9125-3 https://doi.org/10.1007/s11023-008-9125-3 https://doi.org/10.3389/fpsyg.2018.01217 https://doi.org/10.1007/s11920-019-1036-x https://doi.org/10.1016/j.copsyc.2017.06.004 https://doi.org/10.1016/j.copsyc.2017.06.004 https://doi.org/10.1037/a0024244 https://doi.org/10.1017/s0140525x0500004x https://doi.org/10.1177/1754073911410748 https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org emotions as emergent properties 21 lindquist, k. a., wager, t. d., kober, h., bliss-moreau, e., & barrett, l. f. (2012). the brain basis of emotion: a meta-analytic review. behavioral and brain sciences, 35, 161–162. https://doi.org/10.1017/s0140525x11000446 loaiza, j. r. (2021). emotions and the problem of variability. rev.phil.psych., 12, 329–351. https://doi.org/10.1007/s13164020-00492-8 lucky, c., iyai, d., stanley, c., & humphrey, a. (2020). stability and feedback control of nonlinear systems. international journal of control science and engineering, 10(1), 11–15. https://doi.org/10.5923/j.control.20201001.02 marchand, w. r. (2010). cortico-basal ganglia circuitry: a review of key research and implications for functional connectivity studies of mood and anxiety disorders. brain structure and function, 215(2), 73–96. https://doi.org/10.1007/s00429010-0280-y meuleman, b. (2015). computational modeling of appraisal theory of emotion: vols. phd thesis. university of geneva. miller, m., & clark, a. (2018). happily entangled: prediction, emotion, and the embodied mind. synthese, 195(6), 2559–2575. https://doi.org/10.1007/s11229-017-1399-7 mitchell, s. d. (2012). emergence: logical, functional and dynamical. synthese, 185, 171–186. https://doi.org/10.1007/s112 29-010-9719-1 moors, a. (2022). demystifying emotions: a typology of theories in psychology and philosophy. cambridge university press. moors, a., ellsworth, p. c., scherer, k. r., & frijda, n. h. (2013). appraisal theories of emotion: state of the art and future development. emotion review, 5(2), 119–124. https://doi.org/10.1177/1754073912468165 morag, t. (2016). emotion, imagination, and the limits of reason. routledge. packard, n. (1988). adaptation toward the edge of chaos. university of illinois at urbana-champaign, center for complex systems research. pattee, h. h. (1971). physical theories of biological co-ordination. quarterly reviews of biophysics, 4(2-3), 255–276. https://doi.org/10.1017/s0033583500000640 russell, j. a. (2003). core affect and the psychological construction of emotion. psychological review, 110(1), 145–172. https://doi.org/10.1037/0033-295x.110.1.145 sander, d., grandjean, d., & scherer, k. r. (2005). a systems approach to appraisal mechanisms in emotion. neural networks, 18(4), 317–352. https://doi.org/10.1016/j.neunet.2005.03.001 sartenaer, o. (2016). sixteen years later: making sense of emergence (again). journal for general philosophy of science, 47, 79–103. https://doi.org/10.1007/s10838-015-9312-x sarto-jackson, i. (2022). the making and breaking of minds. vernon press. scarantino, a. (2014). the motivational theory of emotions. in j. d’arms & d. jacobson (eds.), moral psychology and human agency (pp. 156–185). oxford university press. https://doi.org/10.1093/acprof:oso/9780198717812.003.0008 scarantino, a. (2015). basic emotions, psychological construction, and variability. in l. f. b. & j. a. russell (ed.), the psychological construction of emotion (pp. 334–376). guilford press. scarantino, a., & sousa, r. de. (2021). emotion. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (summer 2021). https://plato.stanford.edu/archives/sum2021/entries/emotion/; metaphysics research lab, stanford university. scherer, k. r. (1984). on the nature and function of emotion: a component process approach. in p. ekman (ed.), approaches to emotion (pp. 293–316). psychology press. scherer, k. r. (2009a). emotions are emergent processes: they require a dynamic computational architecture. philosophical transactions of the royal society b: biological sciences, 364(1535)(1535), 3459–3474. https://doi.org/10.1098/rstb.2009. 0141 scherer, k. r. (2009b). the dynamic architecture of emotion: evidence for the component process model. cognition and emotion, 23(7), 1307–1351. https://doi.org/10.1080/02699930902928969 scherer, k. r., & moors, a. (2019). the emotion process: event appraisal and component differentiation. annual review of psychology, 70, 719–745. https://doi.org/10.1146/annurev-psych-122216-011854 silvia, p. j., & warburton, j. b. (2006). positive and negative affect: bridging state and traits. in j. c. thomas & d. l. segal (eds.), comprehensive handbook of personality and psychopathology, personality and everyday functioning (pp. 268–285). john wiley & sons. sinha, r., lovallo, w. r., & parsons, o. a. (1992). cardiovascular differentiation of emotions. psychosomatic medicine, 54(4), 422–435. https://doi.org/10.1097/00006842-199207000-00005 verduyn, p., delaveau, p., rotgé, j.-y., fossati, p., & van mechelen, i. (2015). determinants of emotion duration and underlying psychological and neural mechanisms. emotion review, 7(4), 330–335. https://doi.org/10.1177/1754073915590618 walsh, e. (2021). pathological prediction: a top-down cause of organic disease. synthese, 199(1-2), 4127–4150. https://doi.org/10.1007/s11229-020-02972-x walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1017/s0140525x11000446 https://doi.org/10.1007/s13164-020-00492-8 https://doi.org/10.1007/s13164-020-00492-8 https://doi.org/10.5923/j.control.20201001.02 https://doi.org/10.1007/s00429-010-0280-y https://doi.org/10.1007/s00429-010-0280-y https://doi.org/10.1007/s11229-017-1399-7 https://doi.org/10.1007/s11229-010-9719-1 https://doi.org/10.1007/s11229-010-9719-1 https://doi.org/10.1177/1754073912468165 https://doi.org/10.1017/s0033583500000640 https://doi.org/10.1037/0033-295x.110.1.145 https://doi.org/10.1016/j.neunet.2005.03.001 https://doi.org/10.1007/s10838-015-9312-x https://doi.org/10.1093/acprof:oso/9780198717812.003.0008 https://plato.stanford.edu/archives/sum2021/entries/emotion/ https://doi.org/10.1098/rstb.2009.0141 https://doi.org/10.1098/rstb.2009.0141 https://doi.org/10.1080/02699930902928969 https://doi.org/10.1146/annurev-psych-122216-011854 https://doi.org/10.1097/00006842-199207000-00005 https://doi.org/10.1177/1754073915590618 https://doi.org/10.1007/s11229-020-02972-x https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org elena walsh 22 wilkinson, s., deane, g., nave, k., & clark, a. (2019). getting warmer: predictive processing and the nature of emotion. in l. candiotto (ed.), the value of emotions for knowledge (pp. 101–119). springer international publishing. winning, j., & bechtel, w. (2018). rethinking causality in biological and neural mechanisms: constraints and control. minds and machines, 28(2), 287–310. https://doi.org/10.1007/s11023-018-9458-5 winning, j., & bechtel, w. (2019). being vs. pattern emergence: complexity, control, and goal-directedness in biological systems. in s. c. gibb, r. f. hendry, & t. lancaster (eds.), routledge handbook of emergence (pp. 134–144). routledge. woodward, j. (2000). explanation and invariance in the special sciences. the british journal for the philosophy of science, 51(2), 197–254. https://doi.org/10.1093/bjps/51.2.197 woodward, j. (2007). causation with a human face. in h. price (ed.), causation, physics, and the constitution of reality: russell’s republic revisited (pp. 66–105). oxford university press. yigla, m., abassi, z., reisner, s. a., & nakhoul, f. (2006). pulmonary hypertension in hemodialysis patients: an unrecognized threat. seminars in dialysis, 19(5), 353–357. https://doi.org/10.1111/j.1525-139x.2006.00186.x open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. walsh, e. (2023). emotions as emergent properties. philosophy and the mind sciences, 4, 19. https://doi.org/10.33735/phimisci.2023.9965 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11023-018-9458-5 https://doi.org/10.1093/bjps/51.2.197 https://doi.org/10.1111/j.1525-139x.2006.00186.x https://doi.org/10.33735/phimisci.2023.9965 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction dynamical patterns emergent dynamical patterns the boundary problem constraints flexible boundaries conclusion a structural constraint on neural correlates of consciousness a structural constraint on neural correlates of consciousness sascha benjamin finka,b,c (sfink@ovgu.de) lukas kobc,d,e (lukas.kob@ovgu.de) holger lyrea,b,c (lyre@ovgu.de) abstract researchers on the neural correlates of consciousness (ncc) need to distinguish mere statistical nccs from nccs proper. some neural events may be co-occurrent, probabilistically coupled, or coincidental with a type of conscious experience but lack any deeper connection to it, while in other cases, the relation between neural states and a type of experience hints at a strong metaphysical relation, which distinguishes such nccs proper from mere statistical nccs. in order to address this issue of how to distinguish nccs proper from mere statistical nccs, we propose a position we call neurophenomenal structuralism. the position hinges on the uncontroversial idea that phenomenal experiences relate to each other in degrees of similarity and difference. these complex structures are used to identify and individuate experiences in the methods of neuroscience, psychophysics, and phenomenology. such individuation by structure leads to phenomenal holism, which has implications for how to investigate consciousness neuroscientifically and generates a constraint by which we can distinguish nccs proper from mere statistical nccs: the structural similarity constraint. neural activation must preserve the structure governing the domain of experiences it is associated with in order to count as that domain’s ncc proper. any activation that fails to preserve phenomenal structure fails to be an ncc proper. we illustrate how this constraint works with a study by brouwer & heeger (2009) as an example. keywords color ∙ holism ∙ individuation ∙ neural correlates of consciousness ∙ neurophenomenal structuralism ∙ representational geometry ∙ structuralism this article is part of a special issue on “the neural correlates of consciousness,” edited by sascha benjamin fink and ying-tung lin. aphilosophy department, university of magdeburg bcenter for behavioral brain sciences (cbbs), university of magdeburg chumboldt-universität zu berlin, research training group 2386 “extrospection” dbernstein center for computational neuroscience, charité – universitätsmedizin, berlin ehumboldt-universität zu berlin, faculty of philosophy, berlin school of mind and brain fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-7535-1882 https://orcid.org/0000-0002-7704-4373 https://orcid.org/0000-0002-6040-0263 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink, lukas kob, and holger lyre 2 1 introduction: statistical vs. proper correlates of consciousness one of the primary goals in the neuroscience of consciousness is to find the neural correlates of consciousness (nccs). generally speaking, correlation is a measure of the dependence between the values of variables. in ncc research, the variables at issue are neural versus phenomenal goings-on. at its heart, correlation is a statistical notion. it can be positive, negative, and comes in degrees—all the way to where it nearly touches zero. in most samples, one will find variables that correlate to some non-zero degree. in finite and especially in small samples, the correlation may even be statistically strong despite a lack of meaningful connection between the correlating variables. pearson (1897) called these “spurious correlations.” spurious correlations raise a general worry: if statistical correlations between phenomenal and neural goings-on can be found without any meaningful connection between them, how valuable are the contributions of ncc research to finding the neural substrate of consciousness? statistical procedures may help to minimize spurious correlations, but they cannot rule them out completely. beyond spurious correlations, there are other problems too. for instance, some statistical correlates are reliably correlated with a specific conscious experience 𝑐 such that they can be used as indicators of 𝑐 but are nevertheless not 𝑐’s foothold in the observable physical world. this may be because they occur before or after 𝑐 (see also aru et al., 2012), or are simultaneously co-caused by whatever causes 𝑐, or are just a common but not universal or necessary side-effect of 𝑐. so there might be a broad range of neural states that are reliably correlated with 𝑐, but do not qualify for being a neural substrate of 𝑐. their correlation with 𝑐—however strong—is not indicative of any strong metaphysical connection (supervenience, grounding, identity, etc.). the correlation is merely statistical. while these may be useful to identify, we will consider only more-than-merely-statistical correlates as the ncc proper. if ncc research wants to contribute to finding the neural substrate of consciousness, additional constraints beyond correlation become attractive. these allow us to reject certain recorded statistical correlates as candidates for the ncc proper.1 in our view, most of the currently discussed constraints are limited in their applications or have other drawbacks.2 so here, we want to motivate a constraint that stems from the methodology of ncc research itself, inspired by an already prominent emphasis in this field on the structures of phenomenality. 1the main options currently under discussion as additional constraints are: the capacity to account for phenomenal features (seth, 2009; seth & edelman, 2009), synchronous occurrence with the phenomenal experience (aru et al., 2012), being systematically entailed by a theory (hohwy & seth, 2020), being necessary and sufficient (crick, 1995), and being minimally sufficient (chalmers, 2000). 2see, e.g., chalmers (2000) for a critique of crick (1995) and fink (2016) for a critique of chalmers (2000). fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a structural constraint on neural correlates of consciousness 3 the proximity of ncc research to psychophysics3 underpins a position we call neurophenomenal structuralism, which yields a structural similarity constraint on nccs. this constraint eliminates a broad range of statistical correlates as candidates for nccs proper. in section 2, we offer an overview of the emphasis on structures of experience in phenomenology and psychophysics, and show how they can be useful in individuating and identifying experiences. this motivates the position of neurophenomenal structuralism, which we sketch out in section 3. this position rests on the idea that (i) each experience is, in scientific contexts, structurally individuated and that (ii) there is a systematic relation between phenomenal structures and structures in the neural domain. a strong metaphysical reading of these twin ideas is possible: every phenomenal character is exhaustively individuated by its relations and not, as qualia theorists hold, by any intrinsic phenomenal properties. we are sympathetic to this reading, but focus here on a methodological point: in section 4, we propose that a structural morphism of a neural substrate with phenomenal spaces is an additional constraint on nccs proper. if phenomenal character is individuated structurally, and if phenomenal structure is mirrored in neural structure, then nccs proper must preserve structural relations of similarity and dissimilarity between experiences, while mere statistical correlates likely do not. we call this the structural similarity constraint (ssc). in subsection 4.3, we illustrate this constraint in action with an analysis of a study by brouwer & heeger (2009) on color perception. however, we believe our point extends to all phenomenal domains. 2 structures in phenomenology and psychophysics phenomenologically, experiences (e.g., as of colors, odors, tones) relate to each other in terms of similarity and dissimilarity. for example, seeing the color red resembles seeing orange more than it does seeing green; the smell of parmesan resembles the smell of vomit more than it does the smell of mint; the sound of the chord c-maj7 resembles a-min9 more than it does f♯-maj7. for a given domain of experiences (like a sensory modality or parts of it, e.g., colors in vision), we can map out these relative similarities and dissimilarities between types of experience across the domain. if we represent the resulting order geometrically, relative dissimilarity is associated with distance (see the left hand side of fig. 1 in section 4.3 for an example). that is why green is located opposite red on most hue circles: independently of the concrete visual scene, whenever one experiences red, it is the hue most dissimilar to actual or possible green experiences.4 some depictions of such orderings are widely known: the munsell 3to some degree, ncc research can be seen as a descendent of what fechner (1860) called “inner psychophysics.” 4billock et al. (2001) apparently produced reddish-green experiences in experimental settings (see also hardin, 2014). this means that red is not always most distant to green. but this does not speak against our overall point: we are not arguing for a specific ordering of experiences; we are fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink, lukas kob, and holger lyre 4 color spindle and newton’s or goethe’s color wheel are still widely used in art classes. many such attempts to map out a phenomenal domain predate our current scientific approaches in psychology, psychophysics, and neuroscience, especially when it comes to color (kuehni, 2003). within the sciences, psychophysics is the field most prominently concerned with charting the structures that govern the relative similarities between experiences, by investigating “just noticeable differences” (jnds) between them. by developing a wide range of extrospective methods, psychophysicists became able to represent these ordering structures in a finer-grained manner than artists had done introspectively (gescheider, 1997; lawless, 2013). granted, experiences differ—but what makes an experience specifically this one? this is a question of individuation. individuation is differentiation at the limit. by differentiating 𝑎 from 𝑏, we say why these two entities are not identical; by individuating 𝑎, we say what differentiates 𝑎 from all other entities. so if we know the individuation criteria for some 𝑎, we can say why 𝑎 is not identical with any other entity. individuation is thus a pathway to identification: if some 𝑥 fulfills all individuation criteria for 𝑎, then 𝑥 = 𝑎. how reliable this pathway is depends on the extent to which we can know whether 𝑥 fulfills all individuation criteria. some metaphysicians contemplate individuation criteria that are brute, irreducible, or even inaccessible. here, we will bracket such metaphysical debates and focus on a methodological question instead: how can we individuate? given an empirical method 𝑚, we can ask: how does 𝑚 help us to individuate those entities accessible via 𝑚? for psychophysics specifically, we can ask: how do we individuate experiences psychophysically? differentiation of experiences is a key element in psychophysics: psychophysicists measure our ability to match or discriminate experiences in experimental settings in order to extract the relations by which humans order their experiences based on relative similarity judgments implicit in their sorting behavior. for example, 𝑎 is just like 𝑏, but 𝑎 is just noticeably different than 𝑐, while 𝑏 is not, etc. such triangulating judgments allow us to model how a phenomenal domain is ordered, at least relative to the psychophysical methods used. we can call these orderings phenomenal structures5 if they are meant to reflect relations between types of phenomenal characters.6 in experiments, psychophysicists will relate these phenomenal structures to non-phenomenal goings-on. most often, the experiences to be matched or discriminated are elicited by external stimuli. the phenomenal strucsimply claiming that they are ordered. finding out the specific ordering that governs a domain might be a highly complex task. it might even turn out that a 3+-dimensional color ordering, as niederee (2010) suggests, is more adequate than a 3-dimensional one in explaining, e.g., the effects produced by billock et al. (2001) or color relations in tetrachromacy. 5these differ from the structures in an individual phenomenal experience token, such as the distribution of colors and shapes in a single visual impression. 6that is, they are not meant to reflect the relations between the causes of these experiences outside of the organism or between the stimuli outside of how they appear to us. obviously, these do matter for eliciting experiences, but psychophysical spaces are not always interpreted as merely reflecting an ordering in the external world. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a structural constraint on neural correlates of consciousness 5 tures are thereby automatically tracking structures of stimuli as well.7 however, we may also investigate experiences that do not correspond to any external or specific stimulus. chimerical colors (churchland, 2005) are a case in point, i.e., color sensations of afterimages that do not correspond to any reflective stimuli but that can nevertheless be systematically induced in a variety of ways. understood geometrically, these phenomenal structures comprise phenomenal spaces. these capture the difference and similarity structures of all phenomenal characters in a domain. each methodologically accessible phenomenal character has a location in a phenomenal space. such spaces are spanned by certain abstract dimensions. each individual in this space has a location which corresponds to coordinates along these dimensions. the labels we give these dimensions reflect the features in respect to which individuals differ or resemble each other. for colors, such orderings are called color spaces8 (kuehni, 2003; kuehni & schwarz, 2008) and hue, saturation, and lightness are dimensions along which colors can differ.9 in some cases (e.g., hue, saturation, and lightness) the dimensions of variability are introspectively available. however, in psychophysics, the dimensions are usually associated with features that are only extrospectively measurable. in “outer” psychophysics, the dimensions of the phenomenal spaces are related to properties of the stimuli, e.g., wavelengths, amplitudes, intensity. in “inner” psychophysics, these dimensions are associated with internal features of the system, i.e., neural features.10 accordingly, some color spaces are supposedly grounded in neural activation itself, for example the color-opponent processes in ganglion cells or the lgn (on dkl color space, see derrington et al., 1984; kuehni & schwarz, 2008, p. 148), or are linked to the cone responses of the retina (kuehni, 2003, pp. 164–168). it is reasonable to propose (i) that human experience possesses some kind of order (even though this order may be highly complex or otherwise hard to model ad7philosophers such as decock (2006) therefore suggest that these structures are redundant and must be interpreted as purely physical. 8the terms “color space” and “color solid” are both used. in churchland (2005) and kuehni & schwarz (2008), the relation between color solid and space appears to be that psychophysics provides us with highly complex and asymmetrical shapes, the color solids, which represent the order in which all perceivable colors relate to each other. these solids are a subspace inside a larger color space. however, one can think of phenomenal spaces themselves as determined by the structure governing phenomenal characters. in this understanding, the solid need not inhabit a space independent of it, but is simply a finite space. 9such color spaces derived from matching experiments have proven useful for industrial purposes (e.g., cie color space; kuehni & schwarz (2008) 141–142). for example, when it comes to choosing colors for buttons that initiate opposite actions (like start/stop), it is useful to choose colors from opposing extrema of a color space. color spaces have been operationalized in many different ways, and numerous color space models are therefore available, differing in their experimental origin, their metrics, and the explanations they provide. still, in each of these spaces, colors can be individuated structurally, namely by their location inside the overall space. 10this distinction was first drawn by fechner (1860). fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink, lukas kob, and holger lyre 6 equately), and (ii) that psychophysics methodically approximates this ordering.11 at the limit of psychophysical inquiry, an individual type of experience (like red42) is individuated by its relative difference to all other experiences in a domain, i.e., by its location in the overall structure of a phenomenal space, as established by some psychophysical method. this principle will hold not only for color, but for all modalities.12 3 neurophenomenal structuralism in this section we develop our proposal of neurophenomenal structuralism. it rests on two assumptions. first, it is the location in a structure, not an intrinsic property, that individuates a phenomenal character. this view could be called phenomenal structuralism and is introduced in subsection 3.1. second, phenomenal structure is mirrored in neural structure. this view leads to neurophenomenal structuralism and is explored in 3.2. a remarkable consequence of phenomenal structuralism (and therefore also of neurophenomenal structuralism) is phenomenal holism, which we develop in 3.3. 3.1 individuation by location in a structure the core idea of neurophenomenal structuralism is that experiences are structurally individuated. in contrast to this, some philosophers hold that experiences are intrinsically individuated. according to them, what makes an experience have a particular phenomenal character is not reliant on anything else; it is a nonrelational, non-extrinsic phenomenal property. this is the essence of the idea of qualia as introduced by c.i. lewis (1990). according to such intrinsicalists, structures—relations to other experiences, to stimuli, to our cognitive capacities, to our conceptual schema—might help us identify what experiences we are having, but they do not give the experiences their phenomenal character (levine, 1995). a broad range of philosophers (us included) disagree with this metaphysical position for several reasons. but here, we want to focus only on the methodological side. suffice it to say that, in the intrinsicalists’ picture, one must somehow access these intrinsic but individuating properties in order to truly make out the kind of experience one is dealing with. however this is supposed to work, such a kind of access can hardly be made plausible in a scientific context. ultimately, science deals with 11there are complications, however. for example: how do different color spaces relate to each other and to human experience? which color space reflects the true order of human color experience? and so on. these are valuable questions, but we lack the space to expand on them here. instead, we focus on methodology and are happy to accept that color spaces exist relative to methods for assessing phenomenal orderings. 12see also clark (2000) and rosenthal (2015). for a comparable take on olfaction, see young et al. (2014), and for the auditory system, see isaac (2018). color is simply the most studied example by far. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a structural constraint on neural correlates of consciousness 7 structures. in the science of psychophysics, experiences are individuated by their structural location, as stated in the previous section. but what is a structure, specifically? a structure is a set of relata with a set of relations imposed on them.13 a structure in this sense is a tuple σ = ⟨𝑎, 𝑅(𝑎)⟩ consisting of a carrier set or domain 𝑎 = {𝑎1, 𝑎2, ...𝑎𝑘} of 𝑘 elements or entities 𝑎𝑖 equipped with a set of 𝑛-ary relations 𝑅(𝑎). therefore, to capture a domain structurally is to individuate the entities or relata only via the relations in which they stand. now, given two domains, we can map entities in one domain onto entities in the other. some mappings will also preserve the structures in a domain, and such structure-preserving mappings are known as homomorphisms.14 of particular interest are bijective homomorphisms, called isomorphisms. if two structures are isomorphic, then they are structurally indistinguishable. in other words, any structure is defined only up to isomorphy. this matters if we care about individuation, i.e., if we care about the question of what determines that something is specifically this and not something else. do we need more than interrelations to make out what something is, or that it is this? structuralists say that structure is enough. all we can say for a given entity is that it is in this position in the structure. two views contrary to this have been prominent in traditional metaphysics: according to intrinsicalists, entities are individuated by intrinsic properties (properties that an object may have irrespective of the existence of other objects); according to haecceitists, identity is grounded in haecceity (a sort of “identity property” or “primitive this-ness”). structuralists deny both views: the relata in a relational structure possess neither any intrinsicality nor any haecceistic individuality, since they are captured only by their location in the overall structure. there is nothing to specifying what something is over and above stating its location in a structure. this basic structuralist claim accords with the first assumption of neurophenomenal structuralism: in the domain of the phenomenal, insofar as it is captured by psychophysical methods, a phenomenal character is identified by its location in a structure.15 in and of itself, this assumption motivates a phenomenal struc13strictly speaking, this is a relational structure. structured sets may come equipped with either functions or relations or both. if a structure contains no relations, it is an algebraic structure; if it does not contain any functions, it is a relational structure (or system of relations). throughout this paper we will use the term structure in the sense of relational structure. 14two relational structures σ𝑎 = ⟨𝑎, 𝑅(𝑎)⟩ and σ𝑏 = ⟨𝑏, 𝑆(𝑏)⟩ are of the same type, if for every relation 𝑅𝑗 ∈ 𝑅 there is a corresponding relation 𝑆𝑗 ∈ 𝑆 with the same arity. given two such structures, then a mapping ℎ ∶ 𝑎 → 𝑏 is a homomorphic mapping or homomorphism from σ𝑎 to σ𝑏, if for every relation 𝑅𝑗 and any elements 𝑎𝑖 ∈ 𝑎 the following implication holds: 𝑅𝑗(𝑎1, 𝑎2, … , 𝑎𝑘) ⇒ 𝑆𝑗(ℎ(𝑎1), ℎ(𝑎2), … , ℎ(𝑎𝑘)). 15wiese (2018) and tsuchiya et al. (2021), among others, have made remarks which partly mirror our approach. however, we do not straightforwardly share most of the additional constraints made by these authors. for example, we currently remain neutral on whether the structure must be considered metrical (as wiese does) and remain neutral on whether this structure must be captured in category theory (as tsuchiya et al. suggest). fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink, lukas kob, and holger lyre 8 turalism which can be accepted even by anti-physicalists because it does not yet say anything about phenomenality’s relation to any physical stimulus or neural matter. in ncc research, however, we have to consider the neural domain alongside the phenomenal. the connection between the two is the subject of our next section. 3.2 phenomenal structures are mirrored in neural structures scientific methodology captures the structure of a domain. psychophysics attempts to do so for phenomenal characters. but it does more than this: it also relates the structure of experiences either to the structures of external stimuli (in outer psychophysics) or to the structures of neural goings-on (in inner psychophysics). ncc research relates most closely to inner psychophysics because it aims to know which phenomenal characters relate to which neural activations. therefore, we shall consider two domains and their interrelationships:16 • the 𝑃 -domain of phenomenal states and processes (i.e., qualitative experience) • the 𝑁-domain of neural states and processes (i.e., neural activity) because processes can be considered transitions between states, the various domains are sufficiently modeled by the respective state spaces and the dynamics defined on them: • the 𝑃 -space (i.e., phenomenal space or quality space) • the 𝑁-space (typically a neural activation space) finally, since we are adopting a structuralist perspective, we shall focus on the respective structures of the two spaces: • the 𝑃 -structure (governing 𝑃 -space) • the 𝑁-structure (governing 𝑁-space) 16in experimental contexts, experiences are often but not always elicited by specific sensory stimuli. we might therefore consider a third domain of sensory stimuli. there are, however, exceptions. afterimages or metamers can be elicited by various stimuli. additionally, some experiences are elicited solely by interventions on the brain and are not tied to a specific sensory stimulus. these include experiences elicited by pharmacological interventions in psychedelic research, by direct brain stimulation, or by transcranial magnetic stimulation. the connection of experiences to stimuli is thus likely neither exhaustive nor exclusive, but certainly complex. here, we focus on the search for nccs in the context of inner psychophysics and therefore ignore the connection to external stimuli for the moment. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a structural constraint on neural correlates of consciousness 9 now, the second assumption of neurophenomenal structuralism says that 𝑃 structure is mirrored in 𝑁-structure. this needs to be explained in more detail. the notion of mirroring requires us to consider the interrelationships between 𝑃 and 𝑁-structures in terms of their structural similarities. in general, two structures 𝐴 and 𝐵 are structurally similar if the corresponding relations in 𝐴 and 𝐵 have the same number of arguments. paradigmatic cases of structural similarity are between maps, pictures, or sculptures and what they represent. of course, such structural similarity is a secondrather than a first-order similarity. as several authors have pointed out (o’brien & opie, 2004; shepard & chipman, 1970), first-order similarity consists of shared properties in both the source and the target domain. for example, colored chips are used by interior designers to select the intended color for painting a room. in the case of second-order similarity, however, it is only the relations between the relata (i.e., the structural properties) that are shared (or “mirrored”). in a weather map, for instance, the spacing of isobars corresponds to pressure gradients in the atmosphere. in an analogous manner, second-order similarity holds for the structural similarity between the phenomenal and the neural. in the case of color, to say that 𝑃 -structure is mirrored in 𝑁-structure does not, of course, mean that neurons or neural activations are in any (bizarre!) sense colored, but rather that changes in neighbouring color sensations (as encoded into color space) correspond to similar changes in neural activity (as encoded, for example, in the neural activation space). the strongest form of structural similarity between 𝑃 and 𝑁 would be an isomorphism. however, we must account for neural plasticity and so cannot ignore the multiple realizability of phenomenal states by neural states. multiple realizability is also at the heart of the ncc definition as introduced by chalmers (2000) in contrast to crick (1995): any specific neural activation that is an ncc counts only as sufficient, but not as necessary, for its associated phenomenal experience. therefore, rather than isomorphism, we should expect a surjective homomorphism from 𝑁 to 𝑃 , since a surjective mapping is a many-to-one mapping.17 surjective homomorphisms are also known as epimorphisms. the requirement of a surjective homomorphism accounts for the idea that phenomenal structure must be preserved when mapped onto the neural domain. the converse, however, need of course not be true. plenty of structures can certainly be found in the neural domain that do not feature in phenomenology (it is, for instance, of no importance what the structure of the neurons’ physical masses is). still, in order to provide a neural account of consciousness, as ncc researchers aim to do, all phenomenal structures must find a neural correspondence, but the 17given two sets 𝐴 and 𝐵, an injective mapping from 𝐴 to 𝐵 is a one-to-one mapping where we may also have elements of 𝐵 without matching 𝐴-elements. a surjective mapping from 𝐴 to 𝐵 means that every 𝐵-element has at least one matching element of 𝐴, but maybe more than one. hence, in general, a surjective mapping is a many-to-one mapping. a mapping that is both injective and surjective is called bijective. therefore, a bijective mapping amounts to a one-to-one correspondence where each element of one set is paired with exactly one element of the other set. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink, lukas kob, and holger lyre 10 𝑃 → 𝑁 correspondence can be one-to-many.18 conversely, a surjective, manyto-one 𝑁 → 𝑃 mapping accounts for the notorious multiple realizability of the phenomenal by the neural, as neural connections may be functionally replaced by others due to neural plasticity. this means, for instance, that even after certain neurons have died, one may continue to experience colors the way one used to. 3.3 phenomenal holism as we noted above, the phenomenal character of an experience is traditionally considered to be an intrinsic property of an experiencing subject’s mental state (i.e., a quale). qualia enthusiasts like goff (2017) claim that we identify a phenomenal experience by grasping this intrinsic property via introspection. neurophenomenal structuralism, by contrast, claims that to individuate a phenomenal character means to capture a relational rather than an intrinsic property.19 to illustrate this, consider color phenomenology. we should think of a color space not as representing absolute values, but rather as a means to represent relations between colors.20 similar views have been expressed by several other authors. isaac (2014) argues that we use relations between colors to flexibly represent changes and relations occurring in the environment. papineau (2015) reminds us that even though humans are said to be capable of distinguishing millions of colors, we are not able to distinguish them in an absolute fashion. while subjects are typically not able to distinguish between, say, blue-27 and blue-28 if shown one after the other, they might very well be able to discriminate them if shown side by side. gert (2017, p. 6) speaks of “brutely relational processes [that] yield relational information directly, without inferring it from non-relational information about the relata.” finally, clark (2000, p. 13f) says: when one sees a patch of orange, the experience seems to involve an intrinsic monadic quale: the quale orange. but this experience is an illusion. the facts in virtue of which that experience is an experience of orange, and not of some other quality, are all relational facts. […] qualitative content may be relational even though it is not invariably experienced as relational. […] the facts in virtue of which a sensation has the particular qualitative character it has are all relational facts. 18this opens an attractive door for reductionism, but we will leave this out for lack of space. additionally, there may also be a non-reductive reading of neurophenomenal structuralism. 19and because structure on 𝑃 -space has a correspondence in 𝑁-space, we might also do so extrospectively. 20note, however, that in the color realism debate the term ”relationalism” is frequently used to indicate that color is a property emerging in the relation between a physical object and a perceiver (e.g., chirimuuta, 2015, ch. 3). this is not what we have in mind here (although we are also not explicitly opposed to this view). our position is rather that color experiences are relationally individuated among themselves, even if the colors are hallucinated or dreamt of. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a structural constraint on neural correlates of consciousness 11 the “relational facts” that clark has in mind here might pertain to the external regime of color stimuli as well as to goings-on in the neural substrate (e.g., color opponency).21 while this is likely true, we here focus only on the relations governing the color space itself and for the moment ignore its relation to external stimuli22 or a neural substrate: the phenomenal character of each color is what it is due solely to its relation to all other colors. this is what we have called phenomenal structuralism. if a type of experience in a domain is individuated by its relations of graded similarity and difference to all other types of experience in that domain, as structuralists claim, then this has the remarkable implication of phenomenal holism: the 𝑃 -structure as a whole determines each internal relatum. hence, phenomenal holism is implied by phenomenal structuralism. consider again color phenomenology: any single color experience is constituted via its location within the whole color space. as clark (2000, p. 18) puts it: so orange can at best be defined as something like a reddish yellow, equally similar to red as to yellow, complementary to turquoise, more similar to red or to yellow than it is to turquoise, and so on. to draw an analogy: phenomenal structuralism implies phenomenal holism in much the same way as functional role semantics leads to meaning holism (block, 1998). meaning holism is concerned with the meaning of words in a language or the intentional contents of beliefs and thoughts. each word has a specific meaning only when we consider the meaning of all other words in the language because they stand in inferential and semantic relations to each other. in the linguistic context, holism is widely discussed; much less so when it comes to phenomenality. and this is precisely our project. meaning holism has been criticized, but these criticisms do not transfer to phenomenal holism. one standard argument against meaning holism is that meanings may become unstable, may vary among subjects, and thus cannot reliably be shared (fodor & lepore, 1992). this argument is by no means uncontroversial and can be countered (block, 1998). but in any case, this type of argument has no force in relation to holism about phenomenality. on the contrary: we should expect considerable variability in the individual 𝑃 -structure for members of a given species (fink, 2018; hohwy, 2011). from the perspective of neurophenomenal structuralism, this is not surprising and does not pose a problem: the structure governing an individual’s experiences is reliant on the structures of that individual’s nervous 21clark (2000, p. 138) argues that “[…] in sensory systems the root relation from which order derives is not similarity, but difference. instead of qualitative identity as the primitive term, the neural reality would be better reflected by using discriminability: the detection of differences,” where these differences are external. 22of course, we understand that color perception are ultimately anchored in external stimuli. but the relationship between stimuli and phenomenal character is tangled and complex. here, the focus is on ncc research, so we will bracket this complication. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink, lukas kob, and holger lyre 12 system; but neural structures vary inter-individually because the development of biological neural systems depends on individual and environmental circumstances. only by knowing a subject’s full 𝑃 -structure can we know what specifically they are indeed experiencing now. all else is approximation. in the case of color, such variations are known. people vary not only in the make-up of their retinae, but also in the size and interconnections of their visual cortices. neurophenomenal structuralists must therefore predict varying color experiences. and these show themselves in how people diverge in color judgments in everyday scenarios (hofer et al., 2005). additionally, high inter-observer variability in placing the unique hues has been empirically observed (kuehni, 2004). or consider color blindness, for instance the case of red-dichromacy (protanopia): if an entire way of eliciting color experiences is lost, this must show itself in the individual colors experienced.23 or how about adding new cones to the retina? experimental data indicate that there are human tetrachromats who pick a higher number of unique hues than their trichromatic conspecifics (jameson et al., 2001). phenomenal holism adds another twist. subjects with red–green color blindness do not only fail to distinguish red and green. if each color is individuated by its relation to all others, they will experience all colors differently, since their color space structure differs significantly from the color space structure of trichromats. neurophenomenal structuralism is in stark contrast to the classical qualia view. there, we identify experiences by accessing intrinsic properties which are available only via introspection (goff, 2017). phenomenal character can thus be assessed atomistically. all you ever experience could be one single color, but you would still know which color it was. structuralism, instead, leads to phenomenal holism: each experience has its phenomenal character only in virtue of other experiences having specific characters. only by accessing these relations can we identify phenomenal characters. therefore, even if (metaphysically) all one could ever experience were a single color, one could not know which color it was—or that it was a color at all. our method of identification is structural and holistic. let us now turn to the implications of this method for ncc research specifically. 4 from structuralism to structural constraints on nccs we began by pointing out that in psychophysics we capture perceptual experiences relationally in terms of our ability to match and differentiate them, even if the difference is only barely noticeable. we thereby construct phenomenal spaces that represent how similar or different any two basic experiences are, for instance the cie l*a*b* color space (see section 4.3). the fact that phenomenal experiences hang together in different degrees of similarity and difference is uncontroversial—it is common ground even with qualia 23for further details on the empirical claims of this paragraph, see hardin (2014). fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a structural constraint on neural correlates of consciousness 13 enthusiasts who consider qualitative feels to be additional, intrinsic properties that are not individuated by the structure they exist within. the qualia enthusiast may even concede that location within a structure may uniquely identify an experience, but not that this makes an experience what it is (levine, 1995). instead, qualia are supposed to be something additional to this structure. in scientific contexts, it suffices if we can uniquely identify basic experiences by their location in a structure. beyond this, scientists may be agnostic as to whether there is anything that individuates experiences other than their location in a structure. still, the type of structuralism employed in psychophysics has interesting methodological ramifications for how we investigate nccs and which neural areas count as candidates for nccs proper. 4.1 structural constraints on experiments in ncc research the search for nccs is at the heart of contemporary consciousness studies. the ncc of an experience 𝑐 is understood as the neural state 𝑠 that is minimally sufficient for a 𝑐-experience to occur. but if each experience can be determined only by its relations to all other experiences, we cannot search for the ncc of a single experience by itself : an isolated neural state can never be sufficient for a specific experience. without relational context, we cannot say what experience a neural state is a correlate of. therefore, we cannot search for the neural correlate of an individual color like red42. instead we first have to search for the neural correlate of the color space as a whole, i.e. the neural system that can be in states where some are minimally sufficient for each of the perceivable colors. but we have to consider the whole range of states of this neural system, as these are necessary to make any individual state minimally sufficient for this experience. only the neural correlate of the color space as a whole allows us to determine which neural state is the ncc for a specific experience like red42.24 this leads to two methodological suggestions for experiments that deal with stimuli to elicit experiences: (i) intra-individual holistic mapping: in order to specify which conscious state we are finding a correlate for when we record an individual’s brain activity elicited by a specific stimulus (e.g., a photon of 550 nm), we must first psychophysically map out which other perceptual experiences are possible for that individual. only this holistic mapping of experiences allows us to determine which experience is attached to this stimulus. 24here we call neural systems the nccs of a group of phenomenal types. but they are nccs only derivatively, due to the states they can be in, where some of their neural states correspond to types of experiences. whenever such a state is instantiated, this is the ncc of a token of that phenomenal type. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink, lukas kob, and holger lyre 14 (ii) inter-individual difference mapping: we cannot rule out the possibility that other people may experience more or less than we do, e.g., that dior tetrachromats may experience fewer or more shades of color. because (as phenomenal holism states) each phenomenal experience can be individuated only by its relations to all other experiences, this means that such individuals must also have different experiences than trichromats do, thanks to the differences in the structure of their nervous system. therefore, we have to capture inter-individual differences because only then can we determine which experience is attached to a specific stimulus in this specific individual. only if we map holistically can we pool across individuals and ensure that we find a more general neural correlate for a specific type of experience like red42. 4.2 structural constraints on nccs proper we began this paper with the question of how to distinguish mere statistical correlates from the ncc proper. here, neurophenomenal structuralism is helpful. first, let us stress that the second assumption of neurophenomenal structuralism—that phenomenal structure is mirrored in certain neural structures—is an undercurrent in contemporary neuroscience and supported by some empirical evidence (see also section 4.3). for example, the geometry of the visual field maps onto the geometry of the visual cortex (klein et al., 2004). based on such findings, inter-individual differences in experiences can be predicted by inter-individual neural differences. for instance, differences in the size of v1 (genç et al., 2015; schwarzkopf et al., 2011) or the parietal cortex (kanai et al., 2010) allow us to predict how long it will take for images to switch in a binocular rivalry paradigm for a single individual. this is likely not only to hold for the internal structure of individual experiential tokens (like visual impressions), but also to generalize to the structures governing phenomenal types (like colors). we argued in section 3 that due to the mere-sufficiency constraint of the ncc—according to chalmers (2000), a neural state is seen as merely sufficient for a phenomenal experience, and not necessary—this relation will take the form of a surjective homomorphism of the neural onto the phenomenal.25 if the relation between the neural and the phenomenal is generally systematic, we can expect that the neural correlate of a holistically structured phenomenal space should exhibit the structural features of this space. hence, the nccs proper of a phenomenal domain must somehow possess a structure corresponding to the relevant phenomenal space. the color-ncc proper must share the structure of the color space. if phenomenal state 𝑎 is grouped more closely to 𝑏 than to 𝑐, 25interestingly, haynes (2009) has argued in a comparable fashion for a “homeomorphic (i.e. topology-preserving) mapping between a set of experiences and the states of their core ncc,” but focused on spatial patterns in and between objects, not basic experiences like colors. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a structural constraint on neural correlates of consciousness 15 then the neural states corresponding to 𝑎, 𝑏, and 𝑐 must in some respect preserve this grouping.26 we can expect to find structural similarities between phenomenal spaces and neural activation spaces: neural data can be ordered in a space where the dimensions are the units of activation and a point in the spaces represents a particular pattern of activation code (churchland & sejnowski, 1992, p. 167 ff.). within the neural activation space of an ncc, closeness and distance are indicators of degrees of similarity between the associated phenomenal experiences, even if the neural systems instantiating these states are not in anatomical proximity. one illustration of such a mapping can be found in (churchland, 2005), where the structure of the munsell spindle is mapped onto a subspace of the activation space of color-opponent processing cells in the hurvich–james model. proximity in this activation space is here reliant not on spatial proximity of the opponent processing cells, but on similarity of their respective activations. thus the neurophenomenal structuralist’s approach suggests that structure is preserved if we cross over from phenomenal spaces to their neural substrates. conversely, any neural system that lacks structural features of the relevant phenomenal space fails to be an ncc proper. it might statistically co-occur or contribute to proper functioning, but it cannot be the neural substrate of the phenomenal space. we call this the structural similarity constraint on nccs (ssc): (ssc) neural substrates of phenomenal types must share the structure governing the phenomenal types they are associated with, i.e., the structure of a phenomenal space. this, we believe, is an important constraint because it allows for a methodological refinement in empirical research: we can now distinguish mere statistical correlates from the ncc proper. given that most structures patterning perception are highly complex and asymmetrical, we should not expect the same structure to be found in the state space of any arbitrary neural system. we can therefore use it to distinguish mere statistical correlates and artefacts from true neural substrates, since only the true neural substrates will capture all structural features patterning phenomenal experiences. structuralism therefore gives us an exclusion criterion, differentiating merely statistical correlates from substrates in ncc research. the distinction between substrates and statistical correlates might also apply to data collected by methods that are not temporally fine-grained. for example, even if a longer duration is captured in each “frame” of an fmri recording, only part of the collected data-chunk will be maximally isomorphic to the structure of the phenomenal space. neurophenomenal structuralism may therefore lead to an improvement in the methodology and tools used in ncc research. it may also lead to the rejection of several statistically correlating candidates as nccs proper. 26this will not always be as obvious as in the case of the visual field, where visual proximity maps largely onto proximity in v1. some mappings might be much more counterintuitive. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink, lukas kob, and holger lyre 16 these consequences are not mere speculation. indeed, we see this type of reasoning as implicit in some existing empirical studies. in the following we present an illustrative example from contemporary neuroscience. 4.3 an example from neuroscience: brouwer & heeger (2009) the idea that there are similarities between mental domains and their corresponding neural substrates is the second assumption of neurophenomenal structuralism (see section 3.3) and the core of the ssc. it is also a strong undercurrent in contemporary neuroscience. this idea is central to so-called representational geometry methods (kriegeskorte & kievit, 2013), where the assumption is that neural systems encode contents by integrating their (geometrical) structure into neural firing patterns. a research question guiding these approaches is: do activation patterns in relevant neural areas stand in relation to each other in the same way that experiences do? for example, are neural activation patterns linked to perceiving red more similar to neural activations associated with orange than to patterns associated with green?27 cichy et al. (2019) applied representational similarity analysis to investigate the fit between behavioral similarity ratings of objects on different dimensions (color, shape, function, etc.) and fmri as well as meg activation space in visual cortex; they found neural encodings of psychophysical similarity spaces (for further neuroscientific perspectives see haynes, 2009). the distance within such geometries is relative dissimilarity and their dimensions are defined by the specific aspects that are dissimilar. for our example here, representational geometry may measure perceived dissimilarities in the hues and relate it to neural dissimilarities such as how uncorrelated specific firing patterns are. in this way, neuroscientists (such as cichy et al., 2019) analyze whether there is a structural similarity between perceptual and neural structure. we argue that such a structural similarity constraint can be used to refine the concept of the ncc. the core idea is that if we knew the particular structure of experienced color dissimilarities, we would be able to use that structure to home in on the ncc proper of color experience because, according to the ssc, the ncc proper of any phenomenal domain must share the structure of the relevant phenomenal space. in the following, we focus on a study by brouwer & heeger (2009) which we see as illustrative of how the ssc may be used to distinguish mere statistical correlates of consciousness from nccs proper.28 {#cie} brouwer & heeger (2009) focused on color experiences: they investigated the similarity relations between neural activation patterns linked to color stimuli in the visual cortex and compared them with a color space—specifically, 27note that activation pattern similarities might abstract from brain topology, since such patterns can be distributed over large areas. 28we are not suggesting that the authors interpret their results in light of nccs, but we think that their study could be used to do exactly that. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a structural constraint on neural correlates of consciousness 17 figure 1: on the left side: the color stimuli used in brouwer & heeger (2009). on the right side: the result of the authors’ principal component analysis clustering in v1, v4, and mt+ as well as white noise. the similarity in v4 manifests as a closed, non-overlapping loop, corresponding most significantly with the similarity structure of the section in the chosen color space. (copyright 2009, society for neuroscience.) a section from the cie l*a*b* color space. this cie color space is designed to be perceptually uniform, i.e., the distances in this space reflect perceived degrees of dissimilarity between colors (commission internationale de l’eclairage, 1986). brouwer and heeger took a ring-shaped section from this space, consisting of eight individual color stimuli of equal lightness. these eight were chosen such that the similarity between each neighbor was equidistant in cie l*a*b* color space. additionally, each color was equidistant from neutral gray. thus, the colors shown to the subjects possessed a circular similarity structure (see the left side of fig. 1). they then measured fmri activation patterns for these eight colors. the fmri activation patterns in the visual cortex that corresponded to the chosen colors were plotted into an underlying two-dimensional space.29 voxel activations related to the same colors were clustered. for v4, those clusters formed a closed shape. thus, in v4, fmri activation patterns associated with red were more similar to orange-related patterns than to green-related patterns, mirroring the structural organization of the associated experiential domain (see right side of fig. 1). brouwer and heeger’s study offers an example of how researchers might use the structural similarity constraint (ssc) to distinguish statistical correlates from 29this space was obtained by principal component analysis. principal components are underlying dimensions that explain most of the variability within a dataset. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink, lukas kob, and holger lyre 18 nccs proper. according to the structural similarity constraint, some part of the ncc of color experience must encode the structure in the experiential domain as captured by psychophysics. brouwer and heeger used principal component analysis to display neural patterns in a two dimensional neural space in order to extract similarity patterns from the recorded fmri activation signals. according to this method, the activation of v4 displays a structure that resembles the specific hue circle which brouwer and heeger extracted from the cie space. other methods have already shown that v1 does process color but, in contrast to v4, v1 activation does not show the circular, roughly equidistant structure prominent in experiences elicited by the color stimuli. region mt+ exhibited only randomly distributed patterns. so while v4 preserves the closed, non-overlapping loop-structure of the hue circle, v1 fails to show this structure.30 from the perspective of our ssc criterion, we can interpret the above findings in the following way: even though its activation statistically correlates with color experiences, v1 is likely not an ncc proper for color experience because it orders colors in a different way than we experience them. v1 does process color, but likely in a phase preceding color experience. mt+, on the other hand, does not process color in any way. thus, within the visual cortex (and in the context of fmri), the ssc favors v4 as a candidate for an ncc proper of color experience. compared to other areas, it best encodes the structure of a specific phenomenal color space in terms of its circular form and clustering. based on the ssc, we take the study by brouwer & heeger (2009) to indicate that v4 is the currently best candidate for an ncc proper of color experience, while other areas such as v1 are likely mere statistical correlates. 4.4 limitations and applications can the ssc be applied outside of specific experiments and contribute to broadscale theoretical discussions? to a degree. the ssc may usefully contribute to the debate around frontal versus nonfrontal theories of consciousness, where the question is whether nccs extend to the non-sensory areas of the brain, especially the frontal cortex. frontalists claim that they must–in part because the projection of information into the frontal areas may facilitate the formation of higher-order thoughts, which are supposed to make mental qualities conscious. non-frontalists argue that this conflates phenomenal consciousness with access consciousness, which is all that is facilitated by the projection to the frontal areas. frontalists who accept the ssc must expect the prefrontal cortex, a region linked to higher cognitive functions, to re-encode the structural color information available in v4. this is certainly possible. if it does not, the prefrontal cortex will 30interestingly, v1 showed better decoding results than v4. this might indicate that mere information processing diverges from structural content. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a structural constraint on neural correlates of consciousness 19 not count as an ncc proper according to the ssc. still, it could be that prefrontal cortex activation is needed for consciousness even though it might not encode that very structure (haynes, 2009). for instance, it could function as an enabling condition. one interpretation available to frontalists in this scenario would be that the structure-preserving ncc proper of colors in the sensory areas is a kind of content condition, determining what specifically we experience. this would make it similar to the neural correlate of content (ncc) which marvan and polák (2020) suggest needs to be added to a general neural correlate of consciousness (gncc). how decisive is the ssc? again, the ssc is only a comparative constraint: v4 does not perfectly capture the structure of the color space used by brouwer and heeger. still, compared to the other areas, it captures the similarity structure of the experience best. therefore, v4 is the currently best candidate for an ncc proper. this comparison by ssc may, however, reach a limit without yielding one and only one ncc proper. neural areas other than v4 might fulfill the ssc for color equally well if compared against v4. we could then not use ssc to decide whether this hypothetical area or v4 is the ncc of color consciousness. but this is to be expected: the classical ncc definition explicitly allows that several neural areas can bring about the same experience. still, according to the ssc, any area that counts as minimally sufficient would need to encode the corresponding phenomenal structure. the ssc is thus a feature that needs to be shared by all individual nccs, each of which is only minimally sufficient. this aligns with the concept of the ncc 2.0 proposed by fink (2016): any token ncc must share features that make it an ncc of a specific type of experience. manifesting structural similarities with the relevant phenomenal spaces might just be such a feature. what are limitations of the ssc? admittedly, there are some, and they are mainly imposed by the granularity and reliability of the methods we currently have available. first, there are limitations on the degree to which psychophysics captures the one true structure of phenomenal experiences. as a matter of fact, different psychophysical methods lead to phenomenal spaces with slightly different structures (see kuehni, 2003 for a survey on color spaces).31 all of them capture some feature of our experience. which is the right one? this will be hard to determine and more foundational theoretical work is needed. maybe phenomenal structure is so simple that only those features that are invariant across methods truly grasp phenomenal structures, while others are artefacts. maybe phenomenal structure is so complex that each psychophysical method captures only part of it. maybe phenomenal structure is fluid and shifting such that there is no unique structure. currently, it is too early to tell. at the moment, we can only say that insofar as psychophysics captures some structures of experience, these structures must be shared by the relevant ncc. methodologically, the ssc thus makes nccs 31brouwer & heeger (2013) were able to replicate their initial findings with another color space, the dkl space (derrington et al., 1984). again, v4 delivered the best match compared to other visual areas like v1, v2, and v3. this suggests that there is some transferability of these results to other color space models. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink, lukas kob, and holger lyre 20 relative to a method for capturing phenomenal structures. unless we have methods that allow us to distinguish psychophysical artefacts from those structures that are adequately captured, there is hardly a way to avoid this. but this is a general point that applies to all methods capturing first-person phenomena, not only to the ssc: all might produce artefacts. a second limitation of the ssc concerns the methods for measuring neural activation. if these are not fine-grained enough, we may miss structural similarities, for instance on the level of individual neurons rather than columns. or we might see similarities with phenomenal spaces in neural activation patterns that are only artefacts. here, we can only hope for further progress in brain measurement techniques. without it, the ssc may not prove decisive for a large swath of nccs. additionally, earlier exclusions based on the ssc may need to be revised in the light of new methodological approaches. but this is not uncommon in neuroscience. third, taking the two previous points together, if phenomenal structures are captured only in a very simplified fashion (for instance, if the mere cardinality of distinguishable experiences is captured, but not relative similarity or dissimilarity), we might find structural similarities in any area of the brain. the ssc would then never be decisive. this may be the case for some very simple sensory modalities, like our feeling for up versus down or hunger versus satiety. however, for most experiences, and especially colors, the relevant psychophysical spaces show a lot of complexity and asymmetry, which will hardly be preserved by all areas in the brain. again, improvements in methods for capturing phenomenal structures as well as brain structures may make the ssc more and more decisive over time. in summary, we claim simply that the ssc helps to limit ncc candidate areas comparatively. the ssc is not decisive in and of itself, but adds to other constraints that aid us in selecting the areas that serve as true neural correlates of a certain kind of specific experience. the ssc does not pick out a single ncc proper, nor does it pinpoint a general mechanism for all kinds of consciousness, but it does identify specific phenomenal domains with certain structures of sufficient complexity. and here, again, the ssc is constrained by the granularity of our methodology. still, we hope to have illustrated via our reinterpretation of brouwer & heeger’s (2009) findings that the ssc is valuable because it constrains what can count as an ncc proper. 5 conclusion experiences stand in complex relations to each other, which are captured by psychophysics using a variety of methods. in contrast to qualia intrinsicalism, phenomenal structuralism claims that it is this net of relations, the structure governing the types of experience we can have, that helps us to individuate any single experience. this entails a methodological holism, because we need to know the structure as a whole in order to determine the nature of the relata. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a structural constraint on neural correlates of consciousness 21 neurophenomenal structuralism comes with the additional thesis that these phenomenal relations are mirrored in neural structures. motivated by the empirically inspired mere-sufficiency constraint of chalmers’ ncc, we expect a surjective homomorphism of the neural onto the phenomenal. structuralism is very much in tune with empirical methods like representational similarity analysis and the close historical lineage of ncc research to inner psychophysics, and is a good fit for many empirical findings. finally, neurophenomenal structuralism motivates a structural similarity constraint (ssc) that can contribute to ncc research. since phenomenal structure must be mirrored in neural structure, we should exclude as nccs proper all neural activation in areas that fails to capture the relevant phenomenal space structures. this structural similarity constraint on nccs is, we believe, an important step towards reliably distinguishing neural substrates from statistical noise in neural data. acknowledgments we are grateful for the comments of two anonymous reviewers. this research was funded by the deutsche forschungsgemeinschaft (dfg, german research foundation) 337619223 / rtg2386. we also want to thank emily troscianko for her editorial support. references aru, j., bachmann, t., singer, w., & melloni, l. (2012). distilling the neural correlates of consciousness. neuroscience and biobehavioral reviews, 36(2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.12.003 billock, v. a., gleason, g. a., & tsou, b. h. (2001). perception of forbidden colors in retinally stabilized equiluminant images: an indication of softwired cortical color opponency? journal of the optical society of america, 18(10), 2398– 2403. https://doi.org/10.1364/josaa.18.002398 block, n. (1998). semantics, conceptual role. in e. craig (ed.) routledge encyclopedia of philosophy. routledge. https: //doi.org/10.4324/9780415249126-w037-1 brouwer, g. j., & heeger, d. j. (2009). decoding and reconstructing color from responses in human visual cortex. the journal of neuroscience, 29(44), 13992–14003. https://doi.org/10.1523/jneurosci.3577-09.2009 brouwer, g. j., & heeger, d. j. (2013). categorical clustering of the neural representation of color. the journal of neuroscience, 33(39), 15454–15465. https://doi.org/10.1523/jneurosci.2472-13.2013 chalmers, d. j. (2000). what is a neural correlate of consciousness? in t. metzinger (ed.), neural correlates of consciousness (pp. 17–39). the mit press. chirimuuta, m. (2015). outside color: perceptual science and the puzzle of color in philosophy. the mit press. churchland, p. (2005). chimerical colors: some phenomenological predictions from cognitive neuroscience. philosophical psychology, 18(5), 527–560. https://doi.org/10.1080/09515080500264115 churchland, p. s., & sejnowski, t. j. (1992). the computational brain. the mit press. cichy, r. m., kriegeskorte, n., jozwik, k. m., van den bosch, j. j. f., & charest, i. (2019). the spatiotemporal neural dynamics underlying perceived similarity for real-world objects. neuroimage, 194, 12–24. https://doi.org/10.1016/j.neuroimage .2019.03.031 clark, a. (2000). a theory of sentience. oxford university press. https://doi.org/10.1093/acprof:oso/9780198238515.001.0001 commission internationale de l’eclairage. (1986). cie 15.2., colorimetry. (2nd ed.) commission internationale de l’eclairage. crick, f. (1995). the astonishing hypothesis: the scientific search for the soul. scribner. decock, l. (2006). a physicalist reinterpretion of ‘phenomenal’ spaces. phenomenology and the cognitive sciences, 5(2), 197–225. https://doi.org/10.1007/s11097-005-9006-7 fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neubiorev.2011.12.003 https://doi.org/10.1364/josaa.18.002398 https://doi.org/10.4324/9780415249126-w037-1 https://doi.org/10.4324/9780415249126-w037-1 https://doi.org/10.1523/jneurosci.3577-09.2009 https://doi.org/10.1523/jneurosci.2472-13.2013 https://doi.org/10.1080/09515080500264115 https://doi.org/10.1016/j.neuroimage.2019.03.031 https://doi.org/10.1016/j.neuroimage.2019.03.031 https://doi.org/10.1093/acprof:oso/9780198238515.001.0001 https://doi.org/10.1007/s11097-005-9006-7 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink, lukas kob, and holger lyre 22 derrington, a. m., krauskopf, j., & lennie, p. (1984). chromatic mechanisms in lateral geniculate nucleus of macaque. the journal of physiology, 357, 241–265. https://doi.org/10.1113/jphysiol.1984.sp015499 fechner, g. t. (1860). elemente der psychophysik. breitkopf und härtel. fink, s. b. (2016). a deeper look at the "neural correlate of consciousness". frontiers in psychology, 7, 1044. https://doi.org/ 10.3389/fpsyg.2016.01044 fink, s. b. (2018). introspective disputes deflated: the case for phenomenal variation. philosophical studies, 175(12), 3165– 3194. https://doi.org/10.1007/s11098-017-1000-8 fodor, j. a., & lepore, e. (1992). holism: a shopper’s guide. blackwell. genç, e., bergmann, j., singer, w., & kohler, a. (2015). surface area of early visual cortex predicts individual speed of traveling waves during binocular rivalry. cerebral cortex, 25(6), 1499–1508. https://doi.org/10.1093/cercor/bht342 gert, j. (2017). quality spaces: mental and physical. philosophical psychology, 30(5), 525–544. https://doi.org/10.1080/0951 5089.2017.1295303 gescheider, g. a. (1997). psychophysics: the fundamentals. (3rd ed.) routledge. goff, p. (2017). do electrons dream of electric sheep? https://blogs.lse.ac.uk/theforum/go%0b-do-electrons-dream/ hardin, c. l. (2014). more color science for philosophers. in d. stokes, m. matthen, & s. biggs (eds.), perception and its modalities (pp. 379–390). oxford university press. https://doi.org/10.1093/acprof:oso/9780199832798.003.0016 haynes, j.-d. (2009). decoding visual consciousness from human brain signals. trends in cognitive sciences, 13(5), 194–202. https://doi.org/10.1016/j.tics.2009.02.004 hofer, h., singer, b., & williams, d. r. (2005). different sensations from cones with the same photopigment. journal of vision, 5(5), 444–454. https://doi.org/10.1167/5.5.5 hohwy, j. (2011). phenomenal variability and introspective reliability. mind & language, 26(3), 261–286. https://doi.org/10 .1111/j.1468-0017.2011.01418.x hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii). https://doi.org/10.33735/phimisci.2020.ii.64 isaac, a. m. c. (2014). structural realism for secondary qualities. erkenntnis, 79(3), 481–510. https://doi.org/10.1007/s10670013-9519-3 isaac, a. m. c. (2018). prospects for timbre physicalism. philosophical studies, 175(2), 503–529. https://doi.org/10.1007/s110 98-017-0880-y jameson, k. a., highnote, s. m., & wasserman, l. m. (2001). richer color experience in observers with multiple photopigment opsin genes. psychonomic bulletin & review, 8(2), 244–261. https://doi.org/10.3758/bf03196159 kanai, r., bahrami, b., & rees, g. (2010). human parietal cortex structure predicts individual differences in perceptual rivalry. current biology, 20(18), 1626–1630. https://doi.org/10.1016/j.cub.2010.07.027 klein, i., dubois, j., mangin, j.-f., kherif, f., flandin, g., poline, j.-b., denis, m., kosslyn, s. m., & le bihan, d. (2004). retinotopic organization of visual mental images as revealed by functional magnetic resonance imaging. brain research. cognitive brain research, 22(1), 26–31. https://doi.org/10.1016/j.cogbrainres.2004.07.006. kriegeskorte, n., & kievit, r. a. (2013). representational geometry: integrating cognition, computation, and the brain. trends in cognitive sciences, 17(8), 401–412. https://doi.org/10.1016/j.tics.2013.06.007 kuehni, r. g. (2003). color space and its divisions: color order from antiquity to the present. john wiley & sons. kuehni, r. g. (2004). variability in unique hue selection: a surprising phenomenon. color research & application, 29(2), 158–162. https://doi.org/10.1002/col.10237 kuehni, r. g., & schwarz, a. (2008). color ordered: a survey of color order systems from antiquity to the present. oxford university press. lawless, h. t. (2013). quantitative sensory analysis: psychophysics, models and intelligent design. wiley blackwell. levine, j. (1995). qualia: intrinsic, relational, or what? in t. metzinger (ed.), conscious experience (pp. 277–292). ferdinand schöningh. lewis, c. i. (1990). mind and the world-order: outline of a theory of knowledge (unabridged republication of the 1929 edition). dover publications. marvan, t., & polák, m. (2020). generality and content-specificity in the study of the neural correlates of perceptual consciousness. philosophy and the mind sciences, 1(ii). https://doi.org/10.33735/phimisci.2020.ii.61 niederée, r. (2010). more than three dimensions: what continuity considerations can tell us about perceived color. in j. cohen & m. matthen (eds.), color ontology and color science. the mit press. https://doi.org/10.7551/mitpress/7541.00 3.0006 o’brien, g., & opie, j. (2004). notes toward a structuralist theory of mental representation. in h. clapin, p. staines & p. slezak (eds.), representation in mind (pp. 1–20). elsevier. https://doi.org/10.1016/b978-008044394-2/50004-x fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1113/jphysiol.1984.sp015499 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.1007/s11098-017-1000-8 https://doi.org/10.1093/cercor/bht342 https://doi.org/10.1080/09515089.2017.1295303 https://doi.org/10.1080/09515089.2017.1295303 https://blogs.lse.ac.uk/theforum/go%0b-do-electrons-dream/ https://doi.org/10.1093/acprof:oso/9780199832798.003.0016 https://doi.org/10.1016/j.tics.2009.02.004 https://doi.org/10.1167/5.5.5 https://doi.org/10.1111/j.1468-0017.2011.01418.x https://doi.org/10.1111/j.1468-0017.2011.01418.x https://doi.org/10.33735/phimisci.2020.ii.64 https://doi.org/10.1007/s10670-013-9519-3 https://doi.org/10.1007/s10670-013-9519-3 https://doi.org/10.1007/s11098-017-0880-y https://doi.org/10.1007/s11098-017-0880-y https://doi.org/10.3758/bf03196159 https://doi.org/10.1016/j.cub.2010.07.027 https://doi.org/10.1016/j.cogbrainres.2004.07.006. https://doi.org/10.1016/j.tics.2013.06.007 https://doi.org/10.1002/col.10237 https://doi.org/10.33735/phimisci.2020.ii.61 https://doi.org/10.7551/mitpress/7541.003.0006 https://doi.org/10.7551/mitpress/7541.003.0006 https://doi.org/10.1016/b978-008044394-2/50004-x https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org a structural constraint on neural correlates of consciousness 23 papineau, d. (2015). can we really see a million colours? in p. coates & s. coleman (eds.), phenomenal qualities (pp. 274–297). oxford university press. https://doi.org/10.1093/acprof:oso/9780198712718.003.0010 pearson, k. (1897). mathematical contributions to the theory of evolution.—on a form of spurious correlation which may arise when indices are used in the measurement of organs. proceedings of the royal society of london, 60, 489–498. https://doi.org/10.1098/rspl.1896.0076 rosenthal, d. (2015). quality spaces and sensory modalities. phenomenal qualities: sense, perception, and consciousness, 33–65. schwarzkopf, d. s., song, c., & rees, g. (2011). the surface area of human v1 predicts the subjective experience of object size. nature neuroscience, 14(1), 28–30. https://doi.org/10.1038/nn.2706 seth, a. (2009). explanatory correlates of consciousness: theoretical and computational challenges. cognitive computation, 1(1), 50–63. https://doi.org/10.1007/s12559-009-9007-x seth, a. k., & edelman, g. m. (2009). consciousness and complexity. in r. a. meyers (ed.), encyclopedia of complexity and systems science (pp. 1424–1443). springer new york. shepard, r. n., & chipman, s. (1970). second-order isomorphism of internal representations: shapes of states. cognitive psychology, 1(1), 1–17. https://doi.org/10.1016/0010-0285(70)90002-2 tsuchiya, n., phillips, s., & saigo, h. (2021). enriched category as a model of qualia structure based on similarity judgements. osf preprints. https://doi.org/10.31219/osf.io/5peu7 wiese, w. (2018). toward a mature science of consciousness. frontiers in psychology, 9, 693. https://doi.org/10.3389/fpsyg. 2018.00693 young, b. d., keller, a., & rosenthal, d. (2014). quality-space theory in olfaction. frontiers in psychology, 5, 1. https: //doi.org/10.3389/fpsyg.2014.00001 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. fink, s. b., kob, l., & lyre, h. (2021). a structural constraint on neural correlates of consciousness philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1093/acprof:oso/9780198712718.003.0010 https://doi.org/10.1098/rspl.1896.0076 https://doi.org/10.1038/nn.2706 https://doi.org/10.1007/s12559-009-9007-x https://doi.org/10.1016/0010-0285(70)90002-2 https://doi.org/10.31219/osf.io/5peu7 https://doi.org/10.3389/fpsyg.2018.00693 https://doi.org/10.3389/fpsyg.2018.00693 https://doi.org/10.3389/fpsyg.2014.00001 https://doi.org/10.3389/fpsyg.2014.00001 https://doi.org/10.33735/phimisci.2021.79 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction: statistical vs. proper correlates of consciousness structures in phenomenology and psychophysics neurophenomenal structuralism individuation by location in a structure phenomenal structures are mirrored in neural structures phenomenal holism from structuralism to structural constraints on nccs structural constraints on experiments in ncc research structural constraints on nccs proper an example from neuroscience: brouwer & heeger (2009) limitations and applications conclusion self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. self and knowledge in psychedelic therapy reply to commentaries on philosophy of psychedelics chris lethebya,b (chris.letheby@uwa.edu.au) abstract much of my book philosophy of psychedelics is devoted to elaborating and defending two basic claims: that psychedelic therapy works mainly by changing mental representations of the self, and that it has many epistemic benefits consistent with a naturalistic worldview. the commentaries in this symposium generally focus on one or the other of these claims. on the mechanistic front, the commentaries by hoffman and by martin and sterzer seek to supplement my account by drawing attention to factors it does not emphasize. in this reply i agree with most of their claims, but propose some important qualifications. meanwhile, the commentaries by lyon and farennikova and by colombo challenge some core commitments of my account. i agree with many of their claims, but defend some of these core commitments. on the epistemological front, bortolotti and murphy-hollies endorse the overall shape of my account and make some productive connections to ideas concerning agency-first epistemology and self know-how. caporuscio brings into play the idea of self-shaping, which connects to self know-how, but argues against certain core claims of my account. i endorse many of these claims about agency-first, self know-how, and self-shaping, but propose some important qualifications. finally, fink agrees that psychedelic experiences can have some epistemic benefits, but argues that these do not pertain to epistemic justification, and that the main benefit is increased understanding. i am sympathetic to fink’s proposal about understanding, but criticize some of his arguments for the thesis of psychedelic justification impossibilism. keywords ayahuasca ∙ dmt ∙ epistemic agency ∙ epistemic justification ∙ epistemology ∙ hallucinogen ∙ lsd ∙ mescaline ∙ philosophy ∙ predictive processing ∙ psilocybin ∙ psychedelics ∙ self know-how ∙ selfrepresentation ∙ self-shaping ∙ understanding this article is part of a symposium on chris letheby’s book “philosophy of psychedelics” (oup 2021), edited by chiara caporuscio and sascha benjamin fink. athe university of western australia. bthe university of adelaide. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-6293-7873 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 2 1 introduction i would like to begin by thanking all the contributors to this symposium for their generous and constructive engagement with the ideas presented in philosophy of psychedelics (henceforth pop, letheby, 2021). time constraints inherent to the production of a symposium such as this have inevitably limited my reflection on their arguments. to paraphrase bilbo baggins from tolkien’s lord of the rings, i have thought about less than half of these arguments half as deeply as i should like, and here i respond to less than half of them half as adequately as they deserve. the bulk of pop is devoted to elaborating and defending two basic claims: that psychedelic therapy works mainly by changing mental representations of the self, and that it has many epistemic benefits consistent with a naturalistic worldview. in general, the commentaries address one or the other of these claims1; accordingly, i have structured this reply into two main sections. in the interests of brevity, i do not recapitulate the commentators’ arguments in great detail, except where there are substantial points of disagreement. i assume that readers of this reply have read the book (or the précis) and the commentaries. 2 mechanisms the commentaries i discuss in this section vary in how strongly they challenge my claims about the causal mechanisms of psychedelic therapy. the commentaries by hoffman and by martin and sterzer seek to supplement my account by drawing attention to factors it does not emphasize. meanwhile, the commentaries by lyon and farennikova and by colombo challenge core commitments of my account, to one degree or another. 2.1 positive affect hoffman (2022) endorses the general shape of the account of psychedelic therapy sketched in pop, but argues that it does not sufficiently emphasize the role of positive affect, and that a complete account will need to do so. here, i am in agreement. she also argues that exploring the role of affect calls into question some of my claims concerning ego dissolution; on this point, i am less convinced. it is true that i do not strongly emphasize or deeply explore the affective dimensions of psychedelic therapy in pop, an omission that i briefly acknowledge (letheby, 2021, p. 202). hoffman cites several studies linking affective factors to clinical outcomes. these findings suggest that a complete theory of psychedelic therapy must account for affective factors. as hoffman notes, the plausible link between affect and self-representation means this observation is consistent with 1the exception is johnson’s (2022) wonderful introduction, ‘psychedelic science needs philosophy’, with which – of course! – i have no disagreements of note. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org self and knowledge in psychedelic therapy 3 the overall shape of my account: the account needs to be expanded to include affective factors, not fundamentally revised. thus far, i agree entirely. hoffman explores the role of affective factors via a comparison of psychedelic therapy and mdma-assisted therapy, arguing that this comparison reveals some potential weaknesses in my account. i hold that psychedelics, by disrupting selfbinding processes, move the phenomenal self-model along the continuum from phenomenal transparency to opacity. hoffman comments that “this way of describing things requires a commitment to the predictive processing [pp] account, which opens the theory to objections” (hoffman, 2022). but i do not think that it does require such a commitment. the idea that transparent phenomenal selfmodelling results from a binding process, and that psychedelic-induced unbinding renders self-modelling opaque, is independent of pp. pp is one way of spelling out the details – but the self-binding theory of self-representation was developed independently of pp (sui & humphreys, 2015) and internalist “controlled hallucination” views of phenomenal consciousness have been, too (e.g. revonsuo, 2006). pp is one influential basis for this kind of view of phenomenal consciousness and self-consciousness, but the general view can be argued for independently of pp (metzinger, 2004; revonsuo, 2006) – indeed, i think psychedelic experience tends to support it. hoffman also says that my account of the role of phenomenal opacity in psychedelic therapy “hides an ambiguity between cognitive and phenomenal opacity” (hoffman, 2022). while the term “cognitive opacity” is hoffman’s, the distinction originates in metzinger (2004). the distinction is as follows: while phenomenal opacity involves experiencing something as a representation, cognitive opacity involves merely thinking, or judging, that something is a representation (hoffman, personal communication). invoking this distinction, hoffman comments that “metacognitions about the self will always make those parts of the self-model that are their objects more opaque but not typically phenomenally opaque” (hoffman, 2022). thus, she contends, while self-model opacity of some kind is plausibly involved in psychedelic therapy, it may not always be phenomenal opacity. perhaps this is correct; perhaps psychedelic therapy sometimes involve merely cognitive opacity of the self-model. it still seems to me that there is an important difference between experientially-backed cognitive opacity and the kind one can get from, e.g., book-learning. a and b may both believe that their sense of self is continually constructed by a mental modelling process, and is therefore malleable. but if a’s belief results from reading philosophy and watching anil seth’s ted talk, while b’s results from a psychedelic experience of self-unbinding, then b’s is likely, ceteris paribus, to carry deeper conviction and have a greater existential and behavioural impact. of course, this conjecture could be tested empirically, if we could find a way to measure these different kinds of opacity. i also think that mere cognitive opacity, and full-blown phenomenal opacity, can both result from disruption to predictive self-binding. this hypothesis is not letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 4 undermined by the observation that mdma (a) seems to induce therapeutic effects via some kind of self-unbinding, but (b) rarely, if ever, induces full-blown mystical experiences, ego dissolution, and global phenomenal opacity. this observation is consistent with the idea that disruption to self-binding is key to the therapeutic effects of both mdma and psychedelics, even if the latter can induce types and degrees of unbinding that the former rarely does (i.e. disrupt more aspects of the selfmodel, and to greater extents). despite the similarities hoffman notes, there are important differences between not just psychedelics and mdma, but psychedelicassisted and mdma-assisted therapy. for instance, hoffman’s description of “inducing a blissful state to facilitate productive engagement with painful material” (hoffman, 2022) may be a fair description of mdma’s therapeutic mechanisms, but it is less clear that it adequately characterizes typical therapeutic psychedelic experiences. notwithstanding the undeniable role of positive valence in the latter, many parts of these experiences are far from blissful. in some cases an undesired but unavoidable confrontation with painful material is involved, and seems to be facilitated not by a blissful state but by a diminished capacity to avoid engaging with uncomfortable mental contents – plausibly a disruption to self-binding, but not an increase in positive affect. in sum, positive affect is an important factor in psychedelic therapy that is underexplored in pop, and comparing and contrasting psychedelic-assisted and mdma-assisted therapy seems like a fruitful direction for future research. but total ego dissolution may still be the limit case of an unbinding mechanism that underlies the therapeutic effects of both substances – even if taking this mechanism to its limits is not therapeutically necessary. in section 2.4 below i formulate a distinctive empirical prediction of the self-binding account of psychedelic therapy; perhaps formulating predictions of this kind could allow us to contrastively test some of hoffman’s and my claims about similarities and differences between psychedelicand mdma-assisted therapy2. 2.2 context and strengthened beliefs like hoffman, martin and sterzer (m&s) (2022) seek to supplement my account with attention to under-emphasized factors, not to challenge its core tenets. here the under-emphasized factors are (a) the role of strengthened priors and (b) the influence of context on the kinds of alternative self-models that patients encounter during therapeutic experiences. i am almost entirely in agreement with m&s, and have little to add to their discussion. as they note, the importance of strengthened priors can be made compatible with the rebus (relaxed beliefs under psychedelics) model, “so long as these strong prior influences emerge at lower levels in response to the relaxed constraining influence of priors at the highest levels of the system […] the rebus model may involve both ‘strong’ 2i am indebted to matteo colombo for raising this point. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org self and knowledge in psychedelic therapy 5 and ‘weak’ prior influences, which are present at different hierarchical levels to different degrees” (martin & sterzer, 2022). i think this is the correct line to take. it is true that psychedelics strengthen some priors, as well as weakening others (safron, 2020); that this is important to understanding their experiential and therapeutic effects; and that they can be “generally characterised as weakening the constraining influence of priors encoding rigid highly-abstract beliefs of the world and self, and (relatively) strengthening prior beliefs associated with contextual sources, leading to an amplified influence of ‘set’ and ‘setting’ on the mental state of the user” (martin & sterzer, 2022). i also agree that the account i develop under-emphasizes strengthened priors, and that investigating this factor in more detail can help fill out the account by accounting for the crucial role of context in psychedelic therapy. the cognitive playing field is, indeed, not totally level in the psychedelic experience or in the post-acute phase of psychedelic therapy, and something must explain why certain specific (often healthier) forms of self-modelling tend to emerge, rather than others. surely, as m&s say, therapeutic expectations, intentions, and settings will figure importantly in the explanation. i would add, however, that these may not be the only important factors. there are many anecdotal accounts of people discovering and consolidating healthier forms of self-modelling after psychedelic experiences that lacked any therapeutic intent or setting. in this light, consider that many patients do not describe discovering new forms of self-modelling, but rediscovering old, forgotten, or disused forms. as one patient put it: “at [its] most basic, i feel like i used to before the depression” (watts, day, krzanowski, nutt, & carhart-harris, 2017, p. 531). this idea is expressed by in neurobiological terms by nichols et al.: “in essence, the brain is able to restore its network connectivity to a predisease state” (2017, p. 215). this suggests another factor that may explain why specific forms of self-modelling emerge, at least in some cases: they are old, familiar forms that are the most probable alternatives available to the brain when the recent, default model is relaxed. of course, set and setting undeniably play a large role in such cases (as nichols et al. emphasize). but another important factor – which can be subsumed under “set” in a broad sense – consists simply of the ready availability of certain cognitive “materials” (i.e. models) to the brain, based on extensive prior experience. when considering the relationship between therapeutic intentions and therapeutic outcomes, another factor should be kept in mind. this is that some patients’ reports draw attention not to the development of specific new self-models (“being a non-smoker”) but to a loosening of the grip of self-modelling on attention and salience altogether – creating a more flexible and less tyrannical self-model, rather than one with specific contents. in the qualitative study of psilocybin-assisted smoking cessation that m&s discuss, quitting smoking “was often reported as one of the least important effects of the study for participants” (noorani, garciaromeu, swift, griffiths, & johnson, 2018, p. 763). this is exemplified by multiple reports; for example: letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 6 this is about a smoking study, i keep forgetting that. because there’s so much more that happened […] [smoking] just seems so petty compared to some of the stuff that was happening. [participant 405] it was all about searching for answers to questions that are age-old. maybe we have the answer to some of it, maybe we’ll never have the answer to it. but none of it had to do with addiction to cigarettes. it all had to do with stretching space and time, and asking questions like, ‘why is there something rather than nothing?’ and, ‘what happened before the big bang?’ […] all those things that had nothing, absolutely nothing – at least in the conscious thinking of it – with stopping smoking […] every time i think of a cigarette, it brings me back to the three sessions of the psilocybin trip, where you’re trying to ask questions that there may not be answers to. so, that’s more fun than smoking! [participant 422] (noorani et al., 2018, p. 763) reports like this suggest a strong role for disidentification with existing selfnarratives, and disruption to the patterns of attention and valuation that they underpin, as opposed to the strengthening of specific new (or old) self-narratives. clearly a complex interplay of factors is at work. in sum, however, i agree with m&s: strengthened priors are important, and ought to be taken into account; they can help to explain important contextual influences in psychedelic therapy; and their main epistemic benefits probably amount to new knowledge of old facts. 2.3 transparency and opacity while endorsing many claims of my account, lyon and farennikova (l&f) (2022) argue against one of its central tenets. they argue that: (a) my account says increased phenomenal opacity is more important in psychedelic therapy than increased transparency; but (b) increased transparency is at least as important, if not more so; therefore (c) a core claim of my account is false. they take the many different ways of drawing the opacity/transparency distinction to be essentially equivalent. they describe five types of cases in which, they say, psychedelics increase transparency with plausible therapeutic import: (i) increased transparency in representations of nature, (ii) increased transparency in representations of music, (iii) increased transparency of mental imagery, (iv) an amplified sense of reality (e.g. the “noetic quality”), and (v) apparent revelations of a truer or more authentic self. i am not sure these are all cases of increased transparency, as i define that term. i deliberately stick to a definition of transparency as not experiencing representations as representations, since it is unclear whether this phenomenal property results from certain postulated mechanisms, such as the attentional unavailability of earlier processing stages (metzinger, 2004). and while there are important links between transparency/opacity and other dichotomies, such as a sense of reality/unreality, unity/duality, etc., i do not think these links amount to equivalences. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org self and knowledge in psychedelic therapy 7 the link between transparency/opacity and reality/unreality does, in my view. but i am less convinced that transparency/opacity maps neatly, rather than approximately, onto unity/duality, absorption/disconnection, necessity/optionality, etc. l&f’s analysis presupposes such a neat mapping, and bidirectional entailments between these various dichotomies. their first type of case involves an experiential sense of connection or unity with nature. i think these sorts of cases plausibly involve increased transparency, but not necessarily of one’s representations of nature. rather, what becomes more transparent is one’s representations of one’s unity with, or connectedness to, nature. people do not typically describe nature itself seeming more real, but their connection to it seeming more real. in fact, only one of l&f’s examples talks about anything seeming more real, as opposed to simply about experiencing unity/connectedness with nature: “all kinds of other things were coming, too, like feelings of being connected to everything […] and it wasn’t like talking about it, which makes it an idea, it was, like, experiential.” cases like this fit with my overall account: psychedelics (a) disintegrate our default self-models, (b) allowing healthier forms of self-modelling to be discovered and subsequently consolidated. l&f’s second type of case involves music. they say it is clear that these kinds of cases involve increases in transparency – but on my definition of transparency, this is not made clear by the quoted reports. none of these talk explicitly about music seeming more real, or less like a representation. rather, they talk about increased feelings of unity, connection, and absorption with/in the music, and, in some cases, the music seeming to take on different properties, such as agency or being the vehicle for various experiences/realizations. these experiences may feature increased transparency (in my sense) of subjects’ mental representations of music, but the quoted reports do not say as much. similar points apply to l&f’s third type of case: mental imagery. they quote the findings of a qualitative study that the “visuals” experienced by psychedelic subjects often serve as “principle organizing motifs of subjective experience with multifold vectors (e.g., audiovisual, relational, autobiographical, spiritual, epistemological, ontological) [and are] often deeply coded with layers of meaning by participants.” they note that subjects often become deeply absorbed in this imagery. but none of this indicates that subjects do not experience their mental imagery as representational. in some cases, psychedelic visuals certainly do have a high degree of transparency. for example, high-dose dmt subjects often report visiting immersive other worlds that feel “more real than real” (gallimore & strassman, 2016). but it seems equally likely that some participants undergo rich and vivid visual experiences that are deeply meaningful and absorbing, while simultaneously recognizing these experiences to be drug-influenced mental events. the fact that imagery is experienced as vivid, meaningful, important, and absorbing does not entail that it is not experienced as representational (compare artworks of various kinds). l&f’s fourth and fifth types of cases seem to me the clearest and most plausible. the fourth involves the noetic quality which is criterial of the mystical-type letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 8 experience – the feeling of gaining knowledge of ultimate reality. this undeniably involves increased phenomenal transparency as i define it. likewise with the fifth type of case, which involves seeming to discover a better, deeper, or more authentic self. this sort of experience involves increased phenomenal transparency, and is often central to the therapeutic process. however, before one can discover a better, deeper, or more authentic self, one must let go of a worse, shallower, or less authentic self. thus, l&f’s fifth type of case fits well with the mechanism that i posit: unbinding and increased opacity of the default self-model, leading to the discovery and subsequent consolidation of different, healthier forms of self-modelling. l&f anticipate this line of reasoning. they respond that even if transparency increases causally depend on prior opacity increases, this does not mean that the latter are more important. they use the analogy of someone who travels to another country for a relaxing holiday. typically, such a journey causally depends on the prior purchase of a plane ticket. but this does not mean that the purchase of the ticket explains why the person is more relaxed: to explain this, we must look to the experiences they had during the trip. i think this analogy is suggestive, but can be adapted illuminatingly. rather than relaxing holidays, consider a case even closer to our present concerns: transformational travel. imagine several people, each of whom undertakes a meaningful, transformative journey to another country. each person goes to a different country, has different experiences, and is transformed in different (but all salutary) ways. is there a common mechanism? is there any robust, informative generalization we can make? clearly we cannot say things like “they went to australia and hitch-hiked across the nullarbor plain” if only one of them did this, while another went to nepal and undertook a meditation retreat, and yet another went to france, worked in a vineyard, and had a whirlwind romance. however, it clearly would be equally unhelpful to say that each person bought a plane ticket. i think the most informative generalization would be something like this: each person left their home, and with it their normal modes of thought, feeling, and action; each went on a transformative journey to other places in which they had rich, meaningful, and unfamiliar experiences. despite the varying details, each returned home with an expanded sense of possibilities and a vivid, experientially-backed sense of the contingency of their ordinary modes of living and experiencing. each left their ordinary home and ordinary routines, went on a transformative trip, and returned with the profound conviction that ‘there are other ways i can be’ – the conviction, most basically and generally, that ‘things [and i] can be otherwise’. of course, the negative insight that one’s ordinary ways of being are contingent, and the positive but abstract insight that one can be otherwise, are empty without some definite acquaintance with specific, alternative ways of being. but if we want to generalize robustly and informatively, then we need to strike the right level of abstraction. and the details of the specific forms of self-modelling that psychedelics disrupt, and of the specific alternative forms of self-modelling letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org self and knowledge in psychedelic therapy 9 that they can make available, are too heterogeneous to permit much informative generalization at this level. what i think we can say of nearly all therapeutic psychedelic experiences is that people experience different ways of seeing and being, and (equivalently, in my view) different forms of self-modelling, due to the unbinding of their default self-model; that they gain a deep, experientially-backed conviction of the possibility of change and the contingency of their default sense of self; and that these components interact to cause lasting change. true, the discovery of new forms of self-modelling is just as important as the discovery of the contingency of old forms, and i emphasize the latter; but the latter is, nonetheless, typical, generalizable, and important. however, i think l&f are right about one important point: the phenomenon that i call “increased opacity of the self-model” is not always that, given my understanding of opacity. in essence, i make the same sort of mistake with which i earlier charged l&f: i assume that because self-model opacity tends to cause insight into the mutability of one’s self, any insight into the mutability of one’s self must amount to, or result from, self-model opacity. here i was blinded by my relatively idiosyncratic theoretical conviction that there are no selves apart from self-models. in my worldview, a distinction between the mutability of the self and the mutability of the self-model is a distinction without a difference. but to many people who do not share my worldview, the phenomenon i am talking about may be at least as likely to present in the former guise as in the latter. put simply, the objective fact that subjects discover – according to me – is that their ordinary sense of self is malleable, in virtue of its being continuously constructed by a mental modelling process. but often, the way subjects conceive of that fact probably has little to do with self-modelling or self-representation, and more to do with the possibility of change for the “self” or the person. the slogan version of the insight, as it presents to many subjects, is “i can be otherwise” – not “my self-model can be otherwise”. 2.4 serotonin and predictive processing colombo (2022) argues that inconsistency in the pp literature about the neural implementation of precision weighting and the computational functions of serotonin calls into question the reliance of my account on the pp framework. he notes that other computational frameworks for the explanation of psychiatric phenomena exist, and argues that one such framework, reinforcement learning (rl), promises to provide plausible alternative explanations of psychedelic therapy that do a better job of linking different levels of description. actually, i think my main claims are independent of pp. here are some mechanistic claims i make in the book, in descending order of my confidence in them based on current evidence: (1) psychedelic therapy works mainly by changing mental representations of the self; (2) a central function of these representations is to bind self-relevant information across domains and levels of processing; and (3) letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 10 this binding function is performed according to pp principles. in my view, the current evidence concerning psychedelic therapy and self-representation makes quite a strong case for (1). the evidence surveyed by sui and humphreys (2015) seems to provide a fairly strong case for (2). and (3) is one speculative sketch of how this might work. the truth of pp has not been established definitively, and it may turn out that self-representation (and self-binding) do not operate according to pp principles. thus, the specific details of my account should be held fairly lightly, and subjected to scrutiny, such as colombo’s. however, this does not impugn the overarching claims of the account, unless it turns out that there is no plausible way to spell out its details. in essence, i think colombo’s arguments importantly challenge the details, but not the core, of my account. colombo anticipates this line of response. having sketched two possible rl explanations of psychedelic therapy, he comments: one possible [response] is that the truth of either of these rl hypotheses would not change anything in letheby’s key idea […] yet, if different computational hypotheses about serotonin are all compatible with letheby’s idea that psychedelic therapy mainly works by changing self-representations, then one may conclude that this idea is too generic to advance existing understanding of psychedelic therapy, or of serotonin function in relation to self-representations. (colombo, 2022). i concede that the bare idea that psychedelic therapy mainly works by changing self-representations is too generic to advance existing understanding of serotonin function in relation to self-representations. for it to do this, colombo is right: one needs more specific neurocomputational hypotheses, fleshed out with more implementation detail. however, my main goal was not to advance understanding of serotonin function in relation to self-representations – it was to advance understanding of psychedelic therapy. and i do not think the self-representation hypothesis is too generic to do that, even if it is compatible with different computational hypotheses about serotonin. the self-representation hypothesis is a specific, testable proposal about how psychedelic therapy works. it makes different predictions to the molecular neuroplasticity theory, to the metaphysical belief and alief theories, and to other prominent explanatory hypotheses (e.g. hendricks, 2018). one might worry that the self-representation hypothesis and the molecular neuroplasticity theory are pitched at different levels of explanation3, and therefore cannot really be in competition. however, in chapter 4 of pop i argue that at least one version of the molecular neuroplasticity theory – the pure neuroplasticity theory – is genuinely inconsistent with (neuro) cognitive and psychological theories that appeal to changes in self-representation, metaphysical beliefs, and so forth. 3perhaps cognitive vs. neural, psychological vs. neuroscientific, or algorithmic vs. implementational, depending on one’s framework and assumptions. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org self and knowledge in psychedelic therapy 11 when we flesh the self-representation hypothesis out with the self-binding theory, while it remains compatible with multiple implementational hypotheses, it becomes even more specific and testable. let me spell out one prediction of a self-binding account of psychedelic therapy. sui and humphreys (2015) reported on an experiment (wang, humphreys, & sui, 2016) in which participants were asked to associate neutral shapes with themselves and with other people (friends and strangers). later, they were asked to re-associate the same shapes with different people. for shapes that had been linked to the self, this proved difficult – there was a switching cost. sui and humphreys (2015) interpret this in terms of a “sticky trace” from self-binding: since information bound to the self-model is bound preferentially, it is harder to unbind later. this empirically detectable signature of self-binding suggests the following prediction: patients who associate neutral stimuli with self vs. others prior to receiving psilocybin or placebo, and are asked to re-associate that information after the drug session, should show performance differences on the latter task. patients in the psilocybin condition should show less of a “sticky trace” from self-binding, if psilocybin disrupts this function. and if self-unbinding is the main therapeutic mechanism, the extent of this reduction in the “sticky trace” should correlate with symptom reduction in the subsequent days or weeks. this latter correlation would not be predicted by the rl hypotheses that colombo sketches, according to which psychedelic therapy “mainly works not by changing self-representations, but by changing one’s representations of time or value” (colombo, 2022). this is not to say that a self-binding explanation of a diminished “sticky trace” would be the only explanation possible. however, a strong correlation between a reduced “sticky trace” and good clinical outcomes would certainly fit better with a selfrepresentation theory of psychedelic therapy than with a theory that denies that changes to self-representation are the main mechanism. thus, without answering questions about the neural implementation of precision expectation and the computational function of serotonin, the self-representation hypothesis is specific enough to advance existing understanding of psychedelic therapy. there is another point to make: even if psychedelics typically decrease the precision of self-representations via 5-ht2ar agonism, this does not entail that serotonin or the 5-ht2ar have any unique or distinctive involvement in selfrepresentation or in precision estimation. it only entails that 5-ht2ar hyperactivation is capable of inducing some set of neurocognitive effects that typically includes disruption to the substrates of self-representation and/or precision estimation. this is one way to understand the rebus model and its antecedents, such as the entropic brain hypothesis (ebh; carhart-harris et al., 2014): as strictly agnostic about whether serotonin or the 5-ht2ar have any special involvement in computational functions such as precision weighting. (of course, proponents of these hypotheses do make some implementational speculations, but it is not clear that they intend these speculations to constitute core commitments of their accounts.) on the implementationally agnostic reading, the rebus model would only be committed to the following: since the 5-ht2ar is densely expressed in neural systems letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 12 that plausibly have the function of encoding high-level (self-related) priors, hyperactivating this receptor disrupts the functioning of these systems, leading somehow to those priors being assigned lower precision. there are multiple possible ways of filling in the blanks. the overall point is a familiar one in the philosophy of science: theoretical claims (such as the core claims of rebus and ebh) must be conjoined with auxiliary (in this case, implementational) hypotheses in order to yield specific, testable predictions. as colombo notes, the plurality of suggestions in the pp literature about the computational function of serotonin and the neural implementation of precision estimation may plausibly “be taken to highlight that there is a plurality of pathways and mechanisms implementing precision-weighted computing in the brain” (colombo, 2022). equally plausibly, one may suppose that neuromodulators such as serotonin, and receptor subtypes such as the 5-ht2ar, participate in multiple computational functions, such that (i) there is no one-to-one mapping between neuromodulator or receptor subtype and computational function, and (ii) understanding the effects of changes to neuromodulation requires attending to the plausible computational functions of relevant neural structures (dietrich, 2003). these assumptions underlie the account i develop in pop. of course, they may turn out to be false. work should be done to formulate more specific and testable pp hypotheses about psychedelic therapy and pit these against alternative (e.g. rl) hypotheses. it is also worth considering whether the pp framework is genuinely incompatible with the rl hypotheses that colombo outlines (colombo, personal communication). but in any case, the pluralism in the pp literature about serotonin function and precision estimation implementation do not tell strongly against a pp account of psychedelic experience, unless we have reasons to rule out the plausible pluralistic possibility just sketched. moreover, its abstraction from certain implementational details does not make the predictive self-binding account too abstract or generic to advance existing understanding of psychedelic therapy or to generate specific, testable predictions. 3 epistemology the commentaries that focus on epistemological issues also vary in how strongly they challenge my account; again, i discuss these commentaries in ascending order of the degree of challenge. bortolotti and murphy-hollies endorse the overall shape of my account and make some productive connections to ideas concerning agency-first epistemology and self know-how. caporuscio brings into play the idea of self-shaping, which connects to self know-how, but she argues against certain core claims of my account. finally, fink allows that psychedelic experiences can have some epistemic benefits, but argues that these experiences cannot contribute to the justification of beliefs, and that the epistemic benefit i call “new knowledge of old facts” is better conceptualized as increased understanding. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org self and knowledge in psychedelic therapy 13 3.1 agency first or agency also? bortolotti and murphy-hollies (b&m) (2022) endorse the broad outlines of my epistemology of psychedelics, and relate it to a recent trend towards an agency-first approach to epistemology. on such an approach we “move from asking whether the agent’s belief is true and well-grounded to asking whether the belief enables the agent to succeed in their epistemic practices” (bortolotti & murphy-hollies, 2022). as b&m say, this mode of epistemic evaluation is one that i deploy in pop and is necessary to understand the overall epistemic profile of psychedelic experiences. it strikes me as apt and illuminating to relate aspects of my methodology to these broader currents in epistemology. my main disagreement with b&m is a relatively minor point of emphasis and priority: i do not think all the epistemic benefits of psychedelics i postulate in the book can best, or only, be captured by an agential approach. rather than an agency-first approach, i favour an agency-also approach that sees agential and more traditional forms of epistemic evaluation as equally important.4 indeed, i think the value of agential evaluation derives from the value of more traditional evaluation: it is because we care about justification, truth, reliability etc. that we should want to evaluate, not just the justification, truth, etc. of individual cognitions, but also the overall effects of those cognitions on the individual’s propensity to form or acquire cognitions that are justified, true, etc. the fundamental values here are still those of traditional epistemology; we are simply adopting a broader perspective. in chapter 8 of pop i argue that therapeutic psychedelic experiences often have five distinct types of epistemic benefits: the acquisition of important forms of knowledge that (propositional knowledge), knowledge how (ability knowledge), knowledge by acquaintance, and new knowledge of old facts, as well as indirect epistemic benefits. i think types 1 and 5 – propositional knowledge and indirect epistemic benefits – are best captured by an agential approach. the reason why is obvious in the case of type 5, indirect benefits: these precisely have to do with the effects of psychedelic experiences on individuals’ epistemic agency. type 2, knowledge how, probably has beneficial effects on epistemic agency, but it is also epistemically valuable in its own right, and this can be accommodated comfortably by a traditional approach. it is true that traditional epistemology focuses disproportionately on propositional knowledge, but knowledge how is standardly regarded as a form of knowledge in its own right. similar remarks apply to types 3 and 4, knowledge by acquaintance and new knowledge of old facts. the psychedelic-induced acquisition of these plausibly has beneficial effects 4of course, there is more than one way of understanding what an agency-first approach amounts to. it might be a matter of causal or explanatory priority, suggesting that being a successful epistemic agent is necessary to acquire traditional epistemic goods, or it might be a matter of normative or evaluative priority, suggesting that epistemic evaluation ought to focus more strongly on overall epistemic agency than on the truth, justification, etc. of specific mental states. (i am indebted to chiara caporuscio for raising this point.) my reading of b&m is in line with the latter construal, and this is reflected in my discussion. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 14 on epistemic agency – but those indirect effects belong in category five. the intrinsic epistemic value of knowledge by acquaintance and new knowledge of old facts can be captured in traditional epistemological terms, independently of their agential effects. why think that type 1, propositional knowledge, fits well with an agential approach? this has to do with the ideas that b&m introduce about self know-how (bruin & strijbos, 2020). i argue that in some cases psychedelic experience, plus subsequent reflection, can help people to gain new justified true beliefs about their own psychological states and traits. this increases the coherence between their models of themselves and the reality that those models represent (i.e. the person herself and her patterns of thought, feeling, and behaviour). but as b&m, following debruin and strijbos, point out, there is another way that we can increase the coherence between self-models and reality: shaping ourselves in accordance with those models. i think it is plausible that psychedelic experiences can improve this kind of self know-how, and that this epistemic benefit is best captured by an agential lens. these ideas connect with caporuscio’s arguments about self-shaping. 3.2 self shaping and self know-how b&m, following debruin and strijbos (2020), note that one way we can preserve the authority of first-person ascriptions is by making our behaviour line up with them: a world-to-mind, rather than a mind-to-world, direction of fit. this brings to mind the notion of active inference from the pp literature. it also provides a helpful way to think about certain psychedelic phenomena. psychedelic subjects often have apparent insights into the nature of their true or authentic selves, which raises difficult questions, especially when the content of these insights may be at odds with subjects’ recent patterns of thought, feeling, and behaviour. but even when such insights are strictly false, one way to make them true is to act in accordance with them, so that they become self-fulfilling prophecies. perhaps the truth in such insights is not that this is who i really am, but that this is someone i can be. caporuscio’s (2022) motivation to invoke ideas of this kind comes from a critique of my epistemological claims. i argue that, by decreasing the precision of self-related priors, psychedelics probably sometimes lead to accurate insights concerning the self. caporuscio asks: why should we think psychedelic insights concerning the self are any more likely to be accurate than those concerning the external world? after all, if the rebus model is true, then high-level self-related and world-related priors get relaxed equally. thus, if we treat world-related insights as suspect, then why not treat self-related insights the same way? i think there are some reasons to think that self-related insights are likelier to be accurate than (external) world-related insights. for one thing, we know that many of psychedelics’ effects on external-world perception tend in a misrepresentational direction, such as the perception of motion in static objects and geometrical patterns overlaying objects. so there are positive grounds for suspicion in letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org self and knowledge in psychedelic therapy 15 the external world case that are not present to the same degree in the introspective case. (of course, we know that many self-related beliefs are inaccurate, but this is not the same as knowing that psychedelics make them more likely to be inaccurate.) second, it seems plausible to me that some of psychedelics’ effects on introspection simply amount to promoting it, rather than qualitatively altering it. in this light, we ought to remember that introspection, in general, can provide defeasible justification for self-related beliefs, and being on a drug is not, in itself, an automatic defeater (letheby, 2021, p. 163). third, insofar as psychedelics do qualitatively alter introspection, the alterations they induce are ones that seem likely to improve its accuracy. it is plausible that self-related affective biases, of the kind that psychedelics can disrupt via their effects on self-modelling, ordinarily impede self-knowledge (davis & thompson, 2015); thus, psychedelics may reduce obstacles to self-knowledge. (of course, self-related priors, like external world-related priors, are epistemically enabling as well as constraining; thus, relaxing them is sure to have some epistemic costs, at least for the duration of the experience.) in pop i suggest that relaxing self-related priors, such as negative core beliefs, can allow evidence inconsistent with those priors into consciousness. caporuscio agrees that this can happen, but suggests that it is not the whole story. in some cases, she says, there may be little evidence inconsistent with the relevant priors if we have acted in accordance with them for a long time. rather, relaxing these priors allows us to explore new self-conceptions; we can then shape ourselves in accordance with them during the integration period. i think caporuscio is right to draw attention to this kind of case, but i also think many cases of this kind can be accommodated by my account, if we understand the notion of “evidence” sufficiently broadly. take caporuscio’s case, adapted from moran (2001) and mcgeer (1996), in which maria has conceived of herself as self-sufficient and emotionally independent for all of her adult life, and has acted, thought, and felt in accordance with that self-conception, so that behavioural and psychological evidence against it is scant. on psilocybin, she is released from the constraint of this self-conception and explores new ways of being, seemingly discovering a deep, heretofore unknown desire for human connection. one possibility is that the desire was there all along, but was repressed, unattended, or otherwise unknown. i am not sure what sort of psychodynamic posits this possibility requires, but any plausible model of the mind must account for the fact that we often have desires of which we are, in some sense, unaware. another possibility is that the desire did not predate maria’s psilocybin session, but was created anew by it. still, merely by responding positively to the thought of closer human connection, maria gains some information (i.e. evidence) about her own dispositions and capacities – evidence that her ordinary self-model would have prevented her from gaining. the objection might still be raised that maria would be unjustified in concluding that the desire had always been there, i.e., that it antedated the psilocybin session. i agree: without further evidence, she would – though she may well disletheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 16 cover such evidence as she reflects on her past in a new light. in any case, caporuscio and b&m are correct in this much: sometimes, what people discover on psychedelics is not facts about who they really are underneath layers of avoidance and self-deception, but facts about who they can be and who they want to be. however, these are modal and psychological facts for which one can discover evidence, in the form of the availability or emergence of certain patterns of thinking and feeling, when the everyday priors that would normally inhibit those patterns are relaxed. again, too, it is worth noting that many patients do not discover new forms of self-modelling, but instead, old, disused ones: “at its most basic, i feel like i did before the depression.” (watts et al., 2017). patients often describe remembering, or reconnecting with, old passions, pursuits, and hobbies. cases like these seem straightforwardly epistemically beneficial, and allow little room for scepticism: people are not discovering some new, hidden fact about themselves, but being reminded of a fact that they had forgotten, in the banal sense that they had not thought about it for a long time. this is not the same as discovering apparently repressed childhood memories: in the present kind of case, there is usually no reason to distrust the memory. of course, it could have occurred without psychedelics – but whether it would have is unclear. this is a genuine epistemic benefit – simply being reminded of a known, important, but long-forgotten or long-unnoticed fact5. caporuscio argues that the idea of self-shaping casts the epistemic and psychological importance of the integration period in a new light: this period is not merely about scanning putative insights for plausibility, but also about determining which (re-) discovered self-conceptions one wants to commit to. i agree, and caporuscio is right to emphasise this. the benefits of psychedelics for self-knowledge include both (i) discovering facts about who one is, and (ii) discovering facts about who one can be, then shaping oneself accordingly. however, we should note that often, choosing which new self-conceptions to adopt may not be a conscious, deliberate process. as l&f point out, such insights can carry a noetic quality – people feel as though they have discovered who they really are. and in a cognitive dynamic that justifies the appellation “inverse ptsd” (garcia-romeu, r. griffiths, & w. johnson, 2014), the salience and positive valence of this experience compel subjects to revisit it, causing lasting and significant changes to their thoughts, feelings, and behaviour. self-shaping happens during integration, but it need not always be a conscious or deliberate act. 3.3 understanding and justification fink (2022) criticizes some of my views concerning psychedelic epistemology, and advances a positive proposal of his own. his positive proposal is that the epistemic benefit of psychedelic experiences that i call “new knowledge of old facts” can 5i am indebted to matthew nestor for bringing this sort of case to my attention. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org self and knowledge in psychedelic therapy 17 better be conceptualized as increased understanding. this, he notes, has the virtue of generality, covering cases in which the proposition in question is false. i am sympathetic to this line of thought; indeed, i have expressed this sympathy briefly in print before (letheby, 2019). i think that when the proposition in question is true, it is also illuminating to speak of “new knowledge of old facts” – but i will not defend that claim here. instead, i will focus mainly on fink’s arguments that psychedelic experiences cannot contribute to the justification of beliefs. first, however, i will make one other point briefly: fink holds that increased understanding is an indirect epistemic benefit, not a direct one – but this depends on how one defines “direct” and “indirect”. in pop my use of these terms was intended to draw a distinction between (a) epistemic benefits that result from the psychedelic experience itself and (b) epistemic benefits that result, instead, from the lasting psychological effects of the psychedelic experience. on this way of drawing the direct/indirect distinction, increased understanding is, indeed, a direct epistemic benefit of psychedelic experiences. (fink draws this distinction differently; i will return to this point shortly.) what, then, of fink’s arguments concerning justification? to begin with, he agrees with my claim that psychedelics can enhance the context (or processes) of discovery – in other words, even if they cannot give us justification for beliefs, they can help us come up with promising new ideas. however, since there are many processes of discovery, he says, this benefit “cannot be considered unique to psychedelics” (fink, 2022). here, i want to sound a note of caution. it is true that people can come up with promising new ideas in many different ways. but this is consistent with the idea that some of these processes are more powerful or effective than others. and i strongly suspect that psychedelic ingestion may be more effective at helping people to come up with radically new, creative, or interesting ideas than many other methods. this, of course, is an empirical claim, and a body of research already exists examining psychedelics’ effects on creativity (girn, mills, roseman, carhart-harris, & christoff, 2020; sessa, 2008) – but examining this literature is beyond the scope of this paper. nonetheless, it seems plausible that many people who take psychedelics, especially in deliberate, intentional ways for therapeutic, transformative, or problem-solving purposes, thereby come up with important and valuable ideas that they would not have come up with otherwise. fink also argues that, because psychedelic ingestion can lead to the acquisition of false beliefs as well as knowledge, its contribution – even in cases in which knowledge results – may not be epistemic: “because some of the processes altered by psychedelic compounds that lead us to novel mental states may also lead us to comfortable illusions or placebo insights, i would not consider this contribution necessarily epistemic. the contribution of psychedelics to knowledge, if there is any, is coincidental” (fink, 2022). this argument strikes me as invalid. every method of knowledge acquisition – consider, for example, silent mathematical reasoning – is fallible, and capable of generating false beliefs as well as knowledge. but surely it would be too quick to conclude, on this basis, that none of these methods make genuinely epistemic contributions. similar points apply when we letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 18 compare psychedelic ingestion to other processes of discovery. if one engages in some process of discovery in order to come up with a new idea – say, an insight into one’s personality or a solution to some problem – and then comes up with an idea, subsequently tests it, acquires justification, and ends up with new propositional knowledge, this epistemic outcome (i.e. the acquisition of propositional knowledge) can hardly be viewed as a coincidence. true, processes of discovery must be combined with processes of justification; but this does not entail that the former can only make coincidental contributions to our epistemic state. fink’s central critical claim about psychedelic epistemology is the thesis of psychedelic justification impossibilism (pji): psychedelic justification impossibilism: “psychedelics may have psychological benefits, indirect epistemic benefits, or be beneficial for processes of discovery… but they lack direct epistemic benefits for processes of justification” (2022) fink defines “direct” epistemic benefits as those which contribute to a mental state’s status as knowledge. in the case of propositional knowledge, he understands this contribution in purely justificational terms: he says that the psychedelic justification impossibilist may grant indirect epistemic benefits in that psychedelic compounds, experiences, or episodes may contribute to forming states that, as a matter of fact, are knowledge but deny direct epistemic benefits in that they contribute to such states’ status as knowledge, i.e. to processes of justification (fink, 2022) so, in essence, pji is the claim that psychedelic ingestion and psychedelic experiences cannot contribute to the epistemic justification that is required for a true belief to constitute propositional knowledge. before examining fink’s arguments for pji, it is worth noting that the thesis is consistent with most of my epistemological claims about psychedelics. fink suggests that the points he makes may generalize to other types of knowledge besides the propositional. but four of the five types of epistemic benefits i attribute to psychedelic experiences – knowledge how, knowledge by acquaintance, new knowledge of old facts, and indirect epistemic benefits – are ones which are not standardly conceived of as requiring justification. and when it comes to propositional knowledge – the type about which i am most tentative and ambivalent in the book – my most confident claim is that psychedelics have substantial benefits for discovery. as fink notes, this claim would be untouched by the truth of pji. (and, as indicated above, i think benefits for discovery are genuinely epistemic.) so even if pji is true, this does not undermine the idea that therapeutic and transformative psychedelic experiences typically have significant epistemic benefits that are consistent with a naturalistic worldview. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org self and knowledge in psychedelic therapy 19 fink contests the idea that pji will not generalize to other types of knowledge. justification, he argues, can be needed for some other forms of knowledge as well, “e.g., when we need to justify that the success of an act […] is not just lucky but due to know-how or skill” (fink, 2022). but it seems to me that the kind of justification at issue here is not justification for the knowledge how; it is justification for the claim that someone possesses knowledge how – i.e. justification for a proposition. the fact that epistemic justification is needed for our (propositional) attributions of non-propositional types of knowledge does not show that justification is needed to possess those types of knowledge. all this aside, the truth or falsity of pji is clearly a matter of considerable interest and importance. while i suggest in pop that psychedelic experiences alone rarely, if ever, provide sufficient justification for interesting self-related beliefs, i argue that they can make a substantial justificatory contribution. thus, there is a real point of disagreement here. what, then, of fink’s arguments for pji? one argument appeals to the idea that standards of justification are fixed extra-individually, and therefore cannot be affected by an individual’s act of psychedelic ingestion: [epistemic] justification has a social and normative dimension […] what amounts to a good or adequate justification [is] largely independent of any given individual because the standards for justification are social or extra individual. and because psychedelics affect only an individual’s state, leaving the standards of justification untouched, they do not affect which processes can be counted as processes of justification even if processes of justification are in the end psychological processes and indeed affected by psychedelics. the epistemic lifting is done not by psychedelically influenced processes, but by the extra individual circumstances that turn these processes into processes of justification. (fink, 2022). this argument about affecting standards of justification seems to me to be a red herring. of course psychedelic ingestion does not affect which processes can be counted as processes of justification – but this is surely not what any sensible proponent of psychedelic epistemology has in mind. none of our other knowledge acquisition methods – sense perception, reasoning, testimony, etc. – benefit individuals epistemically by affecting which processes can be counted as processes of justification. rather, they benefit individuals epistemically by (inter alia) causing them to engage in processes that are, in fact, processes of justification. when i proposed in pop that psychedelic ingestion contributes somewhat to the context of justification, i did not mean that psychedelic ingestion alters the standards of epistemic justification. rather, i meant that psychedelic ingestion, by affecting individuals’ psychological states, can help those individuals to satisfy the existing standards of epistemic justification. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 20 fink acknowledges that processes of justification may be psychological processes that could turn out to be affected – perhaps for the better – by psychedelics; surely this is as much as we could ask, vis-à-vis justification, from any method of knowledge acquisition. he claims that the “epistemic lifting is done not by psychedelically influenced processes, but by the extra individual circumstances that turn these processes into processes of justification” (fink, 2022) – but this sort of claim seems to me implausible when applied to other epistemic practices. consider, again, silent mathematical reasoning. surely this can do some fairly heavy epistemic (justificatory) lifting, even if it does not alter the standards of justification one whit. fink mentions reliabilism as one position in epistemology that might licence the claim that psychedelics contribute to processes of justification. here, i agree. reliabilism broadly holds that propositional knowledge is true belief formed by a reliable process. part of my strategy in pop was to appeal to certain facts about the psychedelic experience and the integration period to argue that these two, combined, can be a fairly reliable way of attaining accurate insights about the self. fink disagrees, arguing that “the high prevalence of placebo insights and comforting delusions […] makes it implausible that psychedelically altered mechanisms are reliable” (fink, 2022). but when it comes to putative insights concerning the self, we do not know what the prevalence of placebo insights relative to genuine insights is – this is precisely the question at issue. i argue (tentatively) on various grounds that the ratio of genuine self-related insights to placebo self-related insights is likely to be fairly high – perhaps even high enough to pass some relevant reliability threshold for epistemic justification. i do not think we can embrace this conclusion confidently, but nor can we embrace its negation any more confidently. the frequency with which controlled psychedelic ingestion promotes accurate vs. inaccurate psychological or self-related insights is an important matter for future research, both philosophical and empirical. determining this frequency poses formidable methodological difficulties; if any progress can be made, it is likely to require concerted collaborative efforts from experts in multiple disciplines (cf. tulver, kaup, laukkonen, & aru, 2021). finally, i think it is important to consider what exactly pji means, and what it would take to falsify it. suppose, as i suggested in subsection 3.2 above, that one of psychedelics’ main effects on introspection is simply to promote it: i.e. to induce a more introspective state of consciousness than is usual. suppose that you ingest psilocybin, become temporarily more introspective, and thereby acquire a new true belief about yourself that you would not have acquired otherwise. now, introspection in general can provide defeasible justification for self-related beliefs, and i argued in pop that being on a drug is not in itself an automatic defeater. what, then, are we to say about such a case? it seems clear to me that, in a case like this, you have a justified6 true belief such that you would not have formed the belief, nor obtained the justification, had you not taken psychedelics. (similar 6unless some other defeater is present. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org self and knowledge in psychedelic therapy 21 points can be made regarding cases wherein most, or all, of the justification derives from sober reasoning after the experience.) is it not straightforwardly true in such a case that psychedelic use benefited you epistemically – not just by causing you to acquire a true belief, but by causing you to acquire a justified true belief? according to counterfactual theories of causation, at least, your psychedelic use certainly caused you to acquire these epistemic benefits in such a case. in sum, i agree with fink that the epistemic benefit of psychedelics that i call “new knowledge of old facts” can be conceptualized at least as usefully, and with greater generality, in terms of increased understanding. i also agree that, while psychedelics’ benefits for epistemic discovery are fairly clear, it is less clear what contribution, if any, they can make to justification. i argued in pop that they can make a substantive contribution, whereas fink argues that they cannot – this, of course is the thesis of pji. i think it is fair to say that neither of our arguments settles the issue decisively. as such, i conclude by hazarding one final point of agreement: that further research on this intriguing question is required. 4 conclusion all commentators raise important and interesting questions about psychedelics and about the specific account developed in pop. do we need to pay more attention to the roles of affect and strengthened priors to understand psychedelic therapy? is phenomenal opacity more important than transparency in psychedelic therapy? is pp the right framework to explain psychedelic therapy, and how do computational hypotheses relate to neural implementation details? does psychedelic epistemology require an agency-first approach? what roles might self-shaping and self know-how play in psychedelic epistemology? does psychedelic experience provide epistemic justification or increased understanding? here i have offered some initial thoughts on these questions. on the whole, this symposium opens up many intriguing possibilities for the philosophy (and science) of psychedelics. i hope that it stimulates more discussion on these matters both within and across disciplines. i would like to finish by once again thanking all the commentators for their generous contributions. acknowledgments for helpful comments and suggestions on earlier versions, i would like to thank chiara caporuscio, matteo colombo, sascha fink, and sarah hoffman. i would also like to thank chiara caporuscio, sascha fink, and everyone else at philosophy and the mind sciences for generously volunteering to edit/host this symposium and working tirelessly to make it a reality. this research was partially supported by the australian government through the australian research council’s discovery projects funding scheme (project dp190101451). the views expressed herein are those of the author and are not necessarily those of the australian government or australian research council. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 22 references bortolotti, l., & murphy-hollies, k. (2022). the agency-first epistemology of psychedelics. philosophy and the mind sciences, 3. https://doi.org/10.33735/phimisci.2022.9283 bruin, l. de, & strijbos, d. (2020). does confabulation pose a threat to first-person authority? mindshaping, self-regulation and the importance of self-know-how. topoi, 39(1), 151–161. https://doi.org/10.1007/s11245-019-09631-y caporuscio, c. (2022). belief now, true belief later: the epistemic advantage of self-related insights in psychedelic-assisted therapy. philosophy and the mind sciences, 3. https://doi.org/10.33735/phimisci.2022.9310 carhart-harris, r., leech, r., hellyer, p., shanahan, m., feilding, a., tagliazucchi, e., et al. (2014). the entropic brain: a theory of conscious states informed by neuroimaging research with psychedelic drugs. frontiers in human neuroscience, 8. https://doi.org/10.3389/fnhum.2014.00020 colombo, m. (2022). serotonin, predictive processing and psychedelics. philosophy and the mind sciences, 3. https://doi.or g/10.33735/phimisci.2022.9320 davis, j. h., & thompson, e. (2015). developing attention and decreasing affective bias: toward a cross-cultural cognitive science of mindfulness. in handbook of mindfulness: theory, research, and practice (pp. 42–61). new york, ny, us: the guilford press. dietrich, a. (2003). functional neuroanatomy of altered states of consciousness: the transient hypofrontality hypothesis. consciousness and cognition, 12(2), 231–256. https://doi.org/10.1016/s1053-8100(02)00046-6 fink, s. b. (2022). psychedelics favour understanding but not knowledge. philosophy and the mind sciences, 3. https: //doi.org/10.33735/phimisci.2022.9320 gallimore, a. r., & strassman, r. j. (2016). a model for the application of target-controlled intravenous infusion for a prolonged immersive dmt psychedelic experience. frontiers in pharmacology, 7. https://doi.org/10.3389/fphar.2016.0 0211 garcia-romeu, a., r. griffiths, r., & w. johnson, m. (2014). psilocybin-occasioned mystical experiences in the treatment of tobacco addiction. current drug abuse reviews, 7(3), 157–164. https://doi.org/10.2174/1874473708666150107121331 girn, m., mills, c., roseman, l., carhart-harris, r. l., & christoff, k. (2020). updating the dynamic framework of thought: creativity and psychedelics. neuroimage, 213, 116726. https://doi.org/10.1016/j.neuroimage.2020.116726 hendricks, p. s. (2018). awe: a putative mechanism underlying the effects of classic psychedelic-assisted psychotherapy. international review of psychiatry, 30(4), 331–342. https://doi.org/10.1080/09540261.2018.1474185 hoffman, s. (2022). positive affect and letheby’s naturalization of psychedelic therapy. philosophy and the mind sciences, 3. https://doi.org/10.33735/phimisci.2022.9285 johnson, m. w. (2022). psychedelic science needs philosophy. philosophy and the mind sciences, 3. https://doi.org/10.33735 /phimisci.2022.9452 letheby, c. (2019). the varieties of psychedelic epistemology. in n. wyrd, d. luke, a. tollan, c. adams, & d. king (eds.), psychedelicacies: more food for thought from breaking convention. letheby, c. (2021). philosophy of psychedelics. oxford university press. lyon, a., & farennikova, a. (2022). through the psychedelic looking-glass: the importance of phenomenal transparency in psychedelic transformation. philosophy and the mind sciences, 3. https://doi.org/10.33735/phimisci.2022.9323 martin, j. m., & sterzer, p. (2022). how level is the ’cognitive playing field’? context shapes alterations in self-conception during the psychedelic experience. philosophy and the mind sciences, 3. mcgeer, v. (1996). is ”self-knowledge” an empirical problem? renegotiating the space of philosophical explanation. the journal of philosophy, 93(10), 483–515. https://doi.org/10.2307/2940837 metzinger, t. (2004). being no one: the self-model theory of subjectivity. mit press. moran, r. (2001). authority and estrangement: an essay on self-knowledge. https://doi.org/10.1515/9781400842971 nichols, d., johnson, m., & nichols, c. (2017). psychedelics as medicines: an emerging new paradigm. clinical pharmacology & therapeutics, 101(2), 209–219. https://doi.org/10.1002/cpt.557 noorani, t., garcia-romeu, a., swift, t. c., griffiths, r. r., & johnson, m. w. (2018). psychedelic therapy for smoking cessation: qualitative analysis of participant accounts. journal of psychopharmacology, 32(7), 756–769. https://doi.or g/10.1177/0269881118780612 revonsuo, a. (2006). inner presence: consciousness as a biological phenomenon. cambridge, ma, us: mit press. safron, a. (2020). on the varieties of conscious experiences: altered beliefs under psychedelics (albus). psyarxiv preprints. https://doi.org/10.31234/osf.io/zqh4b sessa, b. (2008). is it time to revisit the role of psychedelic drugs in enhancing human creativity? journal of psychopharmacology, 22(8), 821–827. https://doi.org/10.1177/0269881108091597 letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9283 https://doi.org/10.1007/s11245-019-09631-y https://doi.org/10.33735/phimisci.2022.9310 https://doi.org/10.3389/fnhum.2014.00020 https://doi.org/10.33735/phimisci.2022.9320 https://doi.org/10.33735/phimisci.2022.9320 https://doi.org/10.1016/s1053-8100(02)00046-6 https://doi.org/10.33735/phimisci.2022.9320 https://doi.org/10.33735/phimisci.2022.9320 https://doi.org/10.3389/fphar.2016.00211 https://doi.org/10.3389/fphar.2016.00211 https://doi.org/10.2174/1874473708666150107121331 https://doi.org/10.1016/j.neuroimage.2020.116726 https://doi.org/10.1080/09540261.2018.1474185 https://doi.org/10.33735/phimisci.2022.9285 https://doi.org/10.33735/phimisci.2022.9452 https://doi.org/10.33735/phimisci.2022.9452 https://doi.org/10.33735/phimisci.2022.9323 https://doi.org/10.2307/2940837 https://doi.org/10.1515/9781400842971 https://doi.org/10.1002/cpt.557 https://doi.org/10.1177/0269881118780612 https://doi.org/10.1177/0269881118780612 https://doi.org/10.31234/osf.io/zqh4b https://doi.org/10.1177/0269881108091597 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org self and knowledge in psychedelic therapy 23 sui, j., & humphreys, g. w. (2015). the integrative self: how self-reference integrates perception and memory. trends in cognitive sciences, 19(12), 719–728. https://doi.org/10.1016/j.tics.2015.08.015 tulver, k., kaup, k. k., laukkonen, r., & aru, j. (2021). restructuring insight: an integrative review of insight in problemsolving, meditation, psychotherapy, delusions and psychedelics. psyarxiv preprints. https://doi.org/10.31234/osf.io/8f bt9 wang, h., humphreys, g., & sui, j. (2016). expanding and retracting from the self: gains and costs in switching selfassociations. journal of experimental psychology: human perception and performance, 42(2), 247. https://doi.org/10.1 037/xhp0000125 watts, r., day, c., krzanowski, j., nutt, d., & carhart-harris, r. (2017). patients’ accounts of increased “connectedness” and “acceptance” after psilocybin for treatment-resistant depression. journal of humanistic psychology, 57(5), 520–564. https://doi.org/10.1177/0022167817709585 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. letheby, c. (2022). self and knowledge in psychedelic therapy: reply to commentaries on philosophy of psychedelics. philosophy and the mind sciences, 3, 12. https://doi.org/10.33735/phimisci.2022.9642 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.tics.2015.08.015 https://doi.org/10.31234/osf.io/8fbt9 https://doi.org/10.31234/osf.io/8fbt9 https://doi.org/10.1037/xhp0000125 https://doi.org/10.1037/xhp0000125 https://doi.org/10.1177/0022167817709585 https://doi.org/10.33735/phimisci.2022.9642 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction mechanisms positive affect context and strengthened beliefs transparency and opacity serotonin and predictive processing epistemology agency first or agency also? self shaping and self know-how understanding and justification conclusion dendritic integration theory: a thalamocortical theory of state and content of consciousness. dendritic integration theory a thalamocortical theory of state and content of consciousness talis bachmanna (talis.bachmann@ut.ee) mototaka suzukib (mototaka.suzuki@charite.de) jaan arua (jaan.aru@ut.ee) abstract the idea that the thalamocortical system is the crucial constituent of the neurobiological mechanisms of consciousness has a long history. for the last few decades, however, consciousness research has to a large extent overlooked the interplay between the cortex and thalamus. here we revive an integrated view of the neurobiology of consciousness by presenting and discussing several recent major findings about the role of thalamocortical interactions in consciousness. recent findings have demonstrated that there exists a specific cellular mechanism how thalamic nuclei modulate the integration of different processing streams within single cortical pyramidal neurons. we describe the “dendritic integration theory”, which is inspired by recent work done on rodents, but integrates decades of work conducted on various species. we illustrate how this new view readily explains various properties and experimental phenomena associated with conscious experience. we discuss the implications of this idea and some of the experiments that need to be done in order to test it. our view bridges two long-standing perspectives on the neural mechanisms of consciousness and proposes that cortical and thalamocortical processing interact at the level of single cortical pyramidal cells. keywords consciousness ∙ contents of consciousness ∙ phenomenal experience ∙ pyramidal neuron ∙ state of consciousness ∙ thalamus this article is part of a special issue on “the neural correlates of consciousness”, edited by sascha benjamin fink. auniversity of tartu bhumboldt universität zu berlin bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-9595-1594 https://orcid.org/0000-0002-2151-4882 https://orcid.org/0000-0003-3927-452x https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 2 1 introduction how do you usually know whether you are conscious? the only way is to have subjective experience with some kind of content, no matter whether a crisp or a vague one. without entering philosophical debates, we take this to mean that the contents and state of consciousness are intimately related. in some states contents are crisp, vivid, stably present and rich (alert wakefulness), in some other states contents are vivid, but less coherent (e.g. dreams, psychedelic hallucinations), in yet other states contents are less vivid, instable, relatively poor (e.g., sleep or hypnagogic state), in some states they are sporadic, minimally expressed and fully out of deliberate control (minimally conscious state), and in some brain states conscious contents are absent (dreamless sleep, coma). in short, the characteristics of the contents of consciousness depend on the state of consciousness. there is a lot of evidence to show that the contents of consciousness are “encoded” in the cortex. neurobiology and science of clinical anesthesia also have a lot of converging evidence that workings of subcortical structures, especially certain parts of the thalamus are also necessary for consciousness. a theory of consciousness thus has to try to find a way to explain consciousness by simultaneously invoking both cortical and subcortical mechanisms. more than 20 years ago one of the authors of this paper described a theory based on interaction of the nonspecific thalamus and cortical content-encoding neurons so as to explain a number of conscious-perception phenomena (bachmann, 1998). the model of conscious perception was based on excitatory post-synaptic potentials brought about by pre-synaptic afferents from the specific, content-carrying system and afferents from non-specific modulation system. according to the model (bachmann, 1994), for the content to be consciously experienced, the cortical content-carrying neurons had to be modulated by the afferents from the non-specific subcortical neurons. the issue of the state of consciousness remained out of the scope. over these years, however, neurobiology of perception and consciousness has made a serious advance, including discoveries at the (sub)cellular level related to both content processing mechanisms and state control mechanisms (aru, suzuki, rutiku, et al., 2019). the need for an update of the theory became apparent. however, as is often the case, during the updating process the theory itself became a new one and with the help from colleagues essentially a collective one. critically important for this advancement were experimental results from matthew larkum’s research team (larkum, 2013; larkum et al., 1999; suzuki & larkum, 2020; takahashi et al., 2016). in what follows we elaborate on our theory of consciousness – the dendritic integration theory (dit) (aru, suzuki, & larkum, 2020) and show how it reconciles the traditional separation between content vs. state of consciousness and cortexbased vs. thalamocortical foundations of consciousness. this theory takes the view that understanding the pyramidal neurons of the cortex and their dendritic trees is central to explaining the flow and control of information relating to conscious bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dendritic integration theory 3 processing (aru, suzuki, & larkum, 2020). in this endeavour, as the matter to be explained is how conscious experience emerges from the brain, our theory starts by describing the attributes and characteristics of conscious experience in order to find neural mechanisms capable of explaining how these subjective characteristics can be achieved by certain neurobiological mechanisms. we assume that the basic evolutionary blueprint of consciousness neurobiology is similar between humans and other mammalian species and thus the ncc of humans can be studied with a little help from the other animals possessing similar brain structures. indeed, here we aim to show that recent studies done with mice have brought new life into thalamocortical theories of consciousness. in particular, the recent finding that certain thalamic nuclei affect how information is integrated within single cortical neurons (suzuki & larkum, 2020), highlights the tight relationship between thalamic modulation, cortical processing and consciousness. dit was first presented in another recent publication (aru, suzuki, & larkum, 2020). here, we focus on the relationship of dit to thalamocortical theories of consciousness and in particular address the issue of states and contents of consciousness. 2 desiderata: what do we seek to explain? a theory of consciousness should explain properties of consciousness. there are no firmly established properties of consciousness that all agree upon. hence, here we will first posit four phenomenological properties of consciousness that we see as important (see tononi & edelman, 1998; ginsburg & jablonka, 2019 for a similar approach; koch, 2020; tononi & koch, 2015). we do not think that the list is exhaustive or the best one (similarly we do not think that any other list is better). it is simply a list that helps us think more clearly about consciousness and can steer the search for the mechanisms of consciousness. (1) consciousness is mentally rich and multifaceted. conscious experience captures different mental experiences such as feelings, emotions, sensations, perceptions, memories, imagery, agency, etc; also, it embraces sensations, perceptions and imagery in all sensory modalities. (2) conscious experience is integrated (see also tononi & edelman, 1998; tononi et al., 2016). for example, in vision, all individual features differentiating the objects included in the field of subjective experience belong to the same unified field. (however, see bayne, 2010; pinto et al., 2017 for why in some cases phenomenal unity may become disintegrated.) (3) consciousness is continuous. when awake, alert and healthy, consciousness presents itself in a seamless subjective flow. conscious perception is not lost when the eyes blink and conscious experience is continuous despite the saccadic eye movements (aru, 2019). (4) state of consciousness is multidimensional. the state of consciousness can vary along many behavioural and cognitive dimensions, including the measurbachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 4 able subjective dimensions (e.g., arousal, motor reactions, richness and vividness of experienced contents, metacognitive confidence, agency) (bayne et al., 2016). the next six characteristics have been observed empirically and add important constraints to any neurobiological theory of consciousness: (5) there is an interaction between the state and contents of consciousness. the state of consciousness affects the contents of consciousness. the same feature or object of conscious perception can be experienced in different ways (e.g. with different vividness, completeness, stability, subjective duration etc.) in different states of consciousness (bachmann, 2012; bayne et al., 2016; haque et al., 2020; koivisto & neuvonen, 2020). (6) mental information processing can also occur in a subliminal mode, meaning that not all sensory signals that are encoded and adequately responded to are necessarily consciously experienced. only part of the contents represented in the brain becomes conscious. (for reviews see kim & blake, 2005; kouider & dehaene, 2007; mitchell, 2006) (7) emergence of conscious experience of some content in response to a stimulus is delayed in time. there is a temporal latency between stimulation and its experience (for review see bachmann, 2000). (8) conscious experience has a limited temporal resolution, meaning that if the presentation of different stimuli occurs too rapidly then some fail to be consciously experienced. this is evidenced in empirical phenomena like backward masking (for review bachmann & francis, 2013) and attentional blink (for review martens & wyble, 2010). even when an identical stimulus is switched on and off intermittently at very high frequency, the multiplicity of stimulus replications will not be consciously perceived and the stimuli will fuse into a continuous experience (watson, 1986). (9) conscious experience is modulated by attention. attending to an image enhances its subjective contrast (carrasco et al., 2004); attending to an afterimage makes it disappear faster from conscious experience (bachmann & murd, 2010; murd & bachmann, 2011). (10) conscious experience of content depends on prior knowledge and expectations, it is context-dependent. the powerful effects of context-dependence are revealed in the fact that non-veridical representations of objects and events such as illusions or hallucinations are experienced even in the healthy brain (aru et al., 2018; aru & bachmann, 2017; kuhn & rensink, 2016; mack et al., 2016; powers et al., 2017). this list can be extended to incorporate more features of and more results related to consciousness. however, we contend that these are the main characteristics a good scientific theory of consciousness should be able to accommodate. can the dit do this? before assessing this question, we will explore the rich history of studying the influence of certain thalamic nuclei on consciousness. bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dendritic integration theory 5 3 a historical view on the role of thalamus and cortex in consciousness 3.1 two types of thalamic projections to the cortex we will not delve into the neuroanatomy of the thalamus (see jones, 1985; sherman & guillery, 2001 for comprehensive overviews), but rather jump directly into the question how the thalamus is involved in the neurobiological processes underlying consciousness. for that we solely need to understand the difference between two types of thalamic projections to the cortex (fig. 1). the thalamus is by no means only the relay of sensory information to the cortex. it also has other functions and this view is not new. the standard account of the thalamus held in the mid-20th century was that thalamus is a part of the reticulothalamic system with two general types of thalamocortical pathways (see fig. 1): (1) the so-called specific pathways (sp) that carry afferent information to the cortex (e.g., via the lateral geniculate nucleus (lgn) in the case of vision) and (2) the socalled non-specific pathways (nsp) with large receptive fields that do not relay sensory content signals from receptors to cortex, but modulate the state of the cortical sp-neurons through diffuse cortical projections (brazier, 1960; brooks & jung, 1973; hassler, 1978; kimble, 1977; libet, 1993; magoun, 1958; mesulam, 2000; moruzzi & magoun, 1949; newman, 1995). the nsp-thalamus is an evolutionary supplement to the midbrain reticular formation and its role has been seen as the modulator of the state of cortical neural circuits, which allows transition between non-conscious states and alert states of consciousness (brazier, 1960; brooks & jung, 1973; magoun, 1958; mesulam, 2000; moruzzi & magoun, 1949; newman, 1995). the nsp system is universal in that its modulatory effect extends to all sensory modalities and all classes of sp neurons. importantly, receptive fields of sp and nsp can overlap: ascending afferents of both systems, sp and nsp converge on the same set of cortical neurons. a stimulus that drives a response in the sp system also evokes the activity of nsp (brazier, 1960; brooks & jung, 1973; magoun, 1958; mesulam, 2000; moruzzi & magoun, 1949; newman, 1995). in the later years of the 20th century the sp vs nsp distinction underwent important developments (e.g., jones, 1998, 2001). in earlier years the distinction between spand nsp-thalamus was delineated more in terms of clear structural differences and macroscopic circuit system differences. for example, in the visual system sp was more or less stringently associated with the lgn and nsp associated with certain thalamic intralaminar nuclei (e.g., bogen, 1995a, 1995b; laberge, 1997; newman, 1995; purpura & schiff, 1997; ward, 2011). subsequently, this difference became mainly associated with cell types independently of the borders of thalamic nuclei. it turns out that thalamic cells can be characterized by immunoreactivity to different proteins and that this characterization generally captures the functional bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 6 figure 1: according to the earlier influential view, there are two general types of thalamic nuclei that have different projection patterns to the cortex. the specific nuclei (red) receive most of the sensory input and send it mainly to layer 4 of the cortex. the non-specific nuclei (green) also project to layer 1 of the cortex and have a rather non-specific projection pattern, also innervating neighboring cortical columns. adapted with permission from larkum (2013) differences between sp and nsp systems as described above. matrix cells are characterized by immunoreactivity to the calcium-binding protein d28k calbindin and project widely to superficial layers of the cerebral cortex. their receptive field properties are similar to the classic nsp system and they generally receive subcortical inputs that differ from those of the sp pathways. the other type of thalamic neubachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dendritic integration theory 7 rons, the core cells are found in certain nuclei only. they project to middle layers of the cortex and target specific cortical areas. hence they functionally correspond to the sp system of the older literature. core cells are characterized by immunoreactivity to a different calcium-binding protein, parvalbumin (jones, 1998, 2001). however, even this division is a simplification: there exist multi-specific thalamocortical projection neurons that have the specific projection pattern (i.e., to layer 4) in one cortical target area, but a non-specific projection pattern (i.e., to layers 1 and 5/6) in another area (clascá et al., 2016). 3.2 empirical results supporting the role of thalamus in consciousness based on the classic views of the dual function of thalamus, thalamocortical theories of consciousness were introduced (bachmann, 1998, 1984; bogen, 1995a, 1995b; laberge, 1997; llinás et al., 1998; purpura & schiff, 1997; tononi & edelman, 1998; ward, 2011). empirical evidence in support of the thalamic bases of (or at least decisive involvement in) consciousness is rich and varied. whether the brain is asleep or awake depends on nsp. intralaminar thalamocortical neurons as part of the nsp increase their firing rate about 10 s before eeg desynchronization in natural transitions from slow-wave sleep to waking or rem sleep (steriade, 1981). thalamic deactivation is observed at sleep onset (magnin et al., 2010). thalamocortical sources send their excitatory signals up to cortex immediately before the rem sleep episodes, which typically include dream mentation (steriade et al., 1984). neocortical slow oscillations, characteristic of nonconscious states, occur after thalamocortical deafferentation (lemieux et al., 2014). honjoh and colleagues (honjoh et al., 2018) showed that stimulation of matrix cells of the non-specific ventromedial (vm) thalamic nucleus in mice awoke the animal from nrem sleep and anesthesia. this stimulation also caused eeg activation in the high frequency band. there are many known general anesthetics with different neural effects. however, virtually all of them have a common target in the thalamus (alkire et al., 2000, 2008). the fact that under general anesthesia primary sensory pathways remain functional and still communicate specific signals to cortex is adding credibility to the differentiation between spand nsp-functions. electrical stimulation of nsp of the anesthetized experimental animals leads to desynchronization of the eeg with activity-patterns typical of the awake brain (brazier, 1960; moruzzi & magoun, 1949; munk et al., 1996). lesions localized in nsp typically cause absence of consciousness or distortion of conscious experiences in patients despite the fact that the sp system has remained intact (e.g., kinney et al., 1994; bogen, 1995a; newman, 1995). similarly, damage to nsp nuclei like the pulvinar leads to spatial neglect in humans (karnath et al., 2002; ward et al., 2002). ischemia in the area supplied by the thalamo-subthalamic paramedian artery when extending to the paramedian thabachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 8 lamus (ventral posterior nuclei, dorsomedial, centromedian and central medial nuclei) shows itself clinically as impaired consciousness, hypersomnia, aphasia, amnesia and hemineglect (seifert et al., 2004). these three above-described classes of empirical facts show that the thalamus is in a position to hold the prime role in modulating the state and level of consciousness. further research shows that the thalamus can also control the contents of consciousness. for example, it is possible to generate conscious contents by stimulating nsp without the normal sensory stimulation from the receptive field of the sp-units, which manifests in artificial sensations, phosphenes, and sometimes hallucinations (e.g., gellhorn, 1961; tasker et al., 1980). the vividness of these experiences depends directly on the nature of nsp-stimulation. furthermore, timing and subjective vividness of sensation and perception can be improved by nsp functional facilitation by psychopharmacological treatment with alert subjects (brazier, 1960; doty et al., 1973). the temporal latency of the modulatory nsp-system is longer than that of sp. typical input delays in reaching cortical representational neurons are about 40–90 ms slower for nsp as compared to sp (brazier, 1960; brooks & jung, 1973; newman, 1995; steriade et al., 1990). this is in line with the general knowledge about the delay of perceptual conscious experience in response to sensory stimuli (bachmann, 2000). sp-afferents show mainly non-branching axonal projections while axons from nsp-thalamus to cortex are richly branching, thus projecting also to the neighbouring sp-neurons (bogen, 1995a; newman, 1995; purpura & schiff, 1997; steriade et al., 1990). this property puts the nsp modulation in a position to recruit context to support or contrast the sp content. lastly, the content of conscious experience is not only geared at the current environment but may also be retrieved from long-term or short-term memory. it is important to acknowledge that nsp-thalamus has a strong impact on memory formation and retrieval (pereira de vasconcelos & cassel, 2015). to sum up the previous sections, the research done in the last century already had demonstrated a clear relationship between the nsp-thalamus and consciousness. the thalamus was at the center of both experimental and theoretical interest in consciousness. however, despite the quite substantial data about the role of nsp-thalamus in consciousness, since the 1990s more emphasis has been put on the relationship between consciousness and cortical processing when looking for ncc. for example, although the neural global workspace theory (dehaene & changeux, 2011; dehaene & naccache, 2001; mashour et al., 2020) includes thalamocortical processing, it has a minor role compared to the cortical frontoparietal broadcasting system (dehaene & changeux, 2011; mashour et al., 2020). perhaps this change of perspective was caused by the fact that with microelectrode recordings and fmri much evidence could be gathered about the role of cortical areas in consciousness (for review see koch et al., 2016; mashour et al., 2020). also, since the 1990s focus has been on studying and manipulating the contents of consciousbachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dendritic integration theory 9 figure 2: cortical layer 5 pyramidal neurons play a central role in both corticocortical and thalamocortical loops. by being central to both loops, they effectively couple these loops. ness and the respective nccs. the thalamus was somewhat put aside and assumed to provide the “enabling condition” for the contents of consciousness (koch, 2004). the effects in the thalamus were often seen as secondary to those in the cortex (alkire et al., 2008; mashour, 2014). in the following we want to offer a view that reconciles the thalamocortical and corticocortical perspectives (aru, suzuki, rutiku, et al., 2019; aru, suzuki, & larkum, 2020). in a nutshell we claim that thalamocortical and corticocortical processes mechanistically interact at the level of cortical layer 5 pyramidal (l5p) cells where single l5p neurons receive input from both, thalamic and different cortical sources. thalamocortical interaction at the level of l5p cells is not limited to a simple “enabling function”, but forms the basis of igniting contents into consciousness as well as modulating the level of experience of these contents. 4 thalamocortical interactions: an update 4.1 cortical pyramidal cells couple cortical and thalamocortical interactions crucial information for understanding the relationship between the thalamocortical system and consciousness has been obtained from the studies of cortical pyramidal cells (aru, suzuki, rutiku, et al., 2019; aru, suzuki, & larkum, 2020). l5p cells are attractive counterparts for any theory of cortical functioning because of bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 10 their ability to integrate information across the cortical sheet. their cell bodies lie in layer 5, but their dendrites span through all cortical layers. the apical dendrites of pyramidal cells mainly receive input from layer 1. the basal dendrites nearer to the soma mainly collect their inputs from layers 4–6. the layer 5b thick tufted pyramidal cells are the main output of the cortex, sending axons to thalamus, striatum and brainstem (harris & shepherd, 2015). while other neurons (e.g. certain types of layer 6 neurons) also project to the thalamus, layer 5b cells are thought to be the only types of cortical cells that drive thalamic activity (sherman & guillery, 2001). hence, it has long been suspected that these cells also have a central role in consciousness (ramon y cajal, 1894). cortical l5p neurons have two functionally distinct sites of integration, one in the soma (somatic compartment) and one near the top of their apical trunk (apical compartment) (aru, suzuki, & larkum, 2020; larkum, 2013). these compartments are known to be both functionally and anatomically quite distinct (aru, suzuki, & larkum, 2020; larkum, 2013). the apical tuft receives diverse input from higher cortical areas and non-specific thalamic nuclei whereas the basal dendrites receive feedforward information from lower level cortical areas and first-order thalamic nuclei (aru, suzuki, & larkum, 2020; larkum, 2013). it has been proposed that this segregation of two anatomical types of inputs correspond to the segregation of two functional types of inputs: context and sensory data (aru, suzuki, & larkum, 2020; larkum, 2013; phillips, 2017). l5p cells contain specific biophysical mechanisms that amplify the match between these two data streams (larkum, 2013; larkum et al., 1999). hence, the l5p cells are pivotal in linking feedforward and feedback signalling pathways that are necessary for contextually modulated perception (larkum et al., 1999; takahashi et al., 2016; xu et al., 2012). this notion is especially significant bearing in mind the growing theoretical consensus that contents of conscious perception are heavily determined by contextual information and not only by sensory afference (see characteristic 10 in our list of consciousness characteristics). this general view is consistent with some recent research conducted at a single-cell level. using a sensory detection task, takahashi et al. (2016) showed that manipulation of apical dendritic activity of l5p cells had an effect on the behavioral report of the animal. mice learned to detect weak whisker stimuli of different near-threshold magnitudes. psychometric curves for whisker stimulation detection were delineated and correlated with neurometric curves. the activity of the apical tuft dendrites was monitored by performing fast-scanning two-photon ca2+ imaging. the ca2+ signal of apical dendrites was found to correlate with the behavior of the animal. additionally, the apical dendritic signals predicted the behavioral hits and misses of threshold stimuli. most importantly, direct modulation of the dendritic activity through pharmacological intervention or optogenetics had a measurable influence on the detection behavior of the animal, evidenced by a shift of the psychometric curve. optogenetic enhancement of apical dendritic activity also caused false alarms -activating the apical compartment of l5p caused the bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dendritic integration theory 11 animal to lick as if a whisker stimulus had been present (takahashi et al., 2016). we interpret this as a demonstration of illusory conscious perception: the animal experienced an illusory stimulus that was not actually presented, just like humans have been shown to do when expecting a stimulus due to some context (aru et al., 2018; aru & bachmann, 2017). in other related experiments it was found that the positive relationship between dendritic activity and animal behavior is largely constrained to the pyramidal neurons in layer 5b (takahashi et al., 2020). for our present purposes this fact is important because these cells in layer 5b are known to project to thalamic nuclei (harris & shepherd, 2015). while the layer 5a cells possess denser corticocortical projections, they were not so tightly correlated with the perceptual report of the animal (takahashi et al., 2020). hence, these findings add credibility to the idea that when it comes to consciousness, thalamocortical interactions matter more than corticocortical ones. the studies described in the preceding parts suggest that l5p neurons have an important role in conscious perception of the alert animal. remarkably, the larkum team has also demonstrated a direct relationship between l5p cell activity and the state of consciousness. for example, murayama & larkum (2009) performed fiberoptic ca2+ imaging of the apical dendrites in l5p cells to compare the effect of different states of consciousness on the activity of the apical compartment. brief air-puffs were delivered to the hindlimb of the animal. the apical dendritic response to this stimulation was 4-fold stronger in the quiet awake state than in the anesthetized state. thus, the awake state has a strong impact on the activity of l5p apical dendrites (phillips et al., 2018). when the animal moved its hindlimbs in response to the air-puff an indication of perceiving the air-puff -, the apical dendritic response was 14-fold stronger than under anesthesia, this activity stemming from a prolonged response duration. in other words, the change in the state of consciousness was associated with some (4-fold) increase in the activity of apical dendrites, whereas combining the state of consciousness with a particular content (air-puff) led to a massive (14-fold) increase of this activity (murayama & larkum, 2009). hence, the massive increase in the duration of l5p apical dendritic activity reflects the interaction between the state and the contents of consciousness. evidence from research reviewed thus far shows that l5p cells are situated at the strategically central position to integrate corticocortical and thalamocortical processing (aru, suzuki, rutiku, et al., 2019; aru, suzuki, & larkum, 2020). these cells have two functionally distinct compartments. the apical compartment is integrating contextual information and modulations from diverse sources while the somatic compartment is mainly collecting feedforward information. controlling the interaction between these two compartments within a pool of single cortical pyramidal cells could provide one key mechanism for thalamocortical interactions in linking contents and states of consciousness (aru, suzuki, rutiku, et al., 2019; aru, suzuki, & larkum, 2020). bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 12 4.2 non-specific/higher-order thalamus modulates integration within single pyramidal cells recently, suzuki & larkum (2020) reported a relatively simple and elegant mechanism about how non-specific thalamic nuclei could gate the interaction between apical and somatic compartments. they were studying a long-standing mystery about consciousness and anesthetic-induced unconsciousness. namely, how is it possible that consciousness can be disrupted while not substantially affecting the firing properties of cortical neurons in general and not switching off primary feedforward input to cortical areas? the authors optogenetically stimulated the apical compartment of l5p cells and measured the effects of this perturbation on the electrophysiological activity at soma of the same cells, while varying the conscious state of the animal. in the awake state dendritic stimulation had a large effect on the soma and this manipulation led to high frequency firing of the neurons. however, various anesthetics made this effect disappear: the same optogenetic stimulation of the apical compartment did not propagate to the soma. under anesthesia the soma was decoupled from any potential apical influences generated by top-down signals specifying the context. these results could explain why anesthesia leads to loss of consciousness (fig. 3). suzuki & larkum (2020) went on to study the processes underlying this effect: which mechanisms are responsible for this decoupling between the apical and somatic compartments? it turned out that this coupling is gated by the non-specific thalamic nucleus, in this case the posteromedial nucleus (pom). the authors show that blocking metabotropic receptor activation arising from non-specific thalamic input decouples apical and somatic compartments in awake animals. in other words, anesthesia and down-regulation of metabotropic receptor activity along the apical dendrite have the same specific effect. also, inactivation of the pom with muscimol (gabaa receptor agonist) led to a breakdown of this apical-to-soma signal propagation in awake animals, strongly suggesting that non-specific thalamic nuclei can control the interaction between the two compartments of cortical l5p neurons. this is important, as it gives a mechanistic role for the thalamus in controlling cortical processing. in our view, this result demonstrates that corticocortical and thalamocortical theories of consciousness each contain a part of the story and that these two types of interactions are brought together at the level of cortical pyramidal neurons. 5 state and content of consciousness: the dendritic integration theory the key idea we want to stress in this paper is that the mechanisms of content and state of consciousness are basically overlapping and interaction between the bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dendritic integration theory 13 figure 3: the differences between the non-conscious (left) and conscious (right) states can be found both at the thalamic and at the cortical level. according to our hypothesis, these two loops – the thalamocortical and the corticocortical loop – intersect at the level of cortical layer 5 pyramidal neurons. the mechanism for this coupling is found within single pyramidal neurons: in conscious state signals are propagated from the apical to the somatic compartment whereas in the non-conscious state this is not the case. state and the content of consciousness happens on the level of single pyramidal neurons (aru, suzuki, rutiku, et al., 2019). this perspective offers a natural way of understanding how state and content of consciousness interact in the brain. moreover, it suggests that when studying consciousness it is not possible to write off the thalamus simply as an “enabling” factor: the thalamus controls the content represented in the pyramidal neurons by allowing it to “enter” consciousness. l5p cells as the units interacting both with thalamus and other cortical areas are critical for consciousness because of this dual role. we claim that the intertwinement of state and content of consciousness arises at the level of single l5p neurons because they hold a central position in both thalamocortical and corticocortical loops. these principles lay the backbone of the dendritic integration theory (dit) (aru, suzuki, & larkum, 2020). the direct relationship between l5p cell activity and the state of consciousness has been observed in murayama & larkum (2009), described earlier. with regard to the relation of l5p to the contents of perception, takahashi et al. (2016) bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 14 found that upregulation of apical dendritic activity of l5p cells enhanced the detection performance of the animal. hudetz et al. (2020) showed by intracortical in-vivo recordings that anesthesia disrupts cortical layer-specific top-down interactions between layer 3 and layer 5 of content-carrying neurons of the visual cortex. marshel et al. (2019) demonstrated that stimulation of rodent v1 orientationselective neurons caused asymmetric propagation of activity from layer 2/3 to layer 5, with layer 5 stimulation being more effective in facilitating orientationdiscrimination behavior. to put these sets of evidence together, we can say that both corticocortical and thalamocortical theories of consciousness are correct to some extent, but the full characterization of the state and content of consciousness requires the understanding of how these two loops are coupled. supported by the results of suzuki & larkum (2020) we have now more confidence in proposing here that the coupling between the corticocortical and thalamocortical loops happens at the level of single cortical l5p neurons. to recap, these neurons, especially the thick-tufted layer 5b neurons have the following crucial properties. first, the thick-tufted layer 5b neurons are the key cortical output neurons projecting also to non-specific thalamic nuclei (harris & shepherd, 2015; sherman & guillery, 2001). this means that the results of the cortical computations will be propagated from cortex to the non-specific thalamic nuclei through layer 5b neurons. as the non-specific thalamic nuclei have diffuse projection patterns across the whole cortex, a strong signal from a few cortical columns could reach the whole thalamocortical system through the thalamic “hub”. second, the apical compartment of thick-tufted layer 5b neurons is a convergence zone for corticocortical and thalamocortical information (aru, suzuki, & larkum, 2020; larkum, 2013). as the apical processing can have a dramatic effect on the spiking output of these neurons via dendritic ca2+ spikes (larkum, 2013; larkum et al., 1999), anything that is communicated by the layer 5b neurons is contextually modulated (phillips, 2017). also, completing the loop, these layer 5b neurons receive projections from the same non-specific thalamic nuclei they project to. these projections target layers 1 and 5a (larkum, 2013). finally, the research by suzuki & larkum (2020) demonstrates that the input to the apical dendrites of l5p neurons does not propagate to the soma of these cells under anesthesia. hence, under anesthesia the contextually modulated information provided to l5p apical compartment does not reach non-specific thalamic nuclei and hence is also not communicated to the rest of the thalamocortical system. importantly for the theme of the present paper, when suzuki & larkum (2020) artificially suppressed a higher order thalamic nucleus (pom) in the awake state, propagation of activity from the apical compartment to the somatic compartment was also disrupted, similarly to what was observed in the anesthetic state. these observations support the view that consciousness depends on both corticocortical and thalamocortical connectivity. according to dit the corticocortical loop is indeed essential for the computation of the contents of consciousness. yet, without being integrated to the thalambachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dendritic integration theory 15 ocortical system, this corticocortical processing is unconscious. the thalamocortical loop is necessary to integrate the contextually modulated cortical processing into conscious experience. we claim here that conscious perception of the environment or internal thought requires the loop between the apical and the somatic compartments of l5p neurons and the non-specific thalamus, which projects back to the apical dendrites of the same l5p neurons. dit suggests that consciousness is associated with the integration of information streams impinging on the apical and basal compartments of l5p neurons (aru, suzuki, & larkum, 2020). as the apical stream carries context, the result of this integration is contextually modulated experience. such dendritic integration couples thalamocortical and corticocortical loops (see fig. 3) and propagates this contextually modulated experience to the whole thalamocortical system (aru, suzuki, & larkum, 2020). the mechanism discovered by suzuki & larkum (2020) could be the crucial switch for controlling the propagation within this loop (fig. 3). 5.1 dit and the phenomenological properties of consciousness next we will see how dit is compatible with the key phenomenological characteristics of consciousness as listed above. note that we are not claiming that no other theory can explain these properties. we simply demonstrate how dit naturally accounts for them. (1) consciousness is mentally rich and multifaceted. given that specific content is mainly processed in specialized cortical areas (for review koch et al., 2016; mashour et al., 2020), it is important to acknowledge that thalamic efferents, especially those from the nsp-thalamus, innervate all areas of the cortex (jones, 1985; sherman & guillery, 2001) and are thus in a position to modulate such content-specific processing. (2) conscious experience is integrated. the thalamus is in the central position to integrate the various aspects of processing happening at different cortical nodes. it is densely connected to all areas of the cortex between and within modally specified regions (jones, 1985; sherman & guillery, 2001). (3) consciousness, when fully present, is continuous. even though specific contents can change hand-in-hand with the specific l5p neurons that are currently active, consciousness is continuous because at any given moment there are some reverberations in the thalamocortical system (aru, suzuki, & larkum, 2020). in other words, while the combination of l5p neurons participating in conscious experience changes, the integration happening with these neurons is always supported by the nsp-thalamus; the same mechanism works also for the unchanging content. in short, continuity of consciousness comes from the continuity of the non-specific mechanism that modulates specific contents. (4) state of consciousness is multidimensional. as explained above, dit proposes that consciousness arises from the interactions of one data stream impinging bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 16 on the apical compartment, another stream affecting the basal compartment, and the nsp-thalamus that controls the interaction between these streams. these three ingredients (apical compartment, basal compartment and the nsp-thalamus) could be seen as three dimensions of the state of consciousness. for example, it has been proposed that during dreaming the apical compartment can generate conscious experience with a minor contribution from the basal compartment (aru, siclari, et al., 2020). perhaps the dimension of arousal is controlled by the nsp-thalamic input to the l5p neurons. 5.2 dit and the empirical characteristics of consciousness in the introduction we also proposed six empirical characteristics that any theory of consciousness should explain. (5) there is an interaction between the state and contents of consciousness: a theory of consciousness should explain how the state of consciousness affects the contents of consciousness. dit offers a natural explanation: contents are processed by the respective l5p neurons in cortex; nsp-thalamus, a structure traditionally associated with the state of consciousness, modulates the integration of the different compartments of the l5p neurons (aru, suzuki, & larkum, 2020; suzuki & larkum, 2020). thus, the state of consciousness affects the contents of consciousness and this interaction happens along the apical dendrites of l5p neurons (aru, suzuki, rutiku, et al., 2019). (6) processing of mental information can proceed also in its subliminal mode. according to the theory presented here, if the activity in cortical areas is insufficiently propagated to the non-specific thalamic nuclei, this processing will not be conscious. in other words, we make a strong prediction that cortical processing in itself, when not integrated with the non-specific neurons of thalamus, is not conscious. in particular, feedforward cortical processing, where information is mainly flowing on the superficial layers bypassing layer 5 neurons, will be nonconscious. such non-conscious processing can still contribute to our behavior and responses (for reviews see kim & blake, 2005; kouider & dehaene, 2007; mitchell, 2006). this thalamocortical perspective could also shed new light on some debates about which cortical areas are crucial for consciousness. we claim that there are no specific cortical areas that are crucial for consciousness, it is the type of interaction with non-specific thalamic nuclei that matters (aru, suzuki, & larkum, 2020; suzuki & larkum, 2020). some areas like the prefrontal cortex and the parietal association areas have particularly strong connectivity to the non-specific thalamic nuclei (halassa & kastner, 2017; jones, 1985) and hence their activity difference is expected to be more found when the contrast between trials with and without conscious perception is applied (aru et al., 2012). in short, the theory claims that cortical processing without the propagation to non-specific thalamic nuclei will be non-conscious. bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dendritic integration theory 17 (7) there is a temporal latency between the initial cortical representation of a content and the corresponding conscious experience. stimuli need some time to be incorporated into the thalamocortical loop. according to some more classic thalamocortical theories of consciousness, this latency comes from the fact that the non-specific nuclei do not receive direct sensory input and hence respond only after the respective cortical areas have been activated by feedforward routes (bachmann, 2000). according to the present view conscious perception of a content requires the loop between the apical dendrites of l5p neurons, the soma of these cells and the non-specific thalamus, which projects back to the apical dendrites. igniting this loop takes time and therefore there is a latency from the onset of the cortical processing of a particular content to the conscious experience of that content. recent ncc research capitalizing on contrastive analysis sets the lower limit of this latency at about 100-200 ms (förster et al., 2020). (8) conscious experience has a limited temporal resolution. this characteristic is tightly related to the previous one: if any conscious perception is based on the loop we have described, then it is hard for conscious perception to resolve anything that happens faster than the processing time of this loop. in other words we claim that the temporal resolution of conscious experience stems from the propagation time within the thalamocortical loop. an explanation for experimental phenomena like backward masking has also been provided by more classic and more recent thalamocortical theories of consciousness. the novel findings gained over the last decades can help us better understand the specifics of these mechanisms. for example, in masking two successive stimuli are presented at the same spatial location and, under certain conditions, the second stimulus (“the mask”) can completely abolish the first (“the target”) from consciousness (bachmann, 2000). so, although the target is presented 50-70 ms before the mask, it is not consciously perceived. according to the present theory, this effect is obtained as follows: the first stimulus activates l5p neurons and hence starts the loop, but by the time the activity propagates from the l5p neurons to the non-specific thalamus and back to the apical dendrites of l5p neurons, the second stimulus has arrived in cortex and now “steals the fire” ignited by the first stimulus. the first stimulus starts the loop, but the following stimulus benefits from it. (9) conscious experience is modulated by attention. attention has been linked to neurobiological processes happening in thalamus for quite some time (crick, 1984). there are several ways to understand attention in the present framework. first, attention could be controlling the firing activity of cortical pyramidal cells through the modulation of activity at the apical compartment of the selective set of content carrying neurons. second, the main signal for attentional selection could come from subcortical areas like the superior colliculus that project to thalamus (krauzlis et al., 2013; wurtz et al., 2011). finally, it is possible that changes in attention may be associated with changes in the firing activity of l5p by regulating the coupling of two compartments via some non-specific thalamic nuclei (suzuki & larkum, 2020). whichever mechanism turns out to be the best for capturing the efbachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 18 fects of attention, the thalamus is in the position to affect cortical processing in the form of attentional effects. as we and many others have believed that research supports the argument that attention is best conceptualized as a process distinct from consciousness (aru & bachmann, 2013; koch & tsuchiya, 2007; lamme, 2004), we would expect that the neural mechanisms of attention are also different from the loop of consciousness, despite the fact that these mechanisms easily interact. (10) conscious experience relies on prior knowledge and expectations, it is context-dependent. as explained above, contextual information mainly targets the apical compartment of cortical pyramidal cells. by gating the interaction between the apical and the somatic compartment the non-specific thalamus can control whether this prior knowledge is incorporated into conscious experience or not. 5.3 questions for further research while the mechanism demonstrated by suzuki & larkum (2020) is elegant, several questions remain. the key question is whether the demonstrated thalamocortical gating is a condition to regulate the general state of consciousness only or can it be used to selectively gate specific alternative contents – e.g., allow certain activity patterns that carry specific contents to enter or keep from entering consciousness. while it was shown that in the alert state thalamic activity allows for the direct propagation of apical activity into the soma, it is presently unclear whether during alertness this modulation is homogenous all over cortex (i.e., the apical compartments of all pyramidal neurons in the cortex are permitted to influence the somas) or whether this gating mechanism can act more locally (i.e. allow the apical activity to propagate on certain areas of the cortex only). if the latter is the case, work needs to demonstrate how local exactly this modulation can be – can it modulate coupling between the compartments of pyramidal neurons on the level of cortical columns? one earlier model of nonspecific thalamocortical modulation necessary for consciously experiencing the content of conscious perception assumed the overlapping of receptive fields for specific cortical content-neurons and nonspecific subcortical “consciousness-providing” neurons (e.g., bachmann, 1994). it is by no means forbidden to regard overlapping spatial locations of objects as a form of spatial-location context in addition to other forms of context such as categorical associations, priming based on conditioning, or personal relevance related attentional tuning. suzuki & larkum (2020) showed that the (de)coupling mechanism is also sensitive to acetylcholine (ach) blockers that significantly reduced the wakefulnessenhanced somatic responses of the cell whereas noradrenaline (na) blockers did not (although na still modulated dendritic responses). the effect of ach on coupling is consistent with its previously found effects on neuronal properties (goard & dan, 2009; polack et al., 2013). however, the absence of an effect of na blockers on the propagation from the apical dendrites to the soma may seem inconsistent bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dendritic integration theory 19 with the wide set of data on strong involvement of na based neuromodulation in supporting consciousness (carter et al., 2010; gelbard-sagiv et al., 2018; phillips et al., 2018). since the excitability of apical compartment of l5p neurons is enhanced by na (labarrera et al., 2018; suzuki & larkum, 2020), na may contribute to the propagation of slower responses (involving ca2+ spikes) to contextual inputs arriving at the apical compartment. for example, the evidence that na modulates liminal perceptual sensitivity and accuracy (i.e., determining whether the content is consciously perceived or not), as well as late visually evoked eeg and fmri responses, supports an important enabling role for noradrenergic modulation in perceptual awareness (gelbard-sagiv et al., 2018). it is generally known that eeg correlates of conscious perception are relatively slow, emerging after about 100– 200 ms post stimulus (bachmann, 1994; förster et al., 2020; rutiku & bachmann, 2017). na phasic activity timescale nicely corresponds to this perceptual timescale (totah et al., 2019) and often facilitates perception (waterhouse & navarra, 2019). furthermore, na axons in primary cortex of the awake mice show multimodal effects: phasic responses to stimulation in one modality were accompanied by phasic responses in other modalities (e.g., associating somatosensory and visual sensory effects) (deitcher et al., 2019). thus, na neuromodulation seems to be multimodally contextual. on the other hand, involvement of ach in anesthesia vs wakeful perception regulation as well as in mediating fast conscious perception, cannot be overlooked either. for instance, recent work supports the view that cholinergic cortical response to stimulation can be phasic and with temporal resolution allowing reaction after 200 ms (disney & higley, 2020; sarter & lustig, 2020), which roughly corresponds to temporal resolution of the na post-stimulus activity. the critical role of ach as compared to na was recently suggested by pal et al. (2020): prefrontal cholinergic stimulation of the anesthetized rat, but not noradrenergic stimulation, caused awake-like reactions. eeg activation was similar for both of these interventions, however, the righting reflex, a behavioral measure of unconsciousness in rats, and some mobility was restored only by ach. while these authors relate consciousness to ach modulation and not na modulation, there is a clear possibility that behavioral reactions are performed in a zombie mode without the concomitant perceptual experience and eeg changes reflected regaining some sensitivity, but in a state akin to what in humans is called the locked-in state. future research must add more understanding of the exact relative roles of ach and na in the control of conscious states and contents. 6 conclusion in the present work we have attempted to update the classic thalamocortical theories of consciousness in the light of novel insights from circuits neuroscience. we tried to offer a perspective where the corticocortical and thalamocortical theories can be reconciled: these processes interact at the level of single cortical pyramibachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 20 dal cells. non-specific thalamic nuclei affect the coupling between the apical and somatic compartments of the pyramidal cells, thereby coupling or decoupling the thalamocortical loop. by virtue of the present conceptualization, it becomes clear why consciousness can be manipulated in either way – by cortical and by thalamic interventions. it is clear from our analysis that the thalamocortical system possesses the characteristics of a system corresponding to conscious experience. acknowledgments we are thankful to matthew larkum for comments on this manuscript. ja was supported by the european union’s horizon 2020 research and innovation programme under the marie skłodowska-curie grant agreement no. 799411. ms was supported by german research foundation exc 257. references alkire, m. t., haier, r. j., & fallon, j. h. (2000). toward a unified theory of narcosis: brain imaging evidence for a thalamocortical switch as the neurophysiologic basis of anesthetic-induced unconsciousness. consciousness and cognition, 9(3), 370–386. https://doi.org/10.1006/ccog.1999.0423 alkire, m. t., hudetz, a. g., & tononi, g. (2008). consciousness and anesthesia. science, 322(5903), 876–880. https: //doi.org/10.1126/science.1149213 aru, j. (2019). from aliens to invisible limbs: the transitions that never make it into conscious experience. in g. hesselmann (ed.), transitions between consciousness and unconsciousness (pp. 148–162). routledge. aru, j., & bachmann, t. (2013). phenomenal awareness can emerge without attention. frontiers in human neuroscience, 7. https://doi.org/10.3389/fnhum.2013.00891 aru, j., & bachmann, t. (2017). expectation creates something out of nothing: the role of attention in iconic memory reconsidered. consciousness and cognition, 53, 203–210. https://doi.org/10.1016/j.concog.2017.06.017 aru, j., bachmann, t., singer, w., & melloni, l. (2012). distilling the neural correlates of consciousness. neuroscience and biobehavioral reviews, 36(2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.12.003 aru, j., siclari, f., phillips, w. a., & storm, j. f. (2020). apical drive—a cellular mechanism of dreaming? neuroscience & biobehavioral reviews. https://doi.org/10.1016/j.neubiorev.2020.09.018 aru, j., suzuki, m., & larkum, m. e. (2020). cellular mechanisms of conscious processing. trends in cognitive sciences, 24(10), 814–825. https://doi.org/10.1016/j.tics.2020.07.006 aru, j., suzuki, m., rutiku, r., larkum, m. e., & bachmann, t. (2019). coupling the state and contents of consciousness. frontiers in systems neuroscience, 13. https://doi.org/10.3389/fnsys.2019.00043 aru, j., tulver, k., & bachmann, t. (2018). it’s all in your head: expectations create illusory perception in a dual-task setup. consciousness and cognition, 65, 197–208. https://doi.org/10.1016/j.concog.2018.09.001 bachmann, t. (1994). psychophysiology of visual masking: the fine structure of conscious experience. nova science publishers. bachmann, t. (1998). fast dynamics of visibility of brief images: the perceptual-retouch viewpoint. in s. r. hameroff, a. w. kaszniak, & a. c. scott (eds.), toward a science of consciousness ii. the second tucson discussions and debates. (pp. 345–359). the mit press. bachmann, t. (2000). microgenetic approach to the conscious mind. john benjamins publishing company. bachmann, t. (1984). the process of perceptual retouch: nonspecific afferent activation dynamics in explaining visual masking. perception & psychophysics, 35(1), 69–84. https://doi.org/10.3758/bf03205926 bachmann, t. (2012). how to begin to overcome the ambiguity present in differentiation between contents and levels of consciousness? frontiers in psychology, 3. https://doi.org/10.3389/fpsyg.2012.00082 bachmann, t., & francis, g. (2013). visual masking: studying perception, attention, and consciousness. elsevier science. bachmann, t., & murd, c. (2010). covert spatial attention in search for the location of a color-afterimage patch speeds up its decay from awareness: introducing a method useful for the study of neural correlates of visual awareness. vision research, 50(11), 1048–1053. https://doi.org/10.1016/j.visres.2010.03.013 bayne, t. (2010). the unity of consciousness. oxford university press. bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1006/ccog.1999.0423 https://doi.org/10.1126/science.1149213 https://doi.org/10.1126/science.1149213 https://doi.org/10.3389/fnhum.2013.00891 https://doi.org/10.1016/j.concog.2017.06.017 https://doi.org/10.1016/j.neubiorev.2011.12.003 https://doi.org/10.1016/j.neubiorev.2020.09.018 https://doi.org/10.1016/j.tics.2020.07.006 https://doi.org/10.3389/fnsys.2019.00043 https://doi.org/10.1016/j.concog.2018.09.001 https://doi.org/10.3758/bf03205926 https://doi.org/10.3389/fpsyg.2012.00082 https://doi.org/10.1016/j.visres.2010.03.013 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dendritic integration theory 21 bayne, t., hohwy, j., & owen, a. m. (2016). are there levels of consciousness? trends in cognitive sciences, 20(6), 405–413. https://doi.org/10.1016/j.tics.2016.03.009 bogen, j. e. (1995a). on the neurophysiology of consciousness: i. an overview. consciousness and cognition, 4(1), 52–62. https://doi.org/10.1006/ccog.1995.1003 bogen, j. e. (1995b). on the neurophysiology of consciousness: part ii. constraining the semantic problem. consciousness and cognition, 4(2), 137–158. https://doi.org/10.1006/ccog.1995.1020 brazier, m. a. b. (1960). electrical activity of the nervous system. baltimore. brooks, b., & jung, r. (1973). neuronal physiology of the visual cortex. in g. berlucchi, g. s. brindley, b. brooks, o. d. creutzfeldt, e. dodt, r. w. doty, h.-j. freund, c. g. gross, d. a. jeffreys, r. jung, u. kuhnt, d. m. mackay, e. marg, n. negrão, g. rizzolatti, j. m. sprague, g. székely, j. szentágothai, d. whitteridge, & r. jung (eds.), visual centers in the brain (pp. 325–440). springer. https://doi.org/10.1007/978-3-642-65495-4_9 carrasco, m., ling, s., & read, s. (2004). attention alters appearance. nature neuroscience, 7(3), 308–313. https://doi.org/10 .1038/nn1194 carter, m. e., yizhar, o., chikahisa, s., nguyen, h., adamantidis, a., nishino, s., deisseroth, k., & lecea, l. de. (2010). tuning arousal with optogenetic modulation of locus coeruleus neurons. nature neuroscience, 13(12), 1526–1533. https://doi.org/10.1038/nn.2682 clascá, f., porrero, c., galazo, m. j., rubio-garrido, p., & evangelio, m. (2016). chapter 4 anatomy and development of multispecific thalamocortical axons: implications for cortical dynamics and evolution. in k. s. rockland (ed.), axons and brain architecture (pp. 69–92). academic press. https://doi.org/10.1016/b978-0-12-801393-9.00004-9 crick, f. (1984). function of the thalamic reticular complex: the searchlight hypothesis. proceedings of the national academy of sciences, 81(14), 4586–4590. https://doi.org/10.1073/pnas.81.14.4586 dehaene, s., & changeux, j.-p. (2011). experimental and theoretical approaches to conscious processing. neuron, 70(2), 200–227. https://doi.org/10.1016/j.neuron.2011.03.018 dehaene, s., & naccache, l. (2001). towards a cognitive neuroscience of consciousness: basic evidence and a workspace framework. cognition, 79(1), 1–37. https://doi.org/10.1016/s0010-0277(00)00123-2 deitcher, y., leibner, y., kutzkel, s., zylbermann, n., & london, m. (2019). nonlinear relationship between multimodal adrenergic responses and local dendritic activity in primary sensory cortices. biorxiv. https://doi.org/10.1101/814657 disney, a. a., & higley, m. j. (2020). diverse spatiotemporal scales of cholinergic signaling in the neocortex. journal of neuroscience, 40(4), 720–725. https://doi.org/10.1523/jneurosci.1306-19.2019 doty, r. w., wilson, p. d., bartlett, j. r., & pecci-saavedra, j. (1973). mesencephalic control of lateral geniculate nucleus in primates. i. electrophysiology. experimental brain research, 18(2), 189–203. https://doi.org/10.1007/bf00234723 förster, j., koivisto, m., & revonsuo, a. (2020). erp and meg correlates of visual consciousness: the second decade. consciousness and cognition, 80, 102917. https://doi.org/10.1016/j.concog.2020.102917 gelbard-sagiv, h., magidov, e., sharon, h., hendler, t., & nir, y. (2018). noradrenaline modulates visual perception and late visually evoked activity. current biology, 28(14), 2239–2249.e6. https://doi.org/10.1016/j.cub.2018.05.051 gellhorn, e. (1961). cerebral interactions: simultaneous activation of specific and unspecific systems. in electrical stimulation of the brain (pp. 321–328). university of texas press. ginsburg, s., & jablonka, e. (2019). the evolution of the sensitive soul: learning and the origins of consciousness. mit press. goard, m., & dan, y. (2009). basal forebrain activation enhances cortical coding of natural scenes. nature neuroscience, 12(11), 1444–1449. https://doi.org/10.1038/nn.2402 halassa, m. m., & kastner, s. (2017). thalamic functions in distributed cognitive control. nature neuroscience, 20(12), 1669– 1679. https://doi.org/10.1038/s41593-017-0020-1 haque, h., lobier, m., palva, j. m., & palva, s. (2020). neuronal correlates of full and partial visual conscious perception. consciousness and cognition, 78, 102863. https://doi.org/10.1016/j.concog.2019.102863 harris, k. d., & shepherd, g. m. g. (2015). the neocortical circuit: themes and variations. nature neuroscience, 18(2), 170–181. https://doi.org/10.1038/nn.3917 hassler, r. (1978). interaction of reticular activating system for vigilance and the truncothalamic and pallidal systems for directing awareness and attention under striatal control. in cerebral correlates of conscious experience (pp. 111–129). north-holland. honjoh, s., sasai, s., schiereck, s. s., nagai, h., tononi, g., & cirelli, c. (2018). regulation of cortical activity and arousal by the matrix cells of the ventromedial thalamic nucleus. nature communications, 9(1), 2100. https://doi.org/10.1038/ s41467-018-04497-x hudetz, a. g., pillay, s., wang, s., & lee, h. (2020). desflurane anesthesia alters cortical layer-specific hierarchical interactions in rat cerebral cortex. anesthesiology, 132(5), 1080–1090. https://doi.org/10.1097/aln.0000000000003179 jones, e. g. (1985). the thalamus. springer us. bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.tics.2016.03.009 https://doi.org/10.1006/ccog.1995.1003 https://doi.org/10.1006/ccog.1995.1020 https://doi.org/10.1007/978-3-642-65495-4_9 https://doi.org/10.1038/nn1194 https://doi.org/10.1038/nn1194 https://doi.org/10.1038/nn.2682 https://doi.org/10.1016/b978-0-12-801393-9.00004-9 https://doi.org/10.1073/pnas.81.14.4586 https://doi.org/10.1016/j.neuron.2011.03.018 https://doi.org/10.1016/s0010-0277(00)00123-2 https://doi.org/10.1101/814657 https://doi.org/10.1523/jneurosci.1306-19.2019 https://doi.org/10.1007/bf00234723 https://doi.org/10.1016/j.concog.2020.102917 https://doi.org/10.1016/j.cub.2018.05.051 https://doi.org/10.1038/nn.2402 https://doi.org/10.1038/s41593-017-0020-1 https://doi.org/10.1016/j.concog.2019.102863 https://doi.org/10.1038/nn.3917 https://doi.org/10.1038/s41467-018-04497-x https://doi.org/10.1038/s41467-018-04497-x https://doi.org/10.1097/aln.0000000000003179 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 22 jones, e. g. (1998). viewpoint: the core and matrix of thalamic organization. neuroscience, 85(2), 331–345. https://doi.org/ 10.1016/s0306-4522(97)00581-2 jones, e. g. (2001). the thalamic matrix and thalamocortical synchrony. trends in neurosciences, 24(10), 595–601. https: //doi.org/10.1016/s0166-2236(00)01922-6 karnath, h. o., himmelbach, m., & rorden, c. (2002). the subcortical anatomy of human spatial neglect: putamen, caudate nucleus and pulvinar. brain: a journal of neurology, 125(2), 350–360. https://doi.org/10.1093/brain/awf032 kim, c.-y., & blake, r. (2005). psychophysical magic: rendering the visible ’invisible’. trends in cognitive sciences, 9(8), 381–388. https://doi.org/10.1016/j.tics.2005.06.012 kimble, d. p. (1977). psychology as a biological science. scott foresman & co. kinney, h. c., korein, j., panigrahy, a., dikkes, p., & goode, r. (1994). neuropathological findings in the brain of karen ann quinlan. the role of the thalamus in the persistent vegetative state. the new england journal of medicine, 330(21), 1469–1475. https://doi.org/10.1056/nejm199405263302101 koch, c. (2004). the quest for consciousness: a neurobiological approach. roberts & company publishers. koch, c. (2020). the feeling of life itself: why consciousness is widespread but can’t be computed. mit press. koch, c., massimini, m., boly, m., & tononi, g. (2016). neural correlates of consciousness: progress and problems. nature reviews. neuroscience, 17(5), 307–321. https://doi.org/10.1038/nrn.2016.22 koch, c., & tsuchiya, n. (2007). attention and consciousness: two distinct brain processes. trends in cognitive sciences, 11(1), 16–22. https://doi.org/10.1016/j.tics.2006.10.012 koivisto, m., & neuvonen, s. (2020). masked blindsight in normal observers: measuring subjective and objective responses to two features of each stimulus. consciousness and cognition, 81, 102929. https://doi.org/10.1016/j.concog.2020.102929 kouider, s., & dehaene, s. (2007). levels of processing during non-conscious perception: a critical review of visual masking. philosophical transactions of the royal society of london. series b, biological sciences, 362(1481), 857–875. https: //doi.org/10.1098/rstb.2007.2093 krauzlis, r. j., lovejoy, l. p., & zénon, a. (2013). superior colliculus and visual spatial attention. annual review of neuroscience, 36, 165–182. https://doi.org/10.1146/annurev-neuro-062012-170249 kuhn, g., & rensink, r. a. (2016). the vanishing ball illusion: a new perspective on the perception of dynamic events. cognition, 148, 64–70. https://doi.org/10.1016/j.cognition.2015.12.003 labarrera, c., deitcher, y., dudai, a., weiner, b., kaduri amichai, a., zylbermann, n., & london, m. (2018). adrenergic modulation regulates the dendritic excitability of layer 5 pyramidal neurons in vivo. cell reports, 23(4), 1034–1044. https://doi.org/10.1016/j.celrep.2018.03.103 laberge, d. (1997). attention, awareness, and the triangular circuit. consciousness and cognition, 6(2-3), 149–181. https: //doi.org/10.1006/ccog.1997.0305 lamme, v. a. f. (2004). separate neural definitions of visual consciousness and visual attention: a case for phenomenal awareness. neural networks: the official journal of the international neural network society, 17(5-6), 861–872. https: //doi.org/10.1016/j.neunet.2004.02.005 larkum, m. (2013). a cellular mechanism for cortical associations: an organizing principle for the cerebral cortex. trends in neurosciences, 36(3), 141–151. https://doi.org/10.1016/j.tins.2012.11.006 larkum, m. e., zhu, j. j., & sakmann, b. (1999). a new cellular mechanism for coupling inputs arriving at different cortical layers. nature, 398, 338–341. https://doi.org/10.1038/18686 lemieux, m., chen, j.-y., lonjers, p., bazhenov, m., & timofeev, i. (2014). the impact of cortical deafferentation on the neocortical slow oscillation. the journal of neuroscience: the official journal of the society for neuroscience, 34(16), 5689–5703. https://doi.org/10.1523/jneurosci.1156-13.2014 libet, b. (1993). neuronal vs. subjective timing for a conscious sensory experience. in b. libet (ed.), neurophysiology of consciousness (pp. 149–162). birkhäuser. https://doi.org/10.1007/978-1-4612-0355-1_8 llinás, r., ribary, u., contreras, d., & pedroarena, c. (1998). the neuronal basis for consciousness. philosophical transactions of the royal society of london. series b: biological sciences, 353(1377), 1841–1849. https://doi.org/10.1098/rstb.1998.0336 mack, a., erol, m., clarke, j., & bert, j. (2016). no iconic memory without attention. consciousness and cognition, 40, 1–8. https://doi.org/10.1016/j.concog.2015.12.006 magnin, m., rey, m., bastuji, h., guillemant, p., mauguière, f., & garcia-larrea, l. (2010). thalamic deactivation at sleep onset precedes that of the cerebral cortex in humans. proceedings of the national academy of sciences of the united states of america, 107(8), 3829–3833. https://doi.org/10.1073/pnas.0909710107 magoun, h. w. (1958). the waking brain. charles c. thomas. marshel, j. h., kim, y. s., machado, t. a., quirin, s., benson, b., kadmon, j., raja, c., chibukhchyan, a., ramakrishnan, c., inoue, m., shane, j. c., mcknight, d. j., yoshizawa, s., kato, h. e., ganguli, s., & deisseroth, k. (2019). cortical bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/s0306-4522(97)00581-2 https://doi.org/10.1016/s0306-4522(97)00581-2 https://doi.org/10.1016/s0166-2236(00)01922-6 https://doi.org/10.1016/s0166-2236(00)01922-6 https://doi.org/10.1093/brain/awf032 https://doi.org/10.1016/j.tics.2005.06.012 https://doi.org/10.1056/nejm199405263302101 https://doi.org/10.1038/nrn.2016.22 https://doi.org/10.1016/j.tics.2006.10.012 https://doi.org/10.1016/j.concog.2020.102929 https://doi.org/10.1098/rstb.2007.2093 https://doi.org/10.1098/rstb.2007.2093 https://doi.org/10.1146/annurev-neuro-062012-170249 https://doi.org/10.1016/j.cognition.2015.12.003 https://doi.org/10.1016/j.celrep.2018.03.103 https://doi.org/10.1006/ccog.1997.0305 https://doi.org/10.1006/ccog.1997.0305 https://doi.org/10.1016/j.neunet.2004.02.005 https://doi.org/10.1016/j.neunet.2004.02.005 https://doi.org/10.1016/j.tins.2012.11.006 https://doi.org/10.1038/18686 https://doi.org/10.1523/jneurosci.1156-13.2014 https://doi.org/10.1007/978-1-4612-0355-1_8 https://doi.org/10.1098/rstb.1998.0336 https://doi.org/10.1016/j.concog.2015.12.006 https://doi.org/10.1073/pnas.0909710107 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dendritic integration theory 23 layer-specific critical dynamics triggering perception. science (new york), 365(6453), p.eaaw5202. https://doi.org/10.1 126/science.aaw5202 martens, s., & wyble, b. (2010). the attentional blink: past, present, and future of a blind spot in perceptual awareness. neuroscience and biobehavioral reviews, 34(6), 947–957. https://doi.org/10.1016/j.neubiorev.2009.12.005 mashour, g. a. (2014). top-down mechanisms of anesthetic-induced unconsciousness. frontiers in systems neuroscience, 8, 115. https://doi.org/10.3389/fnsys.2014.00115 mashour, g. a., roelfsema, p., changeux, j.-p., & dehaene, s. (2020). conscious processing and the global neuronal workspace hypothesis. neuron, 105(5), 776–798. https://doi.org/10.1016/j.neuron.2020.01.026 mesulam, m.-m. (2000). principles of behavioral and cognitive neurology. oxford university press. mitchell, d. b. (2006). nonconscious priming after 17 years: invulnerable implicit memory? psychological science, 17(11), 925–929. https://doi.org/10.1111/j.1467-9280.2006.01805.x moruzzi, g., & magoun, h. w. (1949). brain stem reticular formation and activation of the eeg. electroencephalography and clinical neurophysiology, 1(4), 455–473. https://doi.org/10.1016/0013-4694(49)90219-9 munk, m. h., roelfsema, p. r., könig, p., engel, a. k., & singer, w. (1996). role of reticular activation in the modulation of intracortical synchronization. science (new york), 272(5259), 271–274. https://doi.org/10.1126/science.272.5259.271 murayama, m., & larkum, m. e. (2009). enhanced dendritic activity in awake rats. proceedings of the national academy of sciences, 106(48), 20482–20486. https://doi.org/10.1073/pnas.0910379106 murd, c., & bachmann, t. (2011). spatially localized motion aftereffect disappears faster from awareness when selectively attended to according to its direction. vision research, 51(10), 1157–1162. https://doi.org/10.1016/j.visres.2011.03.008 newman, j. (1995). thalamic contributions to attention and consciousness. consciousness and cognition, 4(2), 172–193. https://doi.org/10.1006/ccog.1995.1024 pal, d., li, d., dean, j. g., brito, m. a., liu, t., fryzel, a. m., hudetz, a. g., & mashour, g. a. (2020). level of consciousness is dissociable from electroencephalographic measures of cortical connectivity, slow oscillations, and complexity. the journal of neuroscience: the official journal of the society for neuroscience, 40(3), 605–618. https://doi.org/10.1523/jn eurosci.1910-19.2019 pereira de vasconcelos, a., & cassel, j.-c. (2015). the nonspecific thalamus: a place in a wedding bed for making memories last? neuroscience and biobehavioral reviews, 54, 175–196. https://doi.org/10.1016/j.neubiorev.2014.10.021 phillips, w. a. (2017). cognitive functions of intracellular mechanisms for contextual amplification. brain and cognition, 112, 39–53. https://doi.org/10.1016/j.bandc.2015.09.005 phillips, w. a., bachmann, t., & storm, j. f. (2018). apical function in neocortical pyramidal cells: a common pathway by which general anesthetics can affect mental state. frontiers in neural circuits, 12, 50. https://doi.org/10.3389/fncir.20 18.00050 pinto, y., haan, e. h. f. de, & lamme, v. a. f. (2017). the split-brain phenomenon revisited: a single conscious agent with split perception. trends in cognitive sciences, 21(11), 835–851. https://doi.org/10.1016/j.tics.2017.09.003 polack, p.-o., friedman, j., & golshani, p. (2013). cellular mechanisms of brain state–dependent gain modulation in visual cortex. nature neuroscience, 16(9), 1331–1339. https://doi.org/10.1038/nn.3464 powers, a. r., mathys, c., & corlett, p. r. (2017). pavlovian conditioning-induced hallucinations result from overweighting of perceptual priors. science (new york), 357(6351), 596–600. https://doi.org/10.1126/science.aan3458 purpura, k. p., & schiff, n. d. (1997). the thalamic intralaminar nuclei: a role in visual awareness. the neuroscientist, 3(1), 8–15. https://doi.org/10.1177/107385849700300110 ramon y cajal, s. (1894). the croonian lecture.— la fine structure des centres nerveux. proceedings of the royal society of london, 55(331-335), 444–468. https://doi.org/10.1098/rspl.1894.0063 rutiku, r., & bachmann, t. (2017). juxtaposing the real-time unfolding of subjective experience and erp neuromarker dynamics. consciousness and cognition, 54, 3–19. https://doi.org/10.1016/j.concog.2017.05.003 sarter, m., & lustig, c. (2020). forebrain cholinergic signaling: wired and phasic, not tonic, and causing behavior. the journal of neuroscience: the official journal of the society for neuroscience, 40(4), 712–719. https://doi.org/10.1523/jn eurosci.1305-19.2019 seifert, t., enzinger, c., ropele, s., storch, m. k., & fazekas, f. (2004). midbrain ischemia presenting as vertical gaze palsy: value of diffusion-weighted magnetic resonance imaging. cerebrovascular diseases (basel, switzerland), 18(1), 3–7. https://doi.org/10.1159/000078601 sherman, s. m., & guillery, r. w. (2001). exploring the thalamus. elsevier. steriade, m. (1981). eeg desynchronization is associated with cellular events that are prerequisites for active behavioral states. behavioral and brain sciences, 4(3), 489–492. https://doi.org/10.1017/s0140525x00010037 steriade, m., jones, e. g., & llinas, r. r. (1990). thalamic oscillations and signaling. wiley. bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1126/science.aaw5202 https://doi.org/10.1126/science.aaw5202 https://doi.org/10.1016/j.neubiorev.2009.12.005 https://doi.org/10.3389/fnsys.2014.00115 https://doi.org/10.1016/j.neuron.2020.01.026 https://doi.org/10.1111/j.1467-9280.2006.01805.x https://doi.org/10.1016/0013-4694(49)90219-9 https://doi.org/10.1126/science.272.5259.271 https://doi.org/10.1073/pnas.0910379106 https://doi.org/10.1016/j.visres.2011.03.008 https://doi.org/10.1006/ccog.1995.1024 https://doi.org/10.1523/jneurosci.1910-19.2019 https://doi.org/10.1523/jneurosci.1910-19.2019 https://doi.org/10.1016/j.neubiorev.2014.10.021 https://doi.org/10.1016/j.bandc.2015.09.005 https://doi.org/10.3389/fncir.2018.00050 https://doi.org/10.3389/fncir.2018.00050 https://doi.org/10.1016/j.tics.2017.09.003 https://doi.org/10.1038/nn.3464 https://doi.org/10.1126/science.aan3458 https://doi.org/10.1177/107385849700300110 https://doi.org/10.1098/rspl.1894.0063 https://doi.org/10.1016/j.concog.2017.05.003 https://doi.org/10.1523/jneurosci.1305-19.2019 https://doi.org/10.1523/jneurosci.1305-19.2019 https://doi.org/10.1159/000078601 https://doi.org/10.1017/s0140525x00010037 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org talis bachmann, mototaka suzuki, and jaan aru 24 steriade, m., sakai, k., & jouvet, m. (1984). bulbo-thalamic neurons related to thalamocortical activation processes during paradoxical sleep. experimental brain research, 54(3), 463–475. https://doi.org/10.1007/bf00235472 suzuki, m., & larkum, m. e. (2020). general anesthesia decouples cortical pyramidal neurons. cell, 180(4), 666–676.e13. https://doi.org/10.1016/j.cell.2020.01.024 takahashi, n., ebner, c., sigl-glöckner, j., moberg, s., nierwetberg, s., & larkum, m. e. (2020). active dendritic currents gate descending cortical outputs in perception. nature neuroscience, 23(10), 1277–1285. https://doi.org/10.1038/s41593020-0677-8 takahashi, n., oertner, t. g., hegemann, p., & larkum, m. e. (2016). active cortical dendrites modulate perception. science, 354(6319), 1587–1590. https://doi.org/10.1126/science.aah6066 tasker, r. r., organ, l. w., & hawrylyshyn, p. (1980). visual phenomena evoked by electrical stimulation of the human brain stem. applied neurophysiology, 43(3-5), 89–95. https://doi.org/10.1159/000102240 tononi, g., boly, m., massimini, m., & koch, c. (2016). integrated information theory: from consciousness to its physical substrate. nature reviews. neuroscience, 17(7), 450–461. https://doi.org/10.1038/nrn.2016.44 tononi, g., & edelman, g. m. (1998). consciousness and complexity. science (new york), 282(5395), 1846–1851. https: //doi.org/10.1126/science.282.5395.1846 tononi, g., & koch, c. (2015). consciousness: here, there and everywhere? philosophical transactions of the royal society of london. series b, biological sciences, 370(1668). https://doi.org/10.1098/rstb.2014.0167 totah, n. k. b., logothetis, n. k., & eschenko, o. (2019). noradrenergic ensemble-based modulation of cognition over multiple timescales. brain research, 1709, 50–66. https://doi.org/10.1016/j.brainres.2018.12.031 ward, l. m. (2011). the thalamic dynamic core theory of conscious experience. consciousness and cognition, 20(2), 464–486. https://doi.org/10.1016/j.concog.2011.01.007 ward, r., danziger, s., owen, v., & rafal, r. (2002). deficits in spatial coding and feature binding following damage to spatiotopic maps in the human pulvinar. nature neuroscience, 5(2), 99–100. https://doi.org/10.1038/nn794 waterhouse, b. d., & navarra, r. l. (2019). the locus coeruleus-norepinephrine system and sensory signal processing: a historical review and current perspectives. brain research, 1709, 1–15. https://doi.org/10.1016/j.brainres.2018.08.032 watson, a. b. (1986). apparent motion occurs only between similar spatial frequencies. vision research, 26(10), 1727–1730. https://doi.org/10.1016/0042-6989(86)90059-3 wurtz, r. h., mcalonan, k., cavanaugh, j., & berman, r. a. (2011). thalamic pathways for active vision. trends in cognitive sciences, 15(4), 177–184. https://doi.org/10.1016/j.tics.2011.02.004 xu, n.-l., harnett, m. t., williams, s. r., huber, d., o’connor, d. h., svoboda, k., & magee, j. c. (2012). nonlinear dendritic integration of sensory and motor input during an active sensing task. nature, 492(7428), 247–251. https://doi.org/10 .1038/nature11601 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. bachmann et al. (2020). dendritic integration theory: a thalamocortical theory of state and content of consciousness. philosophy and the mind sciences, 1(ii), 2. https://doi.org/10.33735/phimisci.2020.ii.52 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/bf00235472 https://doi.org/10.1016/j.cell.2020.01.024 https://doi.org/10.1038/s41593-020-0677-8 https://doi.org/10.1038/s41593-020-0677-8 https://doi.org/10.1126/science.aah6066 https://doi.org/10.1159/000102240 https://doi.org/10.1038/nrn.2016.44 https://doi.org/10.1126/science.282.5395.1846 https://doi.org/10.1126/science.282.5395.1846 https://doi.org/10.1098/rstb.2014.0167 https://doi.org/10.1016/j.brainres.2018.12.031 https://doi.org/10.1016/j.concog.2011.01.007 https://doi.org/10.1038/nn794 https://doi.org/10.1016/j.brainres.2018.08.032 https://doi.org/10.1016/0042-6989(86)90059-3 https://doi.org/10.1016/j.tics.2011.02.004 https://doi.org/10.1038/nature11601 https://doi.org/10.1038/nature11601 https://doi.org/10.33735/phimisci.2020.ii.52 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction desiderata: what do we seek to explain? a historical view on the role of thalamus and cortex in consciousness two types of thalamic projections to the cortex empirical results supporting the role of thalamus in consciousness thalamocortical interactions: an update cortical pyramidal cells couple cortical and thalamocortical interactions non-specific/higher-order thalamus modulates integration within single pyramidal cells state and content of consciousness: the dendritic integration theory dit and the phenomenological properties of consciousness dit and the empirical characteristics of consciousness questions for further research conclusion why can't we say what cognition is (at least for the time being). why can’t we say what cognition is (at least for the time being) marco facchina (marco.facchin@iusspavia.it) abstract some philosophers search for the mark of the cognitive: a set of individually necessary and jointly sufficient conditions identifying all instances of cognition. they claim that the mark of the cognitive is needed to steer the development of cognitive science on the right path. here, i argue that, at least at present, it cannot be provided. first (§2), i identify some of the factors motivating the search for a mark of the cognitive, each yielding a desideratum the mark is supposed to satisfy (§2.1). i then (§2.2) highlight a number of tensions in the literature on the mark of the cognitive, suggesting they’re best resolved by distinguishing two distinct programs. the first program (§3) is that of identifying a mark of the cognitive capturing our everyday notion of cognition. i argue that such a program is bound to fail for a number of reasons: it is not clear whether such an everyday notion exists; and even if it existed, it would not be able to spell out individually necessary and jointly sufficient conditions for cognition; and even if it were able to spell them out, these conditions won’t satisfy the desiderata a mark of the cognitive should satisfy. the second program is that of identifying a mark of the cognitive spelling out a genuine scientific kind. but the current state of fragmentation of cognitive science, and the fact that it is splintered in a myriad of different research traditions, prevent us from identifying such a kind. and we have no reason to think that these various research traditions will converge, allowing us to identify a single mark. or so, at least, i will argue in (§4). i then conclude the paper (§5) deflecting an intuitive objection, and exploring some of the consequences of the thesis i have defended. keywords cognition wars ∙ definition ∙ embodied cognition ∙ mark of the cognitive ∙ minimal cognition ∙ necessary conditions ∙ representation 1 introduction some philosophers search for the mark of the cognitive (moc): a set of individually necessary and jointly sufficient conditions defining cognition (adams, 2019; adams & aizawa, 2001; rowlands, 2009, 2010). they claim the moc is necessary to allow cognitive science to develop correctly. should cognitive science investigate distributed brain-body-world systems, as argued by the extended mind thesis aistituto universitario di studi superiori pavia. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-5753-9873 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 2 (walter, 2010; wheeler, 2010, 2019)? the answer depends on whether such brainbody-world systems qualify as cognitive systems. are botany and microbiology parts of cognitive science? again, the answer depends on whether plants and bacteria qualify as cognitive systems (cf. adams, 2010, 2018). and to know whether these systems qualify as cognitive, we need to know the moc. here, i claim that these philosophers search in vain: at least at present no moc can be provided. in §2, i examine the literature concerning the moc. i identify (some of) the reasons motivating the search and thus (some of) the desiderata the moc should satisfy (§2.1) and highlight an important tension in the literature (§2.2). anticipating, the tension is that whereas the reasons motivating the search suggest the moc should capture a scientific (or even natural) kind, the role naive intuitions play suggests philosophers are actually after our intuitive notion of cognition. i then tease apart these two projects, and argue that, at least as things stand now, both projects are bound to fail. in §3, i claim that our intuitive notion of cognition (if it exists), cannot be captured by a moc; and, even if it were to be captured by a moc, it wouldn’t satisfy the desiderata motivating the search. in §4, i claim that, as things stand, we cannot identify a moc convincingly capturing cognition as a scientific kind. this is due to the way in which cognitive science is fragmented into numerous research traditions, each suggesting (at least implicitly) a moc. since the mocs thus suggested are often mutually exclusive, we must choose one. yet, since all these research traditions seem equally worthy of pursuit, we lack any principled reason to privilege a moc over the others. §5 considers some objections to, and consequences of, my claim. §6 briefly closes the paper. 2 searching for the mark of the cognitive 2.1 some desiderata… we can all tell apart paradigmatic instances of cognition (e.g. remembering) from paradigmatically non-cognitive processes (e.g. sneezing, cf adams, 2019). why, then, should we seek the moc? this is an important question. answering it in a clear manner makes explicit what we want the moc to do, thereby identifying the desiderata it must satisfy. following akagi (akagi, 2018; akagi, 2016), i identify three motivations fueling the search; thus, three desiderata the moc should satisfy. i focus on these three only because they’re sufficient for my arguments in §3-§4 to work. i don’t want to suggest my list is complete. likely, there are other reasons to seek the moc (and so, other desiderata in addition to the ones i will consider here). motivation #1: as a whole, cognitive science has expanded, and partially shifted, its focus away from “higher thought”, towards skilled sensorimotor interactions (clark, 2001; dennett, 1987). early in its development, cognitive science was mainly interested in “high-level”, perhaps exclusively human, phenomena. early ai researchers, for example, were interested in making computers able to facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 3 play checkers (samuel, 1967). they were sure that their procedures held the key to thought: indeed, they thought that computers in the ’60s, whilst unable to move and perceive appropriately, were able to think (selfridge & neisser, 1960). the zeitgeist seems now inverted: ai researchers focus on sensorimotor interactions (e.g. tani, 2016), and the consensus seems to be that whilst computers might perceive like us, they definitely don’t think like us (mitchell, 2019). and whilst “higher thought” is still an explanandum of cognitive science, the emphasis often is now placed on its sensorimotor roots for instance highlighting the number of ways in which the cortical structures for “higher thought” depend on the ones in charge of our sensorimotor couplings (anderson, 2014; barsalou, 1999; cisek & hayden, 2022). sensorimotor interactions precede “higher cognition” not just in phylogeny and ontogeny, but now also in the order of explanation. note how such a shift in focus generates worries concerning the distribution of cognition. only humans (and some computers) play checkers. only humans (and perhaps some computers) understand natural languages. if these are the central cases of cognition, then cognizers are relatively few: some mammals, maybe some computers. in contrast, if the central cases of cognition consist in some sensorimotor interaction, the number of cognizers is higher, including all multicellular animals, and arguably simple robots (braitenberg, 1984), plants (calvo garzon, 2007) and single celled organisms (lyon, 2015). maybe even some planetary scale processes could be construed as cognitive processes (frank et al., 2022). so, who’s in? which systems should cognitive science study? the moc should enable us to answer. it should give us an extensionally adequate definition of cognition. hence the first desideratum. desideratum #1: the moc should be an extensionally adequate definition of cognition: i.e. a set of individually necessary and jointly sufficient conditions the satisfaction of which identifies all and only cognitive systems (or states, or processes) this seems an important desideratum, whose centrality is greatly emphasized in the literature concerning extended cognition (adams, 2010; adams & aizawa, 2001, 2008; rowlands, 2009, 2010). motivation #2: cognitive science is extremely fragmented. not only the paradigmatic explananda of cognitive science have changed, the explanantia have changed too, and dramatically so. yet, “change” might not be the right word it might suggest a gradual maturation. but, that’s not what one sees when looking at the history of cognitive science. rather, one sees the splintering of a (relatively well defined) research tradition into a myriad of different and competing research traditions, each rhetorically presenting itself as a “kuhnian revolution” replacing all other research traditions and letting cognitive science run free from the shackles of ignorance. 1 notice that here i am using “research tradition” technically, 1as steiner (2019) convincingly argues, the “kuhnian rhetoric” does not capture the relevant conceptual changes in cognitive science. in his view, cognitive science (or, at least, the passage from facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 4 to name a (fairly well-defined) set of theoretical assumptions, modeling tools, experimental procedures and other research practices a group of scientists use to investigate a set of phenomena of interest (laudan, 1977, p. 81). here’s a (simplified, popular and whiggish) history of cognitive science. 2 it all began in the ’50s with the cognitive revolution: a multidisciplinary enterprise guided by an operative definition of cognition as symbolic (digital) computation (newell & simon, 1976). then came the connectionist revolution. connectionists proposed new computational models loosely inspired by the cerebral cortex (rumelhart & pdp group, 1986), inadvertently redefining cognition as subsymbolic computation (churchland, 1992). as these models grew in complexity, cognitive scientists discovered they were often better off using a different branch of math to deal with them; namely dynamical systems theory. hence the dynamicist revolution: computation faded into the background while cognition became the swirl of activity of a self-organizing system (thelen & smith, 1994; van gelder, 1995). this system which might, but need not, be identical with the brain, as the “4e” revolution quickly claimed (clark, 1997). the discovery that such a swirl of activity is a form of bayesian inference (parr et al., 2022) caused another revolution, accompanied by an appropriate redefinition of cognition as inferential prediction (corcoran et al., 2020; kiverstein & sims, 2021). in parallel, “old” computational ideas have been revamped by the cognitive neuroscience revolution (boone & piccinini, 2016). the above are all different and competing research traditions equally worthy of pursuit. they are different, for they all endorse different sets of theoretical assumptions, use different models and modeling techniques, and resort to different explanatory strategies abiding to different explanatory standards (lamb & chemero, 2018; piccinini, 2020). they compete, for they aim (or, at least, publicly declare to aim) at explaining the same thing namely cognitive processes. 3 and they are all equally pursuit-worthy, at least to the extent that none of them is obviously false and they are all able to generate results counting as genuine progress within the boundaries of the tradition. but which is right? the moc should help us answer. by telling us what cognition is, it should identify a scientific (perhaps natural) kind supporting relevant scientific generalizations and principles (adams & aizawa, 2001, 2008; buckner, 2015; newen, 2015). and by so doing it will point us towards the research traditions(s) to pursue: “classic” to “embodied” cognitive science) is best described as the shift in the balance of power between two long-standing and competing research traditions. i agree, and i think the analysis should be expanded to all the supposed “revolutions” in cognitive science. 2for a real history of cognitive science, see boden (2008). 3admittedly, there have been some calls for integration (e.g. eliasmith, 2013; miłkowski et al., 2018) and/or statement that different approaches may be complementary (kaplan & bechtel, 2011). but these are not just few and far between, they also exhibit bias towards privileged research tradition and/or model of explanation. thus, for example, eliasmith suggests a sub-symbolic cognitive level below the symbolic one, and kaplan and bechtel think that the explanatory power of dynamical models depends on them being “mappable” on mechanistic explanations. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 5 desideratum #2: the moc should allow us to identify which research tradition(s) are worth pursuing in the study of cognition motivation #3: disciplinary boundary disputes. all the research traditions mentioned above agree in construing cognitive science as a multidisciplinary enterprise. but the agreement stops here; for, which disciplines should be allowed to take part in the enterprise is a hotly debated matter. sure, “classic” cognitive science had some clear ideas clearly represented by the “cognitive hexagon”. cognitive science was construed as a multidisciplinary enterprise animated by philosophy, psychology, neuroscience, linguistics, anthropology and computer science/a.i. (howard, 1987). yet these ideas hardly translated into practice (núñez et al., 2019): the contribution of anthropology was modest, and “classic” cognitive science was not exactly keen on neuroscience: indeed, “classicism” is often mocked as the view that the best way to study the mind is to systematically ignore the brain (fodor, 1999). as “classic” cognitive science splintered (as sketched above), new disciplines were put in contact with, and included in, the forming research traditions. these include: engineering (pfeifer & bongard, 2008) material science (mcgivern, 2019; tripaldi, 2022), physics and complex system science (ernst, 1978; kelso, 1995), plant biology (calvo garzon, 2007), microbiology (yakura, 2018) archeology (malafuris, 2013) and more. are all of them rightful contributors to cognitive science? this question is important to answer for at least two reasons. first, the regularities and generalization about cognition that cognitive science will discover depend largely on which individual disciplines constitute it. as the number of disciplines constituting cognitive science grows, so does the number of systems cognitive science studies, pushing us towards minimalistic, behavior-based, principles and generalizations (cf. lyon, 2005; sims, 2021). conversely, a cognitive science largely dominated by human psychology will yield demanding, concept-oriented, principles and generalizations (adams, 2016; adams & aizawa, 2008). secondly, it is important to determine the disciplinary boundaries of cognitive science to use our material and intellectual resources correctly. to make the point bluntly: if microbiology was “in”, we should create reliable informational channels connecting microbiologists to, say, psychologists and linguists/psycholinguists, allowing them to share ideas, models, methods of inquiry and results. this isn’t easily done, nor is it something that can be done for free. it will require intellectual elaboration and monetary funds. these are limited resources, which we shouldn’t waste. hence the third desideratum: desideratum #3: the moc should determine the disciplinary boundaries of cognitive science, allowing us to allot our intellectual and nonintellectual resources in an appropriate manner 4 4this might be partially redundant in respect to the second desideratum: knowing which research tradition is right will most likely tell us which disciplines constitute cognitive science and how to allocate resources. yet it was worth making the point explicitly. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 6 a few words about these desiderata. first, as said above, i don’t presume my list is complete. there may be other desiderata in addition to these. secondly, i don’t assume their satisfaction is an all-or-nothing affair: a proposed moc a may satisfy one desideratum better than another proposed moc b. thus, these three desiderata (and others, if the list gets expanded) may function as a metric to determine which proposed moc to accept (and when we should “drop” a proposed moc for a competitor). lastly, notice that all these desiderata indicate that the moc should capture a genuine scientific kind; that is, a kind supporting the genuine generalization and principles of a science of cognition. in fact, extensional adequacy, explanatory power and the capacity of defining the boundaries of a scientific endeavor all seem to be properties of a theoretical term naming a genuine scientific kind (see also adams & aizawa, 2001; wheeler, 2010, 2019). the moc should thus define a theoretical term, used in a theoretical/scientific context (like “energy” in physics), rather than a folk term, used in everyday discourse (like “energy” when we say we woke up full of energy). 2.2 … and a tension and yet, the shadow of the folk looms large over the moc, generating a tension. to feel it, consider the following three features of the search for the moc. feature #1: the appeal to (more or less commonsensical) intuitions 5 is rampant (elpidorou, 2013). examples abound. bermúdez (2014, p. 415) and shapiro (2013, p. 363) simply assert that cognition must involve representations, stating they cannot see how it could be otherwise. adams & garrison (2013) do the exact same thing when they state that personal-level reasons are necessary for cognition. similarly, aizawa (2017, p. 16) claims basically without argument that cognition must in a sense be centrally unified; that is, that a cognitive agent cannot be built out of the interaction of special purpose mechanisms. in all these cases, philosophers rely on their intuitions to indicate an individually necessary condition constituting the moc. this typically isn’t how we go about searching for scientific kinds. indeed, our intuitions often stood in the way of us discovering genuine scientific kinds. our intuitions clumped together jadeite and nephrite as jade. dante’s intuition told him that the sun is a planet (inferno, canto i) and that to move upwards from the center of the earth one must turn 180° (inferno, canto xxxiv). alchemists found it compelling to think that nitric acid and hydrochloride were species of water (called aqua regalia and aqua fortis, (cleland, 2012). yet, when it comes to cognition, the care these examples invite seems to get thrown out of the window. 5i will adopt a very unsophisticated view of intuitions: they are judgments we’re prone to make and report (if asked). thus, for example, most westerners today share the intuition that the earth revolves around the sun, but very few westerners shared that intuition before the “scientific revolution”. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 7 now, one could perhaps adopt a broadly hermeneutical standpoint, arguing naive intuitions (and all sorts of biases) are always informing our scientific practice. they lie in the background, silently skewing our research in certain directions. true. yet notice the intuitions above do not lurk in the background. they are stated in the main text of the papers. their influence is upfront and direct. for they play an essential role in the philosophical literature on the moc, which brings us to the second feature. feature #2: these intuitions fly in the face (and are often intended to counteract) well-established and pursue-worthy research traditions. to continue with the examples above: bermudez and shapiro deem representation necessary despite the presence and successes of anti-representationalist research traditions in cognitive science (beer, 1995, 2000). aizawa takes a “central processor” to be necessary, despite the successes of the massive modularity research tradition (carruthers, 2006). adams and garrison’s case is even more puzzling: it seems to me that no research tradition in cognitive science even mentions personal-level reasons! examples proliferate easily: as chemero (2009, ch. 1) notices, arguments of that sort are fairly common in cognitive science, and indeed pre-date the whole debate on the moc. thus fodor & pylyshyn (1988), finding it intuitive that all cognition must be systematic, claimed that artificial neural networks are how-possibly models depicting the implementation of (independently studied and characterized) cognitive capacities. earlier still, searle (1980) purported to show the untenability of an entire research tradition with a thought experiment; that is, appealing to our intuitive reactions to an imaginary scenario. closer to us, the “dark room” argument against predictive processing views of cognition is based on the intuitive idea that, if all our brain tries to do is to predict the incoming inputs as accurately as possible, our brains (and thus, we) should crave very predictable and boring environments. but we clearly don’t crave them, so predictive processing must be wrong (sims, 2017; smith et al., 2022). notice that this widespread appeal to intuition is far from common. typically, science silences our pretheoretical intuitions. we wouldn’t, for example, trust our intuitions when it comes to discussing matter and energy. and, typically, we would not leverage our pretheoretical intuitions about matter and energy to attack a research tradition in physics. feature #3: the pertinence of certain scientific findings is openly contested. notice: the findings themselves are uncontested. no one claims that a certain experiment never happened, or that such-and-such an observation was not really made, or that certain data have been “rigged” to favor a specific research tradition. to the contrary, the purely factual aspect of discoveries and findings is (typically) left uncontested. what is contested is that such discoveries and findings give us insight about what cognition is and/or how it operates. bluntly: it is contested whether such findings matter for understanding cognition. some examples to clarify. if fodor & pylyshyn (1988) are right, then the study of cognition can ignore artificial neural networks and other neurocomputational facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 8 models. for, strictly speaking, these models stay silent on what cognition is and how cognitive capacities operate. they only illuminate how various cognitive processes may be implemented. similarly, if inspired by searle (1980, 1984) one concludes non-derived content is a necessary ingredient of cognition, then one must conclude that various lines of inquiry concerning human-artifact interactions do not really illuminate cognition. sensory substitution devices (bach-y-rita & kercel, 2003; eriksson, 2018) may be clinically relevant, and studying how tools are used in problem solving may have interesting anthropological or pedagogical implications (bocanegra et al., 2019; risko & gilbert, 2016). yet, if non-derived content is necessary, none of these two lines of inquiry sheds light on cognition. they shed light on something else perhaps in the immediate vicinity of cognition. now, unless one thinks that our intuition is tailored to capture scientific kinds (a highly unconvincing position, see akins, 1996; churchland, 1995), we should regard these features as generating an important tension in the search for the moc. on the one hand, the motivation for the search, and thus the desiderata the moc is called to satisfy, suggest that the moc should define a technical term capturing a scientific kind. on the other hand, the widespread to intuition and the fact that intuition are taken to have the same epistemic standing of scientific result, to the point that they can challenge their evidential status, suggest the moc is aimed to capture something different; namely what we’d normally call “cognition” in our everyday lives. the moc would thus elaborate upon, and make explicit, an important piece of our “manifest image”. 6 whilst both legitimate, the two projects are clearly regulated by different epistemic norms and standards. 7 i thus propose to disentangle them, and consider them separately. so, how are the prospects of these projects? 3 folksy cognition and its moc consider first the project of providing a moc capturing the “everyday” notion of cognition. such a moc aims to define a folk notion i.e., it aims to define what the layperson thinks cognition is. the prospects of this project appear extremely grim. first: “the layperson” is an abstraction. people are different, and intuitions vary. intuitions about cognition vary across cultures: trovato & eyssel (2017), for example, provide data indicating that italian and japanese highschool students attribute mental properties to androids differently including some paradigmatically cognitive properties, such as the capacity to plan and act accordingly. indeed, it appears that italian high school students are much more prone to ascribe mental and cognitive properties to artifacts than their japanese counterparts. now, 6and, in fact, calls to intuition are not evenly distributed in the literature on the moc: they are more frequently made by authors favoring “conservative”, human-centric views of cognition. 7scarantino (2012) seems to notice that two similar projects are also often entangled in affective science. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 9 on the fairly uncontroversial assumption that high school students’ ascriptions are “folk” ascriptions, the data suggest that the folk concept of cognition differs cross-culturally. intuitions about cognition also vary within cultural groups. welleducated westerners provide all sorts of definitions of cognition (cf. bayne et al., 2019, for a sample). swiss people are very divided on what counts as cognitive: 44% of swiss think robots are genuinely intelligent and 56% think they are not (arras & cerqui, 2005). almost a 50/50 split. perhaps one could argue that a moc could be found by looking at smaller cultural subdivisions. whereas swiss people in general have diverging intuitions, perhaps the swiss of a single canton (or of a single city) have more uniform intuitions. this may be the case, and, as far as i can see, there is no data suggesting otherwise. 8 but even if this were the case, there would still be reasons to think such a folk conception of cognition will not be nicely captured by a moc. as hinted at above, mocs are sets of individually necessary and jointly sufficient conditions. and the appeal to intuition is often used to impose necessary conditions. yet, most likely, our folk conception(s) of cognition will not provide us individually necessary conditions. for one thing, attempts at capturing ordinary concepts by sets of individually necessary and jointly sufficient conditions have traditionally been met by a volley of counterexamples (fodor, 1981) and whilst some concepts can be spelled out in that way (e.g. x is a triangle iff x has exactly three sides and exactly three edges) it is typically easy to do so, and the definitions provided are uncontested (machery, 2011). surely this isn’t the case with cognition (cf §2). further, our psychological theories of concepts suggest that “folksy and intuitive” concepts can hardly be adequately captured by sets of individually necessary and jointly sufficient conditions, as they do not seem to include any individually necessary condition (cf. machery, 2009, ch. 4). according to prototype theory (rosch & mervis, 1975) concepts are representations of statistically typical features of a class of items. according to exemplar theory (medin & schaffer, 1978) they represent instead individual members of that class. in both cases, no feature of the concept is individually necessary to categorize an item in a class, what matters is instead the overall similarity between item and concept. other views of concepts are surely possible (e.g. barsalou, 1999; murphy & medin, 1985), but these views do not mention individually necessary conditions either. thus, our currently most credited theories of concepts collectively suggest that our concepts are not constituted by individually necessary features. this gives us solid grounds to think that our “folksy” concept of cognition cannot be adequately captured by a moc consisting of individually necessary features. worse still, even if such a moc were provided, it would most likely not satisfy the desiderata listed above. the search for the moc is driven by the proliferation of research traditions in cognitive science and the consequent uncertainty about the scientific kind cognition (akagi, 2018; rupert, 2013). yet, our folksy intuitions 8however, i suspect this is due to the fact that this hypothesis has never been tested, and so data lacks entirely. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 10 are not tailored to the discovery of scientific kinds and the definition of technical terms. sure, our folk notion of the cognitive does pick up a clump of interesting phenomena (ramsey, 2015). 9 but these are the explananda, not the explanantia, of the cognitive sciences (and perhaps only some of the explananda). and the relevant explanantia actually populating contemporary cognitive science are far removed from our “folk” conception of the mind. for example, we “folksy conceive” vision as a single process, while it most likely consists of at least two different sets of processes (goodale & milner, 1992). our folksy kind “memory” has been subdivided in a myriad of ways (working memory, semantic memory, procedural memory, long term memory, etc). in general, it seems false that our folk psychological categories identify the relevant explanatory building blocks of the mind sciences (buzsáki, 2019, ch. 1; churchland, 1981; pessoa et al., 2021). indeed, if our “folksy intuitive” conceptions about the mind provide us with the right explanantia, it would be very hard to make sense of the history of psychology why did it take so long to become a real science? perhaps i have been unfair. of course providing a moc capturing our commonsensical notion of cognition would do little to aid cognitive science but it does not need to. haven’t i conceded that much at the end of §2.2, when i suggested disentangling two different projects; the first aimed at a moc spelling out a scientific kind, the second aimed at a moc spelling out our folksy intuitions? not quite. whilst i have distinguished these two projects, i have not suggested that they should satisfy different desiderata. indeed, calls to intuitive mocs are typically responsive to the desiderata listed in §2.1. for example, adams and garrison aim at saving cognitive science from the embarrassment of not knowing what cognition is (adams & garrison, 2013, p. 340), searle’s (1980) chinese room wants to identify the right research tradition for artificial intelligence (i.e. weak ai), fodor & pylyshyn (1988) were interested in determining the role of neurocomputational models and aizawa (2017) aimed at evaluating “4e” cognitive science. so, the desiderata these proposed mocs (or parts thereof) are called to satisfy still are the ones highlighted in §2.1, and pointing out that they fail to satisfy them is a fair piece of criticism. summarizing: it is not clear whether there is a single folk notion of cognition for the moc to capture. and even if it were, our best theories of concepts suggest it would not be accurately captured by a set of individually necessary and jointly sufficient conditions. and even if it were so captured, it would not satisfy the relevant desiderata motivating the search. so, at present, the prospects of finding a moc capturing our folk notion of cognition aren’t rosy. but what about a moc capturing a scientific kind? 9which (as noted at the beginning of §2) we typically identify in extension, by listing them. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 11 4 scientific cognition and its moc suppose the moc should now define the technical/theoretical term “cognition”. surely the definition cannot be stipulative. if cognition really is a scientific kind, we want to discover rather than to decide its extension (desideratum #1). and we also want to discover (rather than decide) which are the methodologies and disciplines that probe cognition the best (desideratum #3). thus, we should reject stipulative definitions of cognition, or calls to substitute “cognition” with some other ad-hoc, crisply defined, term (e.g. keijzer, 2020).10 now, a great way to discover what cognition is, is via a dedicated scientific endeavor; namely, cognitive science. and here lies the rub: we’d like the relevant moc to come out of cognitive science, but cognitive science is fragmented in many competing research traditions, at least implicitly suggesting different mocs. indeed, it is precisely because cognitive science is so fragmented that some feel the need for the moc in the first place (§2). notice: crucially, at least some research traditions into which cognitive science is currently splintered implicitly define mutually exclusive mocs. this prevents us from adopting a form of “happy” pluralism according to which cognition itself is so multifaceted and complex that each of these candidate mocs is partially correct.11 cognition, complex as it may be, cannot have contradictory properties. yet some mocs point precisely towards such contradictory properties. for example, methodological solipsists take cognition to be, in an important sense, environmentindependent (chomsky, 1995; fodor, 1980). if the solipsist is right, then ecological psychologists (chemero, 2009), enactivists (e.g. hurley, 2001) and even externalistically minded connectionists (clark, 1993) must be wrong, for they all take cognition to be essentially environment-dependent. one might try to save that form of “happy” pluralism via inclusive disjunctions: cognition is as the solipsist describes, or as the enactivist describes, or as the connectionist describes, etc. but this falls short of the relevant desiderata. maybe this procedure could yield the true extension of cognition as desideratum #1 wishes (though this is actually extremely doubtful). but it surely won’t reveal which research tradition(s) is(are) worth pursuing (desidertaum #2). and, arguably, it tells us little (if anything) about the disciplinary boundaries of cognitive science (desideratum #3). it’s hard to see how it could be used to determine, say, whether microbiology or hematology are parts of cognitive science.12 10importantly, if lyon (2019) is right, “minimal cognition”, being a stipulative term, would be in trouble too. 11importantly, however, this does not altogether exclude that we should be pluralist and let a thousand research traditions bloom (cf. allen, 2017; chemero, 2009). it only excludes (a) that we could “glue together” all these research traditions to obtain a single moc and (b) that all the mocs implicitly suggested cannot be accepted at the same time. 12notice that the moc proposed by akagi (2022) is something of this sort: it consists of a rigid structure of interconnected variables, each of which is able to assume a range of values. each variable represents a locus of contention in regards to the definition of cognition, and each value facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 12 once could contend that a moc created via inclusive disjunctions as hinted above actually satisfies the desiderata. it satisfies desideratum #1 because it gives us the true extension of cognition: everything cognitive scientists study. it satisfies desideratum #2, because it tells us which research traditions to pursue: namely, all of them. and it satisfies desideratum #3 because it tells us the disciplinary boundary of cognitive science: these boundaries include all disciplines one might use to study cognition. yet, it seems to me that arguing in this way leads the moc searcher to a pyrrhic victory (at best). after all, the moc thus provided makes no difference to the current state of cognitive science. the boons the moc should deliver are brought about in name only. since that sort of “happy” pluralism is not an option, we must choose. how? choosing arbitrarily would amount to stipulating a moc. so, we need some principled reason to choose a moc (or at least few mutually consistent ones) over the others. perhaps our choice could be based on our best pieces of empirical evidence. that’s how einstein prevailed over newton. why can’t gibson prevail over gregory the same way? yet, it is hard to see how empirical evidence could decide for one of the many research traditions (and associated mocs) over any other. for, as highlighted in §2.2, what counts as evidence concerning cognition is itself contested. further, it seems that (almost) any piece of evidence can be used to support any moc. consider one of the coarsest divisions in cognitive science; namely the one between representationalist and anti-representationalist research traditions. the former claim cognition requires representations; the latter claim it doesn’t. whilst some think the debate is solved just by noticing that we cannot explain every interesting piece of behavior only in stimulus response terms (churchland, 2002), things are in no way that simple. for, clearly, anti-representationalists are not mad: they do hold that internal states of all sorts matter in the explanation of behaviors, and they do hold that nomically relevant tracking relations hold between these states and external targets. yet they do deny that the former represent the latter in any relevant sense (orlandi, 2014). and, contra (thomson & piccinini, 2018) we cannot simply “take a peek” inside cognitive systems to see whether representations are tokened in there. for, it is easy to interpret (bona fide) nonrepresentational states in representational terms (bechtel, 1998; shapiro, 2013), as well as “deflating” (bona fide) genuine representations as mere causal mediators (facchin, 2021; ramsey, 2007)—even when the cognitive system we’re looking at is the brain (cf. gessell et al., 2021; kriegeskorte & kievit, 2013; ritchie et al., 2019). maybe, then, clever reasoning will succeed where the appeal to evidence fails. we could design sophisticated arguments showing that one, or more, research tradition(s) ought to be abandoned. chomsky managed to identify one such argument represents a position actively engaged in the dispute. but, as akagi notes, such a moc does not yield us the extension of “cognition”, nor does it suggest which scientific endeavors are worth pursuing. what such a moc does is capture in an orderly manner the extent of the disagreement concerning cognition. and that is akagi’s primary purpose. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 13 against behaviorism, and there seems to be no reason as to why, say, enactivism should be immune to such arguments. so we could search the moc by elimination: narrowing down the set of research traditions (and thus candidate mocs) down to one, or few mutually consistent ones. whilst viable in principle, this way of proceeding likely won’t be viable in practice. even chomsky’s famous arguments against behaviorism failed to force a wholesale abandonment of behaviorism (staddon, 2021). minsky’s and papert’s (1969) analysis, whilst rigorous and on the point, (thankfully!) failed to force a wholesale abandonment of connectionism. and the arguments offered by chomsky, minsky and papert are not just strong and well constructed: they are (and have been) persuasive. they impacted the day-to-day research practice of numerous cognitive scientists, and had a sizable impact on cognitive science. most other arguments aimed at motivating the abandonment of a specific research tradition are neither as strong nor as persuasive as these ones (chemero, 2009, p. ch.1). this suggests such a process of elimination is very hard, if not impossible, to translate into practice. and even if it were translated into practice, it might not be translated successfully. even a single research tradition can generate multiple mocs, for the individual disciplines within that tradition would still pull the moc in different directions. for example, microbiology and plant science often focus on the way in which (comparatively simple) biological systems cope with its immediate environment, focusing on relatively small-scale sensorimotor interactions (baluvška & levin, 2016; lyon, 2015). these disciplines actually, their philosophical spokespeople push for fairly minimal and liberal mocs, which can be easily applied to the system they are interested in studying (duijn et al., 2006; lyon, 2005). but robotics and ai push for more restrictive and demanding mocs (cf. webb’s piece in bayne et al., 2019; nolfi, 2002; tani, 2007, 2016; webb, 2006). it’s not hard to understand why: they know that comically simple systems can skillfully interact with the environment (braitenberg, 1984). so they favor demanding mocs justifying their claim that (certain) robots and computers really cognize.13 the general point is nicely exemplified by the exchange between corcoran et al. (2020) and kiverstein & sims (2021). both candidate mocs “came out of” the same research tradition; namely active inference. according to this tradition, cognition is best studied deploying a complex set of modeling tools allowing us to construe cognitive activity as a self-organizing process whereby a system brings about sensory states consistent with (and confirming) its own prolonged existence through time (parr et al., 2022). whilst corcoran and colleagues and kiverstein and sims agree on that much, they still propose different mocs. and these differences matter given the desiderata highlighted above. in short, corcoran and 13note that here i have considered two disciplines that, in the current landscape of cognitive science, are often quite close and willing to cooperate with each other (beer et al., 1997; keijzer, 2001). indeed, these disciplines form the backbone of strongly embodied, enactive approaches to cognition. and yet, they do not seem to agree on the moc. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 14 colleagues suggest that cognition is a rather sophisticated form of counterfactual inference which is not universally possessed by living systems. conversely, kiverstein and sims suggest that cognition is a form of anticipation all (or almost all) living systems exhibit. so, they disagree concerning the extension of “cognition” (desideratum #1). they also disagree on the disciplinary boundaries of cognitive science (desideratum #3): whereas kiverstein and sims suggest that biological sciences are en masse part of cognitive science, corcoran and colleagues resist the suggestion. so, appeals to empirical evidence and “pure” reasoning will not yield a moc. why, then, don’t we ask history? let the research traditions develop and compete. some will blossom, some won’t, and at the end they will deliver a coherent picture of cognition. whilst i am generally sympathetic to this suggestion, i also have some reservations about this sort of “wait and see” strategy. one important reason to be wary of this “wait and see” approach, i think, is that it presupposes that cognitive science will develop in a way that will lead us towards a single moc. this may happen in various ways. on an extreme, it might happen in a “selectionist” fashion: the best research tradition will (eventually) win the day, driving its competitors to extinction and imposing a single moc. on the other extreme, different research traditions may “fuse” by downplaying their differences and/or develop towards a common position, ending up providing a single moc. between these two extremes, all sorts of intermediate developments are possible (e.g. maybe two research traditions ta and tb will “fuse” generating tab which will then prevail in a purely “selectionist” manner). be as it may, the presupposition that somehow cognitive science will develop so as to provide a single moc does not seem to be particularly well-justified. consider first the prospects of a “selectionist” development. some potent reasons as to why such a development is unlikely have been reviewed at length just a few paragraphs above: neither empirical evidence nor arguments seem able, both in principle and in practice, to cause any research tradition to “go extinct”. further, even if a single research tradition were “selected over” its competitors, it could still fail to articulate a single moc: indeed, currently many individual research traditions provide more than one moc. so, it seems unlikely that one single moc will be established through these “selectionist” means. what, then about the other extreme, the “fusion” of various different research traditions? this development seems unlikely too. different research traditions make mutually exclusive claims. for example, whereas classical cognitivists think that cognition is computational (fodor, 1975), enactivists think that our biological nature prevents cognition from being computational in any sense (di paolo & barandarian, 2016). bayesian psychology casts perception as a form of inference (rescorla, 2013), but ecological psychologists claim that perception cannot be understood as a kind of inference (gibson, 1979). it’s hard, to say the least, to see how these different research traditions may “fuse” in a coherent manner. a research tradition cannot be both computationalist and anti-computationalist, inferentialfacchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 15 ist and anti-inferentialist. these are mutually exclusive theoretical stances, which cannot coexist or be “fused” together. when it comes to these matters, no “fusion” seems possible: something has to go. to be clear, this is not to deny that some research traditions might “fuse” in a coherent manner, perhaps because, in the grand scheme of things, their differences are relatively minor.14 but such “fusion friendly” research traditions seem to be the exception, rather than the rule. hence the likelihood of cognitive science marching towards a single moc by progressive “fusions” is extremely low. one might perhaps object that i am overemphasizing the differences between research traditions. the objector makes a fair point: there is some important intellectual work suggesting that it might be possible to coherently “fuse” prima facie mutually inconsistent research traditions. for instance, villalobos and dewhurst (2018, 2017) have tried to build some bridges connecting computationalism and enactivism.15 weinberger & allen (2022) have recently argued that dynamical models of cognition may be less inimical to computational models than initially supposed. whilst “syncretic” works of this sort are still, to my knowledge, few and far between, it is important to explicitly acknowledge their existence here, as they seem to be counterexamples to my claim that certain research traditions might be too theoretically different to fuse coherently. the counterexample is on point. and i don’t want to pose as a fortune teller: it could be the case that some day all the different research traditions of current cognitive science will fuse in one single research tradition yielding a single moc. further, it could be that a single research tradition will be “selected over” all of its competitors. none of these two developments is impossible. so, it is possible to adopt a “wait and see” approach to the moc. at this point, however, it is worth highlighting the tension between the “wait and see” strategy and the current project of searching for the moc. for, a “wait and see” approach is attractive only if one is willing to assume that the development of cognitive science not only can, but also will in fact, develop correctly in the absence of a moc (cf allen, 2017).16 indeed, a “wait and see” approach makes sense only if one thinks that (a) worries about which research traditions should be pursued (desideratum #2) and (b) worries about the disciplinary boundaries of cognitive science (desideratum #3) will eventually take care of themselves if given enough time. but if one thinks that (a) and (b) will eventually take care of themselves, then one won’t be motivated to search for a moc in the first place or, at least, not by the worries discussed in §2.1. simplifying to the extreme: if “wait and see” approaches are right, then we have little reason to search for the moc right now. 14this seems to be happening right now to ecological psychology and enactivism (baggs & chemero, 2021; chemero, 2009). 15yet, notably, these connections would require enactivists to abandon their anti-computational stance. as noted above when it comes to such fusions something must go. 16it’s important to notice that such an approach might underplay a host of “sociological” factors that might impact the development of sciences for the better or the worse (e.g. distribution of funds). facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 16 and, if one is motivated to search for the moc right now, one cannot do so by waiting for cognitive science to take care of itself. so, even conceding that prima facie mutually exclusive research traditions can coherently and productively “fuse” (or that a single research tradition might eventually triumph), there would still be a pragmatic contradiction in searching for the moc and adopting a “wait and see” approach. importantly, no argument in this whole section entails that a moc will not be provided. my arguments are, for the most part, based on the current splintered state of cognitive science. so, as i noticed above, if cognitive science can be re-unified (and the pull of various individual disciplines for different mocs is somehow dealt with), then perhaps a moc defining the scientific kind cognition may be provided. ultimately, then, time will have the last word; we should wait and see what it will say. and, as i have argued just above, adopting this “wait and see” attitude forces us to at least pro tempore abandon our search for the moc. 5 where from here? tying things up: §2 highlighted some of the motivations behind the search for the moc, as well as the desiderata a moc should satisfy (§2.1). it also highlighted a tension in the current search for the moc, which is due to the massive role played by intuitions (§2.2). hence the proposal of distinguishing two different projects: (i) that of providing a moc capturing the “folksy” notion of cognition and (ii) that of providing a moc capturing a scientific kind. i then moved a variety of considerations to claim that, at least at things stand now, none of these two projects can be carried out successfully (§3-§4). am i right? and if yes, what would follow? on the “am i right” point, i want to address a foreseeable objection. the objection is this: asking for individually necessary and jointly sufficient conditions is setting an unreasonably high bar. that sort of definitionism is dead for good (taylor & vickers, 2016). and currently popular accounts of scientific concepts in no way focus on individually necessary and jointly sufficient conditions. according to a first popular account, scientific terms capturing genuine natural kinds pick up homeostatic property clusters; that is the mechanism in virtue of which a number or relevant properties (in this case, cognitive ones) cluster together (boyd, 1991). buckner (2015) proposed a moc of this kind, and his account is safe from many of the problems i raised. according to a second popular account, scientific concepts are patchworks they consist in a richly interconnected series of domain-specific and specific-empirical-technique involving uses of a term (haueis, 2021). while no “patchwork moc” has been provided yet, it might be provided, and may be a serious alternative to more “definitionist” mocs. in reply, notice that the request for individually necessary and jointly sufficient conditions does not come from me (adams & aizawa, 2001; rowlands, 2009, 2010; walter & kästner, 2012). so, while i agree that the bar may be too high, it is not a bar i am setting. notice, importantly, that the request for individually necessary facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 17 and jointly sufficient conditions is non-trivially related to the reasons motivating the search for the moc and the desiderata i have examined in (§2.1). only classic definitions of cognition (i.e. a set of individually necessary and jointly sufficient conditions) identify all and only the instances of cognition (desideratum #1). a cluster-based, or even prototype-based (newen, 2015), approach would leave a “gray area” of uncertain cases. but precisely for this reason, such proposed mocs will not crisply determine which research traditions are worth pursuing and which individual disciplines will be relevant to our cognitive scientific endeavor (desiderata #2 and #3). similarly, a patchwork account would be a descriptive account capturing how the world “cognition” is used in various different scientific contexts. being descriptive, it won’t tell us what we should do to do good cognitive science: hence it will be silent on which research tradition we ought to follow, and on which individual disciplines we ought to practice (desiderata #2 and #3). similarly, it will tell us how “cognition” is used in current cognitive science, rather than when its current usage is correct (desideratum #1). thus, it seems that clusterand patchwork-based approaches are not viable alternatives to a “definitionist” moc at least, if desiderata #1 to #3 set the goals of one’s search. this means that if one does not take meeting these desiderata as the endpoint of one’s own search, one is free to go for a patchworkor cluster-based moc. importantly, however, since these desiderata are non-trivially connected with one’s motivation to search for the moc, this means that one’s reasons to search for the moc will have to be different too. given the difference in motivation, then, it might be smart to construe that kind of project as an altogether different project to construe it as the search for the moc*17 rather than the moc. importantly, this paper is silent on the prospects of the search for a moc*. so, as far as i am concerned, searching for mocs* might be an important and valuable endeavor. my only recommendation when it comes to mocs* is to keep them as distinct as possible from the moc, clarifying that mocs* are supposed to accomplish different epistemic tasks and thus that they are responsive to different desiderata. one could further object my proposal of separating the search for the moc from the search for the moc* is not really coherent with an observation i made in §2.1; namely that the reasons to search for the moc which motivate the adoption of desiderata #1 to #3 are not exhaustive. i have claimed that other reasons could motivate the search for the moc too. doesn’t this assertion run counter my proposal of teasing apart the search for the moc from the search for the moc* based on the different reasons motivating the two searches? no, it doesn’t or at least, not necessarily: there’s a reading of §2.1 according to which no contradiction arises (and that reading, of course, is the intended one, as indicated in §2). in §2.1 i conceded that there may be other reasons to search for the moc in addition to the one i examined. hence, i conceded that desiderata #1 to #3 may be supplemented by 17what is the moc*? i am leaving this issue purposefully undetermined. for now, we can simply think of the moc* as a variable, ready to set a variety of values depending on the reasons motivating one’s search. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 18 desiderata #4 to #n, each corresponding to the reasons that motivate one to search for the moc in addition to the one §2.1 focused on. notice that, in this case, the reasons corresponding to desiderata #1 to #3 would continue to motivate the search for the moc. yet, clearly, the very same reasons cannot motivate the search for a moc*: else, a moc* would still be required to satisfy desiderata #1 to #3, and so patchworkand cluster-based approaches would still not be viable. hence, the different reasons that motivate the search for the moc* (whatever they may be) must be conceived of as an alternative, rather than an addition, to the reasons examined in §2.1. on the “what follows” point, opinions vary. some think the absence of a moc does not matter (clark, 2008) others, instead, paint apocalyptic scenarios. adams and aizawa (2008, pp. 79–83; aizawa, 2017), in particular, contend that, absent a moc, we are all drawn towards a nasty form of operationalism. and that is undesirable for a wide number of reasons. first, it allows us to identify cognition only in reference to some paradigmatic cognitive processing, without knowing what it really is. secondly, it leads us to over-attribute cognition. many outcomes of cognitive processes can be brought about by non-cognitive means. lastly, operationalism leaves the door open to the return of behaviorism, and surely no one wants behaviorism to return, right?18 this pessimism is unjustified. behaviorism is not returning, partially because it never left, and partially because it does not seem to be gaining popularity. antirepresentationalism might be gaining popularity, but anti-representationalism is not behaviorism. reading (anderson, 2014; beer, 2000; chemero, 2009; kelso, 1995) and others, one does not find any reference to classical or operant conditioning, stimulus response chains, or skinner boxes. moreover, the absence of a moc does nothing, as far as i can see, to support behaviorism. on the one hand, the absence of a moc in no way encourages us to try and explain behavior exclusively in terms of stimulus-response chains. indeed, the absence of a moc is entirely compatible with an adoption of computationalism for purely pragmatic reasons (von neumann, 1958). whilst the truth (or appropriateness) of a behavioristic moc would entail the truth (or appropriateness) of behaviorism, the absence of a moc doesn’t. indeed, it’s entirely unclear how the absence of a moc would support a research tradition over any other research tradition. moreover, the charge of operationalism is surely overblown (see also rupert, 2013). operationalism is typically understood as the view that the meaning of theoretical terms consists in observations/measurement outputs. according to operationalism, a statement such as “the temperature of the substance a is x” means roughly “you will read x if you probe a with a thermometer”.19 operationalism is a view on the semantics of theoretical terms. such a semantic hypothesis is surely not entailed by our inability to define cognition. 18these last two worries are more evident in (adams & aizawa, 2001). 19however, this interpretation of operationalism might not capture what psychologists do when they operationalize a term (i.e. providing a working definition, see feest, 2005). facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 19 one may worry that my claim is pushing for mysterianism about cognition. and since mysterianism is bad because it amounts to giving up in our explanatory attempts (dennett, 1991), my claim ought to be rejected. now, mysterianism is the claim that although our phenomenology is physical, our cognitive architecture is wired in a way such that we just cannot understand how it metaphysically depends on the brain (mcginn, 1989). this thesis can be easily applied to cognition. according to the mysterianist about cognition, cognition is a physical phenomenon, but, due to some feature of our cognitive architecture, we cannot figure out how it metaphysically depends on the brain. but i am not pushing for mysterianism in any way. for one thing, nothing in my argument entails that we cannot explain how various cognitive processes relate to various physical systems and processes. understanding how a connectionist model works20, for instance, allows us to understand how a neural network may systematically relate inputs and outputs, and how connections can collectively store a system’s memories. there is nothing mysterious or reason-defying in the workings of connectionist models. sure, the inner workings of complex artificial neural networks with billions of parameters might not be “intuitively graspable” as the simple, three-layered feedforward networks of the 80’s—but there are techniques to track the inner goings-on of such networks (olah et al., 2018), and, at any rate, there is no special mystery concerning how billions of parameters may “store” the complex statistical models governing the functioning of these networks. and in fact, notice that anti-connectionists do not claim that connectionist models fail in making intelligible how cognition “pops out” of the firing of a bunch of interconnected (artificial) neurons. they may object that the models are too static, simple and biologically implausible, or that they are models of implementation rather than cognition properly understood. but, at least to my knowledge, no one denies that such models provide at least an howpossibly explanation of how matter and cognition relates. this gives rise to an important disanalogy between consciousness and cognition21, which significantly deflates the charge of mysterianism. further, my arguments about our inability to define cognition do not depend in any way on the contingent features of our cognitive architecture.22 indeed, if my arguments are correct, in order to find a moc we do not need a genetic mutation altering our cognitive architectures—we only need to more or less significantly alter the way in which we practice cognitive science. so, i am in no way suggesting giving up on our explanatory endeavors. to the contrary, my claim suggests we should keep trying to ameliorate our scientific practices (at least, if we care about providing the moc). 20here, i am using connectionism as an example. analogous considerations hold for different research traditions. 21indeed, that “disanalogy” is precisely what distinguishes easy problems from the hard problem (chalmers, 1997). 22not even the point about our folksy conception of cognition not specifying individually necessary conditions. for, that point was intended to hold regardless of the relevant conceptual format deployed, and thus regardless of the specific architecture relying on them. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 20 lastly, i wish to point out that my conclusion is compatible with the view that “cognition” (and the mentalistic lexicon more generally, see (dennett, 1991; schwitzgebel, 2021) is vague. to be clear: i will not argue that cognition (or the mentalistic lexicon more generally) is vague at least, not here. nor do i claim that my conclusion provides an argument in favor of cognition (or the broader mentalistic lexicon) being vague. i am just claiming that my conclusion is compatible with these forms of vagueness. my aim here is just that of highlighting something interesting for future use. that being said, my conclusion seems immediately compatible with an epistemicist stance on vagueness (williamson, 1994). take a vague term t. the epistemicist claims that: (a) everything is either determinately within the extension of t or not; (b) there’s a sharp and clear cut division between ts and not ts and (c) we’ve no idea where such a division is, and so we’re unable to say, for every x, whether x is t or not t. note the problem is epistemic: there’s a real clear cut division out there—we simply do not know where. it’s intuitive to think my argument licenses a straightforwardly epistemicist conclusion. if i am right, we can’t—at least for the time being—say what cognition is because of our epistemic standing: cognitive science is just too fragmented to allow us to define cognition. i also suspect such an epistemicist conclusion will be appealing to many: indeed, it seems to me that the moc searcher must be “contingently epistemicist” about cognition. searching for the moc seems to presuppose (a*) that every state, process or system is definitely either cognitive or not cognitive, (b*) that there’s a sharp division between the two and (c*) by ignoring what the moc is, we ignore where that division is (d*) that our ignorance is contingent and can be dispelled by finding the moc. endorsing (a*) to (c*) makes the moc seeker an epistemicist about cognition. the addition of (d*) clarifies why this epistemicism is contingent: we can dispel the vagueness around “cognition” by finding the moc. but does my argument support an epistemicism about cognition (or the mentalistic lexicon more generally)? i doubt it. for my arguments here have been blissfully neutral on (a*)—nothing of what i have argued presupposes or entails that every state, process or system is definitely either cognitive or non-cognitive. nor, to be extremely clear, does anything i have argued for here presuppose or entail the falsity of (a*). thus, in order for the arguments provided here to support an epistemicist stance about cognition one must supplement them with a compelling argument for (a*). as things stand, i know of no such argument—further, i must confess i have some troubles even imagining what such an argument might look like. but of course, my ignorance and my lack of imagination are not arguments against (a*). here i want to leave the truth value of (a*) entirely undetermined. for what i want to highlight is something quite different; namely, that my arguments here support an epistemicist stance about cognition only if they’re supplemented with some compelling and independent reasons to accept (a*). this is important to notice, for it seems to place an important argumentative boulder on the shoulders facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 21 of philosophers interested in searching for the moc. for, if, as i have argued, philosophers searching for the moc must be “contingently epistemicist” about cognition, then they owe us some reasons to accept (a*). and, as i have already noticed, these reasons can hardly be found in the current literature on the moc. importantly, since my arguments here are neutral on (a*), they’re also compatible with its negation (and so, a fortiori, they don’t support epistemicism about cognition). notice that, if the negation of (a*) were true, there would be at least one state, process or system that is neither definitely cognitive nor definitely noncognitive. vagueness would thus be a feature of cognition itself, rather than the by-product of our epistemic standing. i think such a view is attractive for several reasons. the falsity of (a*) would provide a plausible explanation accounting for the failures of proposed mocs. the falsity of (a*) could also neatly explain why many of our scientific concepts end up identifying clusters of properties, or “patchworks” of connected uses that can always be extended to novel contexts (waisman, 1968). similarly, the falsity of (a*) would allow us to make sense of the fact that we seem to find cognitive or at least cognitive-like processes, systems and properties everywhere we look (levin, 2022; tripaldi, 2022; yakura, 2018). thus, i think there are several reasons to want (a*) to be false. and whilst i know of no direct argument to the effect that (a*) is false, i can at least imagine one.23 but, to repeat myself for the sake of clarity, here i don’t want to argue that (a*) is false. hence, notice that my arguments here do not lend any inductive support to the claim that (a*) is false. sure, the falsity of (a*) would neatly explain why any attempt at providing a moc has thus far failed. thus, one might be tempted to invoke an inference to the best explanation and conclude for the falsity of (a*). yet, the fact that any attempt to provide a moc has thus far failed is not uniquely explained by (a*) being false it could be explained equally well by other factors (such as, for instance, the complexity of cognition itself, the relatively young age of psychology, or the current fragmentation of the mind sciences). moreover, the falsity of (a*) licenses a conclusion far stronger than the one i have reached here. if (a*) is false, then there’s no moc to be spelled out. but here i have not claimed that there is no moc to be spelled out.24 i have only claimed that the moc cannot be spelled out given the current fragmentation of the mind sciences. for these reasons, my conclusion does not license any inference to the best explanation to the falsity of (a*). summing up: the claim i have here defended is compatible with cognition being vague in one of the two ways seen above yet, my conclusion does not support the claim that cognition is vague in any of the two senses above. 23if you want to imagine it too, read schwitzgebel (2021) and substitute every occurrence of “consciousness” with “cognition”. 24which need not be a problem for cognitive science just like the absence of a definition of life (cleland, 2012; machery, 2011) is not a problem for biology. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 22 6 conclusions in this paper, i have examined the current literature on the moc (§2). i have highlighted the desiderata that the moc should satisfy (§2.1) as well as an important tension that pervades that literature (§2.2). to put it bluntly, it is not clear whether the moc is supposed to capture the folk notion of cognition or the scientific one. i have claimed that no moc capturing our folk notion of cognition can be provided (§3). for, there is likely no single, culturally stable notion of cognition. further, even if there were such a notion, it could hardly be captured in a moc, given that most of our folk notions cannot be captured by sets of individually necessary and jointly sufficient conditions. and, even if such a moc were to be provided, it would fail to satisfy the relevant desiderata it is called to satisfy. thus, i concluded that a moc capturing our folk notion of cognition is useless at best. i have also claimed that a moc capturing the scientific notion of cognition cannot be provided (§4). this is because, at present, cognitive science is splintered in many conflicting and contradictory research traditions. since all these research traditions are equally worth pursuing, we are unable to identify a single notion of cognition for the moc to capture. lastly, i have defended my claims from a number of objections (§5). in closing, i want to indicate some directions for future research. as i have noticed at the end of §5, the quest for the moc is importantly tied to vagueness. moc searchers must be “contingently epistemicist”, about cognition. adversaries of moc searchers may attack their endeavors by claiming that there are systems, states and/or processes that are neither definitely cognitive nor definitely noncognitive. if this is correct, then then vagueness is definitely a new battleground for the “cognition wars”. acknowledgments this paper has been presented at the 5th silfs postgraduate conference in milan, the aisc midterm conference 2022 in parma and the joint espp-spp conference in milan. i wish to thank the audience of all these conferences for their observations, they really improved the paper. a special thanks goes to (in random order) marco viola, giacomo zanotti, bruno cortesi and arianna beghetto for having read and commented upon various previous iterations of this manuscript. lastly, i wish to thank the two anonymous reviewers of philosophy and the mind sciences for their nice and profound comments. references adams, f. (2010). why we still need a mark of the cognitive. cognitive systems research, 11(4), 324–331. https://doi.org/10 .1016/j.cogsys.2010.03.001 adams, f. (2016). information and cognition. in the routledge handbook of philosophy of information (pp. 321–331). routledge. adams, f. (2018). cognition wars. studies in history and philosophy of science part a, 68, 20–30. https://doi.org/10.1016/j. shpsa.2017.11.007 adams, f. (2019). the elusive extended mind. in m. colombo, e. irvine, & m. stapleton (eds.), andy clark and his critics (pp. 21–31). oxford university press. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.cogsys.2010.03.001 https://doi.org/10.1016/j.cogsys.2010.03.001 https://doi.org/10.1016/j.shpsa.2017.11.007 https://doi.org/10.1016/j.shpsa.2017.11.007 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 23 adams, f., & aizawa, k. (2001). the bounds of cognition. philosophical psychology, 14(1), 43–64. https://doi.org/10.1080/09 515080120033571 adams, f., & aizawa, k. (2008). the bounds of cognition. oxford: blackwell. adams, f., & garrison, r. (2013). the mark of the cognitive. minds and machines, 23(3), 339–352. https://doi.org/10.1007/s1 1023-012-9291-1 aizawa, k. (2017). cognition and behavior. synthese, 194(11), 4269–4288. https://doi.org/10.1007/s11229-014-0645-5 akagi, m. (2018). rethinking the problem of cognition. synthese, 195(8), 3547–3570. https://doi.org/10.1007/s11229-0171383-2 akagi, m. (2022). cognition as the sensitive management of an agent’s behavior. philosophical psychology, 35(5), 718–741. https://doi.org/10.1080/09515089.2021.2014802 akagi, m. s. m. (2016). cognition in practice: conceptual development and disagreement in cognitive science [phd thesis]. university of pittsburgh. akins, k. (1996). of sensory systems and the “aboutness” of mental states. the journal of philosophy, 93(7), 337–372. https://doi.org/10.2307/2941125 allen, c. (2017). on (not) defining cognition. synthese, 194(11), 4233–4249. https://doi.org/10.1007/s11229-017-1454-4 anderson, m. l. (2014). after phrenology. mit press. arras, k. o., & cerqui, d. (2005). do we want to share our lives and bodies with robots? a 2000 people survey. technical report, 605. https://doi.org/10.3929/ethz-a-010113633 bach-y-rita, p., & kercel, s. w. (2003). sensory substitution and the human-machine interface. trends in cognitive sciences, 7(12), 541–546. https://doi.org/10.1016/j.tics.2003.10.013 baggs, e., & chemero, a. (2021). radical embodiment in two directions. synthese, 198(9), 2175–2190. https://doi.org/10.100 7/s11229-018-02020-9 baluvška, f., & levin, m. (2016). on having no head: cognition throughout biological systems. frontiers in psychology, 902. https://doi.org/10.3389/fpsyg.2016.00902 barsalou, l. w. (1999). perceptual symbol systems. behavioral and brain sciences, 22(4), 577–660. https://doi.org/10.1017/ s0140525x99002149 bayne, t., brainard, d., byrne, r. w., chittka, l., clayton, n., heyes, c., mather, j., ölveczky, b., shadlen, m., suddendorf, t., et al. (2019). what is cognition? current biology, 29(13), r608–r615. bechtel, w. (1998). representations and cognitive explanations: assessing the dynamicist’s challenge in cognitive science. cognitive science, 22(3), 295–317. https://doi.org/10.1016/s0364-0213(99)80042-1 beer, r. d. (1995). a dynamical systems perspective on agent-environment interaction. artificial intelligence, 72(1-2), 173– 215. https://doi.org/10.1016/0004-3702(94)00005-l beer, r. d. (2000). dynamical approaches to cognitive science. trends in cognitive sciences, 4(3), 91–99. https://doi.org/10 .1016/s1364-6613(99)01440-0 beer, r. d., quinn, r. d., chiel, h. j., & ritzmann, r. e. (1997). biologically inspired approaches to robotics: what can we learn from insects? communications of the acm, 40(3), 30–38. https://doi.org/10.1145/245108.245118 bermúdez, j. l. (2014). cognitive science: an introduction to the science of the mind. cambridge university press. bocanegra, b. r., poletiek, f. h., ftitache, b., & clark, a. (2019). intelligent problem-solvers externalize cognitive operations. nature human behaviour, 3(2), 136–142. https://doi.org/10.1038/s41562-018-0509-y boden, m. a. (2008). mind as machine: a history of cognitive science. oxford university press. boone, w., & piccinini, g. (2016). the cognitive neuroscience revolution. synthese, 193(5), 1509–1534. https://doi.org/10.1 007/s11229-015-0783-4 boyd, r. (1991). realism, anti-foundationalism and the enthusiasm for natural kinds. philosophical studies: an international journal for philosophy in the analytic tradition, 61(1/2), 127–148. braitenberg, v. (1984). vehicles: experiments in synthetic psychology. cambridge, ma. buckner, c. (2015). a property cluster theory of cognition. philosophical psychology, 28(3), 307–336. https://doi.org/10.108 0/09515089.2013.843274 buzsáki, g. (2019). the brain from inside out. oxford university press. calvo garzon, f. (2007). the quest for cognition in plant neurobiology. plant signaling & behavior, 2(4), 208–211. https://doi.org/10.4161/psb.2.4.4470 carruthers, p. (2006). the architecture of the mind. oxford university press. chalmers, d. j. (1997). the conscious mind: in search of a fundamental theory. oxford paperbacks. chemero, a. (2009). radical embodied cognitive science. the mit press. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1080/09515080120033571 https://doi.org/10.1080/09515080120033571 https://doi.org/10.1007/s11023-012-9291-1 https://doi.org/10.1007/s11023-012-9291-1 https://doi.org/10.1007/s11229-014-0645-5 https://doi.org/10.1007/s11229-017-1383-2 https://doi.org/10.1007/s11229-017-1383-2 https://doi.org/10.1080/09515089.2021.2014802 https://doi.org/10.2307/2941125 https://doi.org/10.1007/s11229-017-1454-4 https://doi.org/10.3929/ethz-a-010113633 https://doi.org/10.1016/j.tics.2003.10.013 https://doi.org/10.1007/s11229-018-02020-9 https://doi.org/10.1007/s11229-018-02020-9 https://doi.org/10.3389/fpsyg.2016.00902 https://doi.org/10.1017/s0140525x99002149 https://doi.org/10.1017/s0140525x99002149 https://doi.org/10.1016/s0364-0213(99)80042-1 https://doi.org/10.1016/0004-3702(94)00005-l https://doi.org/10.1016/s1364-6613(99)01440-0 https://doi.org/10.1016/s1364-6613(99)01440-0 https://doi.org/10.1145/245108.245118 https://doi.org/10.1038/s41562-018-0509-y https://doi.org/10.1007/s11229-015-0783-4 https://doi.org/10.1007/s11229-015-0783-4 https://doi.org/10.1080/09515089.2013.843274 https://doi.org/10.1080/09515089.2013.843274 https://doi.org/10.4161/psb.2.4.4470 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 24 chomsky, n. (1995). language and nature. mind, 104(413), 1–61. https://www.jstor.org/stable/2254605 churchland, p. (1981). eliminative materialism and propositional attitudes. the journal of philosophy, 78(2), 67–90. https://doi.org/10.5840/jphil198178268 churchland, p. (1992). brainwise. the mit press. churchland, p. s. (1995). epistemology in the age of neuroscience. the journal of philosophy, 84(10), 544–553. https://doi.org/10.5840/jphil1987841026 churchland, p. s. (2002). brainwise. the mit press. cisek, p., & hayden, b. (2022). neuroscience needs evolution. the philosophical transaction of the royal society b, 377(1844). https://doi.org/10.1098/rstb.2020.0518 clark, a. (1993). associative engines. the mit press. clark, a. (1997). being there. the mit press. clark, a. (2001). reasons, robots and the extended mind. mind & language, 16(2), 121–145. https://doi.org/10.1111/14680017.00162 clark, a. (2008). supersizing the mind. oxford university press. cleland, c. (2012). life without definitions. synthese, 185(1), 125–144. https://doi.org/10.1007/s11229-011-9879-7 corcoran, a. w., pezzulo, g., & hohwy, j. (2020). from allostatic agents to counterfactual cognizers: active inference, biological regulation, and the origins of cognition. biology & philosophy, 35(3), 1–45. https://doi.org/10.1007/s10539020-09746-2 dennett, d. (1987). why not the whole iguana? behavioral and brain sciences, 1(1), 103–104. https://doi.org/10.1017/s014 0525x00059859 dennett, d. (1991). consciousness explained. little brown. di paolo, t. b., ezequiel, & barandarian, x. (2016). sensorimotor life. oxford university press. duijn, m. van, keijzer, f., & franken, d. (2006). principles of minimal cognition: casting cognition as sensorimotor coordination. adaptive behavior, 14(2), 157–170. https://doi.org/10.1177/105971230601400207 eliasmith, c. (2013). how to build a brain. oxford university press. elpidorou, a. (2013). reasoning about the mark of the cognitive: a response to adams and garrison. minds and machines, 24(2), 201–211. https://doi.org/10.1007/s11023-013-9319-1 eriksson, y. (2018). tactile reading: tactile understanding. in assistive technology for blindness and low vision (pp. 211–236). crc press. ernst, h. (1978). synergistics: an introduction. springer. facchin, m. (2021). structural representations do not meet the job description challenge. synthese, 199(3-4), 5479–5508. https://doi.org/10.1007/s11229-021-03032-8 feest, u. (2005). operationism in psychology: what the debate is about, what the debate should be about. journal of the history of the behavioral sciences, 41(2), 131–149. https://doi.org/10.1002/jhbs.20079 fodor, j. a. (1980). methodological solipsism considered as a research strategy in cognitive psychology. behavioral and brain sciences, 3(1), 63–73. https://doi.org/10.1017/s0140525x00001771 fodor, j. a. (1975). the language of thought. harvard university press. fodor, j. a. (1981). the present status of the innateness controversy. in j. a. fodor (ed.), representations: philosophical essays on the foundations of cognitive science (pp. 257–316). cambridge, ma: mit press. fodor, j. a. (1999). why the brain? london review of book, 19(30). https://www.lrb.co.uk/the-paper/v21/n19/jerryfodor/diary fodor, j. a., & pylyshyn, z. w. (1988). connectionism and cognitive architecture: a critical analysis. cognition, 28(1-2), 3–71. https://doi.org/10.1016/0010-0277(88)90031-5 frank, a., grinspoon, d., & walker, s. (2022). intelligence as a planetary scale process. international journal of astrobiology, 21(2), 47–61. https://doi.org/10.1017/s147355042100029x gessell, b., geib, b., & brigard, f. d. (2021). multivariate pattern analysis and the search for neural representations. synthese, 199(5-6), 12869–12889. https://doi.org/10.1007/s11229-021-03358-3 gibson, j. (1979). the ecological approach to visual perception. houghton-mifflin. goodale, m. a., & milner, a. d. (1992). separate visual pathways for perception and action. trends in neurosciences, 15(1), 20–25. https://doi.org/10.1016/0166-2236(92)90344-8 haueis, p. (2021). a generalized patchwork approach to scientific concepts. the british journal for the philosophy of science. https://doi.org/10.1086/716179 facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.jstor.org/stable/2254605 https://doi.org/10.5840/jphil198178268 https://doi.org/10.5840/jphil1987841026 https://doi.org/10.1098/rstb.2020.0518 https://doi.org/10.1111/1468-0017.00162 https://doi.org/10.1111/1468-0017.00162 https://doi.org/10.1007/s11229-011-9879-7 https://doi.org/10.1007/s10539-020-09746-2 https://doi.org/10.1007/s10539-020-09746-2 https://doi.org/10.1017/s0140525x00059859 https://doi.org/10.1017/s0140525x00059859 https://doi.org/10.1177/105971230601400207 https://doi.org/10.1007/s11023-013-9319-1 https://doi.org/10.1007/s11229-021-03032-8 https://doi.org/10.1002/jhbs.20079 https://doi.org/10.1017/s0140525x00001771 https://www.lrb.co.uk/the-paper/v21/n19/jerry-fodor/diary https://www.lrb.co.uk/the-paper/v21/n19/jerry-fodor/diary https://doi.org/10.1016/0010-0277(88)90031-5 https://doi.org/10.1017/s147355042100029x https://doi.org/10.1007/s11229-021-03358-3 https://doi.org/10.1016/0166-2236(92)90344-8 https://doi.org/10.1086/716179 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 25 howard, g. (1987). the mind’s new science. basic books. hurley, s. (2001). perception and action: alternative views. synthese, 129(1), 3–40. https://doi.org/10.1023/a:1012643006930 kaplan, d. m., & bechtel, w. (2011). dynamical models: an alternative or complement to mechanistic explanations? topics in cognitive science, 3(2), 438–444. https://doi.org/10.1111/j.1756-8765.2011.01147.x keijzer, f. (2020). demarcating cognition: the cognitive life sciences. synthese, 198(s1), 137–157. https://doi.org/10.1007/s1 1229-020-02797-8 keijzer, f. (2001). representation and behavior. the mit press. kelso, s. (1995). dynamic patterns. the mit press. kiverstein, j., & sims, m. (2021). is free-energy minimisation the mark of the cognitive? biology & philosophy, 36(2). https://doi.org/10.1007/s10539-021-09788-0 kriegeskorte, n., & kievit, r. a. (2013). representational geometry: integrating cognition, computation, and the brain. trends in cognitive sciences, 17(8), 401–412. https://doi.org/10.1016/j.tics.2013.06.007 lamb, m., & chemero, a. (2018). interacting in the open: where dynamical systems become extended and embodied. in a. newen, l. de bruin, & s. gallagher (eds.), the oxford handbook of 4e cognition (pp. 146–162). oxford university press. https://doi.org/10.1093/oxfordhb/9780198735410.013.8 laudan, l. (1977). progress and its problem: towards a theory of scientific growth. university of california press. levin, m. (2022). technological approach to mind everywhere: an experimentally-grounded framework for understanding diverse bodies and minds. frontiers in systems neuroscience, 16. https://doi.org/10.3389/fnsys.2022.768201 lyon, p. (2015). the cognitive cell: bacterial behavior reconsidered. frontiers in microbiology, 6. https://doi.org/10.3389/fm icb.2015.00264 lyon, p. (2019). of what is minimal cognition the half-baked version? adaptive behavior, 28(6), 407–424. https://doi.org/10 .1177/1059712319871360 lyon, p. (2005). the biogenic approach to cognition. cognitive processing, 7(1), 11–29. https://doi.org/10.1007/s10339-0050016-8 machery, e. (2011). why i stopped worrying about the definition of life… and why you should as well. synthese, 185(1), 145–164. https://doi.org/10.1007/s11229-011-9880-1 machery, e. (2009). doing without concepts. oxford university press. malafuris, l. (2013). how things shape the mind. the mit press. mcginn, c. (1989). can we solve the mind-body problem? mind, 98(july), 349–366. https://doi.org/10.1093/mind/xcviii.39 1.349 mcgivern, p. (2019). active materials: minimal models of cognition? adaptive behavior, 28(6), 441–451. https://doi.org/10 .1177/1059712319891742 medin, d. l., & schaffer, m. m. (1978). context theory of classification learning. psychological review, 85(3), 207–238. https://doi.org/10.1037/0033-295x.85.3.207 miłkowski, m., clowes, r., rucińska, z., przegalińska, a., zawidzki, t., krueger, j., gies, a., mcgann, m., afeltowicz, lukasz, wachowski, w., stjernberg, f., loughlin, v., & hohol, m. (2018). from wide cognition to mechanisms: a silent revolution. frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018.02393 minsky, m., & papert, s. (1969). perceptrons. the mit press. mitchell, m. (2019). artificial intelligence: a guide for thinking humans. penguin. murphy, g. l., & medin, d. l. (1985). the role of theories in conceptual coherence. psychological review, 92(3), 289–316. https://doi.org/10.1037/0033-295x.92.3.289 newell, a., & simon, h. a. (1976). computer science as empirical inquiry. communications of the acm, 19(3), 113–126. https://doi.org/10.1145/360018.360022 newen, a. (2015). what are cognitive processes? an example-based approach. synthese, 194(11), 4251–4268. https://doi.org/10.1007/s11229-015-0812-3 nolfi, s. (2002). power and the limits of reactive agents. neurocomputing, 42(1-4), 119–145. https://doi.org/10.1016/s09252312(01)00598-7 núñez, r., allen, m., gao, r., rigoli, c. m., relaford-doyle, j., & semenuks, a. (2019). what happened to cognitive science? nature human behaviour, 3(8), 782–791. https://doi.org/10.1038/s41562-019-0626-2 olah, c., satyanarayan, a., johnson, i., carter, s., schubert, l., ye, k., & mordvintsev, a. (2018). the building blocks of interpretability. distill. https://doi.org/10.23915/distill.00010 orlandi, n. (2014). the innocent eye. oxford university press. parr, t., pezzulo, g., & friston, k. j. (2022). active inference. the mit press. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1023/a:1012643006930 https://doi.org/10.1111/j.1756-8765.2011.01147.x https://doi.org/10.1007/s11229-020-02797-8 https://doi.org/10.1007/s11229-020-02797-8 https://doi.org/10.1007/s10539-021-09788-0 https://doi.org/10.1016/j.tics.2013.06.007 https://doi.org/10.1093/oxfordhb/9780198735410.013.8 https://doi.org/10.3389/fnsys.2022.768201 https://doi.org/10.3389/fmicb.2015.00264 https://doi.org/10.3389/fmicb.2015.00264 https://doi.org/10.1177/1059712319871360 https://doi.org/10.1177/1059712319871360 https://doi.org/10.1007/s10339-005-0016-8 https://doi.org/10.1007/s10339-005-0016-8 https://doi.org/10.1007/s11229-011-9880-1 https://doi.org/10.1093/mind/xcviii.391.349 https://doi.org/10.1093/mind/xcviii.391.349 https://doi.org/10.1177/1059712319891742 https://doi.org/10.1177/1059712319891742 https://doi.org/10.1037/0033-295x.85.3.207 https://doi.org/10.3389/fpsyg.2018.02393 https://doi.org/10.1037/0033-295x.92.3.289 https://doi.org/10.1145/360018.360022 https://doi.org/10.1007/s11229-015-0812-3 https://doi.org/10.1016/s0925-2312(01)00598-7 https://doi.org/10.1016/s0925-2312(01)00598-7 https://doi.org/10.1038/s41562-019-0626-2 https://doi.org/10.23915/distill.00010 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org marco facchin 26 pessoa, l., medina, l., & desfilis, e. (2021). refocusing neuroscience: moving away from mental categories and towards complex behaviours. philosophical transactions of the royal society b: biological sciences, 377(1844). https://doi.org/ 10.1098/rstb.2020.0534 pfeifer, r., & bongard, j. c. (2008). how the body shapes the way we think. the mit press. piccinini, g. (2020). neurocognitive mechanisms: explaining biological cognition. oxford university press. ramsey, w. (2015). must cognition be representational? synthese, 194(11), 4197–4214. https://doi.org/10.1007/s11229-0140644-6 ramsey, w. (2007). representation reconsidered. harvard university press. rescorla, m. (2013). bayesian perceptual psychology. oxford university press. https://doi.org/10.1093/oxfordhb/978019960 0472.013.010 risko, e. f., & gilbert, s. j. (2016). cognitive offloading. trends in cognitive sciences, 20(9), 676–688. https://doi.org/10.101 6/j.tics.2016.07.002 ritchie, j. b., kaplan, d. m., & klein, c. (2019). decoding the brain: neural representation and the limits of multivariate pattern analysis in cognitive neuroscience. the british journal for the philosophy of science, 70(2), 581–607. https://doi.org/10.1093/bjps/axx023 rosch, e., & mervis, c. b. (1975). family resemblances: studies in the internal structure of categories. cognitive psychology, 7(4), 573–605. https://doi.org/10.1016/0010-0285(75)90024-9 rowlands, m. (2009). extended cognition and the mark of the cognitive. philosophical psychology, 22(1), 1–19. https://doi.org/10.1080/09515080802703620 rowlands, m. (2010). the new sciences of the mind. the mit press. rumelhart, m., david, & pdp group, the. (1986). parallel distributed processing. vol i and ii. the mit press. rupert, r. d. (2013). memory, natural kinds, and cognitive extension; or, martians don’t remember, and cognitive science is not about cognition. review of philosophy and psychology, 4(1), 25–47. https://doi.org/10.1007/s13164-012-0129-9 samuel, a. l. (1967). some studies in machine learning using the game of checkers. ii—recent progress. ibm journal of research and development, 11(6), 601–617. https://doi.org/10.1147/rd.116.0601 scarantino, a. (2012). how to define emotions scientifically. emotion review, 4(4), 358–368. https://doi.org/10.1177/175407 3912445810 schwitzgebel, e. (2021). borderline consciousness, when it’s neither determinately true nor determinately false that experience is present. unpublished manuscript. http://www.faculty.ucr.edu/~eschwitz/schwitzpapers/borderlineconsciousness211220.htm searle, j. r. (1980). minds, brains, and programs. behavioral and brain sciences, 3(3), 417–424. https://doi.org/10.1017/s014 0525x00005756 searle, j. r. (1984). intentionality and its place in nature. dialectica, 38(2-3), 87–99. https://doi.org/10.1111/j.17468361.1984.tb01237.x selfridge, o. g., & neisser, u. (1960). pattern recognition by machine. scientific american, 203(2), 60–69. shapiro, l. a. (2013). dynamics and cognition. minds and machines, 23(3), 353–375. https://doi.org/10.1007/s11023-0129290-2 sims, a. (2017). the problems with prediction. in t. k. metzinger & w. wiese (eds.), philosophy and predictive processing. mind group. https://doi.org/10.15502/9783958573246 sims, m. (2021). a continuum of intentionality: linking the biogenic and anthropogenic approaches to cognition. biology & philosophy, 36(6). https://doi.org/10.1007/s10539-021-09827-w smith, r., ramstead, m. j. d., & kiefer, a. (2022). active inference models do not contradict folk psychology. synthese, 200(2). https://doi.org/10.1007/s11229-022-03480-w staddon, j. (2021). the new behaviorism. taylor & francis. steiner, p. (2019). radical views on cognition and the dynamics of scientific change. synthese, 198(s1), 547–569. https://doi.org/10.1007/s11229-019-02268-9 tani, j. (2007). on the interactions between top-down anticipation and bottom-up regression. frontiers in neurorobotics, 1. https://doi.org/10.3389/neuro.12.002.2007 tani, j. (2016). exploring robotic minds. oxford university press. taylor, h., & vickers, p. (2016). conceptual fragmentation and the rise of eliminativism. european journal for philosophy of science, 7(1), 17–40. https://doi.org/10.1007/s13194-016-0136-2 thelen, e., & smith, l. (1994). a dynamic system approach to the development of cognition and action. the mit press. thomson, e., & piccinini, g. (2018). neural representations observed. minds and machines, 28(1), 191–235. https://doi.org/ 10.1007/s11023-018-9459-4 facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1098/rstb.2020.0534 https://doi.org/10.1098/rstb.2020.0534 https://doi.org/10.1007/s11229-014-0644-6 https://doi.org/10.1007/s11229-014-0644-6 https://doi.org/10.1093/oxfordhb/9780199600472.013.010 https://doi.org/10.1093/oxfordhb/9780199600472.013.010 https://doi.org/10.1016/j.tics.2016.07.002 https://doi.org/10.1016/j.tics.2016.07.002 https://doi.org/10.1093/bjps/axx023 https://doi.org/10.1016/0010-0285(75)90024-9 https://doi.org/10.1080/09515080802703620 https://doi.org/10.1007/s13164-012-0129-9 https://doi.org/10.1147/rd.116.0601 https://doi.org/10.1177/1754073912445810 https://doi.org/10.1177/1754073912445810 http://www.faculty.ucr.edu/~eschwitz/schwitzpapers/borderlineconsciousness-211220.htm http://www.faculty.ucr.edu/~eschwitz/schwitzpapers/borderlineconsciousness-211220.htm https://doi.org/10.1017/s0140525x00005756 https://doi.org/10.1017/s0140525x00005756 https://doi.org/10.1111/j.1746-8361.1984.tb01237.x https://doi.org/10.1111/j.1746-8361.1984.tb01237.x https://doi.org/10.1007/s11023-012-9290-2 https://doi.org/10.1007/s11023-012-9290-2 https://doi.org/10.15502/9783958573246 https://doi.org/10.1007/s10539-021-09827-w https://doi.org/10.1007/s11229-022-03480-w https://doi.org/10.1007/s11229-019-02268-9 https://doi.org/10.3389/neuro.12.002.2007 https://doi.org/10.1007/s13194-016-0136-2 https://doi.org/10.1007/s11023-018-9459-4 https://doi.org/10.1007/s11023-018-9459-4 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why can’t we say what cognition is (at least for the time being) 27 tripaldi, l. (2022). parallel minds. the mit press. trovato, g., & eyssel, f. (2017, august). mind attribution to androids: a comparative study with italian and japanese adolescents. 2017 26th ieee international symposium on robot and human interactive communication (ro-man). https://doi.org/10.1109/roman.2017.8172358 van gelder, t. (1995). what might cognition be, if not computation? journal of philosophy, 92(7), 345–381. https://doi.org/ 10.2307/2941061 villalobos, m., & dewhurst, j. (2018). enactive autonomy in computational systems. synthese, 195(5), 1891–1908. https://doi.org/10.1007/s11229-017-1386-z villalobos, m., & dewhurst, j. (2017). why post-cognitivism does not (necessarily) entail anti-computationalism. adaptive behavior, 25(3), 117–128. https://doi.org/10.1177/1059712317710496 von neumann, j. (1958). the computer and the brain. yale university press. waisman, f. (1968). how i see philosophy. palgrave mcmillan. walter, s. (2010). cognitive extension: the parity argument, functionalism, and the mark of the cognitive. synthese, 177(2), 285–300. https://doi.org/10.1007/s11229-010-9844-x walter, s., & kästner, l. (2012). the where and what of cognition: the untenability of cognitive agnosticism and the limits of the motley crew argument. cognitive systems research, 13(1), 12–23. https://doi.org/10.1016/j.cogsys.2010.10.001 webb, b. (2006). transformation, encoding and representation. current biology, 16(6), r184–r185. https://doi.org/10.1016/ j.cub.2006.02.034 weinberger, n., & allen, c. (2022). static-dynamic hybridity in dynamical models of cognition. philosophy of science, 89(2), 283–301. https://doi.org/10.1017/psa.2021.27 wheeler, m. (2010). in search of clarity about parity. philosophical studies, 152(3), 417–425. https://doi.org/10.1007/s11098010-9601-5 wheeler, m. (2019). breaking the waves: beyond parity and complementarity in the arguments for extended cognition. in m. colombo, e. irvine, & m. stapleton (eds.), andy clark and his critics (pp. 81–95). oxford university press. williamson, t. (1994). vagueness. routledge. yakura, h. (2018). a hypothesis: crispr-cas as a minimal cognitive system. adaptive behavior, 27(3), 167–173. https://doi.org/10.1177/1059712318821102 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. facchin, m. (2023). why can’t we say what cognition is (at least for the time being). philosophy and the mind sciences, 4, 7. https://doi.org/10.33735/phimisci.2023.9664 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1109/roman.2017.8172358 https://doi.org/10.2307/2941061 https://doi.org/10.2307/2941061 https://doi.org/10.1007/s11229-017-1386-z https://doi.org/10.1177/1059712317710496 https://doi.org/10.1007/s11229-010-9844-x https://doi.org/10.1016/j.cogsys.2010.10.001 https://doi.org/10.1016/j.cub.2006.02.034 https://doi.org/10.1016/j.cub.2006.02.034 https://doi.org/10.1017/psa.2021.27 https://doi.org/10.1007/s11098-010-9601-5 https://doi.org/10.1007/s11098-010-9601-5 https://doi.org/10.1177/1059712318821102 https://doi.org/10.33735/phimisci.2023.9664 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction searching for the mark of the cognitive some desiderata… … and a tension folksy cognition and its moc scientific cognition and its moc where from here? conclusions why ncc research is not theory-neutral why ncc research is not theory-neutral michael pauena,b (m@pauen.com) abstract one of the reasons why the neural correlates of consciousness program could appear attractive in the 1990s was that it seemed to disentangle theoretical and empirical problems. theoretical disagreements could thus be sidestepped in order to focus on empirical research regarding the neural substrate of consciousness. one of the further consequences of this dissociation of empirical and theoretical questions was that fundamental questions regarding the mind body problem or the “hard problem of consciousness” could remain unresolved even if the search for the neural correlates had been successful. drawing on historical examples, a widely held consensus in the philosophy of science, and actual ncc research we argue that there is no such independence. moreover, as the dependence between the theoretical and the empirical level is mutual, empirical progress will go hand in hand with theoretical development. thus, contrary to what the original ncc program suggested, we conclude that ncc research may significantly take advantage from and contribute to theoretical progress in our explanation and understanding of consciousness. eventually, this might even contribute to a solution of the hard problem of consciousness. keywords hard problem of consciousness ∙ history of science ∙ neural correlates of consciousness ∙ pain this article is part of a special issue on “the neural correlates of consciousness”, edited by sascha benjamin fink and ying-tung lin. 1 introduction maybe one of the most striking features of the reemerging debate about the problem of consciousness from the 1970s on were growing disagreements particularly among materialists. until the late 1960s almost all materialists agreed that the central issue regarding consciousness was the metaphysical question whether or not mental states are physical states. but from the mid-seventies on, an increasing number of philosophers started to argue for an epistemic understanding of the aberlin school of mind and brain bhumboldt-universität zu berlin, research training group 2386 “extrospection” pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-2136-0772 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 2 problem. according to authors like thomas nagel, frank jackson, joseph levine, or david chalmers the really “hard” problem of consciousness was epistemic: even if mental states were, in fact, physical states, it could still be asked how this fact can be explained and understood. moreover, this hard problem might turn out to be unsolvable. almost at the same time, it became clear that any progress in consciousness research would require an interdisciplinary effort, involving particularly neuroscientists, psychologists, and philosophers. but what could be expected from such a combined effort, if there was no consensus as to what the problem was and whether it was solvable at all? in this situation, an agreement on something like the lowest common denominator could appear extremely attractive, and this is exactly what the proponents of the ncc program had on offer. the basic idea was to set aside the controversies resulting from different theoretical commitments and to focus on “theoretically neutral” data instead, in order allow “researchers of widely different theoretical persuasions” (chalmers, 2000, p. 37) to contribute. empirical facts regarding the neural substrate of conscious experience were needed anyway, no matter what one’s own theoretical commitments were or whether one would believe that these facts were sufficient for a solution of the problem of consciousness. here we will focus on the assumption that the search for the nccs is “theoretically neutral”, such that this search can be pursued independently of the diverging theoretical commitments consciousness researchers might have. our first point will be that theory-neutrality actually is an essential claim of the ncc program. and it is not difficult to see why. theory-neutrality was a response to a specific situation in the science of consciousness where theoretical disagreements seemed to threaten empirical progress. it could thus appear as a reasonable idea to set theoretical disagreements aside for the time being in order to establish the facts. apart from individual remarks particularly in chalmers’s papers on the ncc program, this view also emerges from the idea about the development of consciousness science which is shared by christof koch and david chalmers. according to this idea, consciousness science will develop in three step steps, with ncc research as the first step. while this first step was supposed to be theoretically neutral, the following two steps, by contrast, were described as theoretically loaded: functional explanations that were supposed to come as a second step, and a full-blown theory of phenomenal consciousness as the third one somewhere in the distant future. finally, this interpretation also explains why chalmers and koch talk about the neural correlates of consciousness: our interpretation goes along with the textbook understanding of correlation as a regular co-occurrence of two phenomena that does not imply a causal or a mereological connection between them. all this seems to indicate that the ncc program is clearly defined and well justified, so why is there a need for a discussion, in the first place? the reason why we think that the ncc program with its assumption of theory-neutrality deserves some attention is that it might seem to contradict a widely accepted assumption in the philosophy of science, namely that there is a dependence between empirical pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 3 data and scientific theories. moreover, theory-neutrality is also at variance with the actual scientific practice both in neuroscience and in psychology. thus, our second point is, that this dependence between empirical studies and theoretical commitments holds for ncc research as well, and that it works both ways: while empirical research on consciousness is guided by theoretical presuppositions, new empirical developments may have consequences that lead to a revision of theoretical presuppositions as well. this has an important implication: as empirical findings affect our conceptual, methodological, and theoretical commitments, they may lead to a constant process of revision which makes it highly unlikely that we will be able to chase down the neural correlates without improving our theoretical understanding of consciousness. our claims will get additional support as the history of science shows that the development of identity statements like the famous one on water/h2o did not only require empirical studies. rather, they also involved a competition between opposing theories, in this case between the phlogistonand the oxygen-theory. taken together, all these considerations lead us to conclude that the ncc program which was thought to sidestep problems of explaining and understanding consciousness can make a significant contribution to a solution of these problems, which might even have repercussions on the hard problem of consciousness. we will start with a short sketch of the original ncc program, including a discussion of what theory-neutrality means in part 2. part 3 will provide some historical background regarding the identification of water and h2o which has often been taken as the gold standard for the mind/brain identity claim. the case will illustrate how theoretical presuppositions guide the search for evidence that eventually helps to establish an identity claim. in part 4 we will take the debate between cognitive and non-cognitive theories of consciousness as an example to show that ncc research requires theoretical presuppositions for the design of experiments, the selection of research questions and working hypotheses, and for the interpretation of the data. finally, part 5 will demonstrate that the relation between empirical results and theoretical presuppositions also works the other way round: empirical results can help to decide conflicts between opposing theoretical presuppositions, but they can also help to generate new presuppositions – in this case regarding the phenomenology and function of mental states like pain. we conclude that this mutual interaction between theoretical presuppositions and empirical data may even result in a significant contribution of ncc research to a better understanding of consciousness. 2 neural correlates of consciousness and theoryneutrality according to chalmers (1998, 2000) and koch (1990, 1997; 2002; 2004), nccs are the minimal neural conditions jointly sufficient for a specific conscious experience pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 4 (koch et al., 2016), e.g., for an experience of pain. the most promising candidates for nccs themselves were originally thought to be activity states in specific brain regions, e.g. in the anterior cingulate cortex and the somatosensory cortex in the case of pain (rainville et al., 1997), or activity in the ventral stream of the visual system as a correlate for visual experience (chalmers, 1998; milner & goodale, 2008). this hasn’t changed very much since then with the only exception that – with the advent of multivariate pattern analyses (haxby, 2012; haynes, 2009) – specific distribution patterns of neural activity rather than overall activity states have come into focus. also, certain types of global brain activity, e.g. 40 hertz oscillations (crick & koch, 1997) or so-called “nmda synapse activity” (flohr, 1992) were suggested as ncc candidates in the 1990s, but play a minor role today. 2.1 theory-neutrality as already indicated, theoretical disagreements seemed to severely hamper empirical research and interdisciplinary cooperation probably from the 1980s on. in this situation, the answer of the ncc program was simple and convincing: if we want to make any progress in our understanding of the physical substrate of consciousness, we need neuroscientific data anyway. so, let’s focus on the neural correlates of consciousness and set the methodological and theoretical disagreements aside for the time being. regarding the development of consciousness science, koch and crick envisage a three-step process with the search for the nccs as the first step. it is followed by the description of the causal link between neural activity and conscious experience as a second step, and the explanation of phenomenal experience, that is, the solution of the hard problem as the third and final step (koch & crick, 2002). a similar plan has been brought up by david chalmers. like koch and crick, chalmers envisages a three-step development in the science of consciousness, where the ncc program is the first step. it is “a project that is relatively tractable, clearly defined, and theoretically neutral, one whose goal seems to be visible somewhere in the middle distance” (chalmers, 2000, p. 38, emphasis added). the second step is a functional explanation of the mechanisms underlying consciousness, while the final third step may be a solution of the hard problem (chalmers, 1998). in order to understand the difference between the steps and the extension of the entire program it is worth noting that chalmers expects that “a century or two” will pass before the functional explanations of the second step replace the ncc program. in order to better understand what, exactly, chalmers has in mind when he talks about theory-neutrality, two questions matter: first, what does it mean for a line of research to be theory-neutral? second, what are the theories or theoretical commitments, with respect to which ncc research is supposed to be neutral? pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 5 regarding the first question, theory-neutrality of a certain line of research can mean either that the relevant theoretical commitments are among the presuppositions or that they are among the consequences of that research. as the consequences of a given investigation are difficult to determine and largely beyond the author’s control, it wouldn’t make much sense to make neutrality regarding possible theoretical consequences an essential part of the ncc program. another reason against this move is that it would not contribute to the obvious goal of the ncc program: to allow researchers with a broad spectrum of diverging theoretical commitments to contribute to consciousness research. unlike the theoretical presuppositions of a study, the consequences are independent of the researcher’s commitments. that is why we assume that what matters primarily are the theoretical presuppositions of ncc research, even if it will turn out below in part 5 that there is a mutual relationship between consequences and presuppositions. in fact, it sounds quite sensible to understand theory-neutrality in this way if one wants to establish a program that is open for ‘researchers of widely different theoretical persuasions’: obviously, my theoretical persuasions matter for the presuppositions of my studies, that is, for the methodological commitments i accept, for the research questions i regard as worth investigating, and for my interpretation of the data. by contrast, my persuasions do not (or should not) matter for the data themselves and they certainly do not matter for the consequences that these results may have in the future. 2.2 theoretical presuppositions let’s look at the famous ether wind experiments by michelson and morley (1887) to illustrate this view of theory-neutrality or theory-dependence, respectively. according to our view, michelson and morley’s theoretical commitments regarding the ether wind theory should become apparent in the theoretical presuppositions of their experiments. and this is indeed the case: the ether wind theory shapes the research question (does the relative movement of the ether wind affect the speed of light?), the hypothesis (the ether wind should affect the speed of light) but not the result, which was negative and eventually led to the dismissal of the ether wind theory. the relativity theory, by contrast, is not among michelson and morley’s theoretical commitments, even if it was the most important consequence of their experiments. also note that, on this understanding, theory-ladenness does not mean endorsement: it just means that an experiment cannot be understood without the theory in question, even if the experiment provides evidence against the theory – as michelson and morley’s study did with respect to the ether wind hypothesis. in many cases, theoretical presuppositions concern essential features of a given phenomenon. examples are the assumption that water is an element, that electromagnetic waves require luminiferous ether for transmission, that heat is a subpauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 6 stance (e.g., caloric), or that consciousness is a cognitive phenomenon. while some of the traditional presuppositions (e.g., that water is one of the four elements) may be grounded in speculation or religious belief, scientifically relevant presuppositions today most typically are generalizations grounded in scientific observation. this holds for the assumption that heat is a substance, that electromagnetic waves require ether as a medium, or that consciousness is a cognitive phenomenon, as we will see below. as a consequence, theoretical presuppositions can be substantially modified or even dismissed upon further research – like the assumption that water is an element, that light requires ether, or that heat is a substance. note that our claim is not only that theoretical presuppositions actually play an important role in current neuroscientific research. rather, we argue that they need to play this role because they help to make reasonable decisions based on the best available knowledge, e.g., regarding experimental design, the selection of hypotheses, and the interpretation of data. our second question concerns the theories with respect to which ncc research is supposed to be neutral. let’s first note that empirical research often requires the acceptance of very basic theoretical assumptions. for example, it doesn’t make much sense for consciousness science to investigate the brain, if one does not accept the assumption that, in some way or other, conscious experience is based on brain activity. likewise, you probably accept the bold-effect as a proxy for neural activity, including the underlying theories (logothetis et al., 2001), if you run fmri experiments. we assume that koch and chalmers accept the compatibility of this very basic level of theory-ladenness with their idea of theory-neutrality. this might be one of the reasons why chalmers sometimes relativizes his neutrality claim. as far as the remaining theoretical commitments are concerned, we suggest a threefold distinction between (1) a metaphysical, (2) an epistemological and (3) a neuroscientific level of theory-ladenness. (1) theories on the metaphysical level deal with the identity or non-identity of the mental and the physical. understood in this way, theory-neutrality would sidestep the metaphysical question whether mental states are physical states. while neither chalmers nor crick and koch are very explicit regarding this issue, the definition of nccs they embark on does not seem completely neutral in this respect, though. their definition is incompatible particularly with current forms of interactive dualism as proposed e.g., by eccles or libet. both authors assume that consciousness requires some sort of non-physical activity, e.g., by non-physical “psychons” (eccles, 1994) or “conscious mental fields” (libet, 1994; libet, 1996). this means that, contrary to one of the basic assumptions of the ncc program, neural activities cannot be “minimally sufficient” for conscious experience. however, this incompatibility could be easily removed by minor adjustments of the ncc definition. these adjustments can make room for non-neural conditions, pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 7 and there are reasons to do so because this move would allow researchers with different metaphysical beliefs to participate in ncc studies. in fact, both eccles and libet were neuroscientists who conceded that – their dualist commitments notwithstanding – brain activities are essential for the emergence of consciousness; chalmers (1998, 2000) explicitly refers to libet’s ncc research. this is one of the reasons why such an adjustment of the original definition seems suggestive. let’s conclude then, that ncc research is thought to be (or could easily be made) neutral regarding metaphysical questions, even if this is not reflected by the “official” definition. (2) theories on the epistemological level deal with explanations of phenomenal experience. understood in this way, the most important implication of theoryneutrality would be that ncc research sidesteps issues like the hard problem of consciousness. apparently, this is a likely candidate for the understanding of theoretical neutrality. koch and crick’s (2002) three-step model, that has already been mentioned above, explicitly distinguishes between the search for the nccs as the first step, the investigation of the causal link between neural activities and subjective feelings as a second step and, finally, the hard problem as something that future research may take care of in a third step. a similar view can be found in chalmers who explicitly distinguishes correlation from explanation and stresses that “it certainly does not follow that an ncc will yield an explanation of consciousness” (chalmers, 2000). as already mentioned above, chalmers, like crick and koch, envisages a three-step development in consciousness research. this implies a clear distinction between the ncc program as a first step, the development of functional explanations as a second step, and the solution of the hard problem as a third step. this step is taken as a remote possibility that might come up even later after the completion of the search for functional explanations. “but who knows: somewhere along the line we may be led to the relevant insights that show why the link is there, and the hard problem may then be solved” (chalmers, 1998, p. 25). in a similar vein, chalmers (2000) distinguishes between “theoretically neutral” ncc research and an explanation of consciousness which is understood as a further step: “once we have found an ncc, one might hope that […] it will turn out to yield an explanation of consciousness, but these are further questions” (chalmers, 2000, p. 37). all this shows that both chalmers and koch assume that ncc research is neutral with respect to a full explanation of consciousness, that is a solution of the hard problem which is regarded as an independent project that may be taken care of pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 8 at some point in the future. this means that the search for the nccs can be pursued no matter whether or not they contribute anything to our understanding of consciousness and, conversely, no matter what your theoretical commitments regarding the explanation of consciousness might be. that is, these commitments will not be among the theoretical presuppositions guiding ncc research. (3) but what about the third candidate? does theoretical neutrality extend to the level of neuroscientific theories of consciousness like the global workspace theory, according to koch and chalmers? this is indeed the case. both koch and chalmers assume that neuroscientific theories of consciousness go beyond the purely correlational research that the ncc program focuses on. if we have another look at koch’s three-step model with ncc research as the first, an investigation of the causal link between consciousness and neural activity as a second, and a solution of the hard problem as a third step, it seems obvious that neuroscientific theories of consciousness which actually try to establish causal links between specific neural activities and conscious experience belong to this second step. likewise, chalmers distinguishes between, first, the identification of the nccs, second, “mechanistic explanations” of the functional properties of consciousness and, third, a “full explanation of consciousness” that solves the hard problem. in chalmers’s case, the assignment of neuroscientific theories to the second step is even more obvious. chalmers’s paradigm for a relevant explanandum is global availability, taken from baars’s global workspace theory. one of today’s most important neuroscientific theories of consciousness, dehaene’s neuronal workspace theory which has been derived from baars’s global workspace, does exactly this: it “explains global availability in the brain” (chalmers, 1998). let’s conclude then that the theory-neutrality claim is intended to distinguish research collecting the neural correlates of consciousness (step 1) from theories trying to explain consciousness either on a functional/neuroscientific (step 2) or on a phenomenal level (step 3). this interpretation complies with the familiar textbook distinction between correlates and causal explanations, thus helping us to understand why koch and chalmers chose the term “correlates” in the first place – even if there are numerous fringe cases between mere correlates and full-blown explanations. these fringe cases provide another reason why chalmers slightly1 relativizes the neutrality claim (chalmers, 2000, p. 37) – even if he insists on a “large extent” of neutrality. in fact, his reference to the “standard usage” of the term “correlate” but even more so the explicit distinction between theoretically neutral correlational research in the ncc program and “theoretically loaded” (chalmers, 2000, p. 38) explanations in the later steps makes it clear that theory-neutrality is what sets the ncc program apart from the following stages of consciousness research. 1the qualification is not always included, though: see chalmers (2000, p. 38). pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 9 also note that there is a tradeoff here: of course, a weak interpretation of the neutrality claim, e.g. one that includes metaphysical neutrality only, could make the ncc program less susceptible for objections. the obvious downside, however, is that it would make the ncc program futile because we might end up with what neuroscientists do anyway. conversely, a strong interpretation would help the ncc program to make a real difference but could make it more susceptible for objections. we will come back to this question below. 3 the case of water and h2o so, let’s keep in mind that the original ncc program tried to sidestep theoretical commitments particularly regarding the function and explanation of consciousness. it did so in order to make studies on the neural correlates of consciousness open to researchers with widely diverging theoretical commitments. plausible as this program sounds, it stands in stark contrast to a widespread consensus in the philosophy of science according to which there are strong and mutual connections between theoretical commitments and empirical research. as is fairly well known, epistemologists and philosophers of science like duhem (1991), kuhn (1962), sellars (1956), and quine (1981) have long since argued that empirical work in general depends on theoretical commitments. but what, exactly, does this general epistemological principle of theory dependence mean when the identity of an everyday phenomenon like pain and a scientific kind, e.g., a specific type of neural activity, is at issue? and is it really true that theory dependence as postulated by duhem, kuhn, sellars, and quine is incompatible with the theory-neutrality claim of the ncc program? in order to answer these questions, we will discuss an example that has played a pivotal role in the consciousness debate: the identification of water and h2o which has been taken as a paradigm for an identity claim that is well-understood. that is why knowledge about the behavior of h2o molecules below 0° c can serve to explain the freezing of water. but how did scientists discover this relationship, and what can we learn from this discovery for the science of consciousness and, more precisely, for the questions above? here want to show, first, that the process that led to the identification of water and h2o was not only driven by current experimental results, but also by competing theoretical presuppositions. second, we want to demonstrate that the relationship was mutual. while theoretical presuppositions were instrumental for guiding empirical research, they could also be modified and even dismissed based on empirical evidence, at least in the long run. our first point is that the water/h2o identity claim was not merely a product of empirical evidence. rather, as chang (2012) has shown, the identity claim came out as the result of the competition between two theories, the phlogistonand the oxygen theory which guided the design of experiments and the interpretation of pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 10 the results. moreover, the competition was not decided by empirical evidence only but largely by pragmatic reasons. as the competing theories led to opposing theoretical presuppositions, the case at hand can also help to illustrate what these presuppositions are and how they work. the presupposition of the phlogiston theory defended by priestley and ritter was the traditional idea that water is an element that does not allow for further decomposition. by contrast, lavoisier and the proponents of the oxygen theory argued that water is a compound that can be decomposed into oxygen and hydrogen. as we will see, these presuppositions did not only guide the design of the experiments, and the research questions, they also affected how the results were interpreted, and, eventually, they were decisive for the outcome of the debate (chang, 2012). for example, the famous electrolysis experiment by nicholson and carlisle was obviously motivated by the claim that water is a compound of oxygen and hydrogen. immediately before conducting this experiment, nicholson and carlisle hat tried to construct a battery following volta’s description. in order to improve the contact between the battery and the cables attached to it, they put a few drops of water on the connection. but due to a design failure, there was no direct contact such that cable and battery worked as anode and cathode and thus decomposed water into hydrogen and oxygen, which was realized by the experimenters. and this is where the theoretical presupposition comes in: since nicholson and carlisle knew that this point was critical for the debate between the oxygen and the phlogiston theory, they decided to do another experiment in order to test the decomposition of water into hydrogen and oxygen (chang, 2012). the success of this test seemed to provide support for the oxygen theory and its presupposition that water was a compound of oxygen and hydrogen. but rather than giving in, the proponents of the phlogiston theory came up with an interpretation based on their presupposition, namely that water is an element. eventually though, the oxygen theory prevailed, but, as chang shows, the reason was not support by converging evidence. the proponents of the oxygen theory were even unable for quite some time to overcome a fundamental objection that had been raised by phlogiston theorists like ritter, the so-called distance problem, which was not solved before the next turn of the century: how can oxygen and hydrogen move from one and the same molecule to two completely different locations, namely the anode and the cathode (chang, 2012, p. 73 sqq.)? but if the oxygen theory and the water/h2o identity claim were accepted already four decades earlier, this was mainly for pragmatic reasons: the oxygen theory was a better fit for lavoisier’s analytic chemistry since it was based on the same theoretical presupposition, namely that even basic substances could turn out to be compounds. by contrast, the competing presupposition that water was an element was eventually dismissed. this reveals another important aspect of the relation between empirical studies and theoretical presuppositions: even if presuppositions pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 11 guide the design and interpretation of individual experiments, they may be affected by the outcome of these experiments themselves, at least in the long run. this can happen, if there is a substantial amount of converging evidence that can be easily interpreted by (or strongly supports) one specific presupposition but not by a competing one. we will come back to this point below in part 5. 4 the dependence of theories and data in consciousness research so let us conclude, first, that the identification of water and h2o did not emerge directly from the evidence available, rather it came as a result of the competition between two opposing theories which provided different theoretical presuppositions for empirical research. and second, the success of the oxygen theory and the defeat of the phlogiston theory had a strong effect on the respective theoretical presuppositions as well. this shows that the relation between theoretical presuppositions and empirical evidence is mutual. first, the search for evidence is guided by theoretical presuppositions. second, the evidence, in turn, may affect our theoretical presuppositions. thus, empirical research does not simply produce data that help to decide a pre-defined question about a given phenomenon, rather it results in changing our understanding of the phenomenon itself by way of changing or dismissing even basic assumptions about it, that is, theoretical presuppositions. if any of these observations holds for ncc research as well, then it is highly unlikely that this research can be theory-neutral. but do these observations really hold for consciousness research? in what follows, we will argue that they do. contrary to what the original idea of the ncc program suggests, finding the nccs is not, and cannot be, theory-neutral in the sense mentioned above. this means that very basic assumptions about the nature of consciousness which we have called theoretical presuppositions are needed in order to make progress in empirical research – very much like these principles were needed in the process of determining the “chemical correlate of water”. and as it was the case in the water/h2o debate, empirical research and theoretical presuppositions are mutually related in ncc research as well: on the one hand, this research is guided by certain basic presuppositions regarding the nature of consciousness or some particular kind of conscious experience like pain. these presuppositions most typically come from one of the current theories about consciousness or some of its subfunctions (e.g. pain) and they guide experimental design, the generation of hypotheses, and the interpretation of the data. on the other hand, these presuppositions, in turn, are up for confirmation or revision once new empirical findings come in. they may thus lead to a further development of the underlying theories and help to improve our understanding of consciousness. pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 12 4.1 how theoretical presuppositions guide consciousness research: the debate between cognitive and non-cognitive theories of consciousness one of the probably best examples for how theoretical presuppositions affect ncc research is the debate about cognitive and non-cognitive theories of consciousness (overgaard & grünbaum, 2012). the debate has pitted cognitive theories like the global/neural workspace theory by baars (1996, 1997) and, later, dehaene (2006), or higher order theories of consciousness by rosenthal (2011; 1986) against non-cognitive approaches like block’s account of accessand phenomenal consciousness (block, 1995) or lamme’s recurrent loop theory (lamme, 2006). here we will focus on the global workspace theory as a paradigmatic cognitive approach and on block’s account as an example for a non-cognitive approach. we take this debate to illustrate that the dependence between theories and experimental investigation is essential for ncc research as well. for example, we need a basic, theoretically grounded, idea of what “consciousness” means and what its basic functions are if we want to design ncc experiments or draw conclusions from them regarding the ncc. in other words: we need theoretical presuppositions if we want to collect and interpret new data. moreover, it would be negligent to ignore the current state of knowledge and to refrain from trying to use and apply this knowledge as good as possible in the design and interpretation of empirical studies. as both dehaene’s and block’s approaches are fairly well known, we will focus on the most relevant differences between them and will then try to make clear how the underlying understanding of consciousness affects experimental work on the nccs. dehaene’s neuronal workspace theory goes back to bernard baars’s global workspace theory which is based on a fundamental assumption about what consciousness is and what it does: “consciousness seems to be the publicity organ of the brain. it is a facility for accessing, disseminating and exchanging information, and for exercising global coordination and control” (baars, 1997, p. 299). this understanding of consciousness as the “publicity organ of the brain” is a clear theoretical presupposition, which has an important consequence, namely that consciousness is closely connected to reportability. let’s call this the consciousness/reportability presupposition: if i have a conscious pain experience and consciousness is the publicity organ of the brain, then i should be able to report this experience. this presupposition has guided dehaene’s research on the neural correlates of consciousness for many years (dehaene & naccache, 2001). in the present context, this theoretical presupposition has two important implications for empirical studies. pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 13 first and methodologically speaking, if consciousness is closely connected to reportability, then verbal report should be the gold standard for scientific access to consciousness in experimental design. in fact, this is exactly what dehaene claims explicitly: “subjective reports are the key phenomena that a cognitive neuroscience of consciousness purports to study. as such, they constitute primary data that need to be measured and recorded along with other psychophysiological observations” (dehaene & naccache, 2001, p. 3). dehaene, however, has not only postulated that reportability is a key ingredient of consciousness, rather, against block’s conceptual distinction of access and phenomenal consciousness, he argues that there is even a conceptual connection between phenomenal consciousness and reportability: “if the only support for the existence of phenomenal consciousness comes from conscious reports, then we find no reason to accept a major distinction between these two concepts” (naccache & dehaene, 2007, p. 519). second, dehaene’s consciousness/reportability presupposition directly affects the interpretation of the data in ncc experiments. if reportability is a basic ingredient of consciousness, then the neural correlates of the cognitive processes underlying this reportability which are located primarily in the pfc count as neural correlates of consciousness. and this is, in fact, another main claim of the global workspace theory: “neurophysiological, anatomical, and brain-imaging data strongly argue for a major role of prefrontal cortex, anterior cingulate, and the areas that connect to them, in creating the postulated brain-scale workspace” (dehaene & naccache, 2001, p. 1, sq.). block, by contrast, has developed quite a different view about the nature of consciousness which leads him to a rejection of the consciousness/reportability presupposition. block distinguishes between two sorts of consciousness: access and phenomenal consciousness (block, 1995). while block’s concept of access consciousness is well compatible both with the global workspace theory2 (block, 2005) and the consciousness/reportability presupposition, it is his idea of phenomenal consciousness that leads him to reject this presupposition. in block’s view, a person may well have a specific phenomenal experience without being able to report it. phenomenal experience and, therefore, consciousness may “overflow” reportability (block, 2007). 2in fact, block explicitly endorses the global workspace theory, but only as far as access consciousness is concerned: “access conscious content is content information about which is ‘broadcast’ in the global workspace” (block, 2005). pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 14 the rejection of the consciousness/reportability presupposition does not only set block’s account of phenomenal consciousness apart from the global/neuronal workspace theory, it can also serve to illustrate the consequences that this presupposition has both for experimental design in ncc research and for the interpretation of the data regarding possible nccs themselves. first, as for experimental design, block opposes dehaene’s reliance on report-based methods since – in block’s view – these methods are unable, in principle, to detect phenomenal experiences that come in the absence of reportability or access consciousness: “to insist on introspective reportability as the gold standard is to encourage leaving out cases in which subjects have experiences without higher order thoughts” (block, 2005, p. 50). as a consequence, block insists on methods that are able to provide evidence for conscious experience independent of subjective reports. one promising approach is provided by so called “no report paradigms”, which are able to detect conscious experience without asking for verbal report. we will discuss this method and its implications for the consciousness/reportability presupposition below in part 5. second, as far as the interpretation of experimental data regarding possible nccs is concerned, it would follow that neural activities resulting from verbal reports and the underlying cognitive processes do not count as nccs. rather, they count as confounds (michel, 2017). moreover, because block assumes two distinct sorts of consciousness, he has to expect two different types of nccs as well: phenomenal and access nccs. while block, like dehaene, assumes that the nccs of access consciousness can be found in frontal and parietal areas, phenomenal nccs are most likely located in peripheral areas, e.g., in the occipital lobe if visual experiences are concerned (block, 2005) . these observations have further consequences: theoretical presuppositions are required even in ncc research both regarding the methodology and the interpretation of empirical results. neutrality is not an option here: methodologically speaking, we have to decide whether or not we accept self-reports as the goldstandard in consciousness research. and as far as the interpretation of the results are concerned, we must determine whether we take report-based neural activity as part of an ncc or as a confound. both decisions depend on fundamental theoretical presuppositions regarding the nature of consciousness. obviously, these decisions should be made on the basis of the best knowledge available. and this is what theoretical presuppositions help us to do. 4.2 chalmers’s pre-experimental bridging principles it could be argued, though, that chalmers does not deny the existence of theoretical presuppositions. he does not use this term, but it might be argued that what he calls pre-experimental bridging principles is basically the same as what we call theoretical presuppositions. let’s first try to make clear what chalmers has in mind pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 15 when he talks about pre-experimental bridging principles and then discuss their relationship to theoretical presuppositions. chalmers thinks there is an obvious reason why bridging principles are needed in ncc research: we lack direct access to others’ conscious experience. bridging principles step in and allow us to assign conscious experience to experimental subjects on the basis of proxies like verbal reports or behavioral data. there are two such bridging principles: first, the principle of verbal report which holds that a mental state is conscious if it is subject to verbal report. the second principle holds that a mental state is conscious if it is directly available for global behavioral control. let’s call it the principle of behavioral control. chalmers calls these principles “pre-experimental” because he thinks they do not depend on scientific theories: that is why they are compatible with the theory-neutrality claim. rather, they are based on an analysis of the concept of consciousness and on first-person experience, so they neither call for empirical support nor can empirical evidence lead to substantial revision or improvement of these principles (chalmers, 1998). this is an obvious difference between chalmers’s bridging principles and theoretical presuppositions, because the latter, unlike the former, are generalizations from empirically supported theories. this is why bridging principles are compatible with the demand for theory-neutrality while theoretical presuppositions are not. and since theoretical presuppositions are supported by empirical evidence, they can not only be refined – as chalmers seems to concede for the bridging principles3 –, rather, they can be substantially revised or even dismissed entirely if a sufficient amount of new evidence comes in that is incompatible with a given presupposition. we have shown this above with respect to the debate between the phlogiston and the oxygen theory and we will provide further evidence from ncc research below in part 5. there, we demonstrate in some detail how empirical results may affect theoretical presuppositions like the consciousness/reportability presupposition. finally, chalmers’s pre-experimental principles are dedicated to one very specific issue: the interpretation of verbal reports and behavior with respect to consciousness. but, again, even if the content of the consciousness/reportability presupposition is somewhat similar to chalmers’s principle of verbal report: it is just one of an infinitely large number of potential theoretical presuppositions which may cover all kinds of subjects, e.g. the regions of interest underlying the experience of pain, the role of the amygdala in negative emotions, or the role of working memory in conscious experience: almost any knowledge that may appear helpful in the search for the neural correlates of consciousness can be used as a theoretical presupposition. 3chalmers seems to concede a limited amount of refinement, particularly through first-person experiments, but his position here is unclear. initially he seems to exclude refinements through normal third person experiments altogether, claiming that we “can’t refine them [i.e. the preexperimental principles] experimentally (except perhaps through first-person experimentation)” (chalmers, 1998, p. 3). later he concedes that these “principles may themselves be refined as experimental research goes along” (chalmers, 1998, p. 6). pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 16 in sum, we think that, for all these reasons, chalmers’s understanding of preexperimental bridging principles is significantly different from what we have described as theoretical presuppositions and thus provides another piece of evidence showing that chalmers does not even implicitly accept theoretical presuppositions. let us add as an aside that there are reasons to believe that chalmers’s two bridging principles are not pre-experimental. in other words: experimental evidence does speak to the principle of verbal report and to the principle of behavioral control. first, there is some evidence which might be seen to challenge the principle of verbal report. nisbett and wilson (1977) or gazzaniga and ledoux (1978) have shown already in the 1970s that, under certain conditions, subjects tend to report mental states that they did not experience consciously, e.g., when they would otherwise lack sufficient explanation for their behavior. an even more striking example comes from patients suffering from anton’s syndrome (anton, 1899; maddula et al., 2009; kini & khatker, 2019), a rare case of visual anosognosia. anton’s patients are actually blind, but they claim they have visual perceptions even if there is strong evidence against these claims. thus, contrary to the principle of verbal report, neurologists take these claims as confabulations (carvajal et al., 2012; das & naqvi, 2021). moreover, contrary to the principle of behavioral control, priming experiments show that our behavior can well be affected by pieces of information that escape conscious experience (kristjánsson & campana, 2010). note again that we do not argue for a rejection of chalmers’s bridging principles. rather, we want to show that – contrary to chalmers – empirical evidence does speak to these principles, e.g., because it can help us to specificy the conditions under which these principles hold, but it may also force us to dismiss these principles. all this gives us strong reasons to reject chalmers’s claim that the two principles he mentions are pre-experimental. and as the pre-experimental status of the bridging principles was essential for their compatibility with the theoryneutrality claim of the ncc program, chalmers seems to be forced to make a choice: either, he has to give up the bridging principles in order to save the neutrality claim – this however would make it difficult for ncc research to assign consciousness on the basis of proxies. or he has to give up the theory-neutrality claim, which would jeopardize an essential feature of the ncc program. since the above considerations are decisive for our claim that ncc research cannot be theory-neutral, a brief intermediate summary might be in place. so, let’s ask flat out why can’t chalmers and koch simply accept the observations made above? why can’t they insist that theory-neutrality just means that ncc research sets metaphysical questions and maybe questions related to the hard problem aside, while allowing for theoretical presuppositions resulting from neuroscientific research? there are mainly two reasons why this is not an option for koch and, particularly, for chalmers. first, we have shown that ncc research, as any other empirical investigation, needs theoretical presuppositions in order to make unavoidpauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 17 able decisions about research questions, experimental design, and the interpretation of the data. but if you accept theoretical presuppositions like the consciousness/verbal report presupposition, then you make a commitment regarding the underlying theory, e.g. the global workspace theory that justifies the presupposition. we think that already this commitment, taken by itself, is a clear violation of theory-neutrality. moreover, theories of consciousness like the global workspace theory are explicitly assigned both by chalmers and by koch to the second, “theoretically loaded” stage of consciousness research. that is, these approaches do not count as theory-neutral in chalmers’s and koch’s own view because research in this second phase is supposed to provide functional or mechanistic explanations of consciousness, rather than mere correlations – which is exactly what the global workspace theory does. finally, we have seen that chalmers’s plea for pre-experimental bridging principles cannot be read as an implicit acceptance of theoretical presuppositions. quite the contrary: chalmers goes to a lot of trouble to show that his bridging principles do not imply any theoretical commitments – as our theoretical presuppositions do. bridging principles are thought to follow from conceptual considerations and first-person experience, but not from experimentally based theories and thus are supposed to be compatible with the claim for theory-neutrality. but as we have seen, there are strong reasons to doubt that this is actually the case for the two principles that chalmers identifies. there is still a second, independent reason, why accepting theoretical presuppositions or similar kinds of theoretical commitments is not an option for koch and chalmers. both authors seem to assume that the ncc program really makes a difference compared to the practice in contemporary consciousness research. otherwise, the ncc program would just preach to the converted because it would “demand” what neuroscientists are doing anyway. that is, contrary to what koch and chalmers claim, the ncc program would not help to attract new scientists to the field of consciousness studies, irrespective of their theoretical commitments. for all these reasons, a weak reading of the theory-neutrality claim that would be compatible with the above considerations is not available for koch and chalmers. 5 how empirical results may affect theoretical presuppositions so far, we have argued that – against the claim of theory-neutrality – ncc research like empirical science in general is informed by theoretical presuppositions about the nature of consciousness and it needs to be so because these assumptions guide inevitable choices researchers have to make regarding experimental design, research questions, and the interpretation of their data. in this last part of our paper, we want to show that the relationship between empirical research and theoretical presuppositions is mutual: empirical research, that is, does affect theoretical presuppositions. moreover, we will demonstrate, pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 18 that this process – if successful – can help to improve our understanding of the phenomena under investigation. in order to show this, we will present two examples that illustrate two slightly different ways of how empirical evidence may affect theoretical presuppositions. in the first section, we present the discussion on no-report paradigms in order to demonstrate how empirical studies may help to resolve conflicts between opposing theoretical presuppositions. this is not only a technical or methodological point. rather, given that theoretical presuppositions concern basic features of a phenomenon, dismissing a presupposition that turns out to be unfounded may help to improve our understanding of that phenomenon. our second example will be pain research and pain measurement. we will show how empirical findings can help to generate new theoretical presuppositions, which can then be used in subsequent studies. and again, this may help to improve our understanding of the phenomenon as well. while we don’t want to suggest that chalmers’s understanding of theoretical neutrality is incompatible in general with the possibility that ncc studies have theoretical consequences, we think that a discussion of the effects empirical studies have on theoretical presuppositions is essential for a full understanding of theoretical presuppositions. this holds in particular for our understanding how these presuppositions emerge and how we resolve conflicts between them – even if this requires that we take a look also at the consequences of ncc research. 5.1 no-report paradigms the general assumption underlying chalmers’s two pre-experimental bridging principles is the idea that global availability can be taken as a proxy for consciousness: “when information is directly available for global control in a cognitive system, then it is conscious” (chalmers, 1998). but this assumption is not a priori at all but, rather, it is one of the basic tenets of the global workspace theory of consciousness – as we have already seen above. in fact, chalmers (1998) explicitly refers to the global workspace theory whose proponents insist that their theory “is based entirely on well-established empirical contrasts between pairs of conscious and unconscious events” (baars, 1997; michel & morales, 2020). in doing so they confirm our point that experimental data do speak to those allegedly “pre-experimental” principles. more recently, the principle of global availability has become the subject of a controversial debate about so-called “no-report paradigms” in consciousness studies which is directly related to the debate between cognitive and non-cognitive theories of consciousness mentioned above (block, 2019; frässle et al., 2014; michel, 2017; michel & morales, 2020; naber et al., 2011; tsuchiya et al., 2015). as their name indicates, no-report paradigms try to provide evidence for conscious experience without relying on first-person reports. this is why no-report paradigms speak to the global workspace theory’s basic tenet, according to which global availability is essential for consciousness, a claim pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 19 that has often been used as a theoretical presupposition: if the global workspace theory were right, we would expect that conscious experience cannot be dissociated from global availability. thus, we should find activity in the pfc where the “publicity organ” of the brain is supposed to be located, whenever a participant has a conscious experience, even if they are not requested to report. by contrast, if noncognitive theories were right, we would expect no-report paradigms to provide evidence for conscious experience without reportability, and evidence for activity in the peripheral areas in the absence of pfc activity and global availability. this would support block’s claim that phenomenal and access consciousness should be distinguished. previous evidence for pfc activity from report-based paradigms could then be attributed to a “methodological artefact” (michel, 2017) due to confounds in report paradigms. note that this discussion provides another piece of evidence that chalmers’s principles, no matter whether or not they turn out to be valid, may not be pre-experimental. one no-report paradigm has emerged from binocular rivalry-experiments (frässle et al., 2014; naber et al., 2011). in a typical design, two different optical stimuli moving into opposite directions are projected to the left and the right eye, e.g., a green grid moving to the right is projected to the right eye and a red grid moving to the left to the left eye. the visual system will then switch between the two stimuli every now and then such that, at any given point in time, either the green grid moving to the right or the red grid moving to the left will be dominant in the participant’s experience. normally, participants report which stimulus is dominant and when the switch occurs. frässle et al. (2014) however used the optokinetic nystagmus instead to determine the dominant stimulus. the optokinetic nystagmus is an involuntary eyemovement that follows moving visual stimuli, but also reveals the dominant stimulus in a binocular rivalry experiment. using this paradigm for a comparison between the neural correlates in a report and a no-report condition, frässle et al. could show that the pfc was active in the report condition but not in the no-report condition, although conscious experience did not differ between these two conditions. as essential functions underlying reports (working memory, attention) are located in the pfc, the authors concluded that pfc activity, contrary to what the global workspace theory claims, is not a correlate of conscious experience itself but, rather, just a correlate of introspective reporting. unsurprisingly though, proponents of the global workspace theory reject this interpretation. they see evidence for pfc related activity even in no-report paradigms like frässle et al.’s study (michel & morales, 2020). other authors have tried to show that even more sophisticated “no-cognition paradigms” (block, 2019) are unable, in principle, to rule out the existence of neural correlates of consciousness in the pfc, as proposed by the global workspace theory (phillips & morales, 2020). again, we will not take sides in this debate, nor will we make any commitment as to whether or not pfc activity is an ncc. rather, our point is that the pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 20 discussion about no-report paradigms provides further evidence, that the dependence between empirical evidence and theoretical commitments is mutual – just as it turned out to be in the water/h2o case. here we see that empirical evidence from no-report paradigms speaks to the debate whether or not global availability is an essential feature of consciousness, that is, whether or not we should accept the consciousness/reportability presupposition mentioned above, or chalmers’s – allegedly pre-experimental – principle of verbal report. that is, theoretical presuppositions, even if they are quite abstract, are subject to empirical confirmation and disconfirmation. as these presuppositions most typically express basic assumptions about the phenomenon of consciousness, this continuous process of testing and revisioning of our basic assumptions concerning consciousness should lead to an improved understanding of the phenomenon, at least in the long run. in the case at hand, we should expect a better understanding of the relationship between conscious experience and reportability. 5.2 pain research our second example for the mutual relationship between ncc research and theoretical presuppositions is taken from recent developments in pain research. while the example above has illustrated how empirical data can affect theoretical presuppositions, we will now go one step further and show how theoretical presuppositions that have emerged from empirical research can guide further studies. more specifically, we use pain research as an example that shows more precisely how empirical research on pain can help to develop new theoretical presuppositions, and how these presuppositions guide new experimental work. pain experience is one of the paradigmatic issues in the debate on consciousness since the nineteen-fifties both in philosophy and neuroscience, and it has been an important research subject ever since, also due to its clinical relevance. recent research seems to indicate that the seemingly homogenous phenomenon we call pain can be distinguished further into two sub-components, namely sensorydiscriminative pain on the one hand, and affective pain on the other (gligorov, 2017; gracely et al., 1982; rainville et al., 1997). both components can be differentiated from the first-person phenomenal perspective as well as from the third-person neuroscientific perspective. while sensory-discriminative pain lets you feel the kind and the place of the tissue damage and involves activity in the somatosensory cortex, affective pain is relevant for the aversive character of pain experience and is realized by networks involving the anterior cingulate cortex. the point at issue here is that this differentiation, apart from allowing for more accurate and valid measurements, gives us a more precise idea of the experience of pain and its neural substrate, thus making good on our claim that ncc research can improve our understanding of the phenomenon of pain. pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 21 5.3 neural correlates of experienced and empathic pain in order to show how this improvement plays out in empirical research, we will turn to a more recent study of the neural correlates of experienced and empathic pain. we will demonstrate how the distinction between affective and sensory pain, that is, one of the results from recent pain research, has then been used as a theoretical presupposition in a subsequent study, illustrating one further step in the mutual relationship between theoretical presuppositions and empirical research. our example will be a study by singer and colleagues (2004) that compares the neural correlates of one’s own pain experience with the correlates of the empathic response to someone else’s pain. the study uses also other theoretical presuppositions, e.g., for methodological reasons. like many other pain-researchers (gligorov, 2017; makin, 2016; woo et al., 2017) the authors rely on the so-called pain matrix (legrain et al., 2011; prichep et al., 2011) which specifies the regions were neural correlates of pain can be expected. the important point, however, are the theoretical presuppositions singer et al. use for the interpretation of their data. in a first step, they could show that anterior cingulate cortex and insula are active both in experienced and empathic pain, while the secondary somatosensory cortex is active in experienced pain only. taken by itself, this finding does not tell us very much. but the authors go on to interpret their correlational data with the help of a theoretical presupposition, namely the functional and phenomenal distinction between sensory-discriminative and affective pain, that is one of the theoretical consequences of the empirical studies mentioned above. capitalizing on this presupposition, the authors can show that it is the affective aspect of pain, which is relevant for empathic pain. one’s own pain experience, by contrast, also includes more detailed sensory-discriminative information about the location, quality and intensity of the tissue damage which is less relevant when we empathize with others. note that this interpretation is not just a contingent addendum. rather the theoretical presuppositions about the function and phenomenology of the different areas underlying pain help to make sense of the data, they improve our understanding of pain in general and empathic pain in particular, and they open up new possibilities for behavioral tests: e.g. if singer’s interpretation is correct, we would expect that, compared to first-person pain experience, the epistemic disadvantage of empathic pain primarily concerns the location and quality of the other’s tissue damage but much less so the affective character of the pain experience. but why couldn’t chalmers insist that the above assumptions about the role of the acc, the insula and the secondary somatosensory cortex are compatible with his neutrality claim? most importantly, these assumptions are clearly functional claims which, according to chalmers – and koch – are reserved for the second stage of consciousness research. this is a significant difference since, as we have already seen, we have to wait “a century or two” until we arrive at this stage according to chalmers. and the fact that these functional assumptions are “only” about pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 22 pain and not about consciousness in general can be easily explained by the focus of singer et al.’s paper which is “only” on pain rather than on consciousness in general. this indicates, again not very surprisingly, that there might a correspondence between the scope of a study and the scope of the theoretical presuppositions the study makes use of. 6 conclusion let’s take stock. we have started with the original idea of the ncc program according to which finding the neural correlates of consciousness is an empirical endeavor that remains neutral regarding some of the main theoretical controversies about consciousness. the program suggests that we can accumulate empirical facts which are useful for all the participants in the debate, independently of their divergent theoretical commitments. most importantly, the ncc program is intended to sidestep controversies regarding neuroscientific explanations of consciousness and the hard problem of consciousness: chalmers and koch assign these two issues to subsequent “theoretically loaded” steps in the development of the neuroscience of consciousness, and these steps are clearly separated temporally and systematically from the “theoretically neutral” search for the neural correlates of consciousness as the first step. plausible as it sounds, the neutrality claim is in stark tension with widely accepted epistemological arguments brought forward by authors like duhem, quine, kuhn, or sellars who insist on a strong dependence between theoretical and conceptual commitments on the one hand, and empirical findings on the other. we tried to show that ncc research does not make an exception here. like any other scientific endeavor, the search for the neural correlates of consciousness requires theoretical presuppositions which affect the methodology, help to figure out promising hypotheses and research questions, and support the interpretation of the results. the dependence is mutual: our theoretical and conceptual commitments are affected by empirical results as well. after demonstrating how the interaction between theoretical presuppositions and empirical research influenced the debate about the identity of water and h2o, we took the current debate between cognitive and non-cognitive theories of consciousness as well as pain research as examples for this mutual dependence. in the former debate, the decisive disagreement concerns the presupposition that firstperson reportability is an essential component of consciousness. while cognitive theories accept this presupposition, non-cognitive approaches deny it. this has essential implications for the design of experiments and the interpretation of their data: cognitive theories can accept first-person reports as an experimental method and the correlates of these reports as nccs, while non-cognitive theories deny this. but the dependence between theoretical presuppositions and empirical evidence is mutual: empirical results can and do affect our theoretical commitments as well. this means, first, that, at some point, converging evidence which is inconsistent pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 23 with a certain presupposition may lead us to revise or even dismiss this presupposition, like the claim that heat is a substance or that water is one of the four elements. taking pain research as our last example, we tried to show how the consequences of one series of experiments can become the presuppositions of subsequent studies. this mutual relationship between theoretical presuppositions and empirical studies which already concerns ncc research may result in a continuous process of theory testing and refinement which eventually improves our understanding of pain – or even of consciousness in general. this is of particular importance because one of the most controversial issues the ncc program tried to sidestep, the hard problem, might be affected by an improved understanding of consciousness as well – even if it is impossible to predict when and how this will happen. but if empirical research enhances our understanding of the phenomenon of consciousness – as history witnesses and epistemological considerations show – ncc research may contribute to a solution of the hard problem. if something like this is true, then the ncc program may help to demonstrate that consciousness is not a mystery. rather, it is an ordinary – though extremely complex – problem of scientific study. and ncc research may, somewhat ironically, make a significant contribution to solving a problem that it was intended to sidestep. acknowledgments this research was supported by the german research foundation, grant rtg 2386, „extrospection” and grant exc 2002/1 ”science of intelligence”. i would like to thank elisa filevich, john-dylan haynes, holger lyre, dimitri coelho mollo, sera schwarz, ali yasar, the members of the rtg extrospection, and three anonymous reviewers for helpful suggestions and important comments. references anton, g. (1899). ueber die selbstwahrnehmung der herderkrankungen des gehirns durch den kranken bei rindenblindheit und bindentaubheit. archiv für psychiatrie und nervenkrankheiten, 32(1), 86–127. https://doi.org/10.1007/bf0212 6945 baars, b. j. (1996). understanding subjectivity: global workspace theory and the resurrection of the observing self. journal of consciousness studies, 3(3), 211–217. baars, b. j. (1997). in the theatre of consciousness: global workspace theory, a rigorous scientific theory of consciousness. journal of consciousness studies, 4(4), 292–309. block, n. (1995). on a confusion about a function of consciousness. behavioral and brain sciences, 18(2), 227–247. https: //doi.org/10.1017/s0140525x00038188 block, n. (2005). two neural correlates of consciousness. trends in cognitive sciences, 9(2), 46–52. https://doi.org/10.1016/ j.tics.2004.12.006 block, n. (2007). consciousness, accessibility, and the mesh between psychology and neuroscience. the behavioral and brain sciences, 30(5-6), 481–548. https://doi.org/10.1017/s0140525x07002786 block, n. (2019). what is wrong with the no-report paradigm and how to fix it. trends in cognitive sciences, 23(12), 1003– 1013. https://doi.org/10.1016/j.tics.2019.10.001 pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/bf02126945 https://doi.org/10.1007/bf02126945 https://doi.org/10.1017/s0140525x00038188 https://doi.org/10.1017/s0140525x00038188 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1017/s0140525x07002786 https://doi.org/10.1016/j.tics.2019.10.001 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael pauen 24 carvajal, j., cárdenas, a., pazmiño, g., & herrera, p. (2012). visual anosognosia (anton-babinski syndrome): report of two cases associated with ischemic cerebrovascular disease. journal of behavioral and brain science, 2(3), 394–398. https://doi.org/10.4236/jbbs.2012.23045 chalmers, d. j. (1998). on the search for the neural correlate of consciousness. in s. r. hameroff, a. w. kaszniak, & a. c. scott (eds.), toward a science of consciousness ii (pp. 2–219). mit press. chalmers, d. j. (2000). what is a neural correlate of consciousness? in t. metzinger (ed.) neural correlates of consciousness (pp. 17–39). mit press. chang, h. (2012). is water h2o? evidence, pluralism and realism. springer. crick, f., & koch, c. (1990). towards a neurobiological theory of consciousness. seminars in the neurosciences, 2, 263–275. das, m. j., & naqvi, i. a. (2021). anton syndrome [updated 2021 april 10]. in statpearls, https://www.ncbi.nlm.nih.gov/b ooks/nbk538155/. dehaene, s., changeux, j.-p., naccache, l., sackur, j., & sergent, c. (2006). conscious, preconscious, and subliminal processing: a testable taxonomy. trends in cognitive sciences, 10(5), 204–211. https://doi.org/10.1016/j.tics.2006.03.007 dehaene, s., & naccache, l. (2001). towards a cognitive neuroscience of consciousness: basic evidence and a workspace framework. cognition, 79(1), 1–37. https://doi.org/10.1016/s0010-0277(00)00123-2 duhem, p. m. m., wiener, p. p., vuillemin, j., & broglie, l. de. (1991). the aim and structure of physical theory. princeton university press. eccles, j. c. (1994). how the self controls its brain. springer. flohr, h. (1992). qualia and brain processes. in a. beckermann, h. flohr, & j. kim (eds.), emergence or reduction?: essays on the prospects of nonreductive physicalism (pp. 220–238). w. de gruyter. frässle, s., sommer, j., jansen, a., naber, m., & einhäuser, w. (2014). binocular rivalry: frontal activity relates to introspection and action but not to perception. journal of neuroscience, 34(5), 1738–1747. https://doi.org/10.1523/jneuro sci.4403-13.2014 gazzaniga, m. s., & ledoux, j. e. (1978). the integrated mind. springer. gligorov, n. (2017). don’t worry, this will only hurt a bit: the role of expectation and attention in pain intensity. the monist, 100(4), 501–513. https://doi.org/10.1093/monist/onx025 gracely, r. h., dubner, r., & mcgrath, p. a. (1982). fentanyl reduces the intensity of painful tooth pulp sensations: controlling for detection of active drugs. anesthesia and analgesia, 61(9), 751–755. haxby, j. v. (2012). multivariate pattern analysis of fmri: the early beginnings. neuroimage, 62(2), 852–855. https://doi.or g/10.1016/j.neuroimage.2012.03.016 haynes, j.-d. (2009). decoding visual consciousness from human brain signals. trends in cognitive sciences, 13(5), 194–202. https://doi.org/10.1016/j.tics.2009.02.004 kini, a., khatker, m., lee, a. g., othman, b. a., & ponce, c. p. (2019). anton syndrome. eye wiki: american academy of ophthalmology. https://eyewiki.aao.org/anton_syndrome koch, c. (2004). the quest for consciousness: a neurobiological approach. roberts & co. koch, c., & crick, f. (2002). the neuronal basis of visual consciousness. https://authors.library.caltech.edu/40600/1/465.pdf koch, c., massimini, m., boly, m., & tononi, g. (2016). neural correlates of consciousness: progress and problems. nature reviews neuroscience, 17(5), 307–321. https://doi.org/10.1038/nrn.2016.22 kristjánsson, á., & campana, g. (2010). where perception meets memory: a review of repetition priming in visual search tasks. attention, perception, & psychophysics, 72(1), 5–18. https://doi.org/10.3758/app.72.1.5 kuhn, t. (1962). the structure of scientific revolutions. university of chicago press. lamme, v. a. f. (2006). towards a true neural stance on consciousness. trends in cognitive sciences, 10(11), 494–501. https://doi.org/10.1016/j.tics.2006.09.001 lau, h., & rosenthal, d. (2011). empirical support for higher-order theories of conscious awareness. trends in cognitive sciences, 15(8), 365–373. https://doi.org/10.1016/j.tics.2011.05.009 legrain, v., iannetti, g. d., plaghki, l., & mouraux, a. (2011). the pain matrix reloaded: a salience detection system for the body. progress in neurobiology, 93(1), 111–124. https://doi.org/10.1016/j.pneurobio.2010.10.005 libet, b. (1994). a testable field theory of mind-brain interaction. journal of consciousness studies, 1, 119–126. libet, b. w. (1996). solutions to the hard problem of consciousness. journal of consciousness studies, 3, 33–35. logothetis, n. k., pauls, j., augath, m., trinath, t., & oeltermann, a. (2001). neurophysiological investigation of the basis of the fmri signal. nature, 412(6843), 150–157. https://doi.org/10.1038/35084005 maddula, m., lutton, s., & keegan, b. (2009). anton’s syndrome due to cerebrovascular disease: a case report. journal of medical case reports, 3, 9028. https://doi.org/10.1186/1752-1947-0003-0000009028 pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.4236/jbbs.2012.23045 https://www.ncbi.nlm.nih.gov/books/nbk538155/. https://www.ncbi.nlm.nih.gov/books/nbk538155/. https://doi.org/10.1016/j.tics.2006.03.007 https://doi.org/10.1016/s0010-0277(00)00123-2 https://doi.org/10.1523/jneurosci.4403-13.2014 https://doi.org/10.1523/jneurosci.4403-13.2014 https://doi.org/10.1093/monist/onx025 https://doi.org/10.1016/j.neuroimage.2012.03.016 https://doi.org/10.1016/j.neuroimage.2012.03.016 https://doi.org/10.1016/j.tics.2009.02.004 https://eyewiki.aao.org/anton_syndrome https://authors.library.caltech.edu/40600/1/465.pdf https://doi.org/10.1038/nrn.2016.22 https://doi.org/10.3758/app.72.1.5 https://doi.org/10.1016/j.tics.2006.09.001 https://doi.org/10.1016/j.tics.2011.05.009 https://doi.org/10.1016/j.pneurobio.2010.10.005 https://doi.org/10.1038/35084005 https://doi.org/10.1186/1752-1947-0003-0000009028 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org why ncc research is not theory-neutral 25 makin, s. (2016). imaging: show me where it hurts. nature, 535(7611), 8–9. https://doi.org/10.1038/535s8a michel, m. (2017). methodological artefacts in consciousness science. journal of consciousness studies, 24(11-12), 94–117. michel, m., & morales, j. (2020). minority reports: consciousness and the prefrontal cortex. mind and language, 35(4), 493–513. https://doi.org/10.1111/mila.12264 michelson, a. a., & morley, e. w. (1887). on the relative motion of the earth and the luminiferous ether. american journal of science, s3-34(203), 333–345. https://doi.org/10.2475/ajs.s3-34.203.333 milner, a. d., & goodale, m. a. (2008). two visual systems re-viewed. neuropsychologia, 46(3), 774–785. https://doi.org/10 .1016/j.neuropsychologia.2007.10.005 naber, m., frässle, s., & einhäuser, w. (2011). perceptual rivalry: reflexes reveal the gradual nature of visual awareness. plos one, 6(6), e20910. https://doi.org/10.1371/journal.pone.0020910 naccache, l., & dehaene, s. (2007). reportability and illusions of phenomenality in the light of the global neuronal workspace model. behavioral and brain sciences, 30(5-6), 518–520. https://doi.org/10.1017/s0140525x07002993 nisbett, r. e., & wilson, t. d. (1977). telling more than we can know: verbal reports on mental processes. psychological review, 84(3), 231–259. overgaard, m., & grünbaum, t. (2012). cognitive and non-cognitive conceptions of consciousness. trends in cognitive sciences, 16(3), 137. https://doi.org/10.1016/j.tics.2011.12.006 phillips, i., & morales, j. (2020). the fundamental problem with no-cognition paradigms. trends in cognitive sciences, 24(3), 165–167. prichep, l. s., john, e. r., howard, b., merkin, h., & hiesiger, e. m. (2011). evaluation of the pain matrix using eeg source localization: a feasibility study. pain medicine (malden, mass.), 12(8), 1241–1248. https://doi.org/10.1111/j.15264637.2011.01191.x quine, w. v. (1981). theories and things. belknap press. rainville, p., duncan, g. h., price, d. d., carrier, b., & bushnell, m. c. (1997). pain affect encoded in human anterior cingulate but not somatosensory cortex. science (new york, n.y.), 277(5328), 968–971. https://doi.org/10.1126/science. 277.5328.968 rosenthal, d. m. (1986). two concepts of consciousness. philosophical studies: an international journal for philosophy in the analytic tradition, 49(3), 329–359. http://www.jstor.org/stable/4319832 sellars, w. (1956). empiricism and the philosophy of mind. minnesota studies in the philosophy of science, 1(19), 253–329. singer, t., seymour, b., o’doherty, j., kaube, h., dolan, r. j., & frith, c. d. (2004). empathy for pain involves the affective but not sensory components of pain. science (new york, n.y.), 303(5661), 1157–1162. https://doi.org/10.1126/science. 1093535 tsuchiya, n., wilke, m., frässle, s., & lamme, v. a. f. (2015). no-report paradigms: extracting the true neural correlates of consciousness. trends in cognitive sciences, 19(12), 757–770. https://doi.org/10.1016/j.tics.2015.10.002 woo, c.-w., schmidt, l., krishnan, a., jepma, m., roy, m., lindquist, m. a., atlas, l. y., & wager, t. d. (2017). quantifying cerebral contributions to pain beyond nociception. nature communications, 8(1), 14211. https://doi.org/10.1038/ncom ms14211 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. pauen, m. (2021). why ncc research is not theory-neutral. philosophy and the mind sciences, 2, 10. https://doi.org/10.33735/phimisci.2021.9188 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1038/535s8a https://doi.org/10.1111/mila.12264 https://doi.org/10.2475/ajs.s3-34.203.333 https://doi.org/10.1016/j.neuropsychologia.2007.10.005 https://doi.org/10.1016/j.neuropsychologia.2007.10.005 https://doi.org/10.1371/journal.pone.0020910 https://doi.org/10.1017/s0140525x07002993 https://doi.org/10.1016/j.tics.2011.12.006 https://doi.org/10.1111/j.1526-4637.2011.01191.x https://doi.org/10.1111/j.1526-4637.2011.01191.x https://doi.org/10.1126/science.277.5328.968 https://doi.org/10.1126/science.277.5328.968 http://www.jstor.org/stable/4319832 https://doi.org/10.1126/science.1093535 https://doi.org/10.1126/science.1093535 https://doi.org/10.1016/j.tics.2015.10.002 https://doi.org/10.1038/ncomms14211 https://doi.org/10.1038/ncomms14211 https://doi.org/10.33735/phimisci.2021.9188 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction neural correlates of consciousness and theory-neutrality theory-neutrality theoretical presuppositions the case of water and h2o the dependence of theories and data in consciousness research how theoretical presuppositions guide consciousness research: the debate between cognitive and non-cognitive theories of consciousness chalmers's pre-experimental bridging principles how empirical results may affect theoretical presuppositions no-report paradigms pain research neural correlates of experienced and empathic pain conclusion the many problems with s-representation (and how to solve them). the many problems with s-representation (and how to solve them) jonny leea (leejonathan.cw@gmail.com) daniel calderb (dac@christs-hospital.org.uk) abstract the structural representation (s-representation) account provides an increasingly popular way of understanding the role and value of representation in cognitive science. yet critics remain unconvinced that the account has the resources to rescue representationalism. this paper reviews problems faced by the s-representation account. in doing so, it offers a novel taxonomy that divides objections into two broad camps that ought to be disambiguated: ‘conceptual’ and ‘empirical’. it further shows how these objections can be met, bolstering existing responses in the literature with novel solutions, thus strengthening the s-representation account. finally, this paper suggests that the types of objections identified provide a generalisable taxonomy for a better understanding of any version of anti-representationalism. thus, this review is of value to the representation debate more broadly. keywords cognitive science ∙ cognitive representation ∙ structural representation ∙ eliminativism ∙ content 1 introduction contemporary debates over the role and value of representation in cognitive science revolve around the notion of ‘structural representation’, or ‘s-representation’ for short. the s-representation account characterises cognitive representation as a class of state, structure or mechanism component that guides the behaviour of a cognitive system by structurally resembling features of its task environment, thus playing a mapor model-like role.1 proponents argue this approach idenauniversity of murcia. bchrist’s hospital. 1for simplicity, and in accordance with literature from which we draw (e.g., gładziejewski, 2015) we will refer to ‘cognitive mechanisms’ and ‘representational mechanisms’. however, we do not assume that proponents of the s-representation account are committed to mechanistic explanations of cognition (e.g., machamer et al., 2000), or that s-representations must be cast as mechanisms in the technical sense (e.g., see cummins, 1996). equally, if a mechanistic lens is the best way to view cognitive representation then our defense of s-representation only strengthens this perspective. lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-4453-2247 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 2 tifies a robust notion of representation that is relevant to our best explanations of cognition. despite this promise, there remain several pressing criticisms of the s-representation account. this paper reviews possible objections and organises them within a novel taxonomy that is likely applicable to other concepts of representation. we offer responses to each objection, drawing on the existing literature as well as neglected considerations. in doing so, we both strengthen the s-representation account and illuminate the conceptual landscape surrounding the representation debate more generally. though anti-representationalism is sometimes presented as a monolithic position, we contend that there are at least five independent (albeit overlapping) types of objections to the s-representation account. we label these the ‘a priori’, ‘function’, ‘content’, ‘best theory’, and ‘interpretation’ objections. these fall under two broad categories, which we label ‘conceptual’ (a priori, function and content objections) and ‘empirical’ (best theory and interpretation objections). this taxonomy is both descriptive and prescriptive; it captures a variety of objections advanced by critics as well as distinctions between classes of argument that ought to be disambiguated (we assume the taxonomy is comprehensive but not that it is exhaustive). our central contention is that the s-representation account has the resources to respond to each of these types of objections. we also take the taxonomy to be generalisable to any form of representationalism, beyond the s-representation account. hence, the discussion is pertinent to all debates over the role and value of cognitive representation and will prove useful for proand anti-representationalists alike. the paper proceeds as follows. §2 introduces the s-representation account. we pay special attention to clarifying the commitments of the s-representation account to swerve any potential strawmen. §3 sets out criticisms of the srepresentation account falling under the ‘conceptual’ bracket—the ‘a priori’, ‘function’ and ‘content’ objections—and offers a response to each. §4 turns to ‘empirical’ concerns and how to counter them—covering the ‘best theory’ and ‘interpretation’ objections. we conclude by suggesting that our taxonomy is broadly applicable and caution those within the debate against conflating different types of arguments for anti-representationalism. 2 the s-representation account cognitive scientists ascribe internal representations to parts of cognitive systems. on closer inspection, however, representations are attributed according to different standards across cognitive science (for an overview of some of these inconsistencies, see ramsey, 2007). some have also questioned whether at least some so-called cognitive representations have a recognisably representational role (e.g., facchin, 2021a), whether the concept of cognitive representation is coherent (e.g., bennett & hacker, 2007), and if we might be better off eliminating representation from cognitive science altogether (e.g., hutto & myin, 2012). the s-representation account offers a response to both the historic ambiguity of representation in coglee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 3 nitive science, as well as sustained objections to some or all representation-based theories of cognition. the goal of the s-representation account is to articulate at least one set of jointly sufficient conditions for a theoretical posit or construct to count as a genuine representation, and in doing so, at least partially reflect representation’s role and value in cognitive science. as such, the account need not exclude the possibility of other ways of representing; the criteria for s-representation are not necessary conditions for cognitive representation. according to the s-representation account, representations are a class of mechanism component—at first pass, neural or computational structures within an information processing system—that play a mapor model-like role for a cognitive system. when a mechanism contains a component that determines the behaviour of a system (the ‘user’ or ‘consumer’) via structural correspondence with a taskrelevant item (object, state, process etc.), the thought goes, then it functions like a map or model, i.e., a type of representation. for instance, as a mountaineer might use a cartographic map to navigate a mountain range by exploiting the structural similarities between the artefact and the geographical region, rats appear to use ‘cognitive maps’ to navigate their local environment by exploiting structural similarities between parts of their navigational system (primarily located in the hippocampus) and their local environment (o’keefe, 1976; o’keefe & dostrovsky, 1971; tolman, 1948). the s-representation account thus appeals to an analogy between a class of cognitive mechanism and a class of representational artefact. in ramsey’s (2007) terms, these mechanisms pass the ‘job description challenge’ (jdc)—they play a recognisably representational role—by functioning in a manner that resembles ordinary maps, models and the like. thus, an “if it looks like a duck and quacks like a duck” principle of classification underlies the description of certain mechanisms as representations. this is akin to categorising hearts, certain mechanoreceptor cells and parts of the oceanic carbon cycle as ‘pumps’ because, like those ordinary artefacts from which we derive the term, they function to move fluids by mechanical action. the s-representation account is elucidated through a set of conditions.2 we follow others (particularly gładziejewski, 2015) in identifying four key criteria for s-representation: structural correspondence, action guidance, decouplability and system-detectable error. as these have been discussed at length elsewhere, we restrict ourselves to a brief overview of each (gładziejewski, 2015; gładziejewski & miłkowski, 2017; lee, 2018; e.g., see o’brien & opie, 2004; piccinini, 2020; shea, 2018; williams, 2017). the central characteristic of an s-representation is its structural correspondence with elements in a system’s task environment (broadly understood as the spatiotemporal region containing all variables bearing on the cognitive capacity being executed). as o’brien & opie (2004) summarise, “one system structurally resembles 2we take these conditions to be necessary and jointly sufficient for s-representation and, in turn, jointly sufficient for cognitive representation. again, other conditions may suffice for representation. lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 4 another when the physical relations among the objects that comprise the first preserve some aspects of the relational organization of the objects that comprise the second” (2004 summarise, 15). for instance, in the case of cognitive maps, cells fire selectively in response to particular locations, whilst the firing rate and strength of connections between cells correspond proportionally to the distances between features in the system’s environment. thus, it is assumed that structural resemblance between the cognitive mechanism and the environment is functionally relevant to the rat’s success in navigating. like ordinary maps, correspondence needn’t be absolute for a mechanism to function as an s-representation; the structural correspondence between an ordinary artefact or cognitive mechanism and its target need only be as strong as required to complete the task. structural correspondence is thus only relevant in relation to action guidance. an s-representation corresponds to some task-relevant item and that correspondence is exploited by the system—as when rats exploit the correspondence between firing rates of place cells and features of an environment to aid navigation. furthermore, in keeping with paradigmatic cases of representation, including maps and models, s-representations are potentially decouplable from that which they correspond to in guiding action—they must be usable ‘offline’—as when rats route plan by exploiting cognitive maps. finally, it should be possible for a system to detect error or a mismatch between the s-representation and the item it must correspond to for successful behaviour to occur—for example, an insufficient correspondence between a cognitive map and a rat’s environment. components then ‘update’ to instantiate a stronger correspondence and so raise the probability of success in subsequent tasks. strictly speaking, the capacity to detect error and the capacity to update following error are conceptually independent and so might be understood as two conditions. in practice, however, paradigmatic s-representations, such as cognitive maps, are subject to amendment following the outcome of some behaviour and it is difficult to imagine how one might observe error detection without observing alteration to a mechanism in response to task failure. hence, we treat these conditions as bundled together. having outlined the positive story, we can now highlight connections between the s-representation account and other pertinent ideas in the philosophy of cognitive science. given our concern for objections to the account, we wish to avoid any strawmen lurking in the shadows. thus, we offer three clarifications regarding the s-representation account. we believe more superficial criticism may be avoided by recognising these, and in the process, the concept of s-representation will become clearer. the first clarification is that, though driven by an examination of scientific practice, the s-representation account is not an empirical theory. rather, it is a generic account of what it means to be a type of cognitive representation. in turn, this can be used to understand the role and value of representation in cognitive science. for example, in assessing which theories in cognitive science posit anything resembling genuine representations, ramsey (2007) concludes that the ‘classical comlee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 5 putational theory of cognition’—roughly synonymous with cognitivism—is representational because it implicates a computational architecture that functions to model the world, in a sense analogous to more familiar models. by contrast, ramsey concludes that connectionist theories of cognition employ no such model-like posits, nor anything else sufficiently representation-like (for pushback, see shagrir, 2012). more recently, theorists have offered interpretations of the representational commitments of ‘predictive processing’ (pp) in terms of s-representation (e.g., gładziejewski, 2016; for a partial review see sims & pezzulo, 2021, see §4.2 for discussion). the s-representation account is thus not an empirical theory but a set of conditions for a type of cognitive representation qua a type of theoretical construct or posit.3 secondly, the account is neutral about the range of cognitive phenomena to be explained representationally. sometimes, representations have been taken to be necessary for cognition (e.g., adams & aizawa, 2001). this is what ramsey (2016) calls the ‘representation demarcation thesis’, that is, “the view that cognitive processes necessarily involve inner representations and cognitive theories must thereby be about the representational states and processes” (2016, p. 4). the s-representation account makes no such assumptions (cf. cummins, 1996; ramsey, 2007, 2016). whether cognition is mostly representational, or whether it relies also on non-representational resources within a more mixed cognitive economy, depends on the details of our best theories and our broader conception of cognition. furthermore, the possibility that some cognition within some systems involves s-representations does not preclude the existence of other cognitive systems which are wholly non-representational, nor the existence of systems which represent some other way. for instance, it may be that offline cognitive processes— so-called ‘representation hungry’ problems (clark & toribio, 1994)—that characterise ‘higher cognition’, such as reasoning about counterfactuals, do involve srepresentations whilst online cognitive processes that comprise the minimal cognition of, say, bacterium do not (e.g., duijn et al., 2006). the third clarification concerns the compatibility between the s-representation account and some factions within ‘embodied, embedded, enactive and extended cognition’ (‘4e cognition’ or simply ‘4e’). reviewing the long and complex relationship between representationalism and 4e is beyond the scope of this paper, but we should recognise that some theories and frameworks within 4e revolted against orthodox representationalism. however, the s-representation account is not orthodox representationalism. for one thing, as we saw above, it’s not an empirical theory but an articulation of the generic properties belonging to a type of theoretical posit or construct. for another, as gładziejewski & miłkowski (2017) 3the metaphysical status of s-representations has been contested in recent years. we take the default view to assume that s-representations are a type of theoretical construct or posit but other interpretations claim they are observable entities (in particular, see thomson & piccinini, 2018). a review of possible positions on the metaphysics of cognitive representation must wait for another day. lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 6 point out, the notion of s-representation provides us with “an opportunity to develop, strengthen, and indeed reform the mainstream understanding of what representations are” (2017, p. 338). such reform is driven, in part, by advances in cognitive science. as williams & colling (2018) note, the “cognitive neuroscience revolution”, as outlined by boone & piccinini (2015), hails “a dramatic shift away from thinking of cognitive representations as arbitrary symbols towards thinking of them as icons that replicate structural characteristics of their targets” (2015, p. 1942). at the same time, many of the criticisms of representation found in 4e aim only at a particular understanding of cognitive representation—one that takes it to be discrete, language-like, brain-bound and purely descriptive (e.g., clark, 1997, 2015b, 2015a), in other words, the very notion seemingly discarded by contemporary cognitive neuroscience. we noted already that admitting some processes involve s-representations does not preclude the existence of non-representational processes, of the sort some 4e often concerns itself with.4 furthermore, the notion of s-representation is compatible with many advances within the representation debate which depart from traditional language-based formats and transformations. taking just one example, the s-representation account seems congruent with barsalou (1999) perceptual symbols theory which rejects the notion of amodal symbolic representations bearing arbitrary relations to their referents. perceptual symbols theory stresses the dependence of representing structures underlying more ‘cognitive’ tasks, such as imagination, on the re-activation of sensorimotor processes giving rise to a kind of perceptual simulation. in turn, what a system can represent is constrained by the perceptual modalities embodied by the system. we cannot here offer a complete exploration of how s-representations are productively nested within a 4e-friendly research program [but for some initial attempts at how s-representation supports and vindicates aspects of 4e, (see piccinini, 2022; williams, 2018). for now, without downplaying the genuine tension between cognitive representation and certain quarters of 4e (e.g., see varela et al., 1991), we contend that one’s predilection for 4e approaches does not exclude the possibility of s-representations playing a constructive role within cognitive science (for sample pushback, see hutto, 2013). in closing our introduction to the s-representation account, it is worth acknowledging the possibility of ‘explanatory pluralism’ in cognitive science. according to pluralists, cognition can be understood through a variety of frameworks (e.g., dale et al., 2009). within such an explanatory melting pot, it may be that only some frameworks appeal to s-representations. for example, one may entertain an explanatory pluralism that legitimises both mechanistic and dynamical systems explanations, where the former posit model-like entities deserving of the name 4one recent example of this includes the dual role of ‘model-free’ and ‘model-based’ modes of decision making within pp, which suggests to us a partial but vital role for s-representations within an action-oriented framework (clark, 2016, chapter 8). lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 7 ‘representation’ but the latter do not (see §4.1 for related discussion).5 notice that a version of pluralism in which representation plays a role only in some frameworks is not the same as instrumentalism about representation; pluralists may be realists about representation, denying only that such really existing entities are relevant to all types of explanation. whether such explanatory pluralism is warranted depends on wider issues within philosophy of cognitive science. however, the possibility that representation can be quarantined to certain explanatory frameworks without infecting all cognitive science may relieve some sceptics of representation’s explanatory power from the pressures of a zero-sum game. 3 conceptual objections there are at least five types of objections that arise from discussions of the srepresentation account: the a priori, function, content, best theory and interpretation objections. this taxonomy reflects significant differences in argumentative strategy. there is overlap between these types, as we shall see, and antirepresentationalists needn’t think their position can be divided neatly into the categories we individuate—for instance, because the different types of objections reinforce each other. nevertheless, we should avoid conflating distinctive sorts of arguments. in this section, we examine the a priori, function and content objections. these are chiefly ‘conceptual’ to the extent they question whether the criteria for so-called s-representation is sufficient for representation-hood. in §4, we turn to the best theory and interpretation objections. these are chiefly ‘empirical’ to the extent they question the relevance of s-representations for our best theories in practising cognitive science (and are thus sensitive to the emergence of new evidence), not what it means to be a representation.6 3.1 a priori objections a priori objections fixate on the very notion of a subpersonal representation. for some critics, ‘subpersonal representation’ is a conceptual confusion; how could something like an electrical circuit or neural state literally represent anything? notice that our first type of objection is a general one: it targets all accounts of representation. since this 'a priori eliminativism' is concerned with the possibility of providing an account of cognitive representation it affects the s-representation account as a consequence of targeting every account of representation. as we shall see, there are broad counters available to any representationalist. however, there are also some specific considerations afforded by the s-representation account. in any case, it is important to consider a priori objections because if the s5not everybody agrees that mechanistic and dynamical explanations are different forms of explanation (e.g., brigandt, 2013). we return to this point in §4.1. 6for similar divisions, see chemero (2009); steiner (2014). lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 8 representation account is to be successful it must possess the capacity to respond. they are also important to acknowledge insofar as they help differentiate this class of conceptual concern (about representation generally) from other types of objections (about the s-representation account). a priori eliminativism has its roots in a worry from ordinary language philosophy stemming from the assumption that representation implies some sort of user or consumer. in turn, the thought went, internal representations imply internal homunculi—mini agents who are required to interpret these representations (cf. ryle, 1949; wittgenstein, 1953). hence, attributing psychological predicates to the brain was seen as a kind of category error. perhaps the clearest contemporary expression of this sentiment is found in bennett & hacker’s (2007) attempt to exorcise ‘personal level predicates’ from the subpersonal level. ascribing representations to the brain is, according to bennett & hacker’s view, part of a larger practice of ascribing psychological attributes or predicates to the brain, such as when we talk about the brain constructing hypotheses, estimating probabilities or presenting arguments, as well as hearing, seeing and falling asleep (e.g., bennett & hacker, 2007, pp. 17, 21). moreover, whether such practices are legitimate is principally a conceptual, philosophical question, not a scientific one. their view is that ascribing psychological predicates to the brain is ‘senseless’. the brain, in their words, is “not a logically appropriate subject for psychological predicates” (bennett & hacker, 2007, p. 21). only humans and other agents are the proper domain of such attributes—only agents estimate probabilities or fall asleep. as bennett & hacker put it: “psychological predicates are predicates that apply essentially to the whole living animals, not to its parts” (2007, p. 22). hence, scientists and philosophers commit a ‘mereological fallacy’ when ascribing psychological predicates to parts of persons (bennett & hacker, 2007, p. 22). everyday life provides examples of constraints on literal reference; the wind does not literally whisper nor do fires rage—these are metaphors—and it is hard to imagine how these characteristically human behaviours could function as literal descriptions of the weather. similarly, it is difficult to imagine neurons mourning, celebrating or conniving in any literal sense. the force of views like bennett & hacker’s stems from generalising from such common sense cases of constraints on domains of reference to the conclusion that psychological predicates apply ‘essentially’ (bennett & hacker, 2007, p. 22) and ‘paradigmatically’ (bennett & hacker, 2007, p. 23) to the psychological domain. thus, an account like s-representation is doomed to fail, the thought goes, because the claim that the brain (or any subpersonal aspect of a cognitive system) represents is nonsense. key to the s-representationalist response must be the idea that cognitive science is driving an empirically led change in our understanding of the concept of representation. no a priori constraints prohibit the application of representation to the subpersonal level, and by revealing commonalities between the personal lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 9 and subpersonal levels, cognitive science alters how we attribute representations. even if the implicit rules for the use of representation may have once implied the necessity of agents, science has changed these rules because of what it has discovered about new domains. we suggest two ways of elaborating on this strategy. first is the ‘homuncular functionalist’ approach (e.g., dennett, 1975; fodor, 1968; lycan, 1991). homuncular functionalism can be taken as one version of the ‘technical’ view of applying psychological predicates to the subpersonal level (following figdor, 2018). technical views state that scientific uses of psychological predicates at the subpersonal level are (typically) both literal and meaningful (contra bennett & hacker). however, psychological predicates do not refer to the same set of properties as the original domain—they pick out something different. yet some significant similarity relation between the new and conventional referents is preserved. homuncular functionalism postulates that cognitive capacities at the personal level can be explained by positing levels of sub-systems with capacities that are progressively less sophisticated than the levels above. this allows psychological terms to be attributed to the sub-personal level so long as they are increasingly ‘less intelligent’, and eventually bottom out in brute, non-psychological capacities. in effect, this means that no mereological fallacy is committed when attributing representation to the subpersonal level so long as such representations do not possess all the features of the capacity they explain, and are eventually decomposable into non-psychological parts. put otherwise, any psychological or ‘homuncular’ implications at the subpersonal level are innocuous so long as the cognitive properties in question are progressively attenuated. homuncular functionalism (and other variants of the technical view) remain popular in the philosophy of mind and should be considered a viable strategy for combating a priori objections to the s-representation account. however, among its implications is that attributions of psychological predicates reflect a change in meaning. this is because psychological capacities ascribed to parts cannot be the same capacities as those at the personal level (remember, parts must be progressively stupider). a second way to elaborate the s-representationalist response that avoids this requires us to insist that representation is being used univocally at the personal and subpersonal level but that no fallacy is committed when we attribute the very same capacity of a whole to its parts (figdor, 2018). this is because there is no a priori reason why the same properties cannot be found at multiple levels. planets rotate and this is partly explained by the rotation of their atoms, with no change in the meaning of ‘rotation’. the same can be said of psychological terms. whilst a psychological predicate is so named because it originally described persons (as rotation originally described macroscopic objects), science shows how these terms capture patterns of behaviour in other domains (as it did for rotation at the molecular level). in other words, science applies the same psychological predicates to human agents and the subpersonal level because it has uncovered relevantly similar structures across the two domains. a version of this view is defended by figdor lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 10 (2018) who argues that formal models, in particular, provide grounds for attributing the capacities that we ascribe to human agents using psychological language to other domains, including the subpersonal level.7 there is no formal model for representation (though the structural correspondence condition for s-representation can be at least partially described using formalisms of isomorphism or homomorphism). however, the s-representation account does articulate a set of conditions that are intended to be realisable at the subpersonal level whilst capturing the functional character of a class of ordinary representation. this at least provides the basis for a qualitative analogy that resists the pull of the claim that the subpersonal level cannot instantiate the capacity to represent. this is the case even if such a capacity were quintessentially ‘personallevel’, because the s-representation account, in responding to scientific developments, shows how the functional profile of that capacity is recapitulated in the new domain. to borrow from dennett (2007), in response to bennett & hacker; “it is an empirical fact, and a surprising one, that our brains—more particularly, parts of our brains—engage in processes that are strikingly like guessing, deciding, believing, jumping to conclusions, etc. and it is enough like these personal level behaviours to warrant stretching ordinary usage to cover it” (2007, p. 86). a different flavour of a priori eliminativism is arguably found within strains of enactivism, insofar as they suggest the concept of cognitive representation misconstrues the fundamental ontological relationship between mind and world (e.g., varela et al., 1991). enactivist views on representation are often vague and heterogenous, and we cannot hope to dissect every variant of the enactive approach to representation here. thus, we will settle for noting that at least some enactivist views make clear that they are concerned with the personal level relationship between an agent and the world (particularly perceptual objects), and not subpersonal mechanisms (e.g., see noë, 2004; o’regan, 2011). this paper is concerned with subpersonal mechanisms and so the consequences of the account for claims about the relationship between mind and world at the personal level remain, at worst, ambiguous. we also note that certain ‘radical enactivist’ approaches (hutto & myin, 2012) that do encompass the subpersonal level are principally concerned with the ability of representational theories to demonstrate how a mechanism may function in a way that makes content functionally relevant. we engage with this view in §3.3. 7for example, the drift-diffusion model of two-choice decision-making (ratcliff & mckoon, 2008) applies to the behaviour of both fruit flies and humans (figdor, 2018). this model accounts for relatively automatic two-choice decisions in terms of accumulating information over time at a certain rate (the ‘drift rate’) until evidence for one option exceeds a prespecified threshold and a response to the decision-making task is given. lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 11 3.2 function objections the most straightforward way to undermine the s-representation account is to interrogate whether a mechanism meeting the functional conditions it establishes really qualifies as a representation (for discussion, see facchin, 2021a; gładziejewski & miłkowski, 2017; nirshberg & shapiro, 2020). the clearest expression of this objection, to our knowledge, is offered by facchin (2021a), who argues that s-representations cannot pass the job description challenge (jdc) because their functional profile is satisfied by so-called ‘receptor representations’ which paradigmatically fail the jdc. the idea underlying receptor representations is that if an internal state of some system nomically covaries with some distal event, then that state represents the event. this has its origins in the notion of ‘natural meaning’ elucidated by grice (1957), and paradigm cases such as ‘smoke means fire’ and ‘spots on the face mean measles’. applied to receptors we might intuit that a single neuron, for example, which fires in the presence of a stimulus represents that stimulus. facchin joins others (e.g., ramsey, 2007) in suggesting that receptors do not qualify as real representations. this is because, as basic ‘causal mediators’, their role is not sufficiently representation like, even if we complement the receptor notion with something like a dretskean theory of content, which specifies that some receptors play an important ‘indication’ role for the systems in which they are embedded (dretske, 1988). moreover, many clear-cut cases of non-representation appear to meet the conditions of receptors, for example, states of a firing pin in a gun nomically covary with trigger position. extending the label of ‘representation’ to encompass receptors would weaken the explanatory role of representation ascriptions, and invite a kind of ‘pan-representationalism’. the s-representation account is, in part, motivated by the goal to provide representation ascriptions with a more explanatory significant role than the receptor notion allows. the problem, facchin claims, is that certain receptors meet the criteria for srepresentation, as set out above. to be more precise, all receptors possess actionguiding structural correspondence with the target they are supposed to ‘indicate’ for a system, whilst some receptors can be made decouplable and sensitive to error. in brief, facchin claims that receptors will always meet at least one kind of exploited structural correspondence given the dependence on temporal passage for indication; a system whose functioning depends on receptors to indicate an event is (at minimum) sensitive to the temporal relations holding among features of the receptor mechanism (e.g., the temporal relations holding among different lengths of a bimetallic strip within a thermostat), and the obtaining of a time-dependent structural similarity between the receptor and the target system (e.g., a set of environmental temperatures). for brevity, we cannot discuss facchin’s examples at length, so we will accept for the sake of discussion that these are bona fide counterexamples. but even granting these counterexamples, facchin’s argument can be responded to. lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 12 several possible solutions stand out. the first solution notes that if the reasoning underpinning the function objection is sound, then the four conditions set out above cannot be sufficient to capture the functional profile of ordinary maps and models, and further work is required to identify the property or properties of srepresentations that distinguish them from receptors. after all, the thought goes, if a type of representation is defined by a certain functional profile, then there must be some functional feature of maps and models that we’ve missed. thus, the s-representationalist must identify a fifth condition that characterises such ordinary representations that can be mirrored in cognitive systems. whilst we think this solution is plausible, without any current suggestions on the market, it puts the s-representationalist on the back foot.8 the second solution also acknowledges that a limited number of mechanisms we classify as receptors meet the conditions set out by the s-representation account but clarifies that the issue lies in the coarse-grain nature of the functional profile in most presentations. it is not mere structural correspondence or system-detectable error that is vital to the distinctive functional character of s-representation, but exactly how these broad-brush functional characteristics operate. put differently, such descriptors are coarse-grained placeholders for a more detailed functional specification. like the first solution, this forces the s-representationalist to reconsider how to formulate their account, but it has the benefit of a clearer path to achieving this. for instance, whilst facchin (2021a) may be correct that even the most basic receptor instantiates a non-epiphenomenal structural correspondence with its target, this correspondence alone is clearly of a weak variety compared to the sort of correspondence that is supposed to be at work in, say, a cognitive map underlying counterfactual reasoning during prospective route planning. of course, further work is required to define precisely what the difference-making correspondence is in the latter case, but the strategy is clear enough. a variant of this second solution side steps any need to define different types of structural correspondence, or any other condition for s-representation, by instead claiming the difference between s-representations and (some) receptors is one of degree, not kind (cf. nirshberg & shapiro, 2020). what qualifies s-representations but not receptors for the job of representing is the richness or complexity of functional characteristics like structural correspondence. this may entail that representation is a graded notion with ambiguous cases, but many useful notions are graded, and the existence of borderline cases would not disqualify exemplars (clark & toribio, 1994). this response is developed by rutar et al. (2022), in the context of the predictive processing framework, who argue that at least two conditions for s-representation are gradual features, by appealing to empirical evidence: 8facchin discusses this solution but counters with two problems: (1) there is no obvious candidate for a fifth condition, and (2) the s-representation account, specifically, gładziejewski’s (2015) version, is already demanding. we agree the lack of candidates for a fifth condition is problematic but are less convinced by the second worry; the stringency of the s-representation account is a virtue, and besides, we cannot judge how much more demanding the s-representation account would become before any candidates have been identified. lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 13 structural similarity (in terms of the granularity of state space and the number of exploitable relations) and decouplability (in terms of how much neural structures depend on internal versus external stimulation, and the extent to which they are subject to precision weighting of prediction error). certain developmental considerations support this response; if there is phylogenetic or ontogenetic continuity between some receptors and s-representations within cognitive systems, we might expect a great deal of overlap in their functional profile. the third solution acknowledges that some receptors meet the conditions for srepresentation but bites the bullet and accepts, despite any intuitive classification to the contrary, that we should treat these too as passing the jdc, and thus, as representations. notice, however, that not every apparent receptor meets these conditions, e.g., single neurons firing in response to a stimulus. indeed, facchin’s (2021a) examples of receptors which meet the functional profile for s-representation add several bells and whistles to standard examples. even if we must widen the scope of jdc-passing mechanisms, we are far from trivialising representation. put otherwise, if some receptors satisfy the same profile as s-representations and we allow those receptors to pass the jdc, then there is no fatal threat to the s-representation account. and indeed, some theorists have embraced the conclusion that some receptors pass the jdc (see artiga, 2022; morgan, 2014). in short, facchin’s argument only appears forceful if we refuse to grant some receptors representational status, which the s-representationalist is not obliged to do. these promising solutions provide reasons to think the s-representation account has ways of escaping the function objection. nonetheless, we believe this challenge should be welcomed to the extent that it forces proponents to consider the functional properties of s-representation more carefully. even if opponents are satisfied that the s-representation account establishes a convincing functional profile for cognitive representation, they may still question whether it allows for a plausible story about semantic properties at the subpersonal level. 3.3 content objections representations are about things—they possess ‘content’. representational content implies ‘accuracy conditions’ (encompassing truth, correctness, and other semantic measures of success). this means representations can get things ‘right or wrong’ about what they represent. in other words, representations are ‘semantically evaluable’. representational artefacts acquire their content, at least in part, via the norms surrounding their use by ordinary agents. for instance, a cartographic map represents a geographical region, at least in part, because an individual or community use the object to stand in for that geographical region. there are no agents to fix meaning in the case of subpersonal representation. this leads to a puzzle about how to make sense of the semantic properties of cognitive representations. lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 14 following lee (2018), there are two problems associated with content. the first is the ‘hard problem of content’ (hpc) (hutto & myin, 2012), which concerns why we should think of s-representations as possessing semantic properties in the first place; in other words, what justifies thinking s-representations are in the business of being about anything. the second is the ‘content determination problem’, which concerns the conditions for determining the content of a token representation; in other words, what makes a particular s-representation about x and not y?9 we can ask analogous questions about familiar representations, like cartographic maps. on the one hand, we can ask what properties of maps allow for this type of artefact to function as content-bearing representations. on the other, we can ask what determines a particular map’s content. let’s examine these two problems in turn. hutto & myin (2012) introduce the hpc as a challenge for any naturalistic account of cognitive representation. the thought goes that for something to count as a genuine representation, we must account for its semantic properties in naturalistic terms. and yet, they conclude, there is no satisfactory account of content at the level of ‘basic cognition’. instead, content only appears with the emergence of intersubjective norms that provide standards for determining semantic properties (hutto & myin, 2017; cf. zahnoun, 2021). more recently, segundo-ortin & hutto (2019) have challenged the srepresentation account on similar grounds, suggesting proponents often presuppose but do not explain the origin of content (see also hutto & myin, 2017). they characterise the reasoning underpinning the s-representation account as follows: the properties of a given s-representational vehicle, r, cause it to be structurally similar to some target state of affairs, t. because r can mirror the structure of t more or less accurately, structural similarity entails accuracy conditions. accuracy conditions are taken to entail content. therefore, structural similarity is taken to entail content. thus, s-representationalists conclude, the fact that r structurally mirrors t entails that r contentfully represents t. (segundo-ortin & hutto, 2019, p. 10) however, they go on to conclude that whilst structural similarities might enable semantic evaluation, they are not themselves contentful: [i]t does not follow from the fact that we can make truth evaluable claims based on structural similarities holding between two items, a and b, that a contentfully represents something that might be true or false about b. (segundo-ortin & hutto, 2019, p. 13) 9this partitioning into two sub-problems is also implicit in segundo-ortin & hutto’s (2019) critique of the s-representation account, where they suggest that concerns about content indeterminacy are a distraction from the more serious problem of how structural similarity can ground content in the first place. lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 15 in short, segundo-ortin & hutto accept that structural similarities play a causal role in enabling successful cognition but insist this does not imply structural similarities are contentful. perhaps the most promising way to think about the role of content in the srepresentation account is in terms of how it captures the distinctive relationship between a mechanism which meets the conditions for s-representation, a target item, and behavioural success. segundo-ortin & hutto are correct that structural similarity is insufficient for content, but the s-representation account emphasises that mechanisms represent by virtue of the part they play in cognitive tasks. s-representations explain when the following conditions apply: (1) a system undertakes some task (e.g., navigating to a target location); (2) the outcome of that task depends on a mechanism with a component that structurally resembles task-relevant items (e.g., a cognitive map); (3) success depends on the degree of structural similarity between the mechanism component and those items (e.g., the topographical features of a rat’s local environment). in other words, where an s-representation plays a causal role in realising a cognitive capacity, that capacity causally depends on the degree of structural correspondence between the mechanism and some target item(s). given these features, the mechanism is described as functioning as a stand-in, surrogate or simulation of the target. attributing accuracy conditions that are met if the required correspondence occurs thus captures a facet of the role of the mechanism (its representation-like function). describing a cognitive map as accurate when it corresponds to the rat’s environment (causing success) and inaccurate when it does not correspond to the rat’s environment (causing failure) captures a distinctive and explanatorily relevant relationship between mechanism, target and behaviour. the same story is true, we think, for the familiar representations which inspire the s-representation account; ascribing content to ordinary maps and models is explanatory because of the relationship between vehicle, target and behavioural success. describing a cartographic map as accurate when it corresponds to the mountaineer’s environment (causing success) and inaccurate when it does not correspond to the mountaineer’s environment (causing failure) captures a crucial relationship between map, target and behaviour. if we are right, then content ascriptions are explanatory when behavioural success depends on an item playing a certain role, regardless of whether it is a conventional artefact or an evolved mechanism. the s-representationalist no more ‘presupposes’ content when explaining cognition by appealing to the exploitation of s-representations than we do when explaining the movement of tourists around london by appealing to the exploitation of an underground map; in each case, accuracy conditions fall out of the fact that behavioural success depends on an item playing the role of a stand-in, surrogate or simulation for some target. one might think the difference between the cognitive and ordinary map is that the consumer of the latter is a human agent and this is essential for bona fide content; therefore, so-called cognitive maps are not real representations. we lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 16 see no reason to assume agents are essential for content (see §3.1). however, even if content implied agency, we think cognitive science would be compelled to ascribe something content-like to mechanisms that met the conditions for srepresentation. let’s call this ‘schmontent’; when a mechanism meeting the criteria for s-representation sufficiently corresponds to an item such that it causes successful behaviour, all else being equal, we can call it ‘schmaccurate’, and ‘schminaccurate’ when it does not. if the anti-representationalist would be appeased by distinguishing content from schmontent, the s-representationalist should acquiesce, for the quarrel would transpire to be more-or-less terminological, and moreor-less everything that mattered about s-representations for scientific explanation would be preserved. segundo-ortin & hutto (2019) also raise valid concerns about the rush to draw an analogy between cognitive maps and ordinary maps based on a still somewhat incomplete understanding of the mechanisms involved (principally concerning whether ‘forward sweeps’ of activity in place cells implicated in anticipatory movement are used by the brain as surrogates for available routes). first, this objection amounts to advising caution about assigning representational function to cognitive maps whilst our mechanistic models remain impoverished. it does not demonstrate cognitive maps are not functioning as maps—only that we require more evidence. second, cognitive maps are only one case of purported srepresentation. third, this objection illustrates the shift from more conceptual concerns about the possibility of mechanisms bearing subpersonal content, and towards productive concern for the relative empirical support for the existence and range of s-representations in the brain (see §4).10 even if the hard problem is overcome, the s-representation account must still address the content determination problem. broadly, the worry is that whilst it may be permissible to speak of s-representations bearing content, in the abstract, the s-representation account lacks a satisfactory story about how a token mechanism comes to represent the particular item it does. moreover, many have worried the s-representation account implies a positive but implausible story about how content is fixed: mechanisms represent by virtue of bearing structural similarities to items in the world, the thought goes, so an s-representation must be about what it shares structure with. unfortunately, structural correspondence is cheap. whilst a hippocampal map might correspond to a rat’s environment—which intuitively relates to its content—it also corresponds to very many other things in the universe too. if structural correspondence fixes content, then this leads to ‘massive indeterminacy’ (sprevak, 2011, p. 671). the content determination problem is best addressed by showing how the srepresentation account is compatible with a story of content that does not lead to 10these objections may be mutually supporting. for instance, one’s negative appraisal of the evidence for analogy between brain mechanisms and ordinary maps might support existing suspicions that nothing can naturalise content at the subpersonal level. our adding precision to the debate by taxonomising objections does not imply the objections are not reciprocally reinforcing. lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 17 indeterminacy. there are broadly two ways of achieving this: ‘hybrid’ and ‘taskoriented’ approaches. rather than siding with one or the other, we will provide an overview of both. this is especially appropriate given that the hybrid and taskoriented approaches may be integrated. many theories—from causal dependency theory to asymmetric dependency theory to teleosemantics—have been claimed to resolve the content determination problem for representationalism. nevertheless, proponents of the s-representation account have sometimes formulated their theory, in part, as a response to the perceived inadequacy of these other theories to fully address the role and value of cognitive representation. for instance, ramsey (2007, 2016) claims that theories like teleosemantics fail to identify why representations play a role in cognitive science in the first place; by themselves, traditional theories of content do not establish that anything functions in a recognisably representational manner. however, he also proposes that such theories do plausibly address the content determination problem. hence, a promising strategy is to combine the s-representation account with a traditional theory of content to create a ‘hybrid account’. under this division of labour, the s-representation account specifies which cognitive mechanisms count as representations whilst, say, teleosemantics specifies the conditions that determine what a given representation is about. a second approach to the content determination problem indicates that an account of content determination is already implicated by the causal role of srepresentations in guiding the actions of a system (for related discussion, see lee, 2021; piccinini, 2022). whilst the minutiae vary between theorists, proponents of such task-oriented approaches are often careful to distinguish between two kinds of representational objects, already implicit in much of our discussion so far, which we will refer to here as the ‘target’ and ‘content’ of the representation. the former are, roughly, those conditions the representation is ‘applied to’ (gładziejewski, 2015, p. 80) or ‘used to deal with’ (godfrey-smith, 2006, p. 58). the latter are, roughly, those conditions that must occur for the capacity that depends on the mechanism to succeed. this can be thought of as the difference between what a representation must refer to for behavioural success to occur given the actual conditions of the task versus what a representation does refer to given how its structure guides the system. for example, the target of a cognitive map might be the structure of a novel maze a rat is currently navigating whilst its content refers to the structure of the maze it was previously trained on and which the map developed in response to. the accuracy of a given s-representation is a function of the overlap between target and content. we believe much of the task-oriented story aligns with the response given to the hpc above. however, we also note that the task-oriented and hybrid approaches are not necessarily exclusive. for instance, one may hold that to properly understand what a representation is being applied to—its target—one requires a naturalistic account of what task the system is performing, or what the mechanisms ‘proper function’ is, and this is something only an account such as teleoselee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 18 mantics is equipped to address (for a thorough account which combines elements of both approaches, see shea, 2018). in short, whether adopting the hybrid or task-oriented approach, or some combination, the s-representationalist has the resources to construct an account of content determination that does not lead to massive indeterminacy. in closing our discussion of the first three (‘conceptual’) objections, we should note an emerging theme, namely, that allowing for the possibility of s-representation at the subpersonal level invites the discovery of interesting similarities between aspects of familiar problem-solving and subpersonal cognition. we do not claim the s-representation account wholly reflects the folk usage of ‘representation’ (as if it was well-defined, to begin with). indeed, following a broadly quinean tradition, we take it that much philosophy is not in the business of uncovering the true meaning of clear-cut terms but rather refining how we can or should use a term for an area of discourse that requires high levels of precision. the point of the s-representation account, in our eyes, is not that we are compelled to use representational language de rigeur but that when the four conditions for s-representation (structural correspondence, action-guidance, decouplability, system detectable error) are met by a cognitive mechanism, the representational label effectively captures its functional role, much like labelling the heart as a pump is effective because of the strong resemblance between its activity and those of ordinary pumps. in this way, conceptual precision is aided through the redeployment of folk terms, in turn, refining what those terms mean within a particular (namely, scientific) context. 4 empirical objections suppose we accept both that subpersonal representation is possible, and that srepresentations, if they exist, function as representations with sufficiently determinate content. the anti-representationalist may nevertheless question whether s-representations belong in our best scientific theories. if s-representations do not feature in our best scientific theories, then the mere philosophical coherence of the account is not relevant to explaining cognition. there are two subtly distinct ways in which critics can argue the srepresentation account falters in this regard. first, one can argue that a non-s-representational theory of cognition is better than an s-representational theory of cognition, where everyone more-or-less agrees on which theories are representational and which are not. here, the disagreement concerns which theory is best and not which theories posit s-representations. second, one can argue that our best theory of cognition, thought by opponents to involve s-representations, features nothing of the sort. here, the disagreement concerns how to interpret the posits or constructs of our best theory and not which theory is best. there is an overlap between these objections, as we will see, but the delineation adds precision. the distinction also reflects contours in the literature, lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 19 where independent arguments have been offered for the superiority of a theory that is widely recognised as non-representational or for a non-representational interpretation of a more ambiguous theory. the first line of thought will be explored next as ‘the best theory objection’, and the second in the following section as ‘the interpretation objection’. 4.1 the best theory objection some anti-representationalists are less concerned with the conceptual lucidity of cognitive representation and more with whether representation features in our best scientific theories. here we consider this line of argument, concentrating on one especially prominent way in which the discussion has developed. if a non-representational theory offers a superior explanation of cognition, then via inference to the best explanation, an opponent of s-representation could conclude that even if the existence of s-representations is plausible, they are not in fact instantiated by cognitive systems. (a softer approach might conclude only that they are not useful for understanding real cognitive systems). to illustrate this line of argument, we will introduce a version of ‘radical embodied cognition theory’ as a case study, which draws on dynamical systems theory (dst) (broadly based on the account given by chemero, 2009, sec. 3). we will first present the reasons given for supposing that this anti-representational framework is a genuine alternative to representational theories. one popular reason for dismissing this version of embodied cognition as a genuinely explanatory theory—the question of ‘representation-hunger’—will be considered but found not to be decisive. in response, we will argue that the radically embodied view, which uses dst as a holistic paradigm for explaining cognition, is unpersuasive as an alternative to representationalism, and that the direction of travel in the scientific and philosophical literature shows that dst is better understood as a useful tool for modelling neural dynamics within a paradigm that is amenable towards s-representations. some anti-representationalists, such as chemero (2009), have argued that dst provides a framework for understanding cognition that treats the brain, body, and environment as a complete system whose behaviour can be modelled mathematically and without appeal to representations. dst predicts and explains how the states of a cognitive system will evolve through time. notably, the complex reciprocal relationships between neural states and states of the environment are not characterised by chemero (2009, sec. 2) in terms of representation, but as captured by a set of mathematical functions that govern the dynamics of the system. parallels are drawn between cognitive systems and the watt governor (gelder, 1995). the watt governor is a control system that acts to regulate the speed of an engine by limiting the amount of fuel it receives; as the speed of the engine increases, the centrifugal force it generates is harnessed by a pair of weighted arms to gradually close a valve to slow the consumption of fuel. thus, the governor maintains a constant engine speed as the entire system is engineered to remain lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 20 in a dynamic equilibrium. it is possible to provide a representational description of the governor’s operation by supposing that the angle of the arms represents the speed of the engine for the system, and that representation is ‘used’ by the system to control the aperture of the fuel valve.11 however, the thought goes, the dynamical (non-representational) explanation is more elegant and predictive. for any conditions, the dynamical equations allow us to calculate precisely how the system will evolve. despite sceptical claims that cognition is not the kind of phenomenon that can be explained using a set of dynamical equations, there are good reasons to suppose that this version of radical embodied cognition forms the basis for a progressive research program, in lakatos’ sense (chemero, 2009, p. 207). the haken-kelsobunz (hkb) model, for example, has provided a fruitful experimental paradigm. hkb uses the notion of an oscillator—a system that is stable but moves repeatedly between two or more states in a regular pattern—to characterise aspects of motor control such as walking. hkb can also be extended to be applied to coordination problems (e.g., limb movements, kugler et al., 1980). a common challenge raised against dst, however, is that despite providing an interesting angle for understanding automatic motor control, it cannot explain so-called ‘representation-hungry’ tasks, such as abstract problem-solving, requiring detailed visualisation and subsequent mental manipulation in the imagination (e.g., calculating a line of chess moves, solving a rubik’s cube, or considering which piece of art will best complement your interior design palette). nevertheless, dst has made promising developments in answering this challenge. stephen et al. (2009), for instance, asked experimental subjects to solve a gear-system problem in which an arrangement of gears is displayed, and the subjects are asked to infer the direction of rotation of a particular gear based on information about the rotation of one other gear in the arrangement. novices in the task use an inefficient and cognitively demanding strategy. however, following some experience, subjects develop a more efficient strategy based on an abstraction (counting the gears). the authors found that they could accurately predict this change in cognitive strategy using a dynamical analysis of action during problem-solving. by treating the subject as instantiating a lorenz attractor—a form of unstable attractor that characterises a system which transitions between several well-defined states— they could map the overall entropy of the subjects’ actions to this well-understood trajectory through the state-space of the overarching system. in this way, dynamical systems theorists use behavioural analyses to discover underlying patterns that can be captured by relatively simple mathematical models of our cognitive processes. similar studies use mouse tracking to collect data on hand movements during decision-making to understand subconscious biases (mckinstry et al., 2008) and the structure of language (spivey et al., 2005). 11the representational account of the watt governor is usually expressed in terms of a covariance relation and does not meet the criteria for s-representation (cf. bechtel, 1998). lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 21 citing examples such as these, some advocates of radical embodied cognition argue that dst offers at least some of the tools needed to form a credible alternative to theories of subpersonal cognition that appeal to representational mechanisms. in addition to the success of empirical work and the progressive nature of this research, there appear to be broader theoretical advantages. whilst classical computational theories focus on intracranial processes, dst appeals to facts about the entire brain-body-environment system to provide accurate and robust models. in doing so, it avoids the (apparent) decompositional strategy of some mechanistic theorists, treating cognition as a phenomenon that cannot be decomposed into discrete subsystems. dynamics are often chaotic, with small changes in the structure of a system or its initial conditions leading to consequences that are strictly unpredictable. thus, theoretical commitments may lead us to conclude that representational mechanisms are not part of our best explanations. as radical embodied cognition illustrates, the best theory objection focuses attention on the relative virtues of competing explanatory strategies (pluralism notwithstanding). the objection, taken alone, allows that s-representation is a coherent and potentially explanatorily useful concept, objecting only that a nonrepresentational theory (here, using tools of dst) is preferable. in §2 we clarified that the concept of s-representation is not a theory of cognition, but refers to a type of posit or construct that may be invoked by different theories. the acceptance of srepresentation as a coherent and potentially-useful concept would be a significant concession to its defenders. in turn, those defending the empirical utility of srepresentations may feel unthreatened by ‘best theory’ objections, believing that explanatory strategies involving representation are superior to those that do not, hence their hard work to elucidate the conceptual details of the posit or construct. their confidence may be justified given the continued dominance of model-based theories in cognitive science. the success or failure of the best theory objection will ultimately depend on future scientific developments. in this sense, those involved in the debate can put aside their differences while promoting the empirical work they believe to be most promising. it is worth noting at this point, however, that a significant volume of work on dst embraces the notion of representation and seeks to use system dynamics to unify computational explanations with underlying neural dynamics (e.g., cisek, 2007; clark, 2008; eliasmith & anderson, 2003; piccinini, 2022). in contrast to a more rigidly demarcated pluralism, which denies representation a role within dst but allows for the coexistence of different frameworks, this more ‘inter-theoretic’ approach suggests that dst and s-representation can operate as part of the same explanatory package. one way this might work, in brief, is that representational models make sense of the functional mechanisms within the brain, with dst elucidating how s-representations are instantiated by neural dynamics and, at a greater spatiotemporal scale, the interaction between whole brain-body-environment systems. at the neural level, dynamics may enable us to understand how the behaviour of lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 22 a single cell or set of cells, whose states may vary in many different ways, can be constrained by their organisation to encode discrete computational vehicles. in this way, these analogue systems might even be taken to instantiate classical computational structures (eliasmith & anderson, 2003, p. 55). an example of this is the oculomotor integrator, as analysed by shagrir (2012). the speed and direction of saccades are transformed by this small set of neurons into signals used by the oculomotor system to remember the new position of the eye. thus, the integrator circuit enables certain inputs to represent these properties of the eye, and the output represents the new position and is remembered for later use. plausibly, the entire subsystem encodes an s-representation of possible changes in eye position which is realised by the particular dynamics of the neural activity. higher-level analysis of brain dynamics has primarily been used to motivate a more ‘action-oriented’ understanding of cognitive representation (cisek, 2007; remington et al., 2018). this approach rejects the notion that representations instantiate an agent-neutral model of the world, but rather represent ‘affordances’. for example, rather than representing a water bottle as ‘being 2 feet away’, a water bottle may instead be represented as being ‘reachable’. affordances such as reachability are dependent on the embodiment of the agent and its capacities. this approach to content is not incompatible with the s-representation account, which stresses the importance of action guidance. schöner (2019), for instance, offers a theory-neutral account of how dynamical and representational explanations can be effectively bridged that emphasises features such as decoupleability, stable structural relationships, and system-detectable error. though a complete interpretation of schӧner’s approach to representation through an s-representational lens must wait, we can begin to see how such features are suggestive of action-guiding mapor model-like structures i.e., candidate s-representations. in brief, schöner (2019) illustrates how the dynamics of excitation and inhibition of entire neural populations can realise intentional mental states, together with their conditions of satisfaction (cos). similar to the oculomotor integrator, the dynamical relationships between these populations instantiate a structure that facilitates the use of explanations involving s-representations. in schӧner's example, one neural population (a) is taken to instantiate a particular intentional state. this may play any number of roles within the neural system, being activated by top-down, horizontal, or bottom-up connections originating from other neural populations. another neural population (b) functions to inhibit the activation of a. b’s inputs are taken to covary with the cos for the intentional state instantiated by a. thus, when a and b are active, the system is actively representing a particular state of affairs, and that state of affairs obtains. any errors represented by a will be detected and processed as residual activation flows as a result of imperfect inhibition by b. thus, the possibility of mapping the activation of a neural population encoding one intentional state to another which correlates with the cos for that state enables a basic s-representational interpretation. this is strengthened by further structural lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 23 relationships, however, such as the inhibitory coupling between the two; once the prediction of success encoded by the intention is fulfilled by the activation of the cos population (b), the intentional population (a) is inhibited, and consequently, there is no error to be detected. in this way, inhibitory dynamics can be interpreted as instantiating system-detectable error, coherent with an s-representational understanding of the system. the upshot of such bridging analyses is that we need not adopt a binary approach when considering the explanatory success of s-representational theories and radical embodiment based on dst. there are good reasons to suppose that both strategies can and should be taken together to bring about a full understanding of cognition. rather than treating dst as forming the basis for a complete explanatory paradigm, it should be (and has been) adopted as a useful tool for modelling the dynamics of neural populations in a way that provides further support to representational explanations of cognition. the broader lesson for thinking about the best theory objection is that we must carefully assess whether the supposed rival to a representational theory really does embody a competitor. even if the proand anti-representationalist can agree over which theory in cognitive science is best, the latter may still insist that s-representations are empirically inert because we are mistaken in supposing that our best theory features s-representations to begin with. this brings us to the interpretation objection. 4.2 the interpretation objection if the s-representation account provides a convincing conceptual foundation for subpersonal representation, and all parties agree on our best theory of cognition, disagreement may still be found over whether this theory posits s-representations. whilst the best theory objection is chiefly concerned with which theory of cognition is best, the interpretation objection is concerned with how to construe our best theory’s constituents. at this stage, more overlaps in our taxonomy become evident; if one is convinced by a priori objections to subpersonal representation (see §3.2), then one will be driven to a non-representational interpretation of a theory. however, ‘a priori’ and ‘interpretation’ objections still come apart because one can contest the representational credentials of a theory without thinking subpersonal representation is a category mistake—as evidenced, for instance, by ramsey’s (2007) analysis of representation in different theories of cognition. we should also acknowledge potential ambiguity over whether a disagreement concerns competition between what our best theory is and what our best theory entails. two interpretations of predictive processing (pp), for instance, might diverge so much that we no longer count them as the same theory. regardless, we separate the best theory and interpretation objections for added precision, noting the representationalist and antirepresentationalist may agree on a great many things concerning the best methlee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 24 ods, models, evidence and descriptive tools for understanding cognition, but still disagree over whether these imply s-representation. the interpretation objection can be illustrated with reference to predictive processing. very briefly, pp offers a theory of cognition in which the brain is organised to minimise error in its own internally generated (top-down) predictions of the incoming (bottom-up) sensory input (clark, 2016; for an introduction, see friston, 2009; hohwy, 2013). in orthodox formulations, predictions are measured against sensory input, producing prediction errors that are used to update a multi-level ‘generative model’ which determines future predictions (e.g., clark, 2016). this is achieved by encoding prior expectations (‘priors’) about sensory input, pitched at multiple spatial and temporal scales spread across a processing hierarchy. at higher levels in the generative model, priors consist of a set of ‘hypotheses’ that reflect expectations about the hidden (worldly) causes of stimuli, and an assignment of probability to those hypotheses. according to some versions, hypotheses are selected (assigned a ‘posterior value’) as a function of their prior probability plus their ‘likelihood’—the probability that the state of affairs captured in the hypothesis would cause the received sensory input, were it true—in approximate accordance with bayes’ theorem. the s-representation account has been used to identify the representational posits or constructs of pp (gładziejewski, 2016; kiefer & hohwy, 2018; wiese, 2017). one noteworthy claim found in representational treatments of pp is that the generative model functions as a kind of s-representation (or set of nested srepresentations). the general idea is that for the brain to endogenously generate a prediction of the sensory signal it must embody a causal-probabilistic structure that maps onto the hidden worldly causes of stimuli, accomplished through encoding a multi-level network of updatable priors. gładziejewski (2016) offers such a view when he writes, [c]ognitive systems navigate their actions through the use of a sort of causal–probabilistic “maps” of the world. these maps play the role of representations within the theory. specifically, this map-like role is played by the generative models. it is generative models that, similarly to maps, constitute action-guiding, detachable, structural representations that afford representational error detection. (2016, p. 569) in response, opponents have sought to undermine pp’s representational credentials. there are different ways to attempt this kirchhoff & robertson (2018). however, one common strategy for combating representational interpretations of pp is to downplay the necessity of understanding notions like priors and generative models as representational, and instead recommend we understand such posits or constructs in terms of hardwired constraints, biases, sensory attunement and other lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 25 ‘ecological’ notions.12 for instance, writing about related bayesian approaches to visual perception from an ecological point of view, orlandi writes, [p]riors do not look like contentful states. they do not need to have accuracy conditions to perform their function. they rather look like built-in or evolved functional features of perception that skew visual networks toward certain configurations. (2014, p. 82) engaging with orlandi’s thesis, downey claims, the concepts “prior” and “likelihood” are better understood as referring to mechanisms which pre-dispose brains to configure themselves into specific organisational patterns in response to environmental stimulation […] biases do not pass the job description challenge, and so we have no reason to treat them in terms of representation. (2018, p. 5121) hutto offers a related interpretation of pp, from a ‘radical enactivist’ approach (reworking a passage from clark, 2016, p. 27): nothing here requires [the brain] to engage in processes of … [contentful] prediction or expectation. all that matters is that … [its] systems be able to [anticipate and be adjusted by sensory perturbations] in ways that make the most of whatever regularities … [to which it is attuned, because such attunement has] … proven useful … [in response to such regularities in the past]. (hutto, 2018, p. 2456. original parentheses.) recently facchin (2021b) has suggested that generative models in pp systems do not satisfy the conditions for s-representations, as they either do not satisfy the condition of distality13, or they do not satisfy the condition of exploitable structural similarity. according to facchin, pp systems qua computational networks can feasibly represent only the proximal patterns of input activation, as it is only to this proximal activation that the generative model bears an exploitable structural correspondence. thus, pp systems cannot be interpreted as using s-representations. the general lesson of these perspectives is the possibility of an interpretation of pp that replaces talk of internal models bearing representational content with ecological concepts.14 12though these arguments do not always single out the s-representation account, they are offered in the context of a literature which does interpret the posits or constructs of pp as srepresentations (e.g., gładziejewski, 2016). 13note that this is not a condition included in our definition of s-representations. 14a somewhat different take is offered by kirchhoff & robertson (2018) who target a particular representationalist interpretation of pp that centres kullback-leibler divergence between posterior generative and approximate recognition densities which the authors think relies on a relation of covariance, and covariance does not constitute content (hutto & myin, 2012). we think the authors err in emphasising the role of generalised synchrony (mutual information) in the wider mechanism (meeting the criteria for s-representation) responsible for behavioural success (see lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 26 a satisfactory response to any ‘interpretation objection’ requires closely engaging with the particular details of the scientific theory in question—such as the constructs of pp—and how best to interpret them (and here we are not invested in defending a representational interpretation of pp, in particular). there are, however, some general considerations to bear in mind when developing a defence of s-representation. one important consideration is that the interpretation objection, like the best theory objection, is not itself an objection to the s-representation account qua an account of how cognitive systems might instantiate representing mechanisms, in principle. as noted above, it is possible to maintain the coherency of the s-representation account but object to its theoretical applicability. moreover, the inability of pp (or any other theory) to provide a convincing example of s-representation only threatens the applicability of the account if (a) pp is our best theory, and (b) it is exhaustive i.e., there is no room for theories which do provide effective examples of s-representation alongside pp (this is most obviously possible if pp explains only a subset of cognitive phenomena). like the best theory objection, proponents should welcome any transition from assaults on the very possibility of s-representation to a discussion of how widespread s-representations are in reality. a second consideration is that whilst the interpretation objection relies on evaluating the posits or constructs of a given theory—and in this sense is an ‘empirical’ concern—the quotations above demonstrate that theorists can and do disagree over how best to characterise fundamental theoretical contents; they disagree over whether a posit or construct—such as a system of priors—should be interpreted as a representation (or even as existing at all, following facchin, 2021a!). for such cases, we identify three options for those defending the empirical relevance of the s-representation account. first, one can argue that disagreement is purely terminological. for instance, proponents can argue that a mechanism which predisposes a cognitive system to configure itself into specific organisational patterns in response to stimulation— once the details are hashed out—is just an s-representation without the label (for general consideration of the representation debate as purely terminological, see haselager et al., 2003). so long as the four functional criteria might reasonably be attributed to a mechanism, that mechanism is an s-representation, regardless of how else it is described. however, the extent to which this consideration provides a plausible rebuttal to anti-representationalists is questionable. pp has spawned a quickly evolving debate over what exactly the theory is committed to, and diverging interpretations plausibly reflect different (and mutually exclusive) outcomes. second, one can acknowledge that the sorts of interpretations offered in the quotations above do reflect genuine alternatives to a representational interpreta§3.2). regardless, it’s worth noting the authors do not question the conceptual foundations of s-representation, only the capacity of pp to provide a convincing example. moreover, they recognise their objection only targets a version of a representationalist interpretation of pp and is not a generalised argument against representationalism. lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 27 tion but maintain that the representationalist interpretation of pp is superior. we will not settle the debate here but note that (a) within the philosophy of cognitive science literature, at the very least, the jury is out, with some recent literature defending a representationalist interpretation, and (b) within practising cognitive science, talk of model-like mechanisms remains ubiquitous. therefore, the proponent of the relevance of the s-representation account for cognitive science is, in our estimation, on a firm footing. third, one can argue that the sorts of interpretations offered in the quotations above do reflect genuine alternatives to a representational interpretation, however, such interpretations operate at a different (non-mutually exclusive) level of description. the representationalist does not claim representational descriptions are necessary; one can always explain a cognitive phenomenon in more brute ‘ecological’ terms. the claim is that representational language delivers some explanatory grip, from a certain level of description. for instance, we might describe a mechanism either as a system that anticipates and adjusts itself by sensory perturbations in ways that exploit regularities in its history and as an internal representation. as a general rule, the translatability of representational descriptions into nonrepresentational descriptions does not imply the former are false. in the case of an emerging theory like pp, we may need to wait and see the extent to which progress depends upon understanding systems of updatable priors as action-guiding, decouplable, and error-sensitive models (for discussion on the early developmental stage of pp and the caution this engenders, see dolega, 2017). finally, it is worth mentioning the possibility that pp provides less of a complete theory and more of a mechanistic schema (machamer et al., 2000), in other words, it supplies abstract specification of a mechanism (or set of mechanisms) which must be ‘filled in’ by a more complete mechanistic account (for related discussion see gładziejewski, 2019; miłkowski & litwin, 2022). if correct, then this may help diagnose the ambiguity of what, exactly, pp is committed to: in its most generic sense, pp does not provide enough mechanistic details to determine whether mechanisms meeting something like the s-representation profile are required. regardless, we bet that the persistence of model talk within pp (and across cognitive science) indicates the continued relevance of s-representation for the foreseeable future. 5 conclusion there is growing consensus that the s-representation account offers the most relevant and promising account of representation for contemporary cognitive science. however, there are several types of objections, and these ought to be distinguished. whilst this paper focused on reviewing objections (and their responses) to the srepresentation account, the novel taxonomy of objections we offered will prove useful when assessing criticisms of representationalism more generally. locating an anti-representationalist argument in relation to the categories we set out oflee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 28 fers non-partisan benefits. for the representationalist and anti-representationalist alike, it encourages a refined understanding of logically independent arguments. in turn, it helps identify the relationship between different criticisms, for example, diagnosing shared argumentative strategies among different expressions of the ‘a priori’ objection or distinguishing between purely conceptual and strictly empirical concerns. the protracted representation wars have often suffered a lack of precision which has led to confusion and misunderstanding; representationalism and its repudiation have meant many different things with varying implications for cognitive science. improved clarity should be welcomed by all participants. references adams, f., & aizawa, k. (2001). the bounds of cognition. philosophical psychology, 14(1), 43–64. https://doi.org/10.1080/09 515080120033571 artiga, m. (2022). strong liberal representationalism. phenomenology and the cognitive sciences, 21(3), 645–667. https://doi.org/10.1007/s11097-020-09720-z barsalou, l. w. (1999). perceptual symbol systems. behavioral and brain sciences, 22, 577–609. https://doi.org/10.1017/s014 0525x99002149 bechtel, w. (1998). representations and cognitive explanations: assessing the dynamicist’s challenge in cognitive science. cognitive science, 22(3), 295–317. https://doi.org/10.1207/s15516709cog2203_2 bennett, m., & hacker, p. (2007). selections from philosophical foundations of neuroscience. in m. bennett, d. dennett, p. hacker, & j. searle (eds.), neuroscience and philosophy: brain, mind and language (pp. 3–48). columbia university press. boone, w., & piccinini, g. (2015). the cognitive neuroscience revolution. synthese, 193(5), 1509–1534. https://doi.org/10.1 007/s11229-015-0783-4 brigandt, i. (2013). systems biology and the integration of mechanistic explanation and mathematical explanation. studies in history and philosophy of biological and biomedical sciences, 44, 477–492. https://doi.org/10.1016/j.shpsc.2013.06.002 chemero, a. (2009). radical embodied cognitive science. mit press. https://doi.org/10.7551/mitpress/8367.001.0001 cisek, p. (2007). cortical mechanisms of action selection: the affordance competition hypothesis. philosophical transactions of the royal society of london. series b, biological sciences, 362(1485), 1585–1599. https://doi.org/10.1098/rstb.2007.2054 clark, a. (1997). being there: putting brain, body, and world together again. mit press. https://doi.org/10.7551/mitpress/155 2.001.0001 clark, a. (2008). supersizing the mind: embodiment, action, and cognitive extension. oxford university press. clark, a. (2015a). predicting peace: the end of the representation wars—a reply to michael madary. in t. metzinger & j. m. windt (eds.), open mind: 7(r). mind group. https://doi.org/10.15502/9783958570979 clark, a. (2015b). radical predictive processing. the southern journal of philosophy, 53, 3–27. https://doi.org/10.1111/sjp.12120 clark, a. (2016). surfing uncertainty: prediction, action, and the embodied mind. oxford university press. clark, a., & toribio, j. (1994). doing without representing? synthese, 101(3), 401–431. https://doi.org/10.1007/bf01063896 cummins, r. (1996). representations, targets, and attitudes. mit press. https://doi.org/10.7551/mitpress/5887.001.0001 dale, r., dietrich, e., & chemero, a. (2009). explanatory pluralism in cognitive science. cognitive science, 33(5), 739–742. https://doi.org/10.1111/j.1551-6709.2009.01042.x dennett, d. (1975). why the law of effect will not go away. journal for the theory of social behaviour, 5(2), 169–188. https://doi.org/10.1111/j.1468-5914.1975.tb00350.x dennett, d. (2007). philosophy as naive anthropology: commment on bennett and hacker. in m. bennett, d. dennett, p. hacker, & j. searle (eds.), neuroscience and philosophy: brain, mind and language (pp. 73–94). columbia university press. dolega, k. (2017). moderate predictive processing. in t. k. metzinger & w. wiese (eds.), philosophy and predictive processing. mind group. https://doi.org/10.15502/9783958573116 downey, a. (2018). predictive processing and the representation wars: a victory for the eliminativist (via fictionalism). synthese, 195(12), 5115–5139. https://doi.org/10.1007/s11229-017-1442-8 lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1080/09515080120033571 https://doi.org/10.1080/09515080120033571 https://doi.org/10.1007/s11097-020-09720-z https://doi.org/10.1017/s0140525x99002149 https://doi.org/10.1017/s0140525x99002149 https://doi.org/10.1207/s15516709cog2203_2 https://doi.org/10.1007/s11229-015-0783-4 https://doi.org/10.1007/s11229-015-0783-4 https://doi.org/10.1016/j.shpsc.2013.06.002 https://doi.org/10.7551/mitpress/8367.001.0001 https://doi.org/10.1098/rstb.2007.2054 https://doi.org/10.7551/mitpress/1552.001.0001 https://doi.org/10.7551/mitpress/1552.001.0001 https://doi.org/10.15502/9783958570979 https://doi.org/10.1111/sjp.12120 https://doi.org/10.1007/bf01063896 https://doi.org/10.7551/mitpress/5887.001.0001 https://doi.org/10.1111/j.1551-6709.2009.01042.x https://doi.org/10.1111/j.1468-5914.1975.tb00350.x https://doi.org/10.15502/9783958573116 https://doi.org/10.1007/s11229-017-1442-8 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 29 dretske, f. (1988). explaining behavior: reasons in a world of causes. mit press. duijn, m. van, keijzer, f., & franken, d. (2006). principles of minimal cognition: casting cognition as sensorimotor coordination. adaptive behavior, 14(2), 157–170. https://doi.org/10.1177/105971230601400207 eliasmith, c., & anderson, c. h. (2003). neural engineering: computation, representation, and dynamics in neurobiological systems. mit press. facchin, m. (2021a). predictive processing and anti-representationalism. synthese, 199(3), 11609–11642. https://doi.org/10 .1007/s11229-021-03304-3 facchin, m. (2021b). structural representations do not meet the job description challenge. synthese, 199, 5479–5508. https://doi.org/10.1007/s11229-021-03032-8 figdor, c. (2018). pieces of mind: the proper domain of psychological predicates. oxford university press. https://doi.org/10 .1093/oso/9780198809524.001.0001 fodor, j. (1968). psychological explanation. mit press. friston, k. (2009). the free-energy principle: a rough guide to the brain? trends in cognitive sciences, 13(7), 293–301. https://doi.org/10.1016/j.tics.2009.04.005 gelder, t. van. (1995). what might cognition be, if not computation? the journal of philosophy, 92(7), 345–381. https://doi.org/10.2307/2941061 gładziejewski, p. (2015). explaining cognitive phenomena with internal representations: a mechanistic perspective. studies in logic, grammar and rhetoric, 40(53), 63–90. https://doi.org/10.1515/slgr-2015-0004 gładziejewski, p. (2016). predictive coding and representationalism. synthese, 193(2), 559–582. https://doi.org/10.1007/s112 29-015-0762-9 gładziejewski, p. (2019). mechanistic unity of the predictive mind. theory & psychology, 29(5), 657–675. https://doi.org/10 .1177/0959354319866258 gładziejewski, p., & miłkowski, m. (2017). structural representations: causally relevant and different from detectors. biology & philosophy, 32(3), 337–355. https://doi.org/10.1007/s10539-017-9562-6 godfrey-smith, p. (2006). mental representation, naturalism, and teleosemantics. in d. papineau & g. macdonald (eds.), teleosemantics (pp. 42–68). oxford university press. grice, h. p. (1957). meaning. philosophical review, 66(3), 377–388. https://doi.org/10.2307/2182440 haselager, p., groot, a., & rappard, h. (2003). representationalism vs. anti-representationalism: a debate for the sake of appearance. philosophical psychology, 16(1), 5–24. https://doi.org/10.1080/0951508032000067761 hohwy, j. (2013). the predictive mind. oxford university press. hutto, d. (2013). exorcising action oriented representations: ridding cognitive science of its nazgûl. adaptive behavior, 21(3), 142–150. https://doi.org/10.1177/1059712313482684 hutto, d. (2018). getting into predictive processing’s great guessing game: bootstrap heaven or hell? synthese, 195, 2445– 2458. https://doi.org/10.1007/s11229-017-1385-0 hutto, d., & myin, e. (2012). radicalizing enactivism: basic minds without content. mit press. https://doi.org/10.7551/mitp ress/9780262018548.001.0001 hutto, d., & myin, e. (2017). evolving enactivism: basic minds meet content. mit press. https://doi.org/10.7551/mitpress/9 780262036115.001.0001 kiefer, a., & hohwy, j. (2018). content and misrepresentation in hierarchical generative models. synthese, 195(6), 2387– 2415. https://doi.org/10.1007/s11229-017-1435-7 kirchhoff, m. d., & robertson, i. (2018). enactivism and predictive processing: a non-representational view. philosophical explorations, 21(2), 264–281. https://doi.org/10.1080/13869795.2018.1477983 kugler, p. n., kelso, j. a. s., & turvey, m. t. (1980). on the concept of coordinative structures as dissipative structures i. theoretical lines of convergence. in g. e. stelmach & j. requin (eds.), tutorials in motor behavior (pp. 3–47). north holland. https://doi.org/10.1016/s0166-4115(08)61936-6 lee, j. (2018). structural representation and the two problems of content. mind & language, 34(5), 606–626. https://doi.or g/10.1111/mila.12224 lee, j. (2021). rise of the swamp creatures: reflections on a mechanistic approach to content. philosophical psychology, 34(6), 805–828,. https://doi.org/10.1080/09515089.2021.1918658 lycan, w. (1991). homuncular functionalism meet pdp. in w. ramsey, s. stitch, & d. rumelhart (eds.), philosophy and connectionist theory (pp. 259–286). lawrence erlbaum. machamer, p., darden, l., & craver, c. f. (2000). thinking about mechanisms. philosophy of science, 67(1), 1–25. https://doi.org/10.1086/392759 lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1177/105971230601400207 https://doi.org/10.1007/s11229-021-03304-3 https://doi.org/10.1007/s11229-021-03304-3 https://doi.org/10.1007/s11229-021-03032-8 https://doi.org/10.1093/oso/9780198809524.001.0001 https://doi.org/10.1093/oso/9780198809524.001.0001 https://doi.org/10.1016/j.tics.2009.04.005 https://doi.org/10.2307/2941061 https://doi.org/10.1515/slgr-2015-0004 https://doi.org/10.1007/s11229-015-0762-9 https://doi.org/10.1007/s11229-015-0762-9 https://doi.org/10.1177/0959354319866258 https://doi.org/10.1177/0959354319866258 https://doi.org/10.1007/s10539-017-9562-6 https://doi.org/10.2307/2182440 https://doi.org/10.1080/0951508032000067761 https://doi.org/10.1177/1059712313482684 https://doi.org/10.1007/s11229-017-1385-0 https://doi.org/10.7551/mitpress/9780262018548.001.0001 https://doi.org/10.7551/mitpress/9780262018548.001.0001 https://doi.org/10.7551/mitpress/9780262036115.001.0001 https://doi.org/10.7551/mitpress/9780262036115.001.0001 https://doi.org/10.1007/s11229-017-1435-7 https://doi.org/10.1080/13869795.2018.1477983 https://doi.org/10.1016/s0166-4115(08)61936-6 https://doi.org/10.1111/mila.12224 https://doi.org/10.1111/mila.12224 https://doi.org/10.1080/09515089.2021.1918658 https://doi.org/10.1086/392759 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jonny lee and daniel calder 30 mckinstry, c., dale, r., & spivey, m. j. (2008). action dynamics reveal parallel competition in decision making. psychological science, 19(1), 22–24. https://doi.org/10.1111/j.1467-9280.2008.02041.x miłkowski, m., & litwin, p. (2022). testable or bust: theoretical lessons for predictive processing. synthese, 200(6), 462. https://doi.org/10.1007/s11229-022-03891-9 morgan, a. (2014). representations gone mental. synthese, 191(2), 213–244. https://doi.org/10.1007/s11229-013-0328-7 nirshberg, g., & shapiro, l. (2020). structural and indicator representations: a difference in degree, not in kind. synthese, 198, 7647–7664. https://doi.org/10.1007/s11229-020-02537-y noë, a. (2004). action in perception. in mit press. mit press. o’brien, g., & opie, j. (2004). notes toward a structuralist theory of mental representation. in h. clapin, p. staines, & p. slezak (eds.), representation in mind (pp. 1–20). elsevier. https://doi.org/10.1016/b978-008044394-2/50004-x o’keefe, j. (1976). place units in the hippocampus of the freely moving rat. experimental neurology, 51(1), 78–109. https://doi.org/10.1016/0014-4886(76)90055-8 o’keefe, j., & dostrovsky, j. (1971). the hippocampus as a spatial map. preliminary evidence from unit activity in the freely-moving rat. brain research, 34(1), 171–175. https://doi.org/10.1016/0006-8993(71)90358-1 o’regan, j. k. (2011). why red doesn’t sound like a bell: understanding the feel of consciousness. oxford university press. orlandi, n. (2014). the innocent eye: why vision is not a cognitive process. oxford university press. piccinini, g. (2020). neurocognitive mechanisms: explaining biological cognition. oxford university press. https://doi.org/ 10.1093/oso/9780198866282.001.0001 piccinini, g. (2022). situated neural representations: solving the problems of content. frontiers in neurorobotics, 16, 846979. https://doi.org/10.3389/fnbot.2022.846979 ramsey, w. m. (2007). representation reconsidered. cambridge university press. https://doi.org/10.1017/cbo9780511597954 ramsey, w. m. (2016). untangling two questions about mental representation. new ideas in psychology, 40(a), 3–12. https://doi.org/10.1016/j.newideapsych.2015.01.004 ratcliff, r., & mckoon, g. (2008). the diffusion decision model: theory and data for two-choice decision tasks. neural computation, 20(4), 873–922. https://doi.org/10.1162/neco.2008.12-06-420 remington, e. d., narain, d., hosseini, e. a., & jazayeri, m. (2018). flexible sensorimotor computations through rapid reconfiguration of cortical dynamics. neuron, 98(5), 1005–1019. https://doi.org/10.1016/j.neuron.2018.05.020 rutar, d., wiese, w., & kwisthout, j. (2022). from representations in predictive processing to degrees of representational features. minds and machines, 32, 461–484. https://doi.org/10.1007/s11023-022-09599-6 ryle, g. (1949). the concept of mind. hutchinson. schöner, g. (2019). the dynamics of neural populations capture the laws of the mind. topics in cognitive science, 12, 1257– 1271. https://doi.org/10.1111/tops.12453 segundo-ortin, m., & hutto, d. d. (2019). similarity-based cognition: radical enactivism meets cognitive neuroscience. synthese, 1, 98, 5–23. https://doi.org/10.1007/s11229-019-02505-1 shagrir, o. (2012). structural representations and the brain. the british journal for the philosophy of science, 63(3), 519–545. https://doi.org/10.1093/bjps/axr038 shea, n. (2018). representation in cognitive science. oxford university press. https://doi.org/10.1093/oso/9780198812883.00 1.0001 sims, m., & pezzulo, g. (2021). modelling ourselves: what the free energy principle reveals about our implicit notions of representation. synthese, 199(3), 7801–7833. https://doi.org/10.1007/s11229-021-03140-5 spivey, m. j., grosjean, m., & knoblich, g. (2005). continuous attraction toward phonological competitors. proceedings of the national academy of sciences, 102(29), 10393–10398. https://doi.org/10.1073/pnas.0503903102 sprevak, m. (2011). representation reconsidered by william m. ramsey. british journal for the philosophy of science, 62(3), 669–675. https://doi.org/10.1093/bjps/axr022 steiner, p. (2014). enacting anti-representationalism. the scope and the limits of enactive critiques of representationalism. avant: trends in interdisciplinary studies, 5(2), 43–86. https://doi.org/10.26913/50202014.0109.0003 stephen, d., dixon, j., & isenhowever, r. (2009). dynamics of representational change: entropy, action, and cognition. journal of experimental psychology: human perception and performance, 35(6), 1811–1832. https://doi.org/10.1037/a0 014510 thomson, e., & piccinini, g. (2018). neural representations observed. minds and machines, 28(1), 191–235. https://doi.org/ 10.1007/s11023-018-9459-4 tolman, e. c. (1948). cognitive maps in rats and men. psychological review, 55(4), 189–208. https://doi.org/10.1037/h0061626 lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1111/j.1467-9280.2008.02041.x https://doi.org/10.1007/s11229-022-03891-9 https://doi.org/10.1007/s11229-013-0328-7 https://doi.org/10.1007/s11229-020-02537-y https://doi.org/10.1016/b978-008044394-2/50004-x https://doi.org/10.1016/0014-4886(76)90055-8 https://doi.org/10.1016/0006-8993(71)90358-1 https://doi.org/10.1093/oso/9780198866282.001.0001 https://doi.org/10.1093/oso/9780198866282.001.0001 https://doi.org/10.3389/fnbot.2022.846979 https://doi.org/10.1017/cbo9780511597954 https://doi.org/10.1016/j.newideapsych.2015.01.004 https://doi.org/10.1162/neco.2008.12-06-420 https://doi.org/10.1016/j.neuron.2018.05.020 https://doi.org/10.1007/s11023-022-09599-6 https://doi.org/10.1111/tops.12453 https://doi.org/10.1007/s11229-019-02505-1 https://doi.org/10.1093/bjps/axr038 https://doi.org/10.1093/oso/9780198812883.001.0001 https://doi.org/10.1093/oso/9780198812883.001.0001 https://doi.org/10.1007/s11229-021-03140-5 https://doi.org/10.1073/pnas.0503903102 https://doi.org/10.1093/bjps/axr022 https://doi.org/10.26913/50202014.0109.0003 https://doi.org/10.1037/a0014510 https://doi.org/10.1037/a0014510 https://doi.org/10.1007/s11023-018-9459-4 https://doi.org/10.1007/s11023-018-9459-4 https://doi.org/10.1037/h0061626 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the many problems with s-representation (and how to solve them) 31 varela, f. j., thompson, e., & rosch, e. (1991). the embodied mind: cognitive science and human experience. mit press. https://doi.org/10.7551/mitpress/6730.001.0001 wiese, w. (2017). what are the contents of representations in predictive processing? phenomenology and the cognitive sciences, 16(4), 715–736. https://doi.org/10.1007/s11097-016-9472-0 williams, d. (2017). predictive processing and the representation wars. minds and machines, 28(1), 141–172. https://doi.org/10.1007/s11023-017-9441-6 williams, d. (2018). pragmatism and the predictive mind. phenomenology and the cognitive sciences, 17(5), 835–859. https://doi.org/10.1007/s11097-017-9556-5 williams, d., & colling, l. (2018). from symbols to icons: the return of resemblance in the cognitive neuroscience revolution. synthese, 195, 1941–1967. https://doi.org/10.1007/s11229-017-1578-6 wittgenstein, l. (1953). the philosophical investigations. blackwell. zahnoun, f. (2021). the socio-normative nature of representation. adaptive behavior, 29(4), 417–429. https://doi.org/10.1 177/1059712320922364 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. lee, j., & calder, d. (2023). the many problems with s-representation (and how to solve them). philosophy and the mind sciences, 4, 8. https://doi.org/10.33735/phimisci.2023.9758 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.7551/mitpress/6730.001.0001 https://doi.org/10.1007/s11097-016-9472-0 https://doi.org/10.1007/s11023-017-9441-6 https://doi.org/10.1007/s11097-017-9556-5 https://doi.org/10.1007/s11229-017-1578-6 https://doi.org/10.1177/1059712320922364 https://doi.org/10.1177/1059712320922364 https://doi.org/10.33735/phimisci.2023.9758 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction the s-representation account conceptual objections a priori objections function objections content objections empirical objections the best theory objection the interpretation objection conclusion being for no-one: psychedelic experience and minimal subjectivity being for no-one psychedelic experience and minimal subjectivity chris lethebya (chris.letheby@uwa.edu.au) abstract can there be phenomenal consciousness without self-consciousness? strong intuitions and prominent theories of consciousness say “no”: experience requires minimal self-awareness, or “subjectivity”. this “subjectivity principle” (sp) faces apparent counterexamples in the form of anomalous mental states claimed to lack self-consciousness entirely, such as “inserted thoughts” in schizophrenia and certain mental states in depersonalization disorder (dpd). however, billon & kriegel (2015) have defended sp by arguing (inter alia) that while some of these mental states may be totally selfless, those states are not phenomenally conscious and thus do not constitute genuine counterexamples to sp. i argue that this defence cannot work in relation to certain experiences of ego dissolution induced by potent fast-acting serotonergic psychedelics. these mental states jointly instantiate the two features whose co-instantiation by a single mental state sp prohibits: (a) phenomenal consciousness and (b) total lack of self-consciousness. one possible objection is that these mental states may lack “me-ness” and “mineness” but cannot lack “for-me-ness”, a special inner awareness of mental states by the self. in response i propose a dilemma. for-me-ness can be defined either as containing a genuinely experiential component or as not. on the first horn, for-me-ness is clearly absent (i argue) from my counterexamples. on the second horn, for-me-ness has been defined in a way that conflicts with the claims and methods of its proponents, and the claim that phenomenally conscious mental states can totally lack self-consciousness has been conceded. i conclude with some reflections on the intuitive plausibility of sp in light of evidence from altered states. keywords 5-meo-dmt ∙ consciousness ∙ dmt ∙ for-me-ness ∙ hallucinogen ∙ psychedelic ∙ self ∙ self-awareness ∙ self-consciousness ∙ subjectivity this article is part of a special issue on “radical disruptions of self-consciousness”, edited by thomas metzinger and raphaël millière. auniversity of western australia, university of adelaide letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-6293-7873 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 2 1 introduction can there be phenomenal consciousness without any form of self-consciousness? philosophical intuitions and prominent theories of consciousness say “no”. according to what billon and kriegel call the subjectivity principle (sp), all phenomenally conscious mental states exhibit “subjectivity”, or minimal self-awareness. on this view, not only does every phenomenally conscious mental state belong to a subject, but every phenomenally conscious mental state is experienced as belonging to a subject (billon & kriegel, 2015, p. 29). sp faces apparent counterexamples in the form of anomalous mental states claimed to lack self-consciousness, such as “inserted thoughts” in schizophrenia and disowned mental states in depersonalization disorder (dpd). however, billon and kriegel suggest that, while some of these mental states may totally lack self-consciousness, those states also lack phenomenal consciousness and hence do not constitute genuine counterexamples to sp. the success of this manoeuvre is unclear; a reasonable person could either accept or reject billon and kriegel’s analysis of these cases. thus, the debate over the possibility of totally selfless phenomenal states seems to be at an impasse. here i attempt to break this impasse and advance the debate by introducing and analysing new cases. i argue that psychedelic drug research furnishes clearer counterexamples to sp than depersonalization or thought insertion, and that billon and kriegel’s consciousness* response (see section 2 below) cannot succeed in relation to these new cases. specifically, i argue that certain experiences of ego dissolution induced by the potent and fast-acting psychedelics dmt and 5-meo-dmt jointly instantiate the two features whose co-instantiation by a single mental state sp prohibits: (a) phenomenal consciousness and (b) total lack of self-consciousness or “subjectivity”. my grounds are simple: these mental states exhibit all the features that billon and kriegel deem sufficient for a verdict of total selflessness in other cases, while exhibiting none of the features on which they base their denial of phenomenal consciousness in such cases. one possible objection to my arguments is as follows: while my psychedelic counterexamples may lack “me-ness” (explicit awareness of the self) and “mineness” (explicit awareness of a mental state as owned by the self), they surely cannot lack “for-me-ness” (awareness of the mental state by the self, guillot, 2017). the objection thus asserts universalism about for-me-ness: the claim that all phenomenally conscious mental states, without exception, have this property (farrell & mcclelland, 2017). in response i propose a dilemma for universalists about for-me-ness. to defend universalism against my counterexamples, they need to define for-me-ness precisely. there are two options: either it contains a genuinely experiential component or it does not. if it does, then universalism about it ought to count as refuted by my counterexamples, on pain of unfalsifiability. if it does not, then for-meness has been defined in a way that conflicts with the explicit claims and dialectiletheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 3 cal strategies of its proponents, and the claim that phenomenally conscious mental states can totally lack self-consciousness has effectively been conceded. moreover, universalism about a deflationary, non-experiential for-me-ness faces other problems. notably, it is vulnerable to a debunking argument informed by evidence from altered states. we can satisfactorily explain the intuition that mental states are “given to” or “accessed by” selves or subjects, without assuming that any selves or subjects actually exist. 2 subjectivity theories and the unconsciousness defence in recent decades, philosophers and scientists have proposed many theories of phenomenal consciousness. one group of theories shares a common strategy: explaining how mental states become phenomenally conscious in terms of some relation of access or awareness between subjects and those mental states (o’brien & opie, 2015). billon & kriegel (2015) – henceforth b&k – call these “subjectivity theories” of consciousness. they list three prominent sub-types. acquaintance theories hold that mental states are phenomenally conscious just when their subjects stand to them in a sui generis relation of acquaintance. on this view, of all our mental states, the phenomenally conscious ones are those with which we are acquainted in the relevant way (levine, 2001). higher-order theories hold that mental states are phenomenally conscious just when they are meta-represented by a suitable kind of higher-order mental state (gennaro, 1996). and self-representation theories hold that mental states are phenomenally conscious just when, in addition to their other contents, they represent themselves as being mental states of the subject whose they are (kriegel, 2009). according to b&k, all three types are committed to the following subjectivity principle (sp): (sp) necessarily, a mental state m exhibits phenomenal consciousness only if m exhibits subjectivity. (billon & kriegel, 2015, p. 30) it is possible to question whether all of these theories are in fact committed to sp, as defined by b&k (millière, 2017).1 however, i will not pursue that exegetical question. rather, i will focus on the question whether sp itself is plausible; the answer will have consequences for any theory of consciousness that is in fact committed to sp. what, then, is “subjectivity”? discussing the experiences of drinking apple juice and drinking a banana smoothie, b&k note that these experiences are 1millière points out that the notion of “subjectivity” is ambiguous, and suggests that some subjectivity theories only posit a very minimal form of subjectivity as necessary to all conscious experiences. however, the subjectivity of billon & kriegel (2015) seems to be a richer affair, involving explicit awareness of the self or its ownership of a mental state. these distinctions are discussed in much more detail below. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 4 different in many respects […]. but there is also one respect in which [they] are exactly the same: in both cases it is for you that it is like something to have them. by this we mean not only that both experiences are yours, but more strongly that both are experienced as yours. we call this the subjectivity of experience. your apple-juice and banana-smoothie experiences are different in gustatory and tactile respects, but are the same in respect of subjectivity. (billon & kriegel, 2015, p. 29, italics original, bold mine) if we take this definition at face value, then subjectivity has both a metaphysical and an experiential component. to have subjectivity, a mental state m must both (a) be “yours” (i.e. belong to some subject), and (b) be experienced as “yours” (i.e. be experienced by that subject as belonging to that subject). this hybrid aspect of subjectivity will become important later. sp says that all phenomenally conscious mental states exhibit subjectivity; on b&k’s construal of the latter, this entails that all phenomenally conscious mental states are experienced by their subjects as belonging to, or being had by, those very subjects. thus, a phenomenally conscious mental state whose subject did not experience it as their own in any way would falsify sp (millière, 2017). b&k’s paper is devoted to the question whether psychopathological conditions furnish such counterexamples. they discuss two candidates: “inserted thoughts” in schizophrenia, and disowned mental states of patients with depersonalization disorder (dpd). in thought insertion, patients claim that thoughts which are not their own are occurring within their minds: thoughts are put into my mind, like “kill god”. it is just like my mind working, but it isn’t. they are not my thoughts. they belong to this guy, chris. they are his thoughts. (frith, 1992, p. 66) i look out the window and i think that the garden looks nice and the grass looks cool, but the thoughts of eamonn andrews [a famous television personality] come into my mind. there are no other thoughts there, only his […]. he treats my mind like a screen and flashes thoughts onto it like you flash a picture. (mellor, 1970, p. 17) it is easy to see the apparent problem for sp. these patients are describing mental states occurring within their mind; it seems that those mental states are phenomenally conscious; yet the patients deny that the mental states belong to them, that the thoughts are their thoughts. they report instead that the thoughts belong to some external agency: perhaps “this guy, chris” or the television personality eamonn andrews. it seems that there can be phenomenally conscious mental states lacking subjectivity. of course, matters are not so simple. one response on behalf of sp is to deny the intelligibility or coherence of the patients’ reports, which are extremely difficult for non-patients to understand. after all, these reports are made by people whose ordinary mental functioning is compromised. b&k do not want to take letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 5 this route, however. they accept (as do i) the methodological constraint that we ought to try to make sense of patients’ reports, and take them at face value absent some positive reason to do otherwise; and they note that in many cases of “alienation symptoms”, there is no independent evidence that the patients suffer from generalised irrationality or reasoning deficits. another option is what b&k call the subjectivity* response: denying that the mental states in question lack subjectivity in the sense relevant to sp. a standard version starts from the observation that patients suffering from thought insertion clearly do experience the thoughts as theirs, in the minimal sense of occurring within their minds. it is only because the condition involves experiencing thoughts as not one’s own but occurring within one’s mind that it is remarkable, theoretically puzzling, and troubling to patients. in light of this observation it has become commonplace to distinguish between the phenomenal senses of agency and ownership: the sense that one is the author of a mental state, and the sense that one is undergoing a mental state, respectively. inserted thoughts can then be explained as mental states that lack the sense of agency (prompting the delusional attribution to an external agent) while retaining the sense of ownership (billon & kriegel, 2015). assuming that the sense of ownership is the form of “subjectivity” relevant to sp, this principle can thus be preserved. b&k argue that such accounts of thought insertion cannot do justice to the phenomenological facts – specifically, cannot account for the phenomenal difference between inserted thoughts and other types of thoughts (such as intrusive thoughts in obsessive-compulsive disorder) which arguably lack the sense of agency. instead, b&k question the idea that thought insertion results from the subtraction of some phenomenal feature possessed by non-inserted thoughts. they propose instead that inserted thoughts have an extra phenomenal feature which non-inserted thoughts lack: a positive sense of alienation or phenomenology of insertion. they note that this hypothesis is consistent with the reports of patients, who complain about some positive, extra feature of their thoughts which is distressing to them: the feature of being (or seeming to be) authored by an external agent. however, b&k do not think this “something extra” response will work for the alienation symptoms of depersonalization disorder (dpd), in which many patients emphatically complain of something missing from their experience. dpd is characterized by a persistent and distressing feeling that the self is unreal or non-existent; thoughts, actions, and (to an extent) affective responses persist, but the usual sense that these are happening to me (i.e. to the subject) is drastically reduced or altogether absent. patients describe feeling detached or estranged from the self and its states: i feel some degree of “out of it” all the time […] i can sit looking at my foot or my hand and not feel like they are mine. this can happen when i am writing, my hand is just writing, but i’m not telling it to. it almost feels like i have died, but no one has thought to tell me. so, i’m left living in a shell that i don’t recognize any more (sierra, 2009, p. 27; quoted from billon, 2017, p. 739) letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 6 i suddenly wonder: is it really me here? is it really me walking? then i make enormous efforts in order to apply my consciousness to this unconsciousness […] in order to realize that i am making the walking movements. so at some point during this kind of crisis, before the absolute certainty [of being myself, before the crisis] i am conscious on one side that i am unconscious [sic] on the other side. (séglas & meige, 1895; quotation from billon & kriegel, 2015, p. 45, interpolations original) according to b&k, none of the standard responses on behalf of sp is clearly applicable to such mental states as these. there are no independent grounds for deeming these patients irrational or their reports unintelligible; their explicit insistence that something is missing from their experience precludes a “something extra” response; and their emphatic denial of any form of subjectivity problematizes other types of subjectivity* response. however, b&k propose an ingenious strategy which they call the consciousness* response: they concede that the mental states in question totally lack subjectivity, in the sense relevant to sp, but question whether these mental states are conscious in the sense relevant to sp2 (cf. billon, 2013, 2015). sp precludes the possibility of phenomenally conscious mental states lacking subjectivity. phenomenally conscious mental states are just those that have qualitative feel; in nagel’s (1974) famous phrase, they are states that there is something it is like to be in. b&k suggest that perhaps the totally subjectivity-free mental states of dpd patients are not phenomenally conscious; there is nothing it is like for the patients to have these mental states. at first blush, this seems absurd: how could patients describe these states if there was nothing that it felt like for the patients to be in those states? however, as b&k point out, the ability to report on a mental state does not in fact imply that it is phenomenally conscious. it only implies that the state is access conscious, in block’s (1995) sense: that it is available for verbal report and the guidance of behaviour. phenomenal consciousness and access consciousness are conceptually dissociable: we can understand what it would mean for a mental state to have some qualitative feel or character, despite being so fleeting and evanescent that it could not guide behaviour or verbal report, and we can understand what it would mean for a mental state to be capable of guiding verbalization and behaviour causally or computationally, despite lacking qualitative character. some authors have ar2strictly speaking, b&k do not commit themselves to any views about which response works best for which putative counterexample. as they put it, they simply present a “menu of options” for those who wish to defend sp against pathological counterexamples. however, they make clear that they see the consciousness* response as the most plausible option – more plausible than a “something extra” response, or other subjectivity* responses – for the specific dpd cases that they discuss. this is because patients’ reports of these cases are difficult to interpret in terms of “something extra” rather than something missing, or in terms of some form of subjectivity (e.g. the sense of ownership) being retained. these features are shared (i argue) by my new counterexamples below. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 7 gued that phenomenal and access consciousness in fact dissociate; notably, block (2011) holds that perceptual consciousness contains phenomenal qualities which “overflow” the limits of cognitive access. these issues are controversial. but b&k’s point stands: the fact that dpd patients can report on their mental states does not unproblematically entail that these states are phenomenally conscious; it may be a case of access consciousness without phenomenal consciousness. and, they argue, to the extent that we have good reason to believe sp, this is exactly what we should conclude is occurring. moreover, according to b&k, the consciousness* response is not an ad hoc manoeuvre with nothing to recommend it beyond the preservation of a favoured theory. they quote descriptions from dpd patients which seem to hint at the idea that their disowned mental states indeed lack phenomenal consciousness: i just sink into a kind of unconsciousness. i am just conscious enough to know that things are going on around me but nothing seems to register. i’m like a zombie unable to take in any information. it’s the mental sensibility that is lacking, it is not me who feels. i have no interest in what i appear to be feeling. it is someone else who feels mechanically. (quotations from billon & kriegel, 2015, emphasis theirs) b&k are not claiming that dpd patients totally lack phenomenal consciousness – merely that the specific mental states they describe as lacking subjectivity also lack phenomenality. one possibility is that the disowned mental states, which are themselves phenomenally unconscious, are meta-represented by phenomenally conscious higher-order representations that inform verbal reports. but the phenomenal/access consciousness distinction also raises the possibility that disowned states could influence verbal report “directly”, even while not being phenomenally conscious. in short, b&k argue that neither inserted thoughts nor disowned mental states in dpd constitute conclusive counterexamples to sp. according to their preferred responses, inserted thoughts do not lack subjectivity, but rather have “something extra” – a positive phenomenology of alienation – while disowned mental states in dpd may lack subjectivity, but if so, are not phenomenally conscious, and therefore do not falsify sp. 3 psychedelic mystics are not zombies even in relation to their dpd cases, the plausibility of b&k’s consciousness* response is questionable. the reports from dpd patients quoted above do not unequivocally support a denial of phenomenal consciousness. b&k highlight one patient’s description of themselves as “like a zombie”, which evokes the notion of philosophical zombies: behavioural and functional duplicates of ordinary humans, letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 8 bereft of phenomenal consciousness (chalmers, 1996). however, as b&k themselves emphasise, we should be wary of interpreting patients’ reports through the lens of specialised philosophical and scientific jargon that these patients could not be expected to command. further, use of the language of “feeling” – as in “it is not me who feels… it is someone else who feels mechanically” – can just as readily be interpreted as describing states which are phenomenally conscious – there is something that they feel like – but lack subjectivity, since it is (seemingly) not for the self or “i” that they feel like something. however, the various treatments of thought insertion and dpd offered by b&k have some plausibility; a reasonable person might conclude that they defuse the putative counterexamples to sp. personally, i am agnostic regarding the status of the specific examples discussed by b&k. it is not obvious whether these are genuine cases of phenomenally conscious mental states that totally lack subjectivity. the debate seems to be at an impasse. one way to proceed would be to analyse b&k’s examples more closely, to determine (for instance) whether the consciousness* response really does offer an adequate treatment of disowned mental states in their dpd cases. here i take a different approach: i attempt to advance the debate by introducing new cases. there are other unusual mental states which, in my view, furnish clearer counterexamples to sp than the thought insertion and dpd cases discussed by b&k. these states exemplify all the features on which b&k base their verdict that subjectivity is absent from dpd states: the subjects who undergo these states explicitly describe something missing from their experience, precluding a “something extra” response, and support for intelligibility-denial is lacking, since there is no evidence that these subjects suffer from any deficit in rationality or that their reports are incoherent. indeed, since these cases involve mentally healthy subjects who undergo temporary consciousness alterations voluntarily, and report them retrospectively, the untenability of intelligibility-denial is even clearer than in dpd. however, descriptions of the mental states in question, unlike those of b&k’s dpd cases, provide no support for a denial of phenomenal consciousness.3 indeed, these mental states are typically described as being vividly and memorably conscious (cf. millière, 2017). the states in question are experiences of total ego dissolution induced by injecting or smoking the potent fast-acting serotonergic psychedelic drugs n,n-dimethyltryptamine (dmt) and 5-methoxy-n,n-dimethyltryptamine (5-meo-dmt).4 3my argument does not rely on the claim that dpd states in general are categorically distinct from the types of psychedelic states i discuss. one might worry that there is no sustainable contrast because the psychedelic states are just transient dpd-type episodes. (i am indebted to an anonymous referee for raising this worry.) however, even if this is so, the argument still goes through, as long as the states i describe (a) are equal to b&k’s dpd cases in terms of evidence that subjectivity is lacking, and (b) relevantly different to b&k’s cases in terms of lacking evidence for consciousness-denial. in what follows, i try to show that these conditions are met. 4i believe that my arguments also apply to some experiences of “selflessness” or “emptiness” induced voluntarily by expert practitioners of vipassana meditation; see, e.g., dor-ziderman, letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 9 dmt and 5-meo-dmt are naturally occurring molecules. when smoked or injected in suitable doses they induce a brief but intense altered state of consciousness (sessa, 2012). like other “classic” psychedelics such as lsd and psilocybin, they are believed to alter consciousness primarily via 5-ht2a receptor agonism (nichols, 2016). and like other classic psychedelics, their phenomenological effects are complex and variable, being influenced by the individual’s psychological state and the external environment (“set and setting”, zinberg, 1984). dmt and 5-meo dmt can induce dramatic changes to perception, affect, thinking, and the senses of space, time, and embodiment. most relevantly, at high doses they can occasion experiences described as a total loss or dissolution of the sense of self or ego. here is a research subject administered intravenous dmt: i immediately saw a bright yellow-white light directly in front of me […]. i was consumed by it and became part of it. there were no distinctions – no figures or lines, shadows or outlines. there was no body or anything inside or outside. i was devoid of self, of thought, of time, of space, of a sense of separateness or ego, or of anything but the white light. there are no symbols in my language that can begin to describe that sense of pure being, oneness, and ecstasy. there was a great sense of stillness and ecstasy. (strassman, 2001, pp. 244–245) taken at face value, this report describes a phenomenally conscious mental state completely lacking subjectivity; during the peak of the episode, this volunteer’s phenomenal field seems to have been occupied entirely by visual sensations of brilliant white light, plus feelings of “pure being, oneness, and ecstasy”. such experiences of total ego dissolution are reported relatively uncommonly by users of dmt (strassman, 2001). however, total ego dissolution seems to be a more typical effect of 5-meo-dmt. users of 5-meo-dmt frequently describe an experience of “emptiness,” “nothingness” or “void” which is associated with a cessation of thoughts, extreme sensory deprivation and a complete loss of self-consciousness: […] “i felt that there was nothing to me and there was nothing around me” […]; “the reality around me disintegrated into nothing. i fell into a void [that] i can’t even describe” […]; “i wasn’t anything anymore. i had been broken down into nothingness, into oblivion” […]; “my thoughts ceased to exist, and my senses shut off completely. i could not hear, see, smell, taste or feel anything” […]. (millière et al., 2018, p. 16; for further discussion, see millière, 2020, this issue) a clear and eloquent description of total ego dissolution induced by 5-meo-dmt is provided by michael pollan: berkovich-ohana, glicksohn, & goldstein (2013) and ataria, dor-ziderman, & berkovich-ohana (2015) and, for discussion, millière, carhart-harris, roseman, trautwein, & berkovich-ohana (2018). letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 10 i felt a tremendous rush of energy fill my head […]. terror seized me – and then, like one of those flimsy wooden houses erected on bikini atoll to be blown up in the nuclear tests, “i” was no more, blasted to a confetti cloud by an explosive force. i could no longer locate [myself] in my head, because it had exploded that too, expanding to become all that there was. whatever this was, it was not a hallucination. a hallucination implies a reality and a point of reference and an entity to have it. none of those things remained.5 (pollan, 2018, pp. 276–277) in pollan’s case and that of the dmt subject, we are dealing with mentally healthy volunteers providing lucid and coherent descriptions of utterly unusual states of consciousness. there are no grounds for concluding that these subjects are afflicted by any deficit of rationality at the time of making their retrospective phenomenological reports, nor for deeming these reports incoherent or unintelligible. indeed, despite the longstanding conception of psychedelics as “psychotomimetic” (psychosis-mimicking) drugs (sessa, 2012), a growing body of evidence suggests that psychedelic experiences are associated with improved mental health outcomes (johansen & krebs, 2015), especially in controlled conditions (santos et al., 2016). it is true that acute psychedelic intoxication involves increased suggestibility and tendencies towards magical thinking (carhart-harris, 2013; carhart-harris et al., 2015). however, numerous studies of regular religious users of serotonergic psychedelics find that, in the sober state – which is when such retrospective experiential descriptions are given – their neuropsychological functioning and mental health status is comparable to that of matched controls (barbosa, mizumoto, bogenschutz, & strassman, 2012; barbosa et al., 2016; bouso et al., 2012, 2015). a different reason for mistrusting these descriptions might be that psychedelic experiences are transient and notoriously difficult to describe. how can we be confident that these retrospective reports really reflect the phenomenological reality of the intoxication, rather than subsequent confabulation? a straightforward answer comes from the fact that psychedelic ego dissolution seems to have distinctive neural correlates, in the form of modulation and disintegration of the default mode and salience networks – both neural systems independently implicated in generating the phenomenal sense of self, and modulated by contemplative practices that aim to alter the sense of self (letheby & gerrans, 2017). in multiple studies, ego dissolution under psychedelics has been linked to diminution or modulation of known neural signatures of self-consciousness,6 suggesting that volunteers’ reports reflect a phenomenological reality and are not objectionably theory-contaminated (lebedev et al., 2015; tagliazucchi et al., 2016). 5i am grateful to raphaël millière for bringing the significance of this report to my attention. 6of course, the neurocognitive mechanisms of self-awareness remain controversial, so this argument cannot be regarded as conclusive. however, the fact that systems consistently disrupted by psychedelics have repeatedly been linked to self-awareness by non-psychedelic evidence must carry some weight against the theory-ladenness objection. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 11 of course, it would be logically consistent for a proponent of sp to deny that the mental states being described by these volunteers are phenomenally conscious. a consciousness* response to these cases is not impossible. however, it is implausible and evidentially unwarranted. the only positive evidence supporting such an analysis of b&k’s dpd cases is the fact that patients use relevantly suggestive language, such as talk of being a “zombie” or of “unconsciousness”. no such language is used by these psychedelic subjects, who are very clear that they are describing intense and memorable experiences. there is no positive evidence that these mental states lack phenomenal consciousness. the only grounds for such a conclusion would be an antecedent theoretical conviction that they must lack it, if they lack subjectivity. at this point it might be objected that there are, in fact, positive grounds for thinking that these psychedelic volunteers lost phenomenal consciousness: they describe losing all the mental contents that typically characterise phenomenally conscious states. perhaps, the thought goes, their remarkable experiences amount to the disintegration and subsequent reintegration of the ego or self-model – the transition phases in and out of egolessness – while the “peak” of the experience, totally lacking self-consciousness, is a window of phenomenal unconsciousness.7 this analysis does not seem plausible for strassman’s dmt subject, who emphasises the phenomenal presence of white light, pure being, oneness, and ecstasy, despite the phenomenal absence of self, thought, time, space, and so forth. however, it has more initial plausibility in relation to pollan’s experience, and that of the other 5-meo-dmt subjects quoted by millière et al. (2018), who tend rather to describe their experiences in negative terms, speaking of sensory deprivation, nothingness, or a “void [that] i can’t even describe”. the inference is unwarranted, however. even if (some of) these subjects lost all ordinary phenomenal contents, this does not entail that they lost phenomenal consciousness altogether, unless one assumes that there can be no states of phenomenal consciousness that lack those contents. but that would be questionbegging in the present context. indeed, 5-meo-dmt subjects typically insist that their experiences are remarkable precisely in virtue of (a) lacking all the contents of ordinary experiences but (b) being conscious experiences nonetheless. moreover, some subjects describe an experience of pure bodiless, egoless, non-spatial, non-temporal consciousness that features affective phenomenal content. for example, the subject quoted by millière et al. (2018) as saying that “my thoughts ceased to exist, and my senses shut off completely. i could not hear, see, smell, taste or feel anything” goes on to say: when i say, “i couldn’t see”, you shouldn’t think that i was seeing blackness as a symbol of nothingness. i truly saw nothing. i ceased to have any memory of that sense […]. fortunately, despite the extreme sensory deprivation, i was still having a great experience […]. 7i am grateful to thomas metzinger, raphaël millière, and an anonymous referee for pressing different versions or aspects of this objection. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 12 i entered a state of infinite bliss that was so incredible that it seemed totally impossible […]. the joy and ecstacy [sic] i felt was absolutely infinite. (report #37301 from www.erowid.org) likewise, pollan goes on to say: unfortunately, the terror didn’t disappear with the extinction of my “i”. whatever allowed me to register this experience, the post-egoic awareness i’d first experienced [while intoxicated] on [psilocybin] mushrooms, was now consumed in the flames of terror too. in fact every touchstone that tells us “i exist” was annihilated, and yet i remained conscious. “is this what death feels like? could this be it?” that was the thought, though there was no longer a thinker to have it. (pollan, 2018, p. 277) despite reporting the loss of all sense of self or ego, these subjects are clearly describing conscious experiences; in the first case, an experience of infinite bliss, joy, and ecstasy, and in pollan’s less fortunate case, an experience of “post-egoic” terror. these descriptions do not suggest a temporary loss of phenomenal consciousness. there are no grounds, either, for conceptualizing these states of consciousness as containing “something extra” vis a vis subjectivity, at least not by b&k’s lights: these subjects are at least as explicit as the dpd patients quoted above in describing something missing from their experience8 (“devoid of self [and] a sense of separateness or ego”; “[none] of those things remained”). my conclusion is that these expe8an alternative interpretation of reports like pollan’s is that they describe not ego dissolution, but ego expansion. language such as “expanding to become all that there was” might be taken to suggest that in such states, the felt sense of self does not disappear, but expands to become coextensive with the entire phenomenal field (cf. fink, 2020, this issue). on this view, when subjects describe a “total loss” of self or ego, this is not a theory-neutral phenomenological observation; it is a description which is inferred in dependence on the contentious assumption that a phenomenal sense of self requires a phenomenal self-world distinction. the suggestion is that subjects experience the loss of the phenomenal self-world distinction, or of ego-boundaries, but then engage in a theoretically loaded description of this as a total loss of self or ego. i find it more plausible that the descriptions of ego loss represent a genuine phenomenological reality. it is not clear why, for instance, pollan would say that there was “no longer a thinker to have [the thought]”, if there was still a (boundaryless) sense of self or ego in his experience. however, i am indebted to an anonymous referee for insisting on the difficulty of deciding conclusively between these two interpretations. it is fair to say that more research is required on this point, possibly utilising new psychometric instruments or techniques such as “microphenomenological” interviewing (petitmengin, 2006; cf. millière, 2017; sebastián, 2020, this issue). interestingly, an analogous point can be made about the use of subjective-seeming language in retrospective descriptions, as when pollan says: “every touchstone that tells us ‘i exist’ was annihilated, and yet i remained conscious”. the second “i” might be taken as betraying a phenomenological fact – that some minimal form of self-awareness or subjectivity remained – or merely as reflecting longstanding and deep-seated linguistic and conceptual habits that make it very difficult, in the sober state, to speak or think of consciousness without speaking or thinking of someone who was conscious. this is discussed in more detail in the final section of the present paper. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 http://www.erowid.org https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 13 riences constitute counterexamples to sp: phenomenally conscious mental states that completely lack subjectivity. at least, that is what we should conclude if we accept b&k’s analyses of their cases. 4 the universalist objection i have claimed that there are no grounds for a subjectivity* response in the psychedelic cases since in these cases (unlike in thought insertion) subjects claim explicitly that the relevant mental states lack all forms of phenomenal subjectivity. b&k seem to take a similar view of the dpd states they discuss: that a subjectivity* response is not very plausible, since patients claim quite explicitly that the relevant mental states are totally bereft of all phenomenal subjectivity. they reject agencyand endorsement-based versions of the subjectivity* response as inadequate to thought insertion, and hence adopt the “something extra” response for this condition. rejecting the “something extra” response as inadequate to dpd, with its explicit descriptions of “something missing”, they turn next to the consciousness* response without considering other subjectivity* responses. this suggests that they judge agency, endorsement, and other subjectivity* responses also inadequate to their dpd cases. if one rejects all subjectivity* responses for b&k’s dpd cases, then one ought to reject them for the psychedelic cases i have discussed. the descriptions of these cases as totally selfless are at least as clear as the analogous descriptions of the dpd cases. nonetheless, there is a possible version of the subjectivity* response that might be made. this response starts from the observation, formulated clearly by guillot (2017), that terminological ambiguities plague recent discussions of “subjectivity” and “minimal self-awareness”. guillot distinguishes between three varieties of minimal self-awareness. one, which she calls “me-ness”, is an explicit, “egological” form of self-awareness – the self or subject is encountered as a content of consciousness among others. another, which she calls “mineness”, is an explicit awareness of the ownership relation between the self and its mental states – a phenomenal sense of a mental state as being owned by me. finally, what guillot calls “for-me-ness” is described as a “special inner awareness” of the mental states themselves, by the subject. although zahavi and others do characterise for-me-ness as a form of self-awareness or subjectivity – albeit a non-egological form – it does not therefore amount to an explicit, conscious awareness of the self. rather it consists in the self or subject’s awareness of its mental states – which implies a self or subject who is thus aware, thereby constituting an implicit form of self-awareness. for-me-ness is sometimes characterised as the “first personal givenness” of conscious mental states, or as the “special inner awareness” that each subject has of her own conscious states – or, as zahavi and kriegel (henceforth z&k) put it: letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 14 the “me” of for-me-ness is not in the first instance an aspect of what is experienced but of how it is experienced; not an object of experience, but a constitutive manner of experiencing. to deny that such a feature is present in our experiential life, to deny the for-me-ness or mineness [sic] of experience, is to fail to recognize the very subjectivity of experience […] once anything occurs consciously, it must be given to the subject and thus exhibit for-me-ness. in other words, the “me” of for-me-ness is not a separate and distinct item but rather a pervasive feature of experiential life as such. (zahavi & kriegel, 2015, p. 38, italics original, bold mine) in short, for-me-ness is awareness of mental states (by the subject), me-ness is awareness of the self or subject, and mineness is awareness of the ownership relation between them. as farrell & mcclelland (2017) point out, guillot’s tripartite distinction gives rise to nine target theses. concerning any of the three varieties of subjectivity, one might be a universalist, claiming that it is present in all conscious experiences without exception; a typicalist, claiming that it is present in typical conscious experiences but absent from at least some atypical ones; or an absentist, claiming that it is absent from all conscious experiences.9 within this framework, a defender of sp might offer the following subjectivity* response to my psychedelic cases: granted, your cases refute universalism about me-ness and mineness, since they present clear cases of phenomenally conscious mental states which lack both of these features. however, these cases do not refute universalism about for-me-ness. nothing in the descriptions of these cases indicates that the subjects lack a special, inner awareness of the remarkable mental states they undergo, or that these states are not given to them in a distinctly first-personal way. indeed, they must have such an awareness, otherwise why would it be a remarkable episode in their biography qua conscious creature to undergo these states? how would they know what these states are like without being aware of these states in a way that is special, inner, and direct – a way that differs profoundly from the way in which you or i are aware of their psychedelic experiences? moreover, for-me-ness is the sense of “subjectivity” that is relevant to sp. therefore, while the psychedelic states may lack subjectivity*, i.e. me-ness and mineness, they do not lack subjectivity in the sense relevant to sp, i.e. for-me-ness, and thus do not refute sp. call this the universalist objection to my psychedelic case against sp. it argues that (i) for-me-ness is the sense of subjectivity relevant to sp, and (ii) my examples do not lack for-me-ness, so (iii) my examples do not refute sp.10 the universalist objection has some initial plausibility. in particular, when we look at the examples that b&k give of subjectivity theories – acquaintance theory, higher-order representationalism, and self-representationalism – it seems correct that the relevant sense of subjectivity is more akin to for-me-ness than to me-ness 9strictly speaking, these do not exhaust the possibilities, since one might hold (for instance) that one of the three varieties is present only in atypical, and absent from typical, experiences. 10for a distinct but closely related argument, see sebastián (2020), this issue. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 15 or mineness (millière, 2017). these theories do not attempt to explain phenomenal consciousness in terms of awareness of the self, but in terms of awareness of mental states by the self. in different ways, they each unpack the intuition that conscious mental states are ones that the self or subject is aware of, that are somehow present to or for the subject (levine, 2001, pp. 8–9; o’brien & opie, 2015). thus, it seems correct that the sense of “subjectivity” relevant to sp is for-meness rather than me-ness or mineness. granted this first premise, the success of the universalist objection depends on the second premise that my case studies, even if they refute universalism about me-ness and mineness, do not refute universalism about for-me-ness. to assess this claim, we need to get clearer on what exactly for-me-ness is – and it is a notoriously elusive notion. at a first pass, the proponent of the universalist objection confronts a dilemma. recall that subjectivity, as b&k define it, incorporates both a metaphysical and an experiential component. they say that a mental state has subjectivity when it “not only [is] yours [the subject’s] but [also is] experienced as yours [i.e. experienced by the subject as the subject’s]”. taken literally, this defines subjectivity as a conjunctive property consisting in the co-occurrence of (a) an objective ownership relation between subject and mental state and (b) a phenomenal awareness of that ownership relation by the subject. this would seem to make subjectivity something like a combination of guillot’s (phenomenal) mineness with an objective ownership relation. my purpose is not to fixate unduly on the idiosyncrasies of specific formulations. however, to assess the claim that universalism about for-me-ness survives the putative psychedelic counterexamples, we need to know precisely what sort of thing for-me-ness is supposed to be. specifically, we need to know: is it an experiential, or phenomenal, feature – something that makes a positive difference to what an experience is like? z&k call this a “non-deflationary” construal of for-meness. or is rather it a metaphysical/epistemic feature, consisting in some purely objective relation between a subject and a mental state; for instance, the bare fact that a subject is aware of, or accesses, a given mental state in a way unavailable to no other subject? this, they call a “deflationary” construal. for the purposes of my argument, i will locate conjunctive or hybrid construals of for-me-ness, which define it as a combination of experiential and metaphysical features, on the first horn of the dilemma. thus, the dichotomy is between a non-deflationary (wholly or partly experiential) and a deflationary (wholly non-experiential) construal of for-me-ness. i take the horns in turn. suppose the presence of for-me-ness necessarily involves some specific experiential feature. in this case, it is hard to see what could constitute a clearer counterexample to universalism about it than the psychedelic cases above. descriptions of 5-meo-dmt experiences such as pollan’s are so emphatic in their rejection of “a reality and a point of reference and an entity” that there seems to be no possible phenomenal feature in virtue of which they could instantiate for-me-ness. a universalism about any non-deflationary construal of for-me-ness that does not letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 16 count as falsified by an experience like this runs the risk of becoming unfalsifiable, a declaration by fiat that any phenomenally conscious mental state ipso facto possesses for-me-ness. but “for-me-ness” cannot be merely a synonym for “phenomenal consciousness” – else, whence the controversies over whether the former is a necessary condition of the latter?11 suppose, on the other hand, that for-me-ness is (per the deflationary interpretation) a merely metaphysical/epistemic, in any event non-experiential, property. several problems arise for this supposition. the first is that it is clearly at odds with how prominent advocates of for-me-ness define the term. z&k explicitly adopt a non-deflationary, experiential interpretation, and b&k’s definition of “subjectivity” includes an experiential component. second, the methodological practice of defenders of sp shows that they are concerned with an experiential feature. by conducting detailed analyses of the phenomenology of pathological mental states, b&k and, to a lesser extent, z&k demonstrate that the kind of subjectivity relevant to their interpretation of sp is one that makes a genuinely experiential difference. third, to adopt a deflationary reading of for-me-ness is to change the subject (as it were.) the psychedelic cases were meant to establish that there can be phenomenal consciousness without any kind of self-consciousness. this conclusion stands even if all phenomenally conscious mental states are, as a matter of objective fact, accessed or undergone by some subject. finally, universalism about a deflationary for-me-ness is independently suspect, on theoretical grounds that i will mention briefly later. the second horn – a deflationary interpretation of for-me-ness – is thus a nonstarter. let us return for a closer look at the first horn. how is a universalist likely to respond to my insistence that these psychedelic cases clearly lack any conceivable non-deflationary (phenomenal) form of for-me-ness? as a counterexample to sp, lane (2012) presents the case of patient dp, who experienced visual perceptions allegedly lacking any sense of subjectivity, and could only establish that the visual perceptions were his own via a subsequent process of explicit inference. z&k argue that, despite lacking many more robust forms of self-consciousness, dp’s experience retains for-me-ness: 11a version of universalism which may not fit easily into this taxonomy is that proposed by kriegel (2009): that subjectivity is a determinable of which specific phenomenal characters are determinates. on this view, the relation of for-me-ness to other phenomenal qualities would be akin to the relation of having colour experience to having red experience, having blue experience, and so forth. (i am indebted to an anonymous referee for pointing this out.) i lack the space to explore this possibility here, but will simply note that, if for-me-ness or subjectivity is to be more than a mere synonym for “phenomenal consciousness”, then there must be some possibility of debating whether all phenomenal qualities are in fact determinates of this specific determinable, which is related somehow to self or subjectivity; and in this case, the arguments i offer on each horn of the dilemma would seem to be clearly relevant. if the determinable/determinate view amounts to more than “all phenomenal qualities are determinates of the determinable phenomenally conscious”, then the question whether all phenomenal states are either (a) phenomenally structured in a first-personal way or (b) metaphysically given to a self or subject, must be relevant to its truth. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 17 consider the following variety of overall conscious experiences: being absorbed in a movie; laboriously trying to decipher a menu written in a language you barely know; being suddenly hit in the face by a snowball; being humiliated by your peers; standing on the ten-meter diving board, trying to convince yourself to jump. in addition to the various items such experiences present, they also differ phenomenally with respect to the kind of self-consciousness they instantiate. when comparing such experiences, it should be evident that self-consciousness can vary quite a bit along a spectrum in its experiential acuity or intensity. the kind of experiential for-me-ness we have in mind is a sort of minimum point of self-consciousness. this minimal self-consciousness is present in dp’s experience in the same way it is present in thoughtinsertion patients. (zahavi & kriegel, 2015) the first point to make here is that, if dp’s experience really manifests only the minimum point of self-consciousness, then it should be conceded that pollan’s 5-meodmt experience lacks self-consciousness altogether. the reason is that it seems quite clear that pollan’s experience manifests less in the way of self-consciousness than dp’s.12 dp was still able, via an inferential effort, to establish the existence of a ownership relation between his visual experiences and their subject while these experiences were occurring. in pollan’s case, no phenomenological trace remained of a reality, an entity, or a point of reference; at the peak of his experience, there was no sense, idea, or conception of a subject to whom the experience could possibly belong (even though he was able to infer his ownership of the experiences retrospectively, once the episode of selfless phenomenality had ended.) the second point i want to make connects the foregoing observations about pollan’s experience with an important argument made by z&k. we can call this the descriptive indispensability argument for the reality and ubiquity of for-meness. it responds to a common criticism: that the notion of for-me-ness does no explanatory work, or at least none that cannot be done better by other explanatory posits. z&k deny this, but they also argue that, even if it is true, belief in for-me-ness is still warranted. on what grounds? explanatory adequacy, they say, is not the only reason for believing in an entity or property; descriptive adequacy also matters. in order to try to explain the various properties of phenomenal consciousness, we must first be able to describe them, and “it is impossible to correctly describe the structure of phenomenal consciousness without citing for-me-ness” (zahavi & kriegel, 2015, p. 45). up to a point, z&k are entirely correct. i am no absentist about a nondeflationary, purely experiential construal of for-me-ness – the notion, or something like it, is indeed required to correctly describe the structure of the vast majority of phenomenal states (cf. metzinger, 2003). but the descriptive 12though see millière (2019) for a sceptical perspective on the idea that self-consciousness comes in degrees. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 18 indispensability point cuts both ways. it is true that we cannot adequately and completely describe the structure of my sober experiences of smelling coffee or seeing a rose without mentioning the presence of for-me-ness. nor, however, can we adequately and completely describe the structure of pollan’s 5-meo-dmt experience without mentioning the absence of for-me-ness. if the descriptive indispensability of for-me-ness establishes its existence, then the descriptive indispensability of its negation refutes its universality. to underscore the point, consider how z&k distinguish for-me-ness from other, less controversial phenomenal features: experiential for-me-ness is not a quality or datum of experience […]. [it] is not that in addition to the objects in one’s experiential field […] there is also a self-object. rather the point is that each of these objects, when experienced, is given to one in a distinctly first-personal way, and that this givenness is a pervasive dimension of phenomenal life […]; one grasps such experiential elements as lemon-qualia and mintqualia by appreciating what varies across such phenomenal characters, but grasps what for-me-ness is by appreciating what remains constant across them. (zahavi & kriegel, 2015, p. 38) quite – and one can grasp an experience like pollan’s in its utter uniqueness only by appreciating that it lacks what remains constant across nearly all conscious experiences, including those of inserted thoughts: the apparent fact (the seeming) that experiential objects are “given to [some]one in a distinctly first-personal way”. it is hard to see how this phrase can have any phenomenological content whatsoever without describing precisely that feature of virtually all conscious experiences whose absence is definitive of experiences like pollan’s. even if there is some metaphysical fact that the experience was “given to him” – the objective subject – the single feature that makes the experience unutterably strange is that it was “given”13 in a distinctly non-first-personal way. similar points apply to the claim that “the ‘me’ of for-me-ness is not in the first instance an aspect of what is experienced but of how it is experienced” (zahavi & kriegel, 2015, p. 38). the unusual what of pollan’s experience is the void; the unusual how is the total absence of an apparent entity or perspective to which this void is manifest. as z&k say, it is significant that thought insertion is naturally described in terms of alienated self-consciousness. it is equally significant that pollan’s experience is not. pollan’s description speaks to a fundamental point about self-representation: 13the scare quotes here are intended to indicate that i do not accept the antecedent; i do not think there is any real sense in which experiences are given. for them to be given, there must be someone to whom they are given – but as we will see in the next section, i think there is no such entity. i think that our deep-seated intuition that experiences are “given” is simply an artefact of a mental modelling process – a result of the fact that, in virtually all experiences, the brain represents the existence of an entity to whom the experience is happening. of course i accept that experiences happen spatially within organisms, but this is not the same as accepting that they are given to an experiencing subject in the relevant sense. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 19 that the self is modelled by the brain as a knowing subject, an entity that represents – and thus potentially misrepresents – a distinct world. this is why it is significant that pollan says “[this] was not a hallucination”, on the grounds that a hallucination requires “an entity and a point of reference”. to experience a mental state as a hallucination requires the sense of an entity standing in representational and epistemic relations of truth and error, accuracy and inaccuracy, to a world external to itself – what metzinger (2003) calls the “phenomenal model of the intentionality relation” (pmir). pollan’s denial that his experience was hallucinatory is of a piece with the “noetic quality” – the sense of immediate, direct, undeniable knowledge that is a hallmark of (psychedelic) mystical experiences – in that both can be explained by the absence of the phenomenal self-model (psm) and concomitant pmir. it is precisely because pollan’s experience lacks the basic, near-ubiquitous, usually invisible sense of being an experiencing, knowing entity – a “point of reference” – that he cannot conceptualize it as a hallucination. this is strong evidence that his experience lacked the fundamental structure of (apparent) first-personal givenness that characterizes virtually all conscious experiences. z&k’s descriptive indispensability argument is intended to refute absentism about for-me-ness; its conclusion is that for-me-ness really exists (or that we can rationally believe it does irrespective of its explanatory utility.) i accept their argument, insofar as it concerns a purely experiential form of for-me-ness, and have just supplemented it with a further descriptive indispensability argument intended to refute universalism about experiential for-me-ness. the two arguments, together, establish typicalism about this phenomenal feature;14 both absentism and universalism leave us unable adequately to describe the full variety of conscious human mentality. considerations of phenomenological description lead inexorably, i claim, to typicalism about non-deflationary (genuinely experiential) for-me-ness. what, then, of explanatory considerations and deflationary (purely metaphysical/epistemic) for-me-ness? 5 abandoning the access assumption in their discussion of putative pathological counterexamples to universalism about for-me-ness, z&k consider and roundly reject the identification of minimal selfawareness with some purely “geometrical feature” of experience, such as the structure of the visuospatial first-person perspective. even in pathological cases, they claim, an “experiential perspectival-ness” is retained that goes beyond mere perceptual geometry. their grounds: even in the cases discussed, epistemic asymmetry still obtains: they are available in a special way to the subject in whom they occur. these experiences continue to be characterized by a subjective presence that 14assuming that universalism, absentism, and typicalism are the only options. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 20 makes them utterly unlike public objects, which are accessible in the same way to a plurality of subjects. regardless of how alienated the patient feels vis-à-vis the experience, the experience […] continues to be phenomenally present to the patient in a way that is, in principle, unavailable to others. this is part of what its first-personal character amounts to, and why it remains correct to say that the pathological experience retains its for-me-ness. (zahavi & kriegel, 2015, p. 45) the problem with this argument is that z&k’s premises concern a deflationary interpretation of for-me-ness as some purely metaphysical feature, whereas their conclusion concerns a non-deflationary, experiential interpretation. but to assume that the former entails the latter is to beg the question (guillot, 2017, pp. 34–35). to infer from the relatively uncontroversial fact of epistemic asymmetry (deflationary interpretation) that these pathological experiences are characterized phenomenally by a subjective presence (experiential, non-deflationary interpretation) is to assume that the latter invariably accompanies the former, in the face of counterexamples intended precisely to undermine this generalisation. i do not deny (i) that pollan is better placed than i am to report on his 5-meo-dmt experience, nor (ii) that this is so because the experience occurred within him qua organism and not within me. what i deny is that, in virtue of these epistemic/metaphysical facts, his experience necessarily featured a genuinely phenomenal (non-deflationary) form of for-me-ness – was “given” or structured in a distinctly first-personal way. elsewhere, zahavi expresses puzzlement at the idea that such epistemic asymmetry could obtain without making a phenomenal difference: this is a somewhat surprising claim. is it not rather odd to insist that the difference between my access to my own feeling of nausea (as it is subjectively lived through) and the access i have to your feeling of nausea (as it is displayed in your contorted facial expressions and verbal reports) is a difference with no phenomenal impact? is there not an experiential, i.e. phenomenal, difference between being nauseous oneself and observing somebody else’s nausea? (zahavi, forthcoming, pp. 13–14) the answer to both rhetorical questions is, of course, yes. it would be odd to insist that this difference had no phenomenal impact – but “phenomenal impact” is ambiguous. the fact that my feeling of nausea (n1) is in my brain, and your feeling of nausea (n2) is in your brain, has a phenomenal impact in the following sense: it means that, rather than containing two interoceptive nausea experiences, my total sphere of consciousness contains one interoceptive nausea experience (n1) and one exteroceptive experience of another’s behavioural symptoms of nausea (call it b1). this is the experiential difference between being nauseous oneself and observing somebody else’s nausea: they are different experiences, with different perceptual modalities and representational contents. however, it is far less clear that this epistemic asymmetry or “difference of access” makes a phenomenal difference in the sense of changing the experiential letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 21 character of n1 itself. being in my brain rather than yours does not as such make a difference to the phenomenal properties of an experience.15 so the access differential has a “phenomenal impact” on my total sphere of consciousness: it contains n1 and b1, rather than n1 and n2 (or, etc.) but it does not therefore have a direct phenomenal impact on n1 itself. in short, the metaphysical/epistemic fact that is being described here as “differential access” has a phenomenal impact in the sense that it affects which experiences somebody has (and thus “what it is like to be them”); but this does not show that it has a phenomenal impact in the sense of affecting the character of those experiences themselves. the latter is required for the claim that for-me-ness shows up in all experience. my claim here is that prominent defences of universalism about for-me-ness trade on systematic ambiguities between the deflationary and the non-deflationary interpretations of this term. faced with compelling arguments against universalism about the non-deflationary version, its proponents appeal to seemingly undeniable claims about epistemic asymmetry and differential access to support controversial conclusions about the universality of an experiential feature. but whether such premises warrant such conclusions is precisely the question at issue. if i am correct then what follows, at minimum, is that care must be taken to distinguish metaphysical, epistemological, and phenomenological claims in theoretical debates about self, self-awareness, and phenomenal consciousness. that a given mental state possesses a controversial experiential feature cannot be established merely by citing the existence of a relatively uncontroversial metaphysical or epistemic relation, of ownership or access, between a hypothesized subject and that mental state. at this point a defender of sp may dig in their heels and assert that the weight of evidence and argument favouring this principle outweighs any evidence we might have for a counterexample. this would be analogous to hume’s argument that belief in miracles qua violations of natural law is never warranted, because our evidence for the inviolability of natural laws outweighs any evidence we might have for such a violation. in effect the defender of sp would be rejecting my modus tollens: if sp is true, then there are no states of selfless consciousness. there are states of selfless consciousness. therefore, sp is not true. and advancing the modus ponens: if sp is true, then there are no states of selfless consciousness. sp is true. therefore, there are no states of selfless consciousness. 15it undoubtedly makes an indirect difference since it means that the experience will co-occur and interact with all the other experiences that i have, which will be different from the set of other experiences that you have. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 22 for this to be plausible, there must be overwhelmingly strong reasons to accept sp – sufficiently strong to outweigh the combination of (a) the lucid and detailed testimony of the psychedelic subjects above and (b) the evidential principles laid out explicitly and implicitly by b&k, which classify these as genuine reports of totally selfless consciousness. obviously i cannot engage here with all the evidence and arguments for subjectivity theories and sp. but i will mention one type of consideration that i suspect plays a strong role in motivating a high level of confidence in sp, and indicate briefly why i do not find this consideration persuasive. gerard o’brien and jon opie have suggested that the “most tenacious bad idea bedevilling our attempts to think about consciousness [is] the treatment of consciousness as an access relation between cognitive sub jects and their own mental states” (o’brien & opie, 2015, p. 267). this is the intuition underpinning subjectivity theories: that there are subjects, and there are mental states, and consciousness is what happens when the former access, or become aware, of the latter. this is a metaphysical claim, but it can be seen at work in phenomenological debates, as when universalists about for-me-ness cite seemingly uncontroversial claims about differential access and epistemic asymmetry as evidence of the exceptionless presence of a phenomenal feature. however, the deflationary (metaphysical/epistemic) interpretation of for-meness is itself ambiguous. what do these claims about differential access and asymmetry really mean? in zahavi’s nausea case, the most minimal and uncontroversial reading of the differential access claim is that my feeling of nausea, n1, occurs within the brain of this organism rather than another, and thus is especially wellplaced to influence information processing in various cognitive subsystems of this organism, such as those specialised for autobiographical memory, introspection, verbal report, and behavioural control. any influence that n1 can have over the corresponding systems in other organisms must be much less direct. in that sense certain cognitive subsystems of this organism enjoy a kind of “access” to n1 that cognitive subsystems of no other organisms enjoy. a more substantive and controversial reading of the differential access claim is that there actually exists some entity that can legitimately be described as a self or subject, as someone whose mental state n1 is, and who may at various times be aware or unaware of n1, and access or fail to access it. this is a very natural way to speak and it is a very natural way to think. indeed, it is so natural that it has often been taken to be a necessary or conceptual truth; the idea that there are mental states that are not happening to anyone, or experiences that are like something, but not like something for anyone, has been alleged to be inconceivable or incoherent. but it is an idea that we need to take seriously in light of evidence from altered states. this evidence undercuts accusations of inconceivability. it is true that most mentally healthy adult humans are unable to simulate phenomenally, and thus unable to conceive of, a world in which (a) experiences happen, but (b) there is, properly speaking, no one to whom they happen. this is because the vast majorletheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 23 ity of human conscious experiences feature the subject-object structure of “firstpersonal givenness” that is experiential for-me-ness. but my psychedelic counterexamples show that it is possible for experiences to lack this phenomenal feature. in light of this, there are two ways of explaining the deep-seated intuition about conscious experiences happening for subjects: as a deep and accurate insight into the nature of the mind, or as a mere artefact of an evolved cognitive architecture (metzinger, 2003). it will come as no surprise that i prefer the latter hypothesis. in my view, the most parsimonious explanation of the near-universal human belief in the existence of subjects ontically distinct from mental states, but standing to them in relations of ownership, authorship, and access, is an error theory: brains generate mental representations underpinning phenomenal simulations of the existence of such entities as part of an efficient strategy for prediction and control, but those representations are inaccurate insofar as the posited entities do not really exist (letheby & gerrans, 2017). one reason for thinking that we do, in fact, represent ourselves as entities with such properties is the phenomenal contrast (siegel, 2007) between ordinary experience and totally selfless experience. yet another is the ubiquity of the “access assumption” diagnosed by o’brien and opie. even theorists such as zahavi, who reject the idea of a self ontically distinct from experience, cannot resist using the language of subjects, ontically distinct mental states, and access relations in defending their phenomenological views. this is not the place for a full-blown defence of the error theory of phenomenal selfhood. for my purposes, it is enough to raise the possibility of such a view. thus far i have argued that nearly all conscious experiences have, and some conscious experiences lack, experiential for-me-ness: the structural/phenomenal feature of apparent first-personal givenness, which suggests the existence of an entity to whom experiences are given. one way of resisting my conclusion is to appeal to the allegedly overwhelming antecedent plausibility of sp. this apparent plausibility derives in part, i am suggesting, from conceivability considerations: many of us simply cannot imagine a world in which there are no subjects to whom experiences are given, or for whom experiences happen – as distinct from organisms or brains in which mental states, spatially, occur. (of course, the intuition that all experiences have subjects is metaphysical, but we have already seen that metaphysical and phenomenological claims are often conflated in debates over minimal self-awareness.) for most of us, a world in which experiences simply happen, selflessly, is inconceivable – but this does not entail that such a world is inconceivable simpliciter. to grasp the possibility of such a world conceptually, it helps to grasp its contours experientially, and the evidence suggests that this is possible. for theoretical debates concerning self-awareness and consciousness to make progress, we may need to appreciate that the limits of our sober imaginative capacities do not coincide with those of the suitably stimulated human brain (metzinger, limanowski, & millière, 2018). letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 24 6 conclusion billon and kriegel’s subjectivity principle (sp) holds that “necessarily, a mental state m exhibits phenomenal consciousness only if m exhibits subjectivity” (2015, p. 30). even if not refuted by thought insertion or dpd, sp is refuted by certain psychedelic states which (i) possess all the features necessary for a verdict of total selflessness but (ii) lack any features sufficient for a denial of phenomenal consciousness. the universalist objection to my argument denies premise (i), alleging that my counterexamples may refute universalism about me-ness and mineness but not about for-me-ness. but this objection faces a dilemma: on a non-deflationary construal of for-me-ness, my counterexamples do refute universalism, and on a deflationary construal, universalism is irrelevant to questions of self-consciousness, and dubitable on independent grounds. prominent defences of universalism appeal to descriptive considerations which cut both ways: to the extent that we need an experiential notion of for-me-ness to describe ordinary experience, we need its negation to describe extraordinary experience. moreover, the existence of an epistemic asymmetry or metaphysical access relation does not, on its own, establish the presence of a genuinely phenomenal form of for-me-ness. the psychedelic cases described here show that the apparent first-personal givenness of experience may be an overwhelmingly common, but is not an invariable, feature of the structure of phenomenal consciousness. and conceivability arguments for sp are undercut by evidence that in certain altered states, the human brain can generate vivid phenomenal simulations of a world in which experiences happen to no-one. acknowledgments for helpful feedback and discussions, i would like to thank katherine carter, thomas metzinger, raphaël millière, participants in the 2018 selfless minds workshop in frankfurt and the 2018 psychedelics and the self reading group in adelaide, and two anonymous referees for philosophy and the mind sciences. chris letheby is lecturer in philosophy at the university of western australia and postdoctoral researcher at the university of adelaide. this research was partially supported by an australian research council discovery project (project number dp190101451) funded by the australian government. references ataria, y., dor-ziderman, y., & berkovich-ohana, a. (2015). how does it feel to lack a sense of boundaries? a case study of a long-term mindfulness meditator. consciousness and cognition, 37, 133–147. https://doi.org/10.1016/j.concog.2015.09.002 barbosa, p. c. r., mizumoto, s., bogenschutz, m. p., & strassman, r. j. (2012). health status of ayahuasca users. drug testing and analysis, 4(7-8), 601–609. https://doi.org/10.1002/dta.1383 barbosa, p. c. r., strassman, r. j., silveira, d. x. da, areco, k., hoy, r., pommy, j., et al. (2016). psychological and neuropsychological assessment of regular hoasca users. comprehensive psychiatry, 71, 95–105. https://doi.org/10.1016/j.comppsych.2016.09.003 letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.concog.2015.09.002 https://doi.org/10.1002/dta.1383 https://doi.org/10.1016/j.comppsych.2016.09.003 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org being for no-one: psychedelic experience and minimal subjectivity 25 billon, a. (2013). does consciousness entail subjectivity? the puzzle of thought insertion. philosophical psychology, 26(2), 291–314. https://doi.org/10.1080/09515089.2011.625117 billon, a. (2015). why are we certain that we exist? philosophy and phenomenological research, 91(3), 723–759. https://doi.org/10.1111/phpr.12113 billon, a. (2017). basic self-awareness: lessons from the real world. european journal of philosophy, 25(3), 732–763. https://doi.org/10.1111/ejop.12168 billon, a., & kriegel, u. (2015). jaspers’ dilemma: the psychopathological challenge to subjectivity theories of consciousness. in r. j. gennaro (ed.), disturbed consciousness: new essays on psychopathology and theories of consciousness (pp. 29–54). cambridge, ma: mit press. block, n. (1995). on a confusion about a function of consciousness. behavioral and brain sciences, 18(2), 227–247. https://doi.org/10.1017/s0140525x00038188 block, n. (2011). perceptual consciousness overflows cognitive access. trends in cognitive sciences, 15(12), 567–575. https://doi.org/10.1016/j.tics.2011.11.001 bouso, j. c., gonzález, d., fondevila, s., cutchet, m., fernández, x., barbosa, p. c. r., et al. (2012). personality, psychopathology, life attitudes and neuropsychological performance among ritual users of ayahuasca: a longitudinal study. plos one, 7(8), e42421. https://doi.org/10.1371/journal.pone.0042421 bouso, j. c., palhano-fontes, f., rodríguez-fornells, a., ribeiro, s., sanches, r., crippa, j. a. s., et al. (2015). long-term use of psychedelic drugs is associated with differences in brain structure and personality in humans. european neuropsychopharmacology, 25(4), 483–492. https://doi.org/10.1016/j.euroneuro.2015.01.008%20 carhart-harris, r. (2013). psychedelic drugs, magical thinking and psychosis. journal of neurology, neurosurgery & psychiatry, 84(9), e1. https://doi.org/10.1136/jnnp-2013-306103.17 carhart-harris, r. l., kaelen, m., whalley, m., bolstridge, m., feilding, a., & nutt, d. (2015). lsd enhances suggestibility in healthy volunteers. psychopharmacology, 232(4), 785–794. https://doi.org/10.1007/s00213-014-3714-z chalmers, d. j. (1996). the conscious mind: in search of a fundamental theory. new york: oxford university press. dor-ziderman, y., berkovich-ohana, a., glicksohn, j., & goldstein, a. (2013). mindfulness-induced selflessness: a meg neurophenomenological study. frontiers in human neuroscience, 7, 582. https://doi.org/10.3389/fnhum.2013.00582 farrell, j., & mcclelland, t. (2017). editorial: consciousness and inner awareness. review of philosophy and psychology, 8(1), 1–22. https://doi.org/10.1007/s13164-017-0331-x fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 frith, c. d. (1992). the cognitive neuropsychology of schizophrenia. hillsdale, nj: lawrence erlbaum. gennaro, r. j. (1996). consciousness and self-consciousness: a defense of the higher-order thought theory of consciousness. amsterdam: john benjamins publishing. guillot, m. (2017). i me mine: on a confusion concerning the subjective character of experience. review of philosophy and psychology, 8(1), 23–53. https://doi.org/10.1007/s13164-016-0313-4 johansen, p.-ø., & krebs, t. s. (2015). psychedelics not linked to mental health problems or suicidal behavior: a population study. journal of psychopharmacology, 29(3), 270–279. https://doi.org/10.1177/0269881114568039 kriegel, u. (2009). subjective consciousness: a self-representational theory. oxford: oxford university press. lane, t. (2012). toward an explanatory framework for mental ownership. phenomenology and the cognitive sciences, 11(2), 251–286. https://doi.org/10.1007/s11097-012-9252-4 lebedev, a. v., lövdén, m., rosenthal, g., feilding, a., nutt, d. j., & carhart‐harris, r. l. (2015). finding the self by losing the self: neural correlates of ego‐dissolution under psilocybin. human brain mapping, 36(8), 3137–3153. https://doi.org/10.1002/hbm.22833 letheby, c., & gerrans, p. (2017). self unbound: ego dissolution in psychedelic experience. neuroscience of consciousness, 3(1), 1–11. https://doi.org/10.1093/nc/nix016 levine, j. (2001). purple haze: the puzzle of consciousness. oxford: oxford university press. mellor, c. s. (1970). first rank symptoms of schizophrenia: i. the frequency in schizophrenics on admission to hospital ii. differences between individual first rank symptoms. the british journal of psychiatry, 117(536), 15–23. https://doi.org/10.1192/s0007125000192116 metzinger, t. (2003). being no one: the self-model theory of subjectivity. cambridge, ma: mit press. metzinger, t., limanowski, j., & millière, r. (2018). splendor and misery of self-models: conceptual and empirical issues regarding consciousness and self-consciousness. an interview with thomas metzinger. alius bulletin, 2, 53–73. retrieved from https://www.aliusresearch.org/uploads/9/1/6/0/91600416/metzinger_-_splendor___misery_of_self-models.pdf letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1080/09515089.2011.625117 https://doi.org/10.1111/phpr.12113 https://doi.org/10.1111/ejop.12168 https://doi.org/10.1017/s0140525x00038188 https://doi.org/10.1016/j.tics.2011.11.001 https://doi.org/10.1371/journal.pone.0042421 https://doi.org/10.1016/j.euroneuro.2015.01.008%20 https://doi.org/10.1136/jnnp-2013-306103.17 https://doi.org/10.1007/s00213-014-3714-z https://doi.org/10.3389/fnhum.2013.00582 https://doi.org/10.1007/s13164-017-0331-x https://doi.org/10.33735/phimisci.2020.i.40 https://doi.org/10.1007/s13164-016-0313-4 https://doi.org/10.1177/0269881114568039 https://doi.org/10.1007/s11097-012-9252-4 https://doi.org/10.1002/hbm.22833 https://doi.org/10.1093/nc/nix016 https://doi.org/10.1192/s0007125000192116 https://www.aliusresearch.org/uploads/9/1/6/0/91600416/metzinger_-_splendor___misery_of_self-models.pdf https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org chris letheby 26 millière, r. (2017). looking for the self: phenomenology, neurophysiology and philosophical significance of drug-induced ego dissolution. frontiers in human neuroscience, 11, 245. https://doi.org/10.3389/fnhum.2017.00245 millière, r. (2019). are there degrees of self-consciousness? journal of consciousness studies, 26(3-4), 252–282. retrieved from https://www.ingentaconnect.com/content/imp/jcs/2019/00000026/f0020003/art00012 millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 millière, r., carhart-harris, r. l., roseman, l., trautwein, f.-m., & berkovich-ohana, a. (2018). psychedelics, meditation, and self-consciousness. frontiers in psychology, 9, 1475. https://doi.org/10.3389/fpsyg.2018.01475 nagel, t. (1974). what is it like to be a bat? the philosophical review, 83(4), 435–450. https://doi.org/10.2307/2183914 nichols, d. e. (2016). psychedelics. pharmacological reviews, 68(2), 264–355. https://doi.org/10.1124/pr.115.011478 o’brien, g., & opie, j. (2015). a schizophrenic defense of a vehicle theory of consciousness. in r. j. gennaro (ed.), disturbed consciousness: new essays on psychopathology and theories of consciousness (pp. 265–292). cambridge, ma: mit press. petitmengin, c. (2006). describing one’s subjective experience in the second person: an interview method for the science of consciousness. phenomenology and the cognitive sciences, 5(3-4), 229–269. https://doi.org/10.1007/s11097-006-9022-2 pollan, m. (2018). how to change your mind: the new science of psychedelics. london: penguin. santos, r. g. dos, osorio, f. l., crippa, j. a. s., riba, j., zuardi, a. w., & hallak, j. e. (2016). antidepressive, anxiolytic, and antiaddictive effects of ayahuasca, psilocybin and lysergic acid diethylamide (lsd): a systematic review of clinical trials published in the last 25 years. therapeutic advances in psychopharmacology, 6(3), 193–213. https://doi.org/10.1177/2045125316638008 sebastián, m. á. (2020). perspectival self-consciousness and ego-dissolution: an analysis of (some) altered states of consciousness. philosophy and the mind sciences, 1(i), 9. https://doi.org/10.33735/phimisci.2020.i.44 sessa, b. (2012). the psychedelic renaissance: reassessing the role of psychedelic drugs in 21st century psychiatry and society. london: muswell hill press. séglas, j., & meige, h. (1895). leçons cliniques sur les maladies mentales et nerveuses:(salpêtrière, 1887-1894). paris: asselin et houzeau. siegel, s. (2007). how can we discover the contents of experience? the southern journal of philosophy, 45(s1), 127–142. https://doi.org/10.1111/j.2041-6962.2007.tb00118.x sierra, m. (2009). depersonalization: a new look at a neglected syndrome. cambridge: cambridge university press. strassman, r. (2001). dmt: the spirit molecule: a doctor’s revolutionary research into the biology of near-death and mystical experiences. rochester, vt: simon; schuster. tagliazucchi, e., roseman, l., kaelen, m., orban, c., muthukumaraswamy, s. d., murphy, k., et al. (2016). increased global functional connectivity correlates with lsd-induced ego dissolution. current biology, 26(8), 1043–1050. https://doi.org/10.1016/j.cub.2016.02.010 zahavi, d. (forthcoming). consciousness and (minimal) selfhood: getting clearer on for-me-ness and mineness. in u. kriegel (ed.), the oxford handbook of the philosophy of consciousness. oxford: oxford university press. zahavi, d., & kriegel, u. (2015). for-me-ness: what it is and what it is not. in d. o. dahlstrom, a. elpidorou, & w. hopp (eds.), philosophy of mind and phenomenology: conceptual and empirical approaches (pp. 48–66). new york: routledge. zinberg, n. e. (1984). drug, set, and setting: the basis for controlled intoxicant use. new have, ct: yale university press. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.3389/fnhum.2017.00245 https://www.ingentaconnect.com/content/imp/jcs/2019/00000026/f0020003/art00012 https://doi.org/10.33735/phimisci.2020.i.48 https://doi.org/10.3389/fpsyg.2018.01475 https://doi.org/10.2307/2183914 https://doi.org/10.1124/pr.115.011478 https://doi.org/10.1007/s11097-006-9022-2 https://doi.org/10.1177/2045125316638008 https://doi.org/10.33735/phimisci.2020.i.44 https://doi.org/10.1111/j.2041-6962.2007.tb00118.x https://doi.org/10.1016/j.cub.2016.02.010 https://doi.org/10.33735/phimisci.2020.i.47 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction subjectivity theories and the unconsciousness defence psychedelic mystics are not zombies the universalist objection abandoning the access assumption conclusion the skill of mental health. the skill of mental health towards a new theory of mental health and disorder garson ledera (leder.garson@gmail.com) tadeusz zawidzkib (zawidzki@email.gwu.edu) abstract this paper presents a naturalist skill-based alternative to traditional function-based naturalist theories of mental health and disorder. according to the novel skill view outlined here, mental health is a skilled action of individuals, rather than a question of the functioning of mental mechanisms. mental disorder is the failure or breakdown of this skill. this skill view of mental health is motivated by focusing on the process of mental healing. this paper argues that, when we start with a focus on how and why individuals heal from mental disorders, we gain a better understanding of what mental health is: the exercise of self-regulatory metacognitive skill. keywords mental disorder ∙ mental dysfunction ∙ mental health ∙ skill ∙ stigma this article is part of a special issue on “models and mechanisms in philosophy of psychiatry,” edited by lena kästner and henrik walter. 1 introduction while it is generally agreed that mental disorders are a serious problem, there is far less agreement about what the problem actually is. what exactly is disordered about mental disorders? the answer is normally situated somewhere between the naturalist/normative divide. naturalism, roughly, is the position that mental disorders are value-independent, scientifically identifiable and explainable mental phenomena. normative theories of health conceive of mental disorder as primarily a problem of the meanings attached to behavior and bodies; whether some way of being is a mental disorder is primarily a question of value judgments about acenter for bioethics and medical humanities, medical college of wisconsin. bdepartment of philosophy, the george washington university. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-1350-1579 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 2 norm-transgressing behavior. hybrid theories land somewhere between these poles. naturalism, or hybrid theories with a necessarily naturalistic component, are the received views in academic psychology and psychiatry. 1 despite their wide acceptance, there have been significant problems in formulating naturalist views. notably, it has proven difficult to identify what, exactly, the value-independent basis of mental illnesses is supposed to be. according to the most influential naturalist approaches, mental disorders are constituted, at least in part, by mental dysfunctions. 2 the assumed relationship between function and naturalism is so strong that ‘naturalism’ is often just (mistakenly) defined as a function-based view. 3 function-based naturalist views posit that mental disorders necessarily involve the failure of some mental mechanism, or mechanisms, to function ‘normally’. the most serious difficulty facing these views has been providing an account of normal psychological functioning. the perceived failure of function-based naturalist views to offer a plausible theory of normal mental function has been taken by many to be evidence for the failure of naturalism, full stop, and motivation for the acceptance of normative views. 4 in contrast, this paper argues that mental functions are the wrong point of emphasis in explaining the value-independent nature of mental disorder and that a more plausible naturalist theory can be constructed by grounding the value-independent basis of mental disorder in the concept of skilled action, rather than proper mental function. 5 according to this view, mental health is a type of skilled activity performed by individual persons, not the normal functioning of mental mechanisms. mental disorder is the failure or breakdown of this skill. there are at least two desiderata that a theory of mental health should address: (1) it should offer a plausible theory of the boundary between health and disorder, and (2) it should explain how we might know whether any particular mental state falls on one side or the other of this boundary. the motive for the first desideratum is clear: a naturalist theory of mental disorder needs to explain what differentiates disorder from health. the second desideratum is also crucial: a good theory of mental health and disorder should be useful and provide us with the means to differentiate between the two. the skill view of mental health provides a naturalist answer to both (1) and (2), while avoiding the more serious metaphysical and epistemic problems of traditional function-based views. 1for example, the two primary diagnostic manuals in the west (the dsm-5 and icd-10) both adopt function-based definitions of ‘disorder’ with naturalistic components. 2for example, see: wakefield (1992); boorse (1977); boorse (2014); american psychiatric association (2013); world health organization (1992). we use the term ‘naturalist’ to refer to both ‘pure’ naturalist views (e.g., boorse, 1977) and hybrid views with a necessary naturalist component (e.g., wakefield, 1992). 3for example, see: murphy (2020); kingma (2013). 4e.g.: cooper (2007); fulford (1989). 5this paper does not directly address the tenability of normative views. the focus here is on whether a strong naturalist view can be developed. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 3 the structure of the paper is as follows. section 2 addresses the flaws of traditional function-based views. section 3 introduces the alternative skill-based naturalist theory of mental health and argues for the conceptual and explanatory superiority of the skill view over function-based naturalist theories. section 4 addresses potential concerns with grounding the concepts of mental health and disorder in skill rather than function. section 5 discusses broader theoretical and practical benefits of adopting the skill view of mental health. 2 mental disorders and mental functions mental disorders are traditionally conceptualized in functional terms. for example, both the diagnostic and statistical manual of mental disorders [5th ed.; dsm-5, american psychiatric association (2013)], the primary diagnostic manual in north america, and the dsm’s international counterpart, the world health organization’s international classification of diseases (icd-10, world health organization, 1992), adopt function-based conceptions of disorder. the dsm-5 defines mental disorders as follows: a mental disorder is a syndrome characterized by clinically significant disturbance in an individual’s cognition, emotion regulation, or behavior that reflects a dysfunction in the psychological, biological, or developmental processes underlying mental functioning. (italics added) (american psychiatric association, 2013) both the dsm-5 and icd-10 are intended to serve as theory-neutral diagnostic tools for the identification of symptoms that are jointly sufficient for the diagnosis of a disorder, not to be theories of what disorders are. theories of mental disorder, on the other hand, are attempts to explain what it is about certain ways of being that makes them healthy, disordered, or benign. and, according to the most influential naturalist views, what differentiates ‘genuine’ disorders from normal human variation is that the former, but not the latter, are marked by dysfunctions of some kind. the primary challenge facing function-based naturalist views has been to provide a value-independent account of ‘normal’ mental functioning. naturalist solutions to this problem differ in their specifics, but share the common strategy of defining normal mental function as the contribution of a mental part or process to an ultimate biological norm. the most influential candidates for this ‘ultimate’ norm have been evolutionary design or individual survival and reproduction. 6 according to these ‘biological norms’ views, the normal function of, e.g., the heart is to pump blood, and not make pounding sounds, because the pumping of blood, but not the production of sound, conforms to some ultimate biological norm (e.g., it is what the heart was designed by natural selection to do, it contributes to the 6the touchstones for the evolutionary and longevity/reproduction views are wakefield (1992) and boorse (1977), respectively. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.psychologytoday.com/basics/cognition https://www.psychologytoday.com/basics/emotion-regulation https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 4 ultimate end-goals of individual survival and reproduction, or it conforms to some other biological norm). the problems with function-based views are well known. 7 most notably, dysfunction appears to be neither necessary nor sufficient for the presence of a mental disorder. dysfunction is likely not sufficient for disorder. for example, depending upon what theory of ‘normal’ function one adopts, exclusive homosexuality, asexuality, and a lack of desire to reproduce may all be considered ‘dysfunctions’ (because of their effects on reproduction), but they should not be considered disorders (e.g., cooper, 2007). defenders of the sufficiency view have responded to this concern by arguing that we should divorce our theory of disorder from any normative judgments we have about experiencing a disorder or being labeled as having a disorder (e.g., boorse, 2014). however, even if we accept that the concept of disorder should be value-neutral, we should still be judging a theory of disorder based on the plausibility of its extensional assignments. the sufficiency view is not incoherent, but it is highly implausible. dysfunction is also likely not necessary for disorder. for example, it is an open empirical question whether behaviors and mental phenomena normally considered to be ‘disordered’, such as the phenomena associated with depression, anxiety, and schizophrenia, are in fact the result of dysfunctioning mental mechanisms rather than being the consequences of mechanisms functioning as they normally should. 8 depression, for example, is variously hypothesized to be the result of dysfunctioning mental mechanisms that lead to ‘interlocked’ cognitive-affective cycles of mental processing (teasdale & barnard, 1993), the product of normally functioning mental mechanisms that lead to an increase in critical ruminative thought (andrews & thomson, 2009), the result of dys-functioning mental mechanisms that produce negatively valenced loss-based cognitions (beck & alford, 2009), or the result of normally functioning mental processes that socially signal the need for assistance (watson & andrews, 2002). and we see similar variations in the hypothesized role of ‘dysfunction’ for other paradigmatic disorders such as generalized anxiety disorder, schizophrenia, and bipolar disorder. the take-away point here is not that each of these theories is equally as plausible, nor is it that we are never able to make plausible inferences about the existence or functions of mental mechanisms. rather, the point is that if we accept dysfunction as the grounding of our concept of disorder, then we must accept that it is an open question whether some paradigmatic disorders are in fact disorders. function-based theories commit us to the view that mental health necessarily involves the adherence of mental mechanisms to certain ultimate biological norms. this focus on biological norms (e.g., evolutionary design or individual survival and reproduction), while potentially establishing a value-free foundation for the concept of disorder, also leads to a potentially massively revisionary conception 7the literature here is extensive. e.g.: cooper (2007); kingma (2014). 8this concern has been raised by a number of authors (e.g., lilienfeld & marino, 1995; woolfolk, 1999). leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 5 of mental health. 9 dysfunction views are committed to the claim that it is an open (though possibly unanswerable) empirical question whether any postulated ‘mental disorder’ is in fact the result of a dysfunctioning mental mechanism, and thus a genuine mental disorder. it is possible that paradigmatic disorders such as depression, anxiety, and phobias may turn out not to be mental disorders, but rather products of healthy minds dealing with mere problems of living. while, on the other hand, behavior not commonly considered to be a product of unhealthy minds, such as homosexuality, a lack of desire to reproduce, and the engagement in dangerous ‘thrill-seeking’ activities, may in fact be mental disorders if it turns out that they are the result of dysfunctioning mental mechanisms.10 this is not incoherent, but it is highly counterintuitive. the use of dysfunction, then, to ground the value-neutral basis of mental disorders appears unsuccessful. if function-based theories of health make it possible that depression and anxiety are not mental disorders, but homosexuality is, then this gives us good reason to think that the presence or absence of mental dysfunctions is not necessarily related to the concept of mental health. that the entailments of function-based views may be so revisionary and contrary to common folk conceptions of disorder is prima facie evidence that something has gone wrong with the focus on dysfunctions to explain the distinction between mental health and disorder, and gives us reason to search for a more plausible naturalist conception of mental health. the problem is this: despite their prominence, there have been serious problems in formulating a naturalist conception of mental disorder. dysfunction-based naturalist views are not incoherent but they are seriously flawed. these flaws can either be accepted as part of the cost of formulating a naturalist theory, or taken as evidence for the failure of naturalism. in contrast, this paper argues that a more plausible naturalist view of health can be constructed if we reorient the dialectic away from the concept of normal function and focus instead on the skills constitutive of mental health. skill, rather than function, can provide the objective basis for a naturalist theory of mental disorder. the new naturalist theory of mental health outlined here argues that mental health is best conceived of as a skill. more specifically, this paper argues that mental health is skilled metacognitive self-regulation; mental disorder is a failure or breakdown of this skill. the skill view of mental health provides a naturalist framework for the scientific study and treatment of mental disorders that avoids 9murphy (2006) makes a similar point. 10note that while function-based views agree that mental dysfunctions are necessary for disorders, they differ over whether they are also sufficient. ‘pure’ function-based views posit that the presence of a mental dysfunction is both necessary and sufficient for disorder. ‘hybrid’ views posit that dysfunction is necessary, but not sufficient for disorder; disorders must also be judged to be ‘harmful’ (with the ‘harm’ criterion being a value-based claim that varies between cultures and value systems). so, for example, the ‘pathological’ status of homosexuality depends upon its status as a dysfunction (for both pure and hybrid views) and whether it is judged to be harmful in a given value-system (for hybrid views only). leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 6 the epistemological and conceptual failings of traditional function-based views. the focus on skilled action, rather than adherence to biological norms, allows for a flexible theory of health that avoids pathologizing normal human difference while also providing a non-revisionary accounting of paradigmatic mental disorders. 3 the skill of mental health the skill view of mental health makes the following two claims: 1: mental health is skilled metacognitive self-regulation. 2: mental disorder is a failure or breakdown of this skill. sections 3.1-3.2 outline theory-neutral naturalist accounts of skilled action and self-regulation. sections 3.3-3.4 will address the skill view’s claims in order. 3.1 the value-neutrality of skill the term ‘skill’ has a number of common uses. ‘skill’ can refer to a particular type of possible action (e.g., the skill of playing tennis), to a way of acting (e.g., being a skilled, as opposed to unskilled, tennis server), and to expertise in action (e.g., being a skilled, as opposed to novice, tennis player). this paper is concerned only with the first two conceptions of skill. the focus here is on what differentiates skilled from unskilled action, not what differentiates low skill from high skill, or skill from expertise. this paper adopts a theory-neutral conception of skill: minimally, to be skilled at ɸ-ing requires that one possesses the ability to ɸ intelligently.11 both ‘ability’ and ‘intelligence’ can be disambiguated. following mele (2003), we can distinguish between two senses of ‘ability’:12 ability1: an agent is able to ɸ, given appropriate background conditions. ability2: an agent is able to ɸ and has the appropriate background conditions satisfied to successfully ɸ. the distinction is straightforward. there are many things an agent may be generally able1 to do (such as play tennis), but that she is at some point in time unable2 to do because of something that masks this ability1 (such as a sprained ankle). skill at ɸ-ing necessarily requires the ability1 to ɸ. there is no scenario in which an agent is skilled at performing some action, yet there is no possible circumstance in which she is able1 to do so. for example, whatever the specifics are of a theory of skill, a person cannot be a skilled tennis player if she is not now, nor ever was, able1 to play tennis. however, possessing the ability1 to ɸ does not entail that one is always able2 to ɸ. a skilled tennis server is able1 to strike tennis balls into the 11this definition is consistent with both intellectualist (e.g., stanley & williamson, 2017) and nonintellectualist (e.g., ryle, 1949) theories of skill. 12mele (2003) does not use this exact formulation. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 7 service area, only given that certain background conditions are met (e.g., that she has access to a racket, ball, is free from injury, and so on). the lack of a racket or a sprained ankle might render her unable2 to play tennis on any given day, without necessarily affecting her ability1, or skill, to play. skills may diminish after long periods of being unable2 to perform, but one’s ability1 to ɸ is not necessarily dependent upon whether one is able2 to ɸ at any given time. a skilled action, on the other hand, does require that one is able2 to ɸ. for example, skilled tennis service requires that one is both able1 to strike balls into the opposing service box, and that the enabling conditions are met for her to do so successfully. skill may require ability1, but ability1 is not sufficient for the presence of skill. a person learning to strike tennis balls may be able1 to regularly propel balls into the opposing playing area if the balls are carefully teed up, but this activity is not necessarily a skilled act. the ‘intelligence’ constraint is meant to distinguish skill from reflexive, lucky, or merely successful action. theories of skill differ greatly over how to specify the intelligence clause, but there is broad agreement that, minimally, ‘intelligence’ requires the ability1 to adapt, learn, and intentionally modify one’s behavior in response to novel stimuli.13 the new tennis player learning to strike a ball off a tee lacks the ability1 (at least at that moment) to intentionally modify their behaviors to adapt to new problems such as a change in ball trajectory, speed, or spin. a skilled tennis player, on the other hand, would be able1 to learn from, and adapt to, different shots, opponents, and physical constraints. theories of skill differ over what else, if anything, is needed in addition to intelligence and ability1 to explain skill. as with our analysis of function-based views, we need not take a position here. the aim of this paper is not to argue for a particular theory of skill, but rather to examine whether the concept of skill (given the above constraints) can ground a naturalist theory of mental health. importantly, the concept of skilled action can be formulated in valueindependent terms. judgments of how well one performs some task and how well one modifies one’s behavior clearly are value-judgements. the judgment of skill as expertise also appears to be value-laden (insofar as whether or not some action is done expertly varies depending upon the standards set). in contrast, whether an agent is able2 to intelligently ɸ is not, or at least need not be, a value judgment. for example, metaphysical theories of ability commonly take the conditional form: an agent is able2 to ɸ if she would ɸ if she tried (maier, 2021). theories differ over how to best specify these enabling conditions, but these theoretical differences needn’t be differences in value (just as metaphysical disputes over the existence of midsize objects are not necessarily disputes over values). similarly, theories of the metaphysics of intelligence generally are attempts to precisify some cognitive faculty that things like humans have, and things like modern computers currently lack. there is certainly room for values to enter into the answer to this question, but there is no necessary reason that this need be the case. more generally, insofar as a value-independent metaphysics is possible, the 13for example, both stanley & williamson (2017) and ryle (1949) adopt this view. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 8 metaphysics of skill, like the metaphysics of function, poses no unique problem that necessitates a value-based theory. there are, of course, epistemic challenges in identifying when any particular action is skilled. section 3.4 will address these epistemic constraints, and show that they compare favorably to those facing function-based views. sections 3.2 and 3.3 will outline theoretical commitments of the skill-based view of mental health. 3.2 skill and self-regulation claim (1), defended in the next section, states that mental health is skilled metacognitive self-regulation. ‘self-regulation’, here, refers to the process of altering or controlling one’s responses to align with one’s goals or standards (baumeister et al., 1993; fujita, 2011). the absence of self-regulation is met by purely automatic cognition, emotion, and behavior based on ‘learning, habit, inclination, or even innate tendencies (baumeister et al., 1994).14 everyday examples of self-regulation include regulating emotional responses to perceived losses or gains (e.g., inhibiting the experience of intense emotions), regulating cognitions (e.g., avoiding ruminative patterns of thought, or conversely, focusing one’s attention despite distractors), and regulating impulses (e.g., in regard to food, drink, or self-expression). psychological accounts of self-regulation are traditionally focused on the ability2 to regulate one’s responses towards some ultimate distal goal or idealized conception of self (carver & scheier, 1982). this paper adopts a more restricted conception of self-regulation; we are interested in the ability2 to regulate one’s responses to one’s immediate willings. we are not concerned here with the notion of the idealized self or in identifying individuals’ ultimate goals. rather, the focus is on whether one is able2, at any given time, to skillfully regulate one’s responses to whatever standard one wills. there is also an important difference between regulating one’s responses and completely controlling one’s mental life. self-regulating involves the ‘overriding, stopping, modifying, or otherwise changing’ of one’s responses to exogenously and endogenously generated stimuli (rawn & vohs, 2011). this skill can, but need not, involve controlling the generation of mental content. completely controlling the generation of mental content would require the ability2 to only think, feel, and desire what one wills (e.g., only thinking pleasant thoughts and experiencing positive emotions). self-regulation does not require the (possibly superhuman) ability2 to completely control the generation of one’s cognitions and emotions. regulatory strategies may include the attempt to control the generation of mental content (e.g., by avoiding situational triggers), but attempts to control content can also be part of the problem that requires regulation (as can be the case with obsessive disor14the skill view does not claim that all instances of self-regulation must be conscious and effortful. it is generally accepted that self-regulation may involve automatic processes (though there is debate about how much) (e.g., vohs & baumeister, 2004). this paper uses ‘self-regulation’ to refer to self-regulatory processes (both ‘automatic’ and ‘conscious’) that are directed at regulating one’s responses towards some self-initiated goal (duckworth et al., 2016). leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 9 ders). automatic and unwanted thoughts, emotions, and urges are a normal part of mental functioning, and the failure to completely control the content of one’s mental life is not necessarily a failure of self-regulation. failures of self-regulation involve inability2 to regulate how one responds to these thoughts, emotions, and life-events. 3.3 skill and mental health claim (1) states that mental health is skilled metacognitive self-regulation. the term ‘skilled’ in both claims (1) and (2) refers to the exercise of the skill of self-regulation. mental health requires not just the possession of the skill to self-regulate, but also that the appropriate enabling conditions are met in order to be able2 to exercise this skill. according to this conception of health, mental health is a skilled way of acting, rather than the proper functioning of mental parts. the term ‘metacognitive’ in claim (1) refers to the fact that the skillful selfregulation exercised in mental health is directed at cognitive states, broadly construed, including beliefs, desires, and emotions. as explained in section 3.2, selfregulation involves the ability2 to regulate one’s responses to thoughts, emotions, and life-events, and these responses are, in the first instance, other cognitive states. even if behavioral responses are also in the scope of such self-regulation, such behavioral responses are products of prior cognitive responses, and cannot be skillfully regulated without the skillful regulation of the latter. hence, the kind of skillful self-regulation relevant to mental health is inevitably metacognitive. unlike embodied skills, like athletics or musicianship, or intellectual skills, like chess or advanced mathematics, both of which make use of cognitive states and capacities, metacognitive skills, in addition to making use of cognitive states and capacities, are also directed at cognitive states and capacities. their domain is cognition, unlike embodied skills, the domains of which typically involve some socially constituted, bodily activity, or intellectual skills, the domains of which are more abstract, like chess or mathematics. hence, the skills of self-regulation relevant here are aptly called ‘metacognitive’. the skill view of mental health is motivated by focusing on the process of mental healing. theories of mental health are typically developed without an explicit focus on the process of mental healing. on the one hand, this is understandable. theories of health are normally considered to be the basis from which we can determine what healing is. health is logically prior to healing; if we don’t know what health is, then we don’t know what healing is (i.e., ‘healing’ is healing from what and to what?). on the other hand, the traditional disregard of mental healing in the construction of theories of mental health has missed a significant source of data from which to build a theory. theories of mental health and disorder are not (or at least should not be) developed a priori; they begin with observations about mental phenomena and human behavior (e.g., the ‘symptoms’ of mental disorders) and leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 10 attempt to explain what, if anything, differentiates health from disorder. these observations should include, as important components, what happens when individuals mentally heal.15 in order to figure out what has ‘gone wrong’, theories of health and disorder should look to what happens when things go back to ‘being right’. in support of (1), we should first notice that mental healing and health necessarily involves self-regulation. this is made clear if we consider the process of mental healing. regardless of whether disorders are conceived of as problems of natural function or of value, one of the primary end-goals of psychotherapy is improved self-regulation. in the case of normative theories of mental health, successful psychological treatment just is whatever causes the patient to regulate her responses to fit within the established norms of ‘healthy’ functioning. in the case of naturalist function-based theories of mental health, whatever the mental dysfunction is that is hypothesized to be responsible for a particular mental disorder, the proper function of this hypothesized mechanism involves the ability2 to selfregulate. this is not to say that function-based theories are committed to the claim that self-regulation is the only function of the mind or that self-regulation is the ultimate function of all mental mechanisms. however, if, after successful therapy, the proper functioning of some hypothesized mental mechanism involves persons being unable2 to self-regulate, then we should take this as evidence that we are mistaken about our standard of mental health. there is nothing that should be recognizable as a mental disorder (as opposed to a neurological disorder) that does not involve a problem of self-regulation.16 for example, if a naturalist-inclined therapist claims that a patient seeking treatment for major depression is ’healed’ (i.e., her hypothesized dysfunctional mechanism(s) have returned to normal functioning), yet reports that the patient is still unable2 to regulate her negative emotions and cognitions, we should conclude that the therapist is confused about what depression is. similarly, obsessive compulsive disorder (ocd) involves the experience of recurring intrusive and unwanted thoughts, urges, or impulses, as well as compulsive repetitive behaviors or mental acts to suppress or ignore these obsessions; any psychological intervention that successfully treats ocd necessarily requires that patients become able2 to regulate their responses to their obsessions and compulsions. and we see the same patterns with other paradigmatic disorders. different mental disorders involve different regulatory problems: anxiety disorders are marked by excessive anxiety and worry, and difficulty in controlling these worries; schizophrenia is marked 15of course, a theory of mental health and disorder may end up revising our pretheoretical assumptions about the observed ‘symptoms’ used to construct the theory. 16possible exceptions to this claim are some of the cluster b personality disorders. the skill view entails that at least some personality disorders are not mental disorders if it is the case that they do not involve inabilities2 to skillfully self-regulate. and this is just what we should expect. if a putative nosological entity such as narcissistic personality disorder does not involve any inability2, it seems clear that the behavior being pathologized better reflects disvalued behavior and character, not poor health. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 11 by delusional beliefs or hallucinations as well as disorganized speech and behavior; bi-polar disorders may involve both serious depressive symptoms and manic behaviors and patterns of thought (american psychiatric association, 2013). it is an open empirical question how best to treat these disorders, but the end-goal of psychotherapeutic treatment is the same: no successful treatment leaves patients unable2 to regulate their responses to these symptoms. theories of psychotherapy differ over the causes of the symptoms that are being regulated, and over how to best ameliorate these problems. but, regardless of whether the therapy focuses on identifying and challenging core beliefs, providing insight into subconscious processes, extinguishing learned behavioral responses, or ameliorating any other hypothesized cause of psychopathology, successful psychotherapeutic interventions involve the patient (re)gaining the skill to self-regulate. if successful treatment does not involve the patient (re)gaining some ability2 to intelligently control how she responds to thoughts, emotions, and events, then it is unclear what is supposed to be ‘successful’ about the therapy. for example, no effective psychotherapy for anxiety (measured by the patient no longer fitting the diagnostic criteria for a disorder) leaves the patient unable2 to regulate how she responds to anxious feelings, thoughts, or (formerly) anxietyproducing life events. and no effective therapy for schizophrenia leaves a patient without the ability2 to regulate her mental phenomena. this is not to say that effective therapy needs to completely excise all symptoms from patients’ mental lives; individuals can experience anxious thoughts, intrusive thoughts, compulsions, and hallucinations without fitting the diagnostic criteria of any mental disorder. this is also not to say that improved self-regulation is the only goal of all psychotherapies, or that it is even necessarily an explicit goal. the point here is that if psychotherapy is effective, it is effective insofar as it enables improved self-regulation. if there is no problem of self-regulation, then there should be no problem of mental health. however, some state’s not being a mental disorder does not mean that it is not a disorder. the focus on skilled self-regulation helps explain the common distinction between mental disorders and neurological disorders. neurological disorders, such as amnesia disorders, epilepsy, learning disorders, dementia, and alzheimer’s disease, may result in failures of self-regulation, but they are not best treated as regulatory problems. treatment for neurological disorders such as alzheimer’s disease may include skill training (e.g., learning to cope with the fear and frustration sometimes associated with memory loss), but even if a patient were to (somehow) be able2 to skillfully self-regulate, this would not make it the case that she no longer has a disorder (though it may be less subjectively distressing). in contrast, if an individual seeking treatment for a mental disorder such as depression or anxiety no longer has a problem skillfully self-regulating, then there should no longer be any disorder. what makes a mental disorder mental, rather than somatic, is that mental disorders are best explained and treated as problems of skilled self-regulation while somatic disorders are not.17 17graham (2013) makes a similar point. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 12 both normative and function-based naturalist theories should agree with this; they just disagree about whether individuals with mental disorders are regulating their mental lives to an objective standard of health, and what, if anything we need to posit in addition to self-regulation to explain mental health. for normative views, the additional explanatory piece is a value-judgment. according to these views, while self-regulation plays a role in mental healing, whether the healing is successful, and, more generally, whether an individual is judged to be mentally healthy, depends on a value judgment of some kind. in contrast, function-based naturalist theories posit that we need the extra concepts of normal function and dysfunction in order to establish a value-independent basis for the distinction between health and disorder (and between healthy and disordered self-regulation). the problem, as we’ve seen, is that the use of function to draw this boundary leads to serious epistemic and conceptual problems. the concept of skill can do better. 3.4 skill, mental disorder, and the boundary problem there are two related, but distinct, concerns about boundaries that theories of mental disorder must address: the first is a metaphysical issue (‘what is the difference between mental disorder and health’), the second epistemic (‘how might we know the difference’). the epistemic boundary problem will be addressed in section 3.5. this section is focused on the metaphysical question. claim (2) states that mental disorder is unskilled self-regulation. according to the skill view, the boundary between mental health and mental disorder is a question of ability2 to intelligently self-regulate. this does not entail that any regulatory failure is necessarily a sign of mental disorder. healthy individuals can and do lose control of their thoughts and emotions, and act in ways contrary to their explicit intentions. this is normal and often benign. the difference between normal regulatory lapses and breakdowns of skilled self-regulation (and, thus, the difference between mental health and disorder), is that disorders involve an inability2 to self-regulate, not just the poor exercise of regulatory skill. just as a skilled athlete does not become unskilled when she misstrikes a ball or makes a tactical mistake, mentally healthy individuals do not become disordered merely by engaging in dysregulated behavior. the boundary between mental health and disorder is a question of what individuals are able2 to do, not just what they do. this solution to the boundary problem provides a principled, valueindependent distinction between mental disorders and merely socially disvalued behaviors. mental health requires being able2 to regulate one’s responses to emotions and cognitions, but not necessarily that one regulates the content of these mental states to any particular end. the difference between paradigmatic mental disorders (such as gad or ocd) and instances of normal human variation (such as fringe religious or political beliefs) is that the latter involve the ability2 to skillfully self-regulate, while the former do not. this is not to say that individuals with a mental disorder do not have the capacity to self-regulate (presumably they do), but only that at the time in question they do not have the ability2 to leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 13 intelligently do so.18 a mentally healthy political radical who suffers as a result of her beliefs could (in the sense of having the ability2) choose not to focus her mental life exclusively on her political views, while an individual suffering from gad is not, at that time, able2 to skillfully regulate her anxious emotions and cognitions. similarly, the skill view entails that while some sexual thoughts or behaviors may be symptoms of a mental disorder (e.g., unwanted and intrusive obsessive sexual thoughts and images), there is nothing necessarily disordered about any particular norm-transgressing sexual behavior or thought. mental disorders (such as ocd) can be differentiated from non-disordered states (such as homosexuality) because the former, but not the latter, necessarily involve the inability2 to successfully regulate cognitive and emotional patterns. the content or object of one’s sexual desires, as opposed to the ability2 to intelligently regulate one’s responses to these sexual thoughts or desires, have no necessary relation to mental health. being attracted to some object o, is not necessarily a problem; being unable2 to regulate one’s responses to one’s o-directed obsessions and compulsions is.19 it is possible that individuals with norm-transgressing sexual orientations or preferences may experience distress at their sexual thoughts and desires (e.g., because they run counter to social norms) based on their inability2 to alter or control the content of these thoughts and desires. this felt need to control the object of one’s sexual desires and thoughts may lead to mental health problems, but this does not entail that there is anything pathological about the content of these thoughts or the objects of these desires. these problems, if disordered, would be mood disorders (likely influenced by cultural norms). the focus on skill, rather than function, also explains the difference between disorder and normal human difference. people differ greatly in the content of their mental states. there is nothing necessarily disordered about atypical or normtransgressing beliefs or emotional reactions, and a theory of mental health should reflect this. this normal human variation reflects a disorder only when persons are no longer able2 to regulate these mental states. for instance, both a philosopher and an individual suffering from psychotic episodes may claim that the moon does not exist.20 both persons may present arguments in defense of this claim, and both may also become emotionally agitated when others disagree with them. the difference between a philosopher denying the reality of the moon (and, say, all other non-living composite objects) and a person suffering from delusional beliefs is that, presumably, the philosopher has the ability2 to challenge and regulate her thoughts and beliefs about the moon while the person suffering from delusions cannot. similarly, most people report experiencing intrusive thoughts (e.g., about unwanted sex acts, violence, contamination, etc.), while only a small minority meet the diag18‘capacity’, here, refers to the neurocognitive resources necessary to acquire the relevant ability1. 19this argument extends to desires and behaviors currently falling under the category of ‘paraphilic disorders’ in the dsm-5. 20the example is adapted from van inwagen (1990). leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 14 nostic criteria for obsession-related disorders (rachman & silva, 1978; radomsky, 2014). the difference between normal (if still distressing) intrusive thoughts, and disorders (such as ocd), is demarcated by individuals’ ability2 to intelligently regulate their cognitive and emotional responses to them.21 the difference between mental health and disorder, in general, is drawn by individuals’ ability2 to regulate their mental lives, not by the content or objects of their cognitions and emotions. it is important to note what the skill view is not claiming. the skill view is not claiming that mental content plays no role in the etiology or experience of mental disordersthis is clearly not the case. for example, negatively valenced thoughts and emotions are constitutive of depression; a necessary condition of what it is to be depressed is the presence of depressed mood or diminished pleasure. but the presence and prevalence of negatively valenced mental content is not sufficient for mental disorder. nihilists and existential philosophers may regularly think about the meaninglessness of life, and may even feel intense and persistent angst or sadness in response to these thoughts, but they needn’t be suffering from clinical depression (or any other mental disorder). a predominance of negatively valenced thoughts and emotions may play a causal role in the development of depression, but this need not be the case. one needn’t be disordered to believe that life is meaningless and painful, to devote much of one’s mental life to these beliefs, and to even be profoundly saddened by this. what makes a state marked by negatively valenced mental content a disorder is the failure of skilled self-regulation of this content, not merely the presence of any specific thoughts or emotions. the difference between normal (even if intense) sadness, dysphoria, or hopelessness, on the one hand, and depression on the other, is that depression (and mental disorders in general) involve an inability2 to skillfully regulate how one responds to mental content. mental content, then, plays a role in mental health insofar as responses to content are often what is being regulated, but the presence or absence of any particular mental content does not explain why certain states are disordered and others not. if there is a mental health problem, it is because of a failure of skilled self-regulation, not because of content being regulated. the skill view is also not claiming that any problem of self-regulation is a problem of mental health. there are numerous instances of poor self-regulation that are clearly not disorders. these problems can range from the seemingly trivial (e.g., one can have difficulty resisting a bowl of sweets), to the potentially more serious (e.g., sticking to a diet, or keeping one’s temper while watching a sporting event). the skill view predicts that these cases are, or at least can be, normal problems of living. normal problems of living may be harmful and distressing, but they become mental disorders only if they involve unskilled self-regulation (which requires a genuine inability2 to intelligently modify, override, or change one’s responses). having difficulty regulating one’s diet is not a mental disorder, being 21the skill view provides a similar explanation for the symptoms of schizophrenia. both the positive and negative symptoms associated with schizophrenia can be found in the absence of any mental disorder (van os & reininghaus, 2016). these ‘symptoms’ reflect a disorder only insofar as they involve an inability2 to skillfully regulate how a person responds to them. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 15 unable2 to skillfully regulate one’s food intake is. and this is just what we should expect: mental disorders (such as eating disorders) involve genuine inabilites2 to skillfully self-regulate; normal problems of living (such as having a problem limiting carbohydrates in one’s diet) do not. the skill view does not claim that a lack of the capacity (as opposed to ability2) to skillfully self-regulate entails that an individual has a mental disorder. infants and small children may be unable2 to skillfully self-regulate, yet they are clearly not mentally disordered. the skill view argues that mental disorders are failures to skillfully self-regulate; this assumes that one has the capacity to accomplish this act. in the case of adult mental disorders, the disorders usually involve an inability2, rather than an inability1, to self-regulate (e.g., an adult experiencing clinical depression is likely to be able1 to self-regulate, but may be temporarily unable2 to exercise this skill). in contrast, infants and small children are poor selfregulators not because of a breakdown or failure of ability (1 and 2), but because of the lack of the development of the psychological tools necessary to skillfully selfregulate. small children do not have a problem of ability, but a (normal) lack of capacity. the skill view is also not claiming that the act of self-regulation, by itself, explains mental health. the act of self-regulation may be necessary for mental health, but it is clearly not sufficient; mental health, according to the view defended here, is skilled self-regulation. for example, an individual under the influence of intoxicants or psychotropic drugs may be willing to engage in harmful activities and regulate their responses towards this end. similarly, a person experiencing a manic episode (which is often marked by grandiosity, flight of ideas, and increased goal-directed behavior) may successfully regulate her responses towards some mania-inspired standard. neither case of successful self-regulation is normally considered to be part of healthy mental states. the difference between mere successful self-regulation to a potentially harmful standard, on one hand, and mental health, on the other, is that mental health requires the ability2 to intelligently and flexibly modify one’s responses. this skill is often masked or absent while individuals are experiencing delusions, manic states, or are under the influence of psychotropic drugs. the skill view also does not entail the (false) claim that mental disorders require a global inability2 to self-regulate. individuals suffering from addictions, for example, may be able2 to regulate their responses to their addiction-specific urges and desires in some circumstances (e.g., while in the presence of a police officer). the skill view claims that mental disorders require an inability2 to skillfully selfregulate (which is consistent with individuals being able2 to merely successfully regulate their responses in some situations). the skill view is also not claiming that mental health requires the ability2 to skillfully regulate one’s responses to all possible stimuli. for example, consider someone who lives an utterly untroubled mental life, but who might contingently be unable to skillfully self-regulate their responses to some psychiatric symptoms (say, auditory hallucinations) in other circumstances which, fortunately, do not obleder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 16 tain. intuitively, although this person may lack the skill to regulate her responses to certain psychiatric symptoms (if they were to manifest), this is a person in good mental health. and this is what the skill view predicts. the skill view argues that mental health is the skilled action of regulating one’s responses to one’s mental phenomena (not to all possible mental phenomena that a person may experience). in the hypothetical case noted above, the individual is mentally healthy as long as she is able to regulate her responses to her mental phenomena (which, in this case, does not involve any psychiatric symptoms). this person may certainly develop a mental disorder if she were to develop psychiatric symptoms that she was unable to intelligently self-regulate, but she is not mentally ill just because she lacks the skill to regulate her response to symptoms that she does not have. the skill view also does not entail the (false) claim that all instances of mental healing necessitate the development of new skills. mental healing may involve the development of new self-regulatory metacognitive skills, but it may also involve the unmasking of previously developed skills. mental health requires the exercise of self-regulatory metacognitive skill, and this can be accomplished in many ways (e.g., by acquiring new skills, strengthening old skills to adapt to new conditions, or by removing environmental factors that mask or overwhelm one’s current ability1 to intelligently self-regulate). finally, as noted in section 2, the skill view adopts a restricted definition of self-regulation that focuses on whether individuals are able2 to intelligently alter, modify, or control their responses to their mental phenomena, regardless of the standard that they are attempting to regulate to. this is not to claim that the standard is irrelevant to health; individuals’ goals or standards are often part of the problem that therapy is attempting to ameliorate (e.g., overly high or ‘perfectionist’ standards). in the case of perfectionist or unrealistic goals, therapy may attempt to modify a patient’s high standards, but the high standards are maladaptive only if the patient is genuinely unable2 to intelligently alter or control how she responds to them. impossibly high standards needn’t be maladaptive, and may often serve as useful motivators. the skill view argues that mental health is determined by one’s ability2 to intelligently modify, alter, control their responses towards the achievement of some end; it does not matter (at least in the context of mental health) how well one does this or how achievable the goal is. 3.4.1 skill and function the skill view offers a naturalist definition of mental disorders that is conceptually superior to traditional function-based views. the skill view, in contrast to function-based views, captures what is disordered about mental disorders (i.e., a failure of skilled self-regulation) without revising our conception of paradigmatic disorders and without pathologizing normal human difference. unlike dysfunction views, the skill view predicts that paradigmatic mental disorders such as depression, anxiety, and schizophrenia are clear instances of disorder insofar as they leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 17 involve failures of skilled self-regulation. and, unlike dysfunction views, the skill view predicts that mental states that merely involve norm-transgressing mental content and behavior (or even dysfunctional mental mechanisms), but do not involve problems of skilled self-regulation (such as homosexuality, asexuality, and norm-transgressing gender-identification), are clearly not mental disorders. mental disorders are best conceived of as problems of skilled mental regulation, not mental content or the proper functioning of mental mechanisms. the skill view of health, in contrast to function-based views, is also consistent with divergent theories of psychopathology and psychological healing. the causes of breakdowns in skilled self-regulation are likely various, and may include physiochemical, genetic, cognitive, or environmental factors (and likely some combination of multiple variables). a significant advantage of the skill view over functionbased views is that the identification of mental disorders does not depend upon the truth of any particular theory of psychopathology and psychological functioning. according to function-based views, accuracy about the etiology of hypothesized disorders is crucial, given that whether or not some way of being (such as anxiety or homosexuality) counts as a disorder necessarily depends on whether it is the result of a mental dysfunction. the skill view of health, on the other hand, is compatible with the etiology of mental disorders being an open empirical and conceptual question. mental disorders are failures or breakdowns of skillful selfregulation, whatever the cause. regardless of whether the etiology of a disorder is best described by a cognitive-behavioral theory, psychodynamic theory, or any other theory of psychopathology, if some state is a disorder, it is marked by an inability2 to skillfully self-regulate. 3.5 the epistemic boundary problem the skill view draws the metaphysical boundary between health and disorder at individuals’ ability2 to intelligently self-regulate. there remains the question of how we might know when this is the case. there are two types of epistemic challenges in identifying mental disorders: one is nosological, the other diagnostic. the nosological question is concerned with identifying what types of mental phenomena are and are not mental disorders (i.e., is the mental state normally associated with depression a disorder or a mere problem of living?). the diagnostic question is concerned with how clinicians might identify disorders, given the answers to the nosological question (i.e., how might we know if an individual meets the criteria for any given mental state that we postulate to be disordered?). this section argues that there is a degree of vagueness in the answer to both questions, but, importantly, the epistemic uncertainty facing the skill view is more benign than the vagueness facing function-based views. the skill view identifies mental disorders by identifying inabilities2 to intelligently self-regulate. this requires identifying when individuals cannot flexibly alter or control how they respond to their thoughts, emotions, and behaviors. so, leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 18 identifying specific disorders (such as depression, anxiety, and ocd) is a matter of identifying inabilities to skillfully self-regulate (e.g., to regulate responses to depressive thoughts and emotions, anxious feelings and worry, or obsessions and compulsions). psychometric and diagnostic tools like the dsm-5 and icd-10 are heuristics that should be seen as doing just this. while no one should think that these diagnostic tools are carving nature at its joints, they do a good, but imperfect job of identifying when individuals are unable2 to intelligently self-regulate and are in need of help. diagnostic tools are, of course, imprecise (e.g., there is no metaphysical significance to the dsm-5’s requirement of a minimum of two-weeks of symptom expression for the diagnosis of major depressive disorder, as opposed to three weeks or ten days). and diagnostic tools may be based on mistaken nosological assumptions (e.g., homosexuality was considered a mental disorder in the dsm-ii and was not fully removed until 1973). however, while the diagnostic tools’ lack of precision clearly has practical implications (insofar as diagnoses often entail improved access to resources and aid), this epistemic problem does not reflect a metaphysical shortcoming of the skill view (i.e., the skill view provides us with a framework to differentiate between accurate and mistaken nosological claims: we know that if someone is genuinely unable2 to skillfully self-regulate, then they are experiencing a mental disorder).22 the issue here is one of psychometrics, not metaphysics. the skill view also faces no unique diagnostic challenges. while the skill view and function-based views may differ over their nosological commitments, both views have to rely on the same diagnostic and psychometric tools (such as the dsm-5 and icd-10, and the beck depression inventory and the hospital anxiety and depression scale, respectively) in order to differentiate between health and disorder for any hypothesized nosological entity. for instance, even if one were to adopt a function-based conception of mental health, identifying the precise border between any particular individual’s poor functioning (say, normal, but intense, anxiety) and dys-functioning (say, gad) still comes down to the arbitrary drawing of lines. similarly, whether an individual is unable2 to intelligently self-regulate (as opposed to just being a poorly skilled self-regulator) is not always clear. but in both cases, this is an issue of psychometrics, not metaphysics. providing a specific answer to the diagnostic question is going to be a problem for any naturalist view, and the existence of fuzzy boundaries between health and disorder for any particular nosological entity is problematic only if we take diagnosis to accurately reflect metaphysical distinctions rather than serving as imperfect heuristics to identify breakdowns of skilled action. both dysfunction views and the skill view accept that diagnostic boundaries are fuzzy. this is a serious practical problem, but not a metaphysical one. what is crucial for a naturalist theory of mental disorder is that it provides a plausible value-neutral metaphysical grounding from which we can base the epistemic dif22in contrast, the imprecision of diagnostic tools is a serious problem for function-based views given the etiological commitments noted in section 3.4.1. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 19 ferentiation of genuine disorders from mere problems of living and value-based ‘disorders’. the skill view does this. 4 the skill view and dysfunction the previous section has shown that the skill view offers us a more plausible answer to the boundary problems than dysfunction-based theories. this section addresses some further points of difference. a significant difference between the skill view and function-based theories is that the skill view is a theory of mental health, not health, full stop. traditional function-based views adopt the same conception of health for both somatic and mental health. this uniformity may be considered an advantage in their favor. however, the cost of this uniformity in our conception of somatic and mental disorders is a theory of disorder that is both potentially massively revisionary and epistemically weak. this is a high price. while a thorough defense of this distinction is beyond the scope of this paper, it is important to note that the focus on skill, rather than function, closely aligns with how we normally conceive of mental health and healing.23 mental health is best conceived of as a quality of persons, not of impersonal mechanisms. dysfunction views assume that the same standard of health can be used for any living thing (be it a plant, a bacterium, or a mind). in contrast, the skill view posits different standards of health for minds and bodies. it is far from clear why we would think otherwise. why think that the questions concerning the health of a mind and the health of a plant or a bone are after the same information? two concerns may be that adopting different standards of health assumes some sort of substance dualism about minds and bodies and an anti-scientific conception of mental disorders. but neither worry is legitimate. we needn’t claim that minds are non-material to also posit that their health is best explained by skill rather than function. minds are more complex than bones and plants, and it is not surprising that their health is best judged by different standards. a second significant difference is that many advocates of function-based theories take the concept of normal function to be necessarily connected to our understanding of health. the two most influential naturalist theories of health, boorse’s (1976) ‘pure’ naturalist theory, and wakefield’s (1992) ‘hybrid’ view, both appear to share this view. for example, boorse states that: there is clearly some plausibility in the claim that the history of medical theory is nothing but a record of progressive investigation of normal functioning (1977, p. 560). 23see: leder (2019); zawidzki (2019). leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 20 and wakefield: [it is a] virtual universal tendency to fall back on dysfunction to explain disorder (1992, p. 381). boorse and wakefield may both be correct in claiming that health and disorder (or at least naturalist views of health and disorder) are normally conceptualized in functional terms. but, given the conceptual and epistemic problems facing function-based views, we have little reason to continue to do so. the intuitive strength of the presumed connection between normal function and health, if felt at all, should be dependent upon whether or not function-based conceptions of mental disorder can offer us explanatorily successful theories of health and disorder. the concept of natural function hasn’t been able to do this. skill can. 5 broader advantages of the skill view the previous sections have articulated a precise conception of mental health in terms of metacognitive self-regulatory skill and have shown its superiority to function-based conceptions, relative to a number of core theoretical and practical issues. in this section, we turn to some broader considerations in support of conceptualizing mental health as a skill. in particular, we argue that if mental health consists of metacognitive skills, this (1) shows a way past the historically recurring trilemma of psychiatric reductionism, eliminativism, and dualism; (2) defuses motivations for anti-psychiatry by accommodating cultural variance in mental illness within a value-neutral framework and providing an objective foundation for the scientific study of mental health and healing, and (3) suggests a novel and promising way of striking the delicate balance between stigma reduction and promoting individual agency in mental health. 5.1 avoiding the reductionism-eliminativism-dualism trilemma much of the history of psychiatry can be characterized, with little exaggeration, in terms of a periodic vacillation between three unstable ontological frameworks: reductionism, eliminativism, and dualism. according to reductionists, psychiatric disorders are nothing over and above (neuro-)physiological dysfunctions. according to eliminativists, psychiatric disorders do not really exist; only (neuro-)physiological dysfunctions exist. according to dualists, psychiatric disorders are afflictions of a domain which is, in some sense, autonomous from (neuro-)physiology. each of these frameworks has ultimately proven unstable, for different reasons. reductionism assumes that psychiatric disorders can be mapped directly onto (neuro-)physiological dysfunctions, but, as noted above, this has repeatedly been shown to be untenable. eliminativism makes psychiatry (as commonly practiced) impossible, and jeopardizes the very notion of mental leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 21 health. and dualism makes mysterious the patent dependence of mental health and disorder on physiological and other physical factors. it is plausible that one reason these three unstable, alternative frameworks have proven perennially seductive consists in the underlying assumption that mental disorder must consist in some kind of dysfunction. if it is not a (neuro-)physiological dysfunction, then it must either not exist, or consist in the dysfunction of some non-physical capacity. however, if mental health is conceptualized as a skill, then this seductive trilemma can be avoided. this is best appreciated in terms of an analogy to a typical embodied skill. consider tennis again. nobody is at all tempted by either reductionism, eliminativism, or dualism about skill at tennis. although (neuro-)physiological functionality, and other physical factors are required to successfully acquire and execute skill at tennis, there is no sense in which skill at tennis is the same thing as (neuro-)physiological functionality; this is why lack of skill at tennis is clearly not any kind of (neuro-)physiological dysfunction. however, this in no way supports eliminativism about skill at tennis; it is clearly a real phenomenon. nor does the fact that skill at tennis cannot be identified with (neuro-)physiological functionality in any way make dualism about skill at tennis at all tempting. skill at tennis just seems to belong to the wrong logical category to even raise the question of whether or not it consists in (neuro-)physiological functionality. the question of whether or not, or to what degree, someone is skilled at tennis just seems entirely orthogonal to questions of (neuro-)physiological functionality. the criteria by which and reasons why we make judgments about the former have nothing to do with the criteria by which and reasons why we make judgments about the latter. if we conceive of mental health in terms of self-regulatory metacognitive skills, rather than in terms of biologically normal functionality, then the fruitless vacillation among the horns of psychiatry’s historical, ontological trilemma can be avoided. in many respects, the idea here resembles ryle’s more general perspective on ‘the concept of mind’ (1949). ryle sought to avoid the traditional dilemma between dualism and reductionism (while resisting eliminativism), by arguing that folk psychological concepts are used to track abilities, skills, and dispositions of persons, rather than mechanical facts about our (neuro-)physiology, or “para-mechanical” facts about a “spectral” machine somehow lodged within it. similarly, we claim that the language of mental health and disorder is better understood as tracking metacognitive abilities, skills, and dispositions, rather than facts about (dys-)functioning (neuro-)physiological mechanisms, and that this reconceptualization can help us avoid the persistent ontological trilemma. on our view, like any other skill, self-regulatory, metacognitive skills are valued in specific cultural contexts, can be acquired to greater or lesser degrees by persons inhabiting those contexts, and depend on (neuro-)physiological functionality and other physical facts for their successful acquisition and exercise. none of this requires either reductionism, eliminativism, or dualism about them, any more than about skill at tennis. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 22 5.2 respecting cultural variance while avoiding antipsychiatry mental illnesses may manifest different symptomologies in different cultures and across time (radden, 2003). if we think of mental illnesses as dysfunctions of biological mechanisms, then this kind of cultural variance is hard to explain. the functionalist view is largely motivated by the desire to assimilate psychiatry to the rest of medicine (wakefield, 1992). but this kind of cultural variance in symptomology seems far less pronounced in the case of somatic illness: covid-19, cancer, heart disease, etc., all display remarkably homogeneous manifestations across cultures. if the functionalist view is the only possible foundation for scientific psychiatry, cultural variance in mental illness can naturally motivate skeptical and anti-psychiatry views. in contrast, if we think of mental illness as a deficit in self-regulatory metacognitive skill, then cultural variance can be accommodated without embracing antipsychiatry or sacrificing the objectivity of mental illness. although cultures can differ in the sorts of goals relative to which their members want to self-regulate, it remains an objective fact whether or not they are able2, skillfully, to do this. perhaps the task of managing grief differs across cultures. in one culture, it may require a stoic return to normal life within a few weeks, while in another, more expressive culture, it might require public expressions of grief over the course of months. such differences might yield different judgments of mental illness with respect to individuals incapable of meeting such goals. an individual may be judged mentally ill relative to a stoic culture, even if they are fully within the norm relative to a more expressive culture, and vice versa. however, on the skill view of mental health, there remains an objective fact of the matter as to whether or not they are mentally ill: if they cannot skillfully self-regulate relative to one of their (sometimes culturally determined) goals, e.g., appropriate grieving, then they count as objectively mentally ill. thus, the skill view predicts that cultural norms may influence the standard one attempts to regulate their responses to, but they do not influence whether any way of being should be considered disordered. once again, a comparison to embodied skills is pertinent here. consider speech therapy. this clearly involves a degree of cultural relativity, since different languages place different phonetic demands on their speakers. hence, someone may count as speech disabled relative to one language and not relative to another. nevertheless, there are objective facts about whether or not someone is speech disabled relative to some language, about the relevant (neuro-)physiological capacities, and about the sorts of behavioral interventions that enable them to overcome their disability. speech therapy is an entirely objective enterprise, based on investigation of such facts and the implementation of relevant interventions. just as such facts and interventions may partially but not wholly overlap across languages, our suggestion here is that facts and interventions relevant to psychiatric care may partially but not wholly overlap across cultures. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 23 the skill view offers a robust antidote to anti-psychiatry; self-regulatory metacognitive skills, like embodied skills, involve objective, (neuro-)physiological capacities that psychiatrists can study. this perspective suggests the following taxonomy of objective, psychiatric sciences. neuropsychiatry can be seen as the study of the neural mechanisms underlying self-regulatory, metacognitive skills. clinical psychiatry can be seen as the study of the sorts of interventions required to develop or unmask such skills (including therapies involving mindfulness and other forms of metacognitive practice). and applied clinical psychiatry can involve the implementation of such therapies and regimes of practice. the latter might be analogized to such well-established health professions as occupational therapy, which focus on (re-)training of patients wanting to re-enter the workforce after debilitating traumas. 5.3 striking a delicate balance: agency without stigma perhaps the greatest advantage of the skill view is that it suggests a promising understanding of agency with regard to mental illness which avoids the kind of stigma that typically accompanies failures of rational agency, as traditionally conceived. one often touted, practical advantage of function-based views concerns stigma reduction. if mental illness is no different from somatic illness (i.e., resulting from a biological dysfunction), then, so the argument goes, the mentally ill are no more to blame for their afflictions than those suffering from somatic illnesses. a person suffering from bipolar disorder, for example, should be subject to no more stigma than a person who has caught covid-19: these are caused by biological dysfunctions for which they are not responsible; they deserve medical care, not blame or stigmatization. thus, on function-based views, stigma reduction is bought at the price of the near-elimination of agency: like patients with somatic illnesses, the most patients with mental illnesses can do is consult medical experts, and follow their orders and prescriptions. and for this reason, they should not be blamed or stigmatized. in contrast, the skill view offers a more promising means of stigma reduction, which does not depend on such an extreme diminution of agency in the mentally ill. one traditional reason individuals with mental illness have suffered from stigma has to do with a particular conception of rational agency. the idea is that when a rational agent knows the rational course of action, they are, in virtue of this knowledge alone, appropriately motivated to engage in it. according to the most stringent, traditional versions of this principle, any individual who fails to match this pattern fails to qualify as a rational agent, and hence, does not deserve the regard to which only rational agents are entitled.24 but such impairments are typical of those experiencing mental disorders: in most mental disorders, individuals know that avoiding the patterns of behavior to which they are prone 24this understanding of practical rationality, and its relationship to the central category of moral status, i.e., personhood, has dominated the western philosophical canon, at least since kant. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 24 is the rational course of action, yet they are unable2 or inappropriately motivated to avoid them. viewing mental illness as a dysfunction, rather than such a lapse in rationality, can offer some respite from stigmas that attend the latter. however, these are not the only sorts of stigma that potentially accompany mental illness; biological dysfunctions, like somatic illnesses, carry their own forms of stigma. the evidence for the association between acceptance of dysfunction views and stigma reduction is mixed. while the acceptance of dysfunction views is associated with a reduction of perceived blame and responsibility of people with mental disorders (phelan, 2002), it is also associated with an increase in essentialist conceptions of illness (i.e., beliefs that illness is caused by biological processes over which individuals have little control), increased social rejection (e.g., beliefs about the dangerousness and ‘otherness’ of the mentally ill) (schomerus, 2012), and prognostic pessimism about recovery (deacon & baird, 2009). so, while the ‘dysfunction’ narrative may reduce blame, it may also promote social rejection, hinder recovery, and diminish patients’ sense of agency. hence, conceiving of mental illnesses as dysfunctions, while protecting against some forms of stigma, may foster others. in contrast, the skill view suggests a conception of agency different from the traditional understanding of rational agency, which has the potential to reduce the kind of stigma that attends failures of the latter, to avoid the stigma that attends dysfunction, and to promote a kind of agency among individuals with mental illness that is necessary for recovery. the skill view avoids two extreme conceptions of metacognitive agency: (1) taking metacognitive agency entirely for granted, as on some traditional views of rational agency, and (2) diminishing metacognitive agency, as on functionalist conceptions of mental health and illness. on the traditional view, to be practically rational just is to exercise metacognitive agency: one is rational in virtue of willing appropriate means to one’s goals; if one cannot do that, one just is not rational. on the functionalist view, there is limited metacognitive agency: choosing appropriate means to one’s goals is a matter of normal biological functioning, over which one has minimal control. the skill view provides a third alternative, inspired by notions of agency evident in embodied skill development, a notion of agency that neither ignores nor takes for granted metacognitive agency: on the skill view, metacognitive agency is something that persons can develop and maintain through practice, supported by appropriate social and physical scaffolds (e.g., supportive environments, protection against trauma, and undamaged (neuro-)physiological capacities). athletes are rightfully praised for developing and exercising their skills; they have a kind of agency. however, this is not a kind of agency that they have automatically, in virtue of being athletes. it is an agency that they have to develop, through an arduous and gradual process. similarly, we want to suggest, the metacognitive agency required for skilled self-regulation, in terms of which we understand mental health, is something that can be developed only gradually, through diligent practice (though much of this practice occurs during childhood leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 25 development). it is not a form of agency that we can just take for granted in virtue of being rational. hence, there are steps that individuals experiencing mental disorders can take to improve their lot, just as there are steps that most persons can take to improve athletic skills. they retain agency. however, it is not the kind of agency presumed by traditional views of practical rationality: it is fragile, and contingent on adequate physical and social scaffolding. a failure to demonstrate this kind of agency does not motivate stigma, in the way that a failure to demonstrate the kind of agency traditionally deemed to be necessary for practical rationality can. with the exception of some outlier communities, nobody is stigmatized for not having had the opportunity to develop tennis skills, for example, or for losing the ability2 to play tennis as the result of some trauma. similarly, we want to suggest, if we think of mental health as a skill that takes practice, together with contingent physical and social scaffolds, to develop and maintain it, there is no reason to stigmatize those who lack this ability or these opportunities. thus, we have here a way of allowing for agency in individuals with mental disorder (the same kind of agency that persons in athletics have: the ability1 to acquire or regain skills), while avoiding the stigma that accompanies lapses in traditional rationality and avoiding the stigma that accompanies dysfunction. it might be objected that the adoption of the skill view may increase negative stigma sometimes associated with mental disorders. the concern may be that the conception of mental disorders as failures or breakdowns of skilled action could encourage the idea that mental disorders are the results of personal weakness. if you want to become more skilled at some mode of functioning, the thought might go, you just need to practice and put in the work to develop it. while it is certainly true that avoiding and recovering from mental illness is to some degree a matter of effort (this is something that most theories of psychological healing accept), conceiving of mental health as a skill does not entail that mental health is merely a question of willpower or effort. there are constraints to the development of any skill that have nothing to do with strength of character or volition. individuals differ in their potential to develop specific skills due to biological, cognitive, and environmental differences. the skill of mental health, like skill in sports and intellectual activities, depends in large part on one’s physical and cognitive capacities. while it is likely that most individuals can improve their ability to perform skilled actions through practice and study, many factors constrain one’s development of skills; practice and diligence cannot overcome basic anatomical or cognitive constraints such as visual impairments or neurological deficits. the causes of variance in skill are likely very complex (including, but in no way limited to, a person’s prenatal environment, genetics, upbringing, and socio-economic standing), but clearly, they are not just a matter of will-power. the skill of mental health, like embodied or intellectual skills, is something that must be developed and maintained, and is significantly constrained or enabled by physiological and environmental factors. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org garson leder and tadeusz zawidzki 26 6 conclusion we have shown that a focus on skill, rather than function, provides us with a more explanatorily robust and pragmatically useful theory of mental health. on a theoretical level, this conception of mental illness allows for more intuitive and tenable assignment of extensions to our mental illness categories. on the pragmatic level, the focus on metacognitive skill rather than function enables a conception of mental disorder that is consistent with the scientific study of psychopathology while also promoting human agency. mental health, according to the skill view developed here, is something that individuals do. references american psychiatric association. (2013). diagnostic and statistical manual of mental disorders (5th ed.). andrews, p. w., & thomson, j. a. (2009). the bright side of feeling blue: depression as an adaptation for analyzing complex problems. psychological review, 116(3), 620–554. https://doi.org/10.1037/a0016242 baumeister, r. f., heatherton, t., & tice, m. (1993). when ego threats lead to self-regulation failure: negative consequences of high self-esteem. journal of personality and social psychology, 64(1), 141–156. https://doi.org/10.1037//00223514.64.1.141 baumeister, r. f., heatherton, t., & tice, m. (1994). losing control. academic press. beck, a. t., & alford, b. a. (2009). depression: causes and treatment. university of pennsylvania press. boorse, c. (1976). what a theory of mental health should be. journal for the theory of social behaviour, 6, 61–84. https://doi.org/10.1111/j.1468-5914.1976.tb00359.x boorse, c. (1977). health as a theoretical concept. philosophy of science, 44(4), 542–573. https://doi.org/10.1086/288768 boorse, c. (2014). a second rebuttal on health. journal of medicine and philosophy, 39(6), 683–724. https://doi.org/10.1093/ jmp/jhu035 carver, c. s., & scheier, m. f. (1982). control theory: a useful conceptual framework for personality social, clinical, and health psychology. psychological bulletin, 92, 111–135. https://doi.org/10.1037/0033-2909.92.1.111 cooper, r. (2007). psychiatry and the philosophy of science. mcgill-queens university press. deacon, b. j., & baird, g. l. (2009). the chemical imbalance explanation of depression: reducing blame at what cost? journal of social and clinical psychology, 28, 415–435. https://doi.org/10.1521/jscp.2009.28.4.415 duckworth, a. l., gendler, t. s., & gross, j. j. (2016). situational strategies for self-control. perspectives on psychological science: a journal of the association for psychological science, 11(1), 35–55. https://doi.org/10.1177/1745691615623247 fujita, k. (2011). on conceptualizing self-control as more than the effortful inhibition of impulses. personality and social psychology review, 15(4), 352–366. https://doi.org/10.1177/1088868311411165 fulford, k. w. m. (1989). moral theory and medical practice. cambridge university press. graham, g. (2013). the disordered mind: an introduction to philosophy of mind and mental illness (2nd revised edition). routledge. kingma, e. (2013). naturalist accounts of disorder. in k. w. m. fulford, m. davies, g. graham, j. sadler, g. stanghellini, & t. thorton (eds.), oxford handbook of philosophy and psychiatry (pp. 363–384). oxford university press. kingma, e. (2014). naturalism about health and disease: adding nuance for progress. journal of medicine and philosophy, 39(6), 590–608. https://doi.org/10.1093/jmp/jhu037 leder, g. (2019). what does it mean to have a meaning problem? meaning, skill, and the mechanisms of change in psychotherapy. philosophy, psychiatry, & psychology, 26(3), e-35-e-50. https://doi.org/10.1353/ppp.2019.0027 lilienfeld, s. o., & marino, l. (1995). mental disorder as a roschian concept: a critique of wakefield’s “harmful dysfunction” analysis. journal of abnormal psychology, 104, 411–420. https://doi.org/10.1037//0021-843x.104.3.411 maier, j. (2021). abilities. in e. n. zalta (ed.), the stanford encyclopedia of philosophy. metaphysics research lab, stanford university. https://plato.stanford.edu/entries/abilities/ mele, a. r. (2003). agents’ abilities. nous, 37(3), 447–470. https://doi.org/10.1111/1468-0068.00446 murphy, d. (2006). psychiatry in the scientific image. mit press. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1037/a0016242 https://doi.org/10.1037//0022-3514.64.1.141 https://doi.org/10.1037//0022-3514.64.1.141 https://doi.org/10.1111/j.1468-5914.1976.tb00359.x https://doi.org/10.1086/288768 https://doi.org/10.1093/jmp/jhu035 https://doi.org/10.1093/jmp/jhu035 https://doi.org/10.1037/0033-2909.92.1.111 https://doi.org/10.1521/jscp.2009.28.4.415 https://doi.org/10.1177/1745691615623247 https://doi.org/10.1177/1088868311411165 https://doi.org/10.1093/jmp/jhu037 https://doi.org/10.1353/ppp.2019.0027 https://doi.org/10.1037//0021-843x.104.3.411 https://plato.stanford.edu/entries/abilities/ https://doi.org/10.1111/1468-0068.00446 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the skill of mental health 27 murphy, d. (2020). concepts of disease and health. in e. n. zalta (ed.), the stanford encyclopedia of philosophy. metaphysics research lab, stanford university. https://plato.stanford.edu/entries/health-disease/ phelan, j. c. (2002). genetic bases of mental illness a cure for stigma? trends in neurosciences, 25(8), 430–431. https://doi.org/10.1016/s0166-2236(02)02209-9 rachman, s., & silva, p. de. (1978). abnormal and normal obsessions. behaviour research and therapy, 16, 411–423. https://doi.org/10.1016/0005-7967(78)90022-0 radden, j. (2003). is this dame melancholy?: equating today’s depression and past melancholia. philosophy, psychiatry, & psychology, 10(1), 37–52. https://doi.org/10.1353/ppp.2003.0081 radomsky, et al., a. s. (2014). part 1 you can run but you can’t hide: intrusive thoughts on six continents. journal of obsessive-compulsive and related disorders, 3(3), 269–279. https://doi.org/10.1016/j.jocrd.2013.09.002 rawn, c., & vohs, k. (2011). people use self-control to risk personal harm: an intra-interpersonal dilemma. personality and social psychology review, 15(3), 267–289. https://doi.org/10.1177/1088868310381084 ryle, g. (1949). the concept of mind. chicago university press. schomerus, et al., g. (2012). evolution of public attitudes about mental illness: a systematic review and meta-analysis. acta psychiatrica scandinavica, 125, 440–452. https://doi.org/10.1111/j.1600-0447.2012.01826.x stanley, j., & williamson, t. (2017). skill. nous, 51(4), 713–726. https://doi.org/10.1111/nous.12144 teasdale, j. d., & barnard, p. j. (1993). affect, cognition, and change: remodelling depressive thought. psychology press. van inwagen, p. (1990). material beings. cornell university press. van os, j., & reininghaus, u. (2016). psychosis as a transdiagnostic and extended phenotype in the general population. world psychiatry, 15(2), 118–124. https://doi.org/10.1002/wps.20310 vohs, k. d., & baumeister, r. f. (2004). understanding self-regulation. in r. f. baumeister & k. d. vohs (eds.), handbook of self-regulation (pp. 1–12). guilford. wakefield, j. c. (1992). the concept of mental disorder: on the boundary between biological facts and social values. american psychologist, 47, 373–388. https://doi.org/10.1037//0003-066x.47.3.373 watson, p. j., & andrews, p. w. (2002). toward a revised evolutionary adaptationist analysis of depression: the social navigation hypothesis. journal of affective disorders, 72(1), 1–14. https://doi.org/10.1016/s0165-0327(01)00459-1 woolfolk, r. l. (1999). malfunction and mental illness. the monist, 82(4), 658–670. https://doi.org/10.5840/monist199982429 world health organization. (1992). the icd-10 classification of mental and behavioural disorders: clinical descriptions and diagnostic guidelines. zawidzki, t. w. (2019). metacognitive skill and the therapeutic regulation of emotion. philosophical topics, 47(2), 27–51. https://doi.org/10.5840/philtopics201947214 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. leder, g., & zawidzki, t. (2023). the skill of mental health. philosophy and the mind sciences, 4, 3. https://doi.org/10.33735/phimisci.2023.9684 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://plato.stanford.edu/entries/health-disease/ https://doi.org/10.1016/s0166-2236(02)02209-9 https://doi.org/10.1016/0005-7967(78)90022-0 https://doi.org/10.1353/ppp.2003.0081 https://doi.org/10.1016/j.jocrd.2013.09.002 https://doi.org/10.1177/1088868310381084 https://doi.org/10.1111/j.1600-0447.2012.01826.x https://doi.org/10.1111/nous.12144 https://doi.org/10.1002/wps.20310 https://doi.org/10.1037//0003-066x.47.3.373 https://doi.org/10.1016/s0165-0327(01)00459-1 https://doi.org/10.5840/monist199982429 https://doi.org/10.5840/philtopics201947214 https://doi.org/10.33735/phimisci.2023.9684 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction mental disorders and mental functions the skill of mental health the value-neutrality of skill skill and self-regulation skill and mental health skill, mental disorder, and the boundary problem skill and function the epistemic boundary problem the skill view and dysfunction broader advantages of the skill view avoiding the reductionism-eliminativism-dualism trilemma respecting cultural variance while avoiding anti-psychiatry striking a delicate balance: agency without stigma conclusion exploring the range of reported dream lucidity. exploring the range of reported dream lucidity remington malletta michelle carrb,c martin freegardb karen konkolyd ceri bradshawb michael schredle abstract dream lucidity, or being aware that one is dreaming while dreaming, is not an all-or-none phenomenon. often, subjects report being some variant of “a little lucid” as opposed to completely or not at all. as recent neuroimaging work begins to elucidate the neural underpinnings of lucid experience, understanding subtle phenomenological variation within lucid dreams is essential. here, we focus on the variability of lucid experience by asking participants to report their awareness of the dream on a 5-point likert scale (from not at all to very much). participants implemented a combination of mnemonic training lucid dream induction methods at home for one week and provided detailed reports about their dream experiences each morning. consistent with previous research, cognitive induction methods led to about half of participants reporting at least one lucid dream and about half of all dreams including some level of lucidity. however, we also show that induction success rate varies significantly depending on the minimum criteria for lucidity. participants also reported how much they adhered to specific components of each induction method, and the amount of mnemonic rehearsal during a brief early awake period was predictive of lucidity level. furthermore, lucidity levels were positively correlated with dream control, dream bizarreness, and next-morning positive affect. lastly, we asked participants open-ended questions about why they chose particular levels of lucidity. we focus a qualitative discussion on responses to those “semi-lucid” dreams (rated just a little, moderately, or pretty much lucid) to explore why participants rate their dreams as having intermediate levels of awareness. together, the present study explores the frequency of semi-lucid dreams, what they are, why they might arise, their correlates, and how they impact methodological concerns in lucid dreaming research. keywords consciousness ∙ dreaming ∙ lucidity ∙ lucid dream induction ∙ phenomenology auniversity of texas at austin; corresponding author (mallett.remy@gmail.com) bswansea university cuniversity of rochester medical center dnorthwestern university ecentral institute of mental health, medical faculty mannheim/heidelberg university mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org r. mallett, m. carr, m. freegard, k. konkoly, c. bradshaw, and m. schredl 2 1 introduction in 1913, dutch psychologist frederik van eeden coined the term “lucid dream” when he introduced the modern scientific community to the phenomenon of being aware of a dream in real-time: […] the type of dreams which i called “lucid dreams,” seems to me the most interesting and worthy of the most careful observation and study[…] in these dreams the reintegration of the psychic functions is so complete that the sleeper remembers day-life and his own condition, reaches a state of perfect awareness, and is able to direct his attention, and to attempt different acts of free volition. yet the sleep, as i am able to confidently state, is undisturbed, deep and refreshing. (van eeden, 1913, p. 446) this general description of lucid dreaming has been consistent from the earliest objective verifications of lucid dreaming (hearne, 1978; laberge et al., 1981) up to the widely-accepted modern definition of “becoming aware that one is dreaming while dreaming” (baird et al., 2019, p. 305). while not explicitly stating so, this definition implies a binary nature of lucidity – either to be or not to be aware of the dream. however, only pages after introducing the scientific definition that still holds today, van eeden hints at the variability of lucid experience by observing in one instance that “[t]he lucidity had not been very intense, and i had some doubts about my real condition” (1913, p. 445). in this case, it seems that van eeden expresses only a partial “level” of lucidity. despite not having full realization of the dream, he was aware of his condition beyond the typical single-mindedness of non-lucid dreaming (rechtschaffen, 1978). online surveys of lucid dream experiences suggest there is large variation in how individuals pursue, experience, and utilize lucid dreams (lemyre et al., 2020; mota-rolim et al., 2013; stumbrys, erlacher, johnson, et al., 2014). being fully lucid and aware of the dream state implies that the dreamer can execute predetermined actions within the dream (erlacher & schredl, 2008), yet participants report often not being able to complete their intended actions due to “insufficient clarity” (stumbrys, erlacher, johnson, et al., 2014). others have reported incidences where lucidity wanes amidst lucid dream task completion, or even where lucid dream tasks are completed without specific dream awareness (worsley, 1984). moss (1986) proposed a “dream lucidity continuum” ranging from minor dream awareness to fullfledged lucidity and suggested that these lucidity levels are often traversed within a single dream. similarly, alan worsley, one of the most studied and documented modern lucid dreamers, noted “rapid change in the level of lucidity” of his dreams and that “how well [he knows he is] dreaming varies from moment to moment within the same dream” (1988, p. 327). brooks & vogelsong (1999) provide a more direct example of how awareness of the dream can be present even without a complete grasp of the situation. in a lucid dream report, the dreamer “realized he had mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org exploring the range of reported dream lucidity 3 left his keys and wallet inside a dream house and made the keys appear in his hand so he could unlock the door to retrieve the wallet” (1999, p. 28). of course, keys aren’t strictly necessary to enter a dream house, which should be even more obvious to the dreamer since they recognized their ability to will the spontaneous presence of keys. though the dreamer might have other motivation to follow the existing dream layout (e.g., dream stability), such contradictory logic is common even after achieving the analytical feat of dream awareness. there are many early descriptive accounts of lucid dreaming variation (gebremedhin, 1987; gillespie, 1984; kellog iii, 1989; moss, 1986; tart, 1984, 1985; worsley, 1984). initial attempts to formalize the lucidity continuum proposed a single “pre-lucid” stage between non-lucid and lucid dreams (green, 1968). in such pre-lucid dreams, the dreamer questions the nature of the dream, but might conclude (incorrectly) that they are not dreaming. still others have argued for the necessity of a fourth “incipient/implicit pre-lucid” dream that captures cases where the dreamer observes bizarreness without asking the question of whether they are dreaming (sparrow et al., 2013, 2018). pre-lucidity was implemented as a level of awareness in subsequent scales (leslie & ogilvie, 1996; ogilvie et al., 1982; purcell et al., 1986; stewart & koulack, 1989). some of these scales offered even more within-lucidity variation, based on length/stability of lucidity (ogilvie et al., 1982) or the amount of control in the lucid dream (stewart & koulack, 1989). more modern approaches to measuring lucidity include a variety of questionnaires based on indirect measures of dream awareness, such as levels of insight or self-reflectiveness (dresler et al., 2014; kahan & laberge, 2011; kahan & sullivan, 2012; lee & kuiken, 2015; voss et al., 2013). to-date, the current focus of these scales has not been to quantify the variation within lucid dreams, but rather to strictly contrast lucid versus non-lucid dreams or waking. voss et al. (2013) used a 6-point likert scale ranging from strongly disagree (0) to strongly agree (5), with probes such as “while dreaming, i often asked myself whether i was dreaming.” factor analysis on these 28 probes revealed that lucid dreams contained more insight, control, thought, memory, dissociation, and positive emotion than non-lucid dreams. additionally, some lucidity-related factors – control, thought, memory, and positive emotion – varied significantly depending on whether the dreams were reported from home or the laboratory. notably, the scale’s range between strongly disagree and strongly agree suggests a neutral point (at 2.5), and although some factors were higher in lucid dreams, they were still below neutral (e.g., control and dissociation). therefore, it’s unclear if they should be determined as characteristics of lucidity (voss et al., 2018). contrasting views have resulted in different lucid dreaming definitions across studies, for example some including control as a defining characteristic (schädlich & erlacher, 2012; see also horton, 2020) and others dissociation (voss et al., 2014). these results highlight the difficulty in measuring lucidity and the characteristics that may or may not encompass dream awareness (see also windt & voss, 2018), particularly through statements that do not directly probe dream awareness. mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org r. mallett, m. carr, m. freegard, k. konkoly, c. bradshaw, and m. schredl 4 only a few questionnaires designed to quantify dream lucidity ask the explicit question about how aware of the dream the participant was. kahan (1994) asked participants to self-report their lucidity directly on a likert scale (1-7 with n/a option as 0) by asking them “were you aware of dreaming while in the dream?” using the dream rating scale. rather than most responses clustering at 1 (not at all) or 7 (very much), 33% of dreams included a middle level of lucidity (response options 2-6). she concluded that dreams contain a variety of lucidity levels as opposed to a distinct grouping. the dream lucidity questionnaire (dlq; stumbrys et al., 2013) is a 12-item questionnaire that asks participants how much they agree with certain statements related to lucid dreaming (e.g., how much control they had over the dream, or how sure they were about being asleep). responses are made on a 5-point likert scale (0 = not at all, 1 = just a little, 2 = moderately, 3 = pretty much, 4 = very much), and the first question is a direct assessment of lucidity on a continuous scale (i was aware that i was dreaming). dyck et al. (2017) administered the dlq to investigate the efficacy of a variety of lucidity induction methods and found that while 10% of dreams were pretty much or very much lucid, still 33% of dreams were just a little or moderately lucid. the use of these continuous scales has provided critical insight into the non-binary nature of dream awareness, but the lack of dream reports in the previous studies prevents an understanding of the phenomenology that underlie such ratings. the goal of the current study was to explore the frequency and phenomenology of semi-lucid dreams, or those that are within the extreme bounds of a likert lucidity probe. towards this aim, we took the simplest method of probing a lucidity continuum (how aware of the dream were you, 0-4) and investigated (1) how the full range of the scale was utilized, (2) how measures of induction success were impacted by varying the minimum likert criterion for lucidity, (3) the relationship between the range of lucidity and other dreaming and waking characteristics, and (4) the phenomenology of semi-lucidity. the notion of lucid dream variability has important implications for how we measure lucidity, which is a crucial topic in this developing field of research (baird et al., 2019). 2 methods 2.1 participants thirteen undergraduate psychology students from swansea university participated in exchange for course credit. all participants were female within the age range of 18 and 20 years (m = 19.3, sd = 0.7). participants responded to an advertisement for a study about lucid dreaming. ethics were approved by swansea university. mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org exploring the range of reported dream lucidity 5 2.2 procedure and lucid dream induction methods participants were emailed an information packet that included general information about lucid dreaming and specific lucid dream induction methods that they were to practice daily for the following week (modeled after aspy et al., 2017). the email also included a link to a web-based survey that they were asked to complete each morning (see measures). there is a large variety of existing induction methods (price & cohen, 1988; stumbrys, erlacher, schädlich, et al., 2012), yet those that require only behavioral/cognitive training are the easiest to implement in field studies. therefore, we asked participants to perform a combination of cognitive lucid dream induction methods: both the mnemonic induction of lucid dreams (mild) and reality checking (rc). mild – as implemented in the current study – is an induction method that consists of waking up early (after about five hours of sleep), staying awake for a short period of time (generally 5-60 minutes), and returning to bed while mentally rehearsing the intention of becoming lucid during the next dream (laberge, 1980). rc is an induction method that is based on brief but regular “reality checks” throughout the day, which are moments where one contemplates the question of whether they are currently dreaming or not (tholey, 1983). note that mild and rc are complementary in that they operate at different timescales. while mild is a specific practice that occurs during early morning hours and is expected to have more immediate consequences, rc occurs throughout the day and presumably has a more longitudinal impact. participants were asked to complete both practices to the extent they were comfortable doing so. 2.3 measures participants were asked to, on each morning, complete a web survey through psytoolkit (stoet, 2017, 2010). the survey consisted of a series of questionnaires and open-ended questions. it began with questions regarding their sleep (e.g., bedtime, sleep quality) and adherence to the induction methods for the previous twentyfour hours. following an open dream report, participants completed a custom 8-item questionnaire regarding dream characteristics (e.g., bizarreness, sensory vividness, emotionality) and mood upon awakening. questions were framed, for example, as please note the intensity of bizarreness in the dream. response options followed a 9-point likert scale (1 = very little, 9 = a lot, with an additional option 0 = no recall). the next questionnaire was an extended (19-item) version of the dream lucidity questionnaire (dlq). the original dlq (stumbrys et al., 2013) is a 12-item questionnaire assessing dream awareness and various aspects that often coincide with lucidity (e.g., dream control and access to waking memories). response options of the dlq follow a 5-point likert scale (0 = not at all, 1 = just a little, 2 = moderately, 3 = pretty much, 4 = very much). we extended the original dlq to include additional probes that assess further aspects of lucidity and dream control (as in dyck et al., 2017, 2018), the latter being modeled after the lucid dreaming skills questionnaire (lusk; schredl et al., 2018). mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org r. mallett, m. carr, m. freegard, k. konkoly, c. bradshaw, and m. schredl 6 though we administered the entire extended dlq, our analyses focus on the first probe (dlq-1), i was aware that i was dreaming, because this is the most straightforward and direct assessment of lucidity under the current literature definition. the only other dlq responses used in the current study were those related to control (dlq probes 4, 6, 8, and 10), which were averaged together for a single measure of dream control. the extended dlq was followed by the positive and negative affect schedule (panas; watson et al., 1988), which was administered to assess the impact of lucidity on morning affect. the panas consists of 20 probes about present moment feeling, equally split across positive and negative affect (e.g., how enthusiastic or scared one feels). panas response options follow a 5-point likert scale (0 = not at all, 4 = very much), and each of the positive and negative affect probes were summed for individual measures of positive and negative morning affect. finally, participants completed a series of open-ended questions designed specifically to probe why a lucidity level (i.e., dlq-1 response) was selected. participants were asked (1) why did you rate your awareness at the value you did (2) what kind of experience(s) gave you an impression of your selected level of awareness, and (3) what prevented you from attributing full awareness to your dream. we found the first question to ultimately be the most useful for interpreting why participants chose different levels of lucidity, and thus focused our results on responses to only that question. 2.4 analyses 2.4.1 lucidity induction first, we simply counted the frequency of each level of lucidity (dlq-1 response) reported by participants for each night of their diary. second, we combined all participants to get frequency counts for each lucidity level aggregated across all nights in the sample. third, we calculated induction success rates at the group level according to different lucid dream cutoffs. there are many ways to measure lucidity, as well as many ways to measure lucid dream induction success. to highlight this, we report lucid dream frequency (i.e., induction success) in three different ways. each participant’s lucid dreaming frequency was calculated as the fraction of nights that include a lucid dream according to each cutoff point (from a minimal cutoff of just a little lucidity up to a strict cutoff of very much). we averaged this lucid dreaming frequency across all participants to calculate induction success rate of the sample at each cutoff point. we repeated this process first using all diary nights, then using only those nights with dream recall. to compare the effects of varying the lucidity cutoff and changing the inclusion of nights without recall, we ran a 2-way repeated measures anova using the python package pingouin (vallat, 2018), with one factor as ‘nights included’ (all vs. only those with dream recall) and the other factor as ‘lucidity cutoff’ (just a little, moderately, pretty much, very much). lastly, we also report a binarized measure of lucid dream frequency that mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org exploring the range of reported dream lucidity 7 represents the fraction of participants that became lucid at least one night, again at each lucidity cutoff. to investigate how adherence to the prescribed induction methods influenced lucidity level, we ran a mixed effects ordinal regression model using the r package ordinal (christensen, 2019) with mild rehearsal length, mild awake time, and the number of reality checks performed the previous day as predictors of dlq-1 response. thus, each model predictor was evaluated for its independent contribution to variation in lucidity level. this analysis was first run including all nights regardless of whether a dream was recalled. we also report the same analysis after dropping nights without recall. 2.4.2 semi-lucid interrogation a main goal of the current study was to qualitatively assess the reasons for denoting a dream as semi-lucid. we interpreted responses 1-3 (just a little, moderately, and pretty much) to the dlq-1 as being semi-lucid, or semi-aware of the dream as it was occurring. to interrogate participants’ reason for a semi-lucid response selection, we focused on the open-ended question: why did you rate your awareness the way you did? answers were grouped according to dlq-1 response and evaluated qualitatively. 2.4.3 correlates of lucidity with a continuous measure of lucidity, we were able to investigate its relationship with other dream characteristics and morning affect. all correlations were run using kendall’s tau correlation measure, preferred for ordinal data (somers, 1962). because each participant reported a unique amount of dreams, we ran a resampling method where a random night with recall was sampled from each participant and then a single tau value was computed. this process was repeated 1000 times, and then all tau values were fisher z scored. for two-tailed significance tests, the smaller of the two proportions of z scored values above and below zero was doubled. 3 results 3.1 lucidity induction though all participants were asked to report their dream each morning for one week, participants contributed varying amounts of morning reports (figure 1a). when aggregating across all participants, each of the four nonzero lucidity options appears to be chosen roughly equally (figure 1b). the frequency of dreams reported as having nonzero lucidity (i.e., at least just a little) appear similar to the aggregate frequency of those with no lucidity (figure 1b, inner panel). across a variety of cutoffs used to measure lucid dream frequency, induction success varied mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org r. mallett, m. carr, m. freegard, k. konkoly, c. bradshaw, and m. schredl 8 from 5% to 69% (figure 1c). unsurprisingly, reporting lucid dream frequency as a function of all nights rather than only nights with dream recall resulted in lower frequencies (f = 8.2, p = .014), and implementing a more stringent criterion for a lucid dream resulted in lower frequencies (f = 10.0, p < .001). there was also an interaction, such that a more stringent lucidity criterion was less impacted by the choice of restricting to nights with dream recall (f = 5.0, p = .006). it’s clear visually in figure 1c that the rate of participants who became lucid at least once according to each cutoff is higher than the induction rates across all nights in a sample. these results suggest that how lucid dream frequency is measured has significant impacts on how induction method results are reported and interpreted. mild rehearsal length was a significant predictor of lucidity level (odds ratio = 1.2, odds ratio ci = [1.0, 1.4], p = .027; figure 2), but this was not the case for mild awake time (p = .133) nor the number of reality checks performed (p = .713). the effect of mild rehearsal length was consistent when we only included nights with dream recall (p = .062). these results suggest that the mental exercises involved in mild have a strong influence on lucidity level, even after controlling for the length of the mild waking period and recent reality check frequency. 3.2 interrogation of semi-lucidity participants reported dreams across all semi-lucid levels of dlq-1 (figure 1), and thus it seems the whole range of lucidity offered to participants was utilized. our qualitative assessment of the open question why did you rate your awareness the way you did? supports this notion as well (figure 3). participants tended to defend their selection of just a little lucidity with comments that suggested they were not lucid by the strict definition (e.g., “i only knew the situation was odd”; “i had no idea i was asleep”). the selection of moderately lucid seemed to consist of dreams in which there was a tendency towards observing non-realness but without explicit dream awareness, perhaps in line with notions of implicit pre-lucidity (e.g., “the dream at moments felt real”; “[…] the things occurring in my dream were too bizarre to be real life”). dreams reported as pretty much lucid start to include responses that appear to fit the modern criterion for lucidity (e.g., “because i realized i was dreaming”), but with deficits in features of lucidity such as dream length (e.g., “[…] it took me a while to realize i was asleep”), stability (e.g., “i was mostly aware that i was dreaming but at times things felt more real”), or control (e.g., “[…] i was able to partially control it”). in summary, when asked about their motivation for selecting semi-lucid levels, low-end semi-lucid dreams (i.e., just a little) tended to not include any level of “awareness” but rather just skepticism (similar to the existing definition of prelucid or implicit pre-lucid dreams). in contrast, high-end semi-lucid dreams (i.e., pretty much) tended to include awareness but not control. these results are important given the current lack of agreement over methods of measuring lucidity (baird et al., 2019). with a formal definition of just awareness and not specifically mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org exploring the range of reported dream lucidity 9 figure 1: lucidity induction. a) display of individual differences of dlq-1 range utilization. b) roughly half of nights with dream recall included nonzero lucidity, and all nonzero scale options were utilized similarly. upper right inset aggregates all the nonzero lucidity response options together, shown on the same scale. c) lucidity induction success varies across a variety of measurement approaches. evaluations include the frequency of lucid dreams across all nights (circles), across only nights with recall (squares), and the frequency of participants that had 1 or more lucid dreams (triangles). error bars represent sem. mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org r. mallett, m. carr, m. freegard, k. konkoly, c. bradshaw, and m. schredl 10 figure 2: induction method adherence and lucidity level. a) only mild rehearsal length was significantly predictive of attained lucidity level. note that slight variation around each lucidity level (y-axis gridlines) is to show all data points and does not represent variation in values, and participants might contribute multiple datapoints to each plot. b) the significant effect of mild rehearsal length plotted differently, as continuous model predictions after being fit with empirical data. figure 3: semi-lucidity interrogation. representative defenses of why participants picked each dlq-1 response (i.e., lucidity level). control, it seems our low-end semi-lucid dreams do not meet this definition, but our high-end semi-lucid dreams do. these results might be interpreted as lucidity levels binning into “low” and “high” lucidity, yet this brief assessment seems consistent with the use of a 5-point likert scale to assess lucidity, as each response level reveals a unique pattern of response profiles. mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org exploring the range of reported dream lucidity 11 figure 4: lucidity level correlates. a) dream control and dream bizarreness were the strongest dream characteristic predictors of lucidity level. positive, but not negative, morning affect increased along with lucidity. violin shadings represent the full distribution of resampled z values, with error bars highlighting the 95% confidence intervals. stars indicate p < .05, triangles indicate p < .10. b) the strongest relationships with lucidity plotted with full dataset. note that slight variation around each lucidity level (y-axis gridlines) is to show all data points and does not represent variation in values, and participants might contribute multiple datapoints to each plot. slope lines are averaged across all resampled correlations. 3.3 correlates of lucidity participants reported dream characteristics and morning affect upon awakening, all of which were tested for correlations with reported lucidity (figure 4). lucidity level was correlated with dream control (mean z = .89, p < .001) and dream bizarreness (mean z = .61, p = .002). there was a positive trending relationship between lucidity and dream sensory vividness (mean z = .37, p = .084), as well as positive dream body sensations (mean z = .37, p = .080). we were also interested in how dream lucidity was related to affect upon awakening. we found that lucidity level was positively correlated with positive morning affect (mean z = .38, p = .032), but showed no relationship with negative morning affect (mean z = .06, p = .754). mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org r. mallett, m. carr, m. freegard, k. konkoly, c. bradshaw, and m. schredl 12 4 discussion now that there is little debate about the existence of lucid dreaming as a real phenomenon (baird et al., 2019), the field’s focus has moved from the goal of verifying lucid dreams (laberge et al., 1981) to their neural correlates (dresler et al., 2012; voss et al., 2009), how to induce them (blanchette-carrière et al., 2020; carr et al., 2020; laberge et al., 2018; stumbrys, erlacher, schädlich, et al., 2012; voss et al., 2014), what sleep stages they occur in (stumbrys & erlacher, 2012), and how they can be used for therapeutic purposes (ellis et al., 2020; macêdo et al., 2019; spoormaker & van den bout, 2006). with the current study, we propose that an additional line of investigation should be into the non-binary expression of dream awareness within lucid dreams (see also mota-rolim et al., 2010; noreika et al., 2010; stumbrys & erlacher, 2014). though several existing questionnaires use a continuous scale in measuring lucidity (e.g., stumbrys et al., 2013; voss et al., 2013), they are almost exclusively used to bin dreams into a dichotomous lucid or non-lucid categorization. emphasizing the full distribution of scores on these or novel scales might provide more sensitive insight into what induces lucidity and how lucidity impacts waking life. by focusing on a 5-point likert scale of specific awareness of the dream, we were able to observe a variety of novel features of the proposed lucidity continuum (moss, 1986). 4.1 semi-lucid dreams all nonzero lucidity levels (anything above not at all aware of the dream) were utilized equally across our sample. this finding suggests that lucidity follows a natural continuum. further, when we interrogated participants’ reasons for selecting semi-lucidity (just a little, moderately, or pretty much), responses were consistent with this notion. participants reported just a little lucidity as most frequently containing only skepticism about the dream or observations of bizarreness, consistent with the previous categorization of pre-lucid dreams (green, 1968), or more specifically, implicit pre-lucid dreams (sparrow et al., 2013, 2018). some such dreams were by all definitive criteria non-lucid, as participants were not aware of dreaming. moderately lucid dreams were similarly full of skepticism, albeit to a stronger degree and still fitting some criteria for implicit pre-lucidity. it was not until the pretty much lucid dreams that participants began to report explicit awareness of the dream. why then were these dreams not reported as very much lucid? most responses included certain limitations of lucidity, such as a lack of control, or a fleeting moment of lucidity that appeared only at the beginning or end of the dream. the current definition of lucid dreaming does not include dream control (baird et al., 2019), although dream control is typically higher in lucid than non-lucid dreams (laberge et al., 2018; voss et al., 2013). also note that van eeden’s (1913) original description of lucid dreams included dream control elements. our results warrant future discussion about how control fits into the strict definition of lucid dreaming (see also horton, 2020; windt & voss, 2018). mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org exploring the range of reported dream lucidity 13 viewing only the likert responses to i was aware that i was dreaming suggests that lucidity falls along a true continuum, and that further variation in response options might capture more of its variability. while our results suggest a lucidity continuum, they also do not refute a lucidity spectrum containing clear boundaries or quantized sections on a subscale. it is also possible to interpret the open-ended responses as evidence for a binary nature of lucidity. that is, under a strict definition of being fully aware of the dream as it is occurring, our results suggest that only pretty much and very much lucid dreams are truly lucid, while just a little and moderately lucid dreams might actually be preor non-lucid. the criterion of being fully lucid is not included in the current literature definition of a lucid dream, and so future consideration of this is important. others have argued that a good criterion of gauging lucidity is to simply ask the participant if they were lucid, offering just yes and no as response options (baird et al., 2019). while on the surface this seems the most straight-forward way of determining lucidity, it is possible that forcing participants into a binary categorization of something that might be non-binary could result in false positives and misses. another approach to lucidity variation, not exclusive to the above-mentioned, is that even within fully lucid dreams lies a variety of cognitive profiles (barrett, 1992; laberge & degracia, 2000; lee, 2018; sparrow, 2019). interesting new studies have investigated the degree to which self-reflectiveness relates to within-dream memory (e.g., during my dream, i remembered what happened earlier in the dream; lee, 2017, 2018). future work might investigate whether such high-level cognitive processes such as access to long-term memory vary as a function of reported lucidity. given the unique neural profile of lucid dreaming (baird et al., 2019), teasing apart the specific cognitive components of lucid dreaming can make broader contributions to how the waking brain effectively implements higher-order cognition. our interrogation of semi-lucid dreams also suggests that there are individual differences in what one qualifies as lucidity. how someone chooses to report their lucidity on a continuous scale is likely to be dependent on their previous lucid dreaming experience. if a participant’s first experience with lucidity is momentary and without dream control, they might rate it as very much lucid. but then after a subsequent lucid dream that includes dream control, the same participant might reevaluate their use of the lucidity response options and drop the same experience to pretty much lucid. a longitudinal dataset collected with a methodology similar to that presented here might be able to address how the use of the reported lucidity range changes as a function of experience with lucid dreams. in a similar vein, it seems clear that lucid dreaming can be learned (laberge, 1980; price & cohen, 1988; stumbrys, erlacher, schädlich, et al., 2012), yet whether it is a proper skill (i.e., stable after learning) or an ability (i.e., decreases after training) is still undetermined (schredl et al., 2018; stumbrys & erlacher, 2014). collecting continuous lucidity reports before, during, and after lucid dream training might help to answer this and related questions. mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org r. mallett, m. carr, m. freegard, k. konkoly, c. bradshaw, and m. schredl 14 4.2 methodological decisions in reporting induction success entangled with the issue of measuring lucidity is the question of how to measure lucid dream induction success (stumbrys & erlacher, 2014). the immense promise of lucid dreaming is limited by the capacity to experimentally induce lucidity (appel et al., 2018), and thus a leading goal of the field is to develop reliable lucid dream induction methods. the method of quantifying induction success rate is often inconsistent across experiments, and might be one of the contributing factors to literature discrepancies (stumbrys, erlacher, schädlich, et al., 2012). induction success is often reported as a proportion of dreams that are lucid, but this might be a proportion of all attempts, a proportion of all reported dreams (i.e., excluding attempts without dream recall), or a proportion of participants that became lucid (i.e., not accounting for multiple attempts within each participant). our results suggest that the success rate varies significantly depending on this selection. unsurprisingly, our results also show that the success rate is further dependent on the operational definition of dream lucidity; success rate decreases as the criterion for what constitutes a lucid dream becomes increasingly stringent along the range of reported lucidity. importantly, these success rates differ dramatically and have a significant impact on the interpretation of a given induction method. while we show that the choice of induction success measure matters, it is difficult to advise one over another. the optimal metric to use for success rate should differ across study motivations. on the one hand, a therapeutic approach might be more concerned with only how many participants become lucid, since the primary motivation is to increase lucidity at some point across repeated attempts. on the other hand, laboratory investigations, due to time and effort, might be more concerned with how effective an induction method is at inducing lucid dreams on a single attempt. similarly, despite the common practice of removing nights without dream recall (since it is possible a lucid dream was forgotten), knowing induction success across all attempts would be critical to how effective an induction method is. thus, a best-practices approach might be to include all success rates to aid in cross-study comparisons. another possible reason for discrepancies in induction success across studies and induction methods is that they might induce different profiles of lucidity (laberge & degracia, 2000; mota-rolim et al., 2010). as we highlight here, lucidity lies along a continuum, and different induction methods might induce different but predictable levels of lucidity. despite the combination of mild and reality checking inducing varying levels of lucidity in the current study, it’s possible that a given induction method might be effective in so far as it induces strictly low or strictly high levels of lucidity. often the goal of inducing lucid dreams is to induce fully lucid dreams – even dream control specifically – for experimental control over dream actions or for nightmare sufferers to overcome negative dream content (gieselmann et al., 2019; macêdo et al., 2019; zadra & pihl, 1997). thus, it will be crucial for future research to clarify the level of lucidity induced with a given induction method. mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org exploring the range of reported dream lucidity 15 4.3 adherence to lucidity induction protocols the mnemonic induction of lucid dreams method (mild), as implemented in the current study, involves waking up in the middle of the night and staying awake for a brief period of time while performing a mental rehearsal task (laberge, 1980). the mental rehearsal task to be performed during this brief awake period involves setting an intention to remember to become aware during the next dream and imagining the moment of lucidity. a related lucid dream induction method is known as wake-back-to-bed (wbtb), which entails waking up for a period of time during the night and then returning to sleep (stumbrys, erlacher, schädlich, et al., 2012; stumbrys & erlacher, 2014). because wbtb is primarily a behavioral technique and does not specify what action is performed while awake (e.g., appel et al., 2020; erlacher & stumbrys, 2020), it is often used in conjunction with cognitive techniques such as mild. to simplify instructions for our participants, we used the term “mild” in its originally proposed implementation (laberge, 1980), a combination of wbtb’s behavioral aspect and mild’s cognitive aspect. while both methods are effective at inducing lucidity, the mechanism behind their efficacy is unclear. wbtb is presumed to aid in “catching” a rem cycle, where lucid dreams are more likely to occur (laberge, 1988; laberge et al., 1981), a hypothesis that has recently received preliminary support (gott et al., 2020). another contributor to wbtb’s efficacy, not mutually exclusive from such rem re-entry, is the cognitive activity of the brief awake period (erlacher & stumbrys, 2020). mild was originally devised based on the notion of prospective memory, or setting an intention during waking to remember one is dreaming during the next dream (laberge, 1980; tholey, 1983), and our data suggest that the mental rehearsal component of mild also serves as a primary catalyst for inducing lucidity. aspy et al. (2017) found contrasting results in that longer mild practice led to a decreased chance of achieving lucidity, although they note that this effect was likely a result of mild practice impacting sleep habits. it is possible that our approach of relating induction method adherence to the degree of lucidity, rather than strictly lucid or not, can help to uniquely reveal the efficacy of certain lucid dream induction methods. length of time awake during mild (a.k.a. wbtb duration) is presumably an important factor in induction success, yet the existing literature is mixed. when looking at longer intervals, 30-60 minutes of awake time during the morning is optimal for inducing lucidity compared to shorter intervals of approximately 10 minutes (laberge et al., 1994), but a comparison within shorter intervals showed ~5 minutes awake time to be more effective than ~10 minutes (aspy et al., 2017). we found no effect of awake time on lucidity level, leaving this still an open question. another popular method of inducing lucidity is reality checks (stumbrys, erlacher, schädlich, et al., 2012; tholey, 1983). the likely reason for this method’s broad popularity and appeal is that it doesn’t incur any modification to typical sleep habits, which is a concern for other induction methods (soffer-dudek, 2020; vallat & ruby, 2019). the proposal behind the efficacy of reality checks is rooted in the continuity hypothesis of dreams, which states that we dream about themes from our waking life (schredl & hofmann, 2003), even if not specific episodic remallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org r. mallett, m. carr, m. freegard, k. konkoly, c. bradshaw, and m. schredl 16 play (fosse et al., 2003; malinowski & horton, 2014; mallett, 2020). if one asks oneself if one is dreaming throughout the day, then the theory of reality checks proposes that one will also ask oneself while one is dreaming, and ideally come to the correct conclusion that indeed, one is. notably, it is also possible to ask oneself if one is dreaming during a dream and incorrectly resolve that one is not. the efficacy of reality checks alone as an induction method is contentious (stumbrys, erlacher, schädlich, et al., 2012) and has recently shown no singular effect at increasing lucidity (aspy et al., 2017; dyck et al., 2017), however seems to work when used in combination with other induction methods (aspy et al., 2017; stumbrys, erlacher, schädlich, et al., 2012). in the current study, we found no relationship between the amount of reality checks performed in a single day and the amount of lucidity in the subsequent night, which is consistent with recent work (aspy et al., 2017). however, waking memories are frequently incorporated into dream content after a few days rather than on the subsequent night (blagrove et al., 2011; van rijn et al., 2015). future work might account for this “dream-lag effect” with the specific prediction that the number of reality checks performed a few days before might be more predictive of dream lucidity. 4.4 lucidity and bizarreness lucid and nonlucid dreams might be similar in bizarreness (voss et al., 2013), yet when rated by external judges, non-lucid dreams show higher bizarreness (yu & shen, 2020). our results suggest that increased dream bizarreness is related to increased dream lucidity. another recent study found lucid dreams to be higher in bizarreness than non-lucid dreams (laberge et al., 2018), although these results are conflated with the ingestion of galantamine, which might have influenced dream bizarreness independently of lucidity (although see sparrow et al., 2016). there is an intuitive notion that naturally occurring lucidity occurs as a result of a bizarre dream event (sparking the dreamer’s realization “how strange, i must be dreaming”), but this is inconsistent with the high frequency of bizarre events that do not provoke lucidity. as with our finding of a positive relationship between lucidity and sensory vividness, we are unable to determine if bizarreness induced lucidity. among other possibilities, perhaps increased lucidity allowed for a more direct/reflective evaluation of dream content, lucidity made dreams more bizarre, or a third variable (e.g., cortical arousal) induced both lucidity and bizarreness. a further consideration is that different levels of lucidity might be initiated by different dream characteristics. 4.5 positive impact of lucidity a promising benefit of lucid dreaming is its potential for nightmare therapy (abramovitch, 1995; aurora et al., 2010; garfield et al., 1988; gieselmann et al., 2019; holzinger et al., 2015; macêdo et al., 2019; mota-rolim & araujo, 2013; payne, 2014; taitz, 2014; zadra & pihl, 1997). the initially proposed idea behind mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org exploring the range of reported dream lucidity 17 lucid dreaming therapy is that lucid dreamers have control over their dream content, and thus can actively change the narrative of a dysphoric dream. however, recent studies have shown that lucid dreaming training sometimes improves nightmare symptomology despite a lack of induced lucidity (spoormaker et al., 2003; spoormaker & van den bout, 2006), or might reduce only nightmare-related symptoms (holzinger et al., 2020). our results suggest the possibility that in such situations, lucid dreaming training might induce low levels of lucidity potentially undetectable using a binary lucidity outcome measure. our semi-lucid interrogations are also in line with others suggesting that not all lucid dreams include dream control (mota-rolim et al., 2013; schädlich et al., 2017; schredl et al., 2018; stumbrys, erlacher, johnson, et al., 2014; windt & voss, 2018). but without inducing dream control, how could lucid dreaming therapy be effective? one possibility is that the moment of lucidity brings with it a sense of relief, even if the content can’t be changed. while this “just a dream” realization typically occurs when one awakens fully from a nightmare, there might be a particular impact of having this realization within the same environment of the nightmare event. for example, the emotional response to the ongoing nightmare might be different if the dreamer is aware of its non-reality. another way to assess the clinical benefit of lucid dreaming is to quantify affect in the morning following a lucid dream (konkoly & burke, 2019; stocks et al., 2020). our results suggest that increased lucidity leads to increased morning affect, which might be another potential clinical benefit to lucid dreaming beyond the strict control of nightmares. recent work suggests that characteristics of the dream state carry over into wakefulness (lee & kuiken, 2015; sikka et al., 2018), and lucid dreams contain more positive emotions than non-lucid dreams (mallett, 2020; stocks et al., 2020; voss et al., 2013). thus, the positivity of lucid dreams carrying over into the waking state might offer a reason for why lucid dreams without control could benefit nightmare sufferers (see also stocks et al., 2020). we found a clear positive relationship between lucidity and dream control, and so our results are unable to determine whether dream control was responsible for these positive waking impacts. notably, lucid dreaming therapy has recently been scrutinized for its potential negative impact due to occasional sleep interruptions and/or reality-fantasy questioning (soffer-dudek, 2020; vallat & ruby, 2019). our data shows no relationship between lucidity and negative affect or subjective sleep quality (as in aspy, 2020; ribeiro et al., 2020; schadow et al., 2018; schredl et al., 2020; stocks et al., 2020), however we did not implement measures of psychosis, which will be important for future lucid dream induction studies. while bodily sensations are under-appreciated within the discussion and description of lucid (garrett, 2017; kühle, 2015) and non-lucid (windt, 2010) dreams, the same logic of carry-over effects might apply to the positive body sensations that increased with lucidity levels in the current study. lucid dreamed actions share the neural substrate of waking actions (dresler et al., 2011; erlacher & schredl, 2008) and might have the same carry-over effects from dream experience into waking (stumbrys et al., 2016). mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org r. mallett, m. carr, m. freegard, k. konkoly, c. bradshaw, and m. schredl 18 4.6 the continuity of consciousness not only dream awareness, but also waking self-reflection fluctuates amongst a continuum (fazekas & overgaard, 2018, 2016; kahan & laberge, 2011; kahan & sullivan, 2012; smallwood & schooler, 2015). the unique state of non-lucid dreaming – and its variants (nielsen, 2017, 2004; windt et al., 2016) – can offer unique insight into the study of consciousness more broadly (fazekas & nemeth, 2018; hunt, 1986; revonsuo & valli, 2010; windt et al., 2016; windt & noreika, 2011). lucid dreaming is often referred to as a point along a continuum (fazekas & nemeth, 2018; hobson et al., 2000), and our results support other arguments that further studying variation within lucid dreams can offer further insight into studies of consciousness (baird et al., 2019; hobson, 2009). future research into the neurophysiology of lucidity variation might aid in understanding the (waking) neural basis of cognitive components that are specific to lucid dream sub-types (motarolim et al., 2010). the characterization of the continuous nature of self-reflective awareness in waking has important implications for the diagnosis and treatment of clinical disorders of consciousness (fernández-espejo & owen, 2013). 4.7 conclusion in summary, the current study was a largely exploratory investigation into the varieties of dream awareness. by focusing our analyses on a likert-scale probe aimed directly at dream lucidity, we explored specific relationships between reported lucidity level and induction adherence, dream characteristics, morning affect, and phenomenology. our goal was to contribute to the current methodological discussion of how lucidity should properly be measured and described in future research. in doing so, we showed that participants report dreams as existing along a continuum of lucidity, but how dream researchers interpret and apply the use of this continuum is still open for discussion. references abramovitch, h. (1995). the nightmare of returning home: a case of acute onset nightmare disorder treated by lucid dreaming. israel journal of psychiatry and related sciences, 32(2), 140–145. appel, k., füllhase, s., kern, s., kleinschmidt, a., laukemper, a., lüth, k., steinmetz, l., & vogelsang, l. (2020). inducing signal-verified lucid dreams in 40% of untrained novice lucid dreamers within two nights in a sleep laboratory setting. consciousness and cognition, 83, 102960. https://doi.org/10.1016/j.concog.2020.102960 appel, k., pipa, g., & dresler, m. (2018). investigating consciousness in the sleep laboratory – an interdisciplinary perspective on lucid dreaming. interdisciplinary science reviews, 43(2), 192–207. https://doi.org/10.1080/03080188.2017.1380 468 aspy, d. j. (2020). findings from the international lucid dream induction study. frontiers in psychology, 11. https://doi.org/ 10.3389/fpsyg.2020.01746 aspy, d. j., delfabbro, p., proeve, m., & mohr, p. (2017). reality testing and the mnemonic induction of lucid dreams: findings from the national australian lucid dream induction study. dreaming, 27(3), 206–231. https://doi.org/10.103 7/drm0000059 aurora, r. n., zak, r. s., auerbach, s. h., casey, k. r., chowdhuri, s., karippot, a., maganti, r. k., ramar, k., kristo, d. a., bista, s. r., lamm, c. i., & morgenthaler, t. i. (2010). best practice guide for the treatment of nightmare disorder in adults. journal of clinical sleep medicine, 06(04), 389–401. https://doi.org/10.5664/jcsm.27883 mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.concog.2020.102960 https://doi.org/10.1080/03080188.2017.1380468 https://doi.org/10.1080/03080188.2017.1380468 https://doi.org/10.3389/fpsyg.2020.01746 https://doi.org/10.3389/fpsyg.2020.01746 https://doi.org/10.1037/drm0000059 https://doi.org/10.1037/drm0000059 https://doi.org/10.5664/jcsm.27883 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org exploring the range of reported dream lucidity 19 baird, b., mota-rolim, s. a., & dresler, m. (2019). the cognitive neuroscience of lucid dreaming. neuroscience & biobehavioral reviews, 100, 305–323. https://doi.org/10.1016/j.neubiorev.2019.03.008 barrett, d. (1992). just how lucid are lucid dreams? dreaming, 2(4), 221–228. https://doi.org/10.1037/h0094362 blagrove, m., fouquet, n. c., henley-einion, j. a., pace-schott, e. f., davies, a. c., neuschaffer, j. l., & turnbull, o. h. (2011). assessing the dream-lag effect for rem and nrem stage 2 dreams. plos one, 6(10), e26708. https://doi.org/ 10.1371/journal.pone.0026708 blanchette-carrière, c., julien, s.-h., picard-deland, c., bouchard, m., carrier, j., paquette, t., & nielsen, t. (2020). attempted induction of signalled lucid dreaming by transcranial alternating current stimulation. consciousness and cognition, 83, 102957. https://doi.org/10.1016/j.concog.2020.102957 brooks, j. e., & vogelsong, j. a. (1999). the conscious exploration of dreaming: discovering how we create and control our dreams. 1st books library. carr, m., konkoly, k., mallett, r., edwards, c., appel, k., & blagrove, m. (2020). combining presleep cognitive training and rem-sleep stimulation in a laboratory morning nap for lucid dream induction. psychology of consciousness: theory, research, and practice. https://doi.org/10.1037/cns0000227 christensen, r. h. b. (2019). ordinal: regression models for ordinal data. https://cran.r-project.org/package=ordinal dresler, m., eibl, l., fischer, c. f., wehrle, r., spoormaker, v. i., steiger, a., czisch, m., & pawlowski, m. (2014). volitional components of consciousness vary across wakefulness, dreaming and lucid dreaming. frontiers in psychology, 4. https: //doi.org/10.3389/fpsyg.2013.00987 dresler, m., koch, s. p., wehrle, r., spoormaker, v. i., holsboer, f., steiger, a., sämann, p. g., obrig, h., & czisch, m. (2011). dreamed movement elicits activation in the sensorimotor cortex. current biology, 21(21), 1833–1837. https: //doi.org/10.1016/j.cub.2011.09.029 dresler, m., wehrle, r., spoormaker, v. i., koch, s. p., holsboer, f., steiger, a., obrig, h., sämann, p. g., & czisch, m. (2012). neural correlates of dream lucidity obtained from contrasting lucid versus non-lucid rem sleep: a combined eeg/fmri case study. sleep, 35(7), 1017–1020. https://doi.org/10.5665/sleep.1974 dyck, s., kummer, n., könig, n., schredl, m., & kühnel, a. (2018). effects of lucid dream induction on external-rated lucidity, dream emotions, and dream bizarreness. international journal of dream research, 74–78. https://doi.org/10.11588/ijo dr.2018.1.43867 dyck, s., schredl, m., & kühnel, a. (2017). lucid dream induction using three different cognitive methods. international journal of dream research, 10(2), 151–156. https://doi.org/10.11588/ijodr.2017.2.37498 ellis, j. g., koninck, j. d., & bastien, c. h. (2020). managing insomnia using lucid dreaming training: a pilot study. behavioral sleep medicine, 0(0), 1–11. https://doi.org/10.1080/15402002.2020.1739688 erlacher, d., & schredl, m. (2008). do rem (lucid) dreamed and executed actions share the same neural substrate? international journal of dream research, 7–14. https://doi.org/10.11588/ijodr.2008.1.20 erlacher, d., & stumbrys, t. (2020). wake up, work on dreams, back to bed and lucid dream: a sleep laboratory study. frontiers in psychology, 11. https://doi.org/10.3389/fpsyg.2020.01383 fazekas, p., & nemeth, g. (2018). dream experiences and the neural correlates of perceptual consciousness and cognitive access. philosophical transactions of the royal society b: biological sciences, 373(1755), 20170356. https://doi.org/10.1 098/rstb.2017.0356 fazekas, p., & overgaard, m. (2018). a multi-factor account of degrees of awareness. cognitive science, 42(6), 1833–1859. https://doi.org/10.1111/cogs.12478 fazekas, p., & overgaard, m. (2016). multidimensional models of degrees and levels of consciousness. trends in cognitive sciences, 20(10), 715–716. https://doi.org/10.1016/j.tics.2016.06.011 fernández-espejo, d., & owen, a. m. (2013). detecting awareness after severe brain injury. nature reviews neuroscience, 14(11), 801–809. https://doi.org/10.1038/nrn3608 fosse, m. j., fosse, r., hobson, j. a., & stickgold, r. j. (2003). dreaming and episodic memory: a functional dissociation? journal of cognitive neuroscience, 15(1), 1–9. https://doi.org/10.1162/089892903321107774 garfield, p., fellows, p., halliday, g., & malamud, j. r. (1988). clinical applications of lucid dreaming. in j. gackenbach & s. laberge (eds.), conscious mind, sleeping brain: perspectives on lucid dreaming (pp. 289–319). springer new york. https://doi.org/10.1007/978-1-4757-0423-5_12 garrett, l. (2017). utilising dreambody choreutics to integrate somatics with lucid dream criteria. body, movement and dance in psychotherapy, 12(2), 145–158. https://doi.org/10.1080/17432979.2016.1231134 gebremedhin, e. (1987). problems at refining the ”lucid” label: shooting at a moving target. lucidity letter, 6(1), 80. gieselmann, a., aoudia, m. a., carr, m., germain, a., gorzka, r., holzinger, b., kleim, b., krakow, b., kunze, a. e., lancee, j., nadorff, m. r., nielsen, t., riemann, d., sandahl, h., schlarb, a. a., schmid, c., schredl, m., spoormaker, v. i., steil, r., … pietrowsky, r. (2019). aetiology and treatment of nightmare disorder: state of the art and future perspectives. journal of sleep research, 28(4), e12820. https://doi.org/10.1111/jsr.12820 mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neubiorev.2019.03.008 https://doi.org/10.1037/h0094362 https://doi.org/10.1371/journal.pone.0026708 https://doi.org/10.1371/journal.pone.0026708 https://doi.org/10.1016/j.concog.2020.102957 https://doi.org/10.1037/cns0000227 https://cran.r-project.org/package=ordinal https://doi.org/10.3389/fpsyg.2013.00987 https://doi.org/10.3389/fpsyg.2013.00987 https://doi.org/10.1016/j.cub.2011.09.029 https://doi.org/10.1016/j.cub.2011.09.029 https://doi.org/10.5665/sleep.1974 https://doi.org/10.11588/ijodr.2018.1.43867 https://doi.org/10.11588/ijodr.2018.1.43867 https://doi.org/10.11588/ijodr.2017.2.37498 https://doi.org/10.1080/15402002.2020.1739688 https://doi.org/10.11588/ijodr.2008.1.20 https://doi.org/10.3389/fpsyg.2020.01383 https://doi.org/10.1098/rstb.2017.0356 https://doi.org/10.1098/rstb.2017.0356 https://doi.org/10.1111/cogs.12478 https://doi.org/10.1016/j.tics.2016.06.011 https://doi.org/10.1038/nrn3608 https://doi.org/10.1162/089892903321107774 https://doi.org/10.1007/978-1-4757-0423-5_12 https://doi.org/10.1080/17432979.2016.1231134 https://doi.org/10.1111/jsr.12820 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org r. mallett, m. carr, m. freegard, k. konkoly, c. bradshaw, and m. schredl 20 gillespie, g. (1984). problems related to experimentation while dreaming lucidly. lucidity letter, 3(2/3). gott, j., rak, m., bovy, l., peters, e., hooijdonk, c. f. m. van, mangiaruga, a., varatheeswaran, r., chaabou, m., gorman, l., wilson, s., weber, f., talamini, l., steiger, a., & dresler, m. (2020). sleep fragmentation and lucid dreaming. consciousness and cognition, 84, 102988. https://doi.org/10.1016/j.concog.2020.102988 green, c. (1968). lucid dreams (1st edition). institute of psychophysical research. hearne, k. m. (1978). lucid dreams: an electro-physiological and psychological study [phd thesis]. university of liverpool. hobson, j. a. (2009). the neurobiology of consciousness: lucid dreaming wakes up. international journal of dream research, 41–44. https://doi.org/10.11588/ijodr.2009.2.403 hobson, j. a., pace-schott, e. f., & stickgold, r. (2000). dreaming and the brain: toward a cognitive neuroscience of conscious states. behavioral and brain sciences, 23(6), 793–842. https://doi.org/10.1017/s0140525x00003976 holzinger, b., klösch, g., & saletu, b. (2015). studies with lucid dreaming as add-on therapy to gestalt therapy. acta neurologica scandinavica, 131(6), 355–363. https://doi.org/10.1111/ane.12362 holzinger, b., saletu, b., & klösch, g. (2020). cognitions in sleep: lucid dreaming as an intervention for nightmares in patients with posttraumatic stress disorder. frontiers in psychology, 11. https://doi.org/10.3389/fpsyg.2020.01826 horton, c. l. (2020). key concepts in dream research: cognition and consciousness are inherently linked, but do no not control “control”! frontiers in human neuroscience, 14. https://doi.org/10.3389/fnhum.2020.00259 hunt, h. t. (1986). some relations between the cognitive psychology of dreams and dream phenomenology. the journal of mind and behavior, 7(2/3), 213–228. https://www.jstor.org/stable/43853215 kahan, t. l. (1994). measuring dream self-reflectiveness: a comparison of two approaches. dreaming, 4(3), 177–193. https://doi.org/10.1037/h0094411 kahan, t. l., & laberge, s. p. (2011). dreaming and waking: similarities and differences revisited. consciousness and cognition, 20(3), 494–514. https://doi.org/10.1016/j.concog.2010.09.002 kahan, t. l., & sullivan, k. t. (2012). assessing metacognitive skills in waking and sleep: a psychometric analysis of the metacognitive, affective, cognitive experience (mace) questionnaire. consciousness and cognition, 21(1), 340–352. https://doi.org/10.1016/j.concog.2011.11.005 kellog iii, e. w. (1989). mapping territories: a phenomenology of lucid dream reality. lucidity letter, 8(2). konkoly, k., & burke, c. t. (2019). can learning to lucid dream promote personal growth? dreaming, 29(2), 113–126. https://doi.org/10.1037/drm0000101 kühle, l. (2015). insight: what is it, exactly? a commentary on ursula voss & allan hobson. in t. k. metzinger, j. m. windt, & j. m. windt (eds.), open mind (pp. 1–13). mind group. https://doi.org/10.15502/9783958570696 laberge, s. (1988). the psychophysiology of lucid dreaming. in j. gackenbach & s. laberge (eds.), conscious mind, sleeping brain: perspectives on lucid dreaming (pp. 135–153). springer new york. https://doi.org/10.1007/978-1-4757-0423-5_7 laberge, s. (1980). lucid dreaming as a learnable skill: a case study. perceptual and motor skills, 51(3_suppl2), 1039–1042. https://doi.org/10.2466/pms.1980.51.3f.1039 laberge, s., & degracia, d. j. (2000). varieties of lucid dreaming experience. in r. g. kunzendorf & b. wallace (eds.), individual differences in conscious experience (pp. 269–307). john benjamins publishing. laberge, s., lamarca, k., & baird, b. (2018). pre-sleep treatment with galantamine stimulates lucid dreaming: a doubleblind, placebo-controlled, crossover study. plos one, 13(8), e0201246. https://doi.org/10.1371/journal.pone.0201246 laberge, s., nagel, l. e., dement, w. c., & zarcone, v. p. (1981). lucid dreaming verified by volitional communication during rem sleep. perceptual and motor skills, 52(3), 727–732. https://doi.org/10.2466/pms.1981.52.3.727 laberge, s., phillips, l., & levitan, l. (1994). an hour of wakefulness before morning naps makes lucidity more likely. nightlight, 6(3). http://www.lucidity.com/nl63.ru.naps.html lee, m.-n. (2018). a phenomenological study of reflective awareness in dreams: characteristics of attention, memory, and anticipation. international journal of dream research, 6–12. https://doi.org/10.11588/ijodr.2018.1.40338 lee, m.-n. (2017). reflective awareness and cognitive abilities in dreams: implications for lucid dream research. international journal of dream research, 157–163. https://doi.org/10.11588/ijodr.2017.2.40294 lee, m.-n., & kuiken, d. (2015). continuity of reflective awareness across waking and dreaming states. dreaming, 25(2), 141–159. https://doi.org/10.1037/a0039147 lemyre, a., légaré-bergeron, l., landry, r. b., garon, d., & vallières, a. (2020). high-level control in lucid dreams. imagination, cognition and personality, 0276236620909544. https://doi.org/10.1177/0276236620909544 leslie, k., & ogilvie, r. (1996). vestibular dreams: the effect of rocking on dream mentation. dreaming, 6(1), 1–16. https: //doi.org/10.1037/h0094442 macêdo, t. c. f. de, ferreira, g. h., almondes, k. m. de, kirov, r., & mota-rolim, s. a. (2019). my dream, my rules: can lucid dreaming treat nightmares? frontiers in psychology, 10. https://doi.org/10.3389/fpsyg.2019.02618 mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.concog.2020.102988 https://doi.org/10.11588/ijodr.2009.2.403 https://doi.org/10.1017/s0140525x00003976 https://doi.org/10.1111/ane.12362 https://doi.org/10.3389/fpsyg.2020.01826 https://doi.org/10.3389/fnhum.2020.00259 https://www.jstor.org/stable/43853215 https://doi.org/10.1037/h0094411 https://doi.org/10.1016/j.concog.2010.09.002 https://doi.org/10.1016/j.concog.2011.11.005 https://doi.org/10.1037/drm0000101 https://doi.org/10.15502/9783958570696 https://doi.org/10.1007/978-1-4757-0423-5_7 https://doi.org/10.2466/pms.1980.51.3f.1039 https://doi.org/10.1371/journal.pone.0201246 https://doi.org/10.2466/pms.1981.52.3.727 http://www.lucidity.com/nl63.ru.naps.html https://doi.org/10.11588/ijodr.2018.1.40338 https://doi.org/10.11588/ijodr.2017.2.40294 https://doi.org/10.1037/a0039147 https://doi.org/10.1177/0276236620909544 https://doi.org/10.1037/h0094442 https://doi.org/10.1037/h0094442 https://doi.org/10.3389/fpsyg.2019.02618 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org exploring the range of reported dream lucidity 21 malinowski, j. e., & horton, c. l. (2014). memory sources of dreams: the incorporation of autobiographical rather than episodic experiences. journal of sleep research, 23(4), 441–447. https://doi.org/10.1111/jsr.12134 mallett, r. (2020). partial memory reinstatement while (lucid) dreaming to change the dream environment. consciousness and cognition, 83, 102974. https://doi.org/10.1016/j.concog.2020.102974 moss, k. (1986). the dream lucidity continuum. lucidity letter, 5(2), 25–28. mota-rolim, s. a., & araujo, j. f. (2013). neurobiology and clinical implications of lucid dreaming. medical hypotheses, 81(5), 751–756. https://doi.org/10.1016/j.mehy.2013.04.049 mota-rolim, s. a., erlacher, d., tort, a. b. l., araujo, j. f., & ribeiro, s. (2010). different kinds of subjective experience during lucid dreaming may have different neural substrates: commentary on “the neurobiology of consciousness: lucid dreaming wakes up” by j. allan hobson. international journal of dream research, 3(1), 33–35. https://doi.org/ 10.11588/ijodr.2010.1.596 mota-rolim, s. a., targino, z. h., souza, b. c., blanco, w., araujo, j. f., & ribeiro, s. (2013). dream characteristics in a brazilian sample: an online survey focusing on lucid dreaming. frontiers in human neuroscience, 7. https://doi.org/ 10.3389/fnhum.2013.00836 nielsen, t. a. (2017). microdream neurophenomenology. neuroscience of consciousness, 2017(1). https://doi.org/10.1093/nc /nix001 nielsen, t. a. (2004). chronobiological features of dream production. sleep medicine reviews, 8(5), 403–424. https://doi.or g/10.1016/j.smrv.2004.06.005 noreika, v., windt, j. m., lenggenhager, b., & karim, a. a. (2010). new perspectives for the study of lucid dreaming: from brain stimulation to philosophical theories of self-consciousness. international journal of dream research, 36–45. https://doi.org/10.11588/ijodr.2010.1.586 ogilvie, r. d., hunt, h. t., tyson, p. d., lucescu, m. l., & jeakins, d. b. (1982). lucid dreaming and alpha activity: a preliminary report. perceptual and motor skills, 55(3), 795–808. https://doi.org/10.2466/pms.1982.55.3.795 payne, j. d. (2014). the (gamma) power to control our dreams. nature neuroscience, 17(6), 753–755. https://doi.org/10.103 8/nn.3727 price, r. f., & cohen, d. b. (1988). lucid dream induction. in j. gackenbach & s. laberge (eds.), conscious mind, sleeping brain: perspectives on lucid dreaming (pp. 105–134). springer new york. https://doi.org/10.1007/978-1-4757-0423-5_6 purcell, s., mullington, j., moffitt, a., hoffmann, r., & pigeau, r. (1986). dream self-reflectiveness as a learned cognitive skill. sleep, 9(3), 423–437. https://doi.org/10.1093/sleep/9.3.423 rechtschaffen, a. (1978). the single-mindedness and isolation of dreams. sleep, 1(1), 97–109. https://doi.org/10.1093/sleep/ 1.1.97 revonsuo, a., & valli, k. (2010). dreaming as a model system for consciousness research. in e. k. perry, d. collerton, f. e. n. lebeau, & h. ashton (eds.), new horizons in the neuroscience of consciousness. john benjamins publishing. ribeiro, n., gounden, y., & quaglino, v. (2020). is there a link between frequency of dreams, lucid dreams, and subjective sleep quality? frontiers in psychology, 11. https://doi.org/10.3389/fpsyg.2020.01290 schadow, c., schredl, m., rieger, j., & göritz, a. s. (2018). the relationship between lucid dream frequency and sleep quality: two cross-sectional studies. international journal of dream research, 11(2), 154–159. https://doi.org/10.11588/ijodr.2 018.2.48341 schädlich, m., & erlacher, d. (2012). applications of lucid dreams: an online study. international journal of dream research, 134–138. https://doi.org/10.11588/ijodr.2012.2.9505 schädlich, m., erlacher, d., & schredl, m. (2017). improvement of darts performance following lucid dream practice depends on the number of distractions while rehearsing within the dream – a sleep laboratory pilot study. journal of sports sciences, 35(23), 2365–2372. https://doi.org/10.1080/02640414.2016.1267387 schredl, m., dyck, s., & kühnel, a. (2020). lucid dreaming and the feeling of being refreshed in the morning: a diary study. clocks & sleep, 2(1), 54–60. https://doi.org/10.3390/clockssleep2010007 schredl, m., & hofmann, f. (2003). continuity between waking activities and dream activities. consciousness and cognition, 12(2), 298–308. https://doi.org/10.1016/s1053-8100(02)00072-7 schredl, m., rieger, j., & göritz, a. s. (2018). measuring lucid dreaming skills: a new questionnaire (lusk). international journal of dream research, 11(1), 54–61. https://doi.org/10.11588/ijodr.2018.1.44040 sikka, p., pesonen, h., & revonsuo, a. (2018). peace of mind and anxiety in the waking state are related to the affective content of dreams. scientific reports, 8(1), 12762. https://doi.org/10.1038/s41598-018-30721-1 smallwood, j., & schooler, j. w. (2015). the science of mind wandering: empirically navigating the stream of consciousness. annual review of psychology, 66(1), 487–518. https://doi.org/10.1146/annurev-psych-010814-015331 soffer-dudek, n. (2020). are lucid dreams good for us? are we asking the right question? a call for caution in lucid dream research. frontiers in neuroscience, 13. https://doi.org/10.3389/fnins.2019.01423 mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1111/jsr.12134 https://doi.org/10.1016/j.concog.2020.102974 https://doi.org/10.1016/j.mehy.2013.04.049 https://doi.org/10.11588/ijodr.2010.1.596 https://doi.org/10.11588/ijodr.2010.1.596 https://doi.org/10.3389/fnhum.2013.00836 https://doi.org/10.3389/fnhum.2013.00836 https://doi.org/10.1093/nc/nix001 https://doi.org/10.1093/nc/nix001 https://doi.org/10.1016/j.smrv.2004.06.005 https://doi.org/10.1016/j.smrv.2004.06.005 https://doi.org/10.11588/ijodr.2010.1.586 https://doi.org/10.2466/pms.1982.55.3.795 https://doi.org/10.1038/nn.3727 https://doi.org/10.1038/nn.3727 https://doi.org/10.1007/978-1-4757-0423-5_6 https://doi.org/10.1093/sleep/9.3.423 https://doi.org/10.1093/sleep/1.1.97 https://doi.org/10.1093/sleep/1.1.97 https://doi.org/10.3389/fpsyg.2020.01290 https://doi.org/10.11588/ijodr.2018.2.48341 https://doi.org/10.11588/ijodr.2018.2.48341 https://doi.org/10.11588/ijodr.2012.2.9505 https://doi.org/10.1080/02640414.2016.1267387 https://doi.org/10.3390/clockssleep2010007 https://doi.org/10.1016/s1053-8100(02)00072-7 https://doi.org/10.11588/ijodr.2018.1.44040 https://doi.org/10.1038/s41598-018-30721-1 https://doi.org/10.1146/annurev-psych-010814-015331 https://doi.org/10.3389/fnins.2019.01423 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org r. mallett, m. carr, m. freegard, k. konkoly, c. bradshaw, and m. schredl 22 somers, r. h. (1962). a new asymmetric measure of association for ordinal variables. american sociological review, 27(6), 799–811. https://doi.org/10.2307/2090408 sparrow, g. s. (2019). fading light and sluggish flight: a two-dimensional model of consciousness in lucid dreams. international journal of dream research, 12(30), 82–88. https://doi.org/10.11588/ijodr.2019.1.57801 sparrow, g. s., carlson, r., & hurd, r. (2016). assessing the perceived differences in post-galantamine lucid dreams vs. nongalantamine lucid dreams. international journal of dream research, 71–74. https://doi.org/10.11588/ijodr.2016.1.25606 sparrow, g. s., hurd, r., carlson, r., & molina, a. (2018). exploring the effects of galantamine paired with meditation and dream reliving on recalled dreams: toward an integrated protocol for lucid dream induction and nightmare resolution. consciousness and cognition, 63, 74–88. https://doi.org/10.1016/j.concog.2018.05.012 sparrow, g. s., thurston, m., & carlson, r. (2013). dream reliving and meditation as a way to enhance reflectiveness and constructive engagement in dreams. international journal of dream research, 6(2), 84–93. https://doi.org/10.11588/ijo dr.2013.2.10151 spoormaker, v. i., & van den bout, j. (2006). lucid dreaming treatment for nightmares: a pilot study. psychotherapy and psychosomatics, 75(6), 389–394. https://doi.org/10.1159/000095446 spoormaker, v. i., van den bout, j., & meijer, e. j. g. (2003). lucid dreaming treatment for nightmares: a series of cases. dreaming, 13(3), 181–186. https://doi.org/10.1023/a:1025325529560 stewart, d. w., & koulack, d. (1989). a rating system for lucid dream content. imagination, cognition and personality, 9(1), 67–74. https://doi.org/10.2190/nnp2-22v0-6k7h-a6j8 stocks, a., carr, m., mallett, r., konkoly, k., hicks, a., crawford, m., schredl, m., & bradshaw, c. (2020). dream lucidity is associated with positive waking mood. consciousness and cognition, 83, 102971. https://doi.org/10.1016/j.concog.202 0.102971 stoet, g. (2017). psytoolkit: a novel web-based method for running online questionnaires and reaction-time experiments. teaching of psychology, 44(1), 24–31. https://doi.org/10.1177/0098628316677643 stoet, g. (2010). psytoolkit: a software package for programming psychological experiments using linux. behavior research methods, 42(4), 1096–1104. https://doi.org/10.3758/brm.42.4.1096 stumbrys, t., & erlacher, d. (2014). the science of lucid dream induction. in r. hurd & k. bulkeley (eds.), lucid dreaming: new perspectives on consciousness in sleep (vol. 1, pp. 77–102). praeger. stumbrys, t., & erlacher, d. (2012). lucid dreaming during nrem sleep: two case reports. international journal of dream research, 151–155. https://doi.org/10.11588/ijodr.2012.2.9483 stumbrys, t., erlacher, d., johnson, m., & schredl, m. (2014). the phenomenology of lucid dreaming: an online survey. the american journal of psychology, 127(2), 191–204. https://doi.org/10.5406/amerjpsyc.127.2.0191 stumbrys, t., erlacher, d., schädlich, m., & schredl, m. (2012). induction of lucid dreams: a systematic review of evidence. consciousness and cognition, 21(3), 1456–1475. https://doi.org/10.1016/j.concog.2012.07.003 stumbrys, t., erlacher, d., & schredl, m. (2016). effectiveness of motor practice in lucid dreams: a comparison with physical and mental practice. journal of sports sciences, 34(1), 27–34. https://doi.org/10.1080/02640414.2015.1030342 stumbrys, t., erlacher, d., & schredl, m. (2013). testing the involvement of the prefrontal cortex in lucid dreaming: a tdcs study. consciousness and cognition, 22(4), 1214–1222. https://doi.org/10.1016/j.concog.2013.08.005 taitz, i. y. (2014). clinical applications of lucid dreaming therapy. in r. hurd, k. bulkeley, & k. bulkeley (eds.), lucid dreaming: new perspectives on consciousness in sleep (vol. 1, pp. 167–192). praeger. tart, c. (1984). terminology in lucid dream research. lucidity letter, 3(1), 4–6. tart, c. (1985). what do we mean by lucidity? lucidity letter, 4(2), 12–17. tholey, p. (1983). techniques for inducing and manipulating lucid dreams. perceptual and motor skills, 57(1), 79–90. https: //doi.org/10.2466/pms.1983.57.1.79 vallat, r. (2018). pingouin: statistics in python. journal of open source software, 3(31), 1026. https://doi.org/10.21105/joss. 01026 vallat, r., & ruby, p. m. (2019). is it a good idea to cultivate lucid dreaming? frontiers in psychology, 10. https://doi.org/10 .3389/fpsyg.2019.02585 van eeden, f. (1913). a study of dreams. proceedings of the society for psychical research, 26, 431–461. van rijn, e., eichenlaub, j. b., lewis, p. a., walker, m. p., gaskell, m. g., malinowski, j. e., & blagrove, m. (2015). the dreamlag effect: selective processing of personally significant events during rapid eye movement sleep, but not during slow wave sleep. neurobiology of learning and memory, 122, 98–109. https://doi.org/10.1016/j.nlm.2015.01.009 voss, u., d’agostino, a., kolibius, l., klimke, a., scarone, s., & hobson, j. a. (2018). insight and dissociation in lucid dreaming and psychosis. frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018.02164 mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.2307/2090408 https://doi.org/10.11588/ijodr.2019.1.57801 https://doi.org/10.11588/ijodr.2016.1.25606 https://doi.org/10.1016/j.concog.2018.05.012 https://doi.org/10.11588/ijodr.2013.2.10151 https://doi.org/10.11588/ijodr.2013.2.10151 https://doi.org/10.1159/000095446 https://doi.org/10.1023/a:1025325529560 https://doi.org/10.2190/nnp2-22v0-6k7h-a6j8 https://doi.org/10.1016/j.concog.2020.102971 https://doi.org/10.1016/j.concog.2020.102971 https://doi.org/10.1177/0098628316677643 https://doi.org/10.3758/brm.42.4.1096 https://doi.org/10.11588/ijodr.2012.2.9483 https://doi.org/10.5406/amerjpsyc.127.2.0191 https://doi.org/10.1016/j.concog.2012.07.003 https://doi.org/10.1080/02640414.2015.1030342 https://doi.org/10.1016/j.concog.2013.08.005 https://doi.org/10.2466/pms.1983.57.1.79 https://doi.org/10.2466/pms.1983.57.1.79 https://doi.org/10.21105/joss.01026 https://doi.org/10.21105/joss.01026 https://doi.org/10.3389/fpsyg.2019.02585 https://doi.org/10.3389/fpsyg.2019.02585 https://doi.org/10.1016/j.nlm.2015.01.009 https://doi.org/10.3389/fpsyg.2018.02164 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org exploring the range of reported dream lucidity 23 voss, u., holzmann, r., hobson, a., paulus, w., koppehele-gossel, j., klimke, a., & nitsche, m. a. (2014). induction of self awareness in dreams through frontal low current stimulation of gamma activity. nature neuroscience, 17(6), 810–812. https://doi.org/10.1038/nn.3719 voss, u., holzmann, r., tuin, i., & hobson, a. j. (2009). lucid dreaming: a state of consciousness with features of both waking and non-lucid dreaming. sleep, 32(9), 1191–1200. https://doi.org/10.1093/sleep/32.9.1191 voss, u., schermelleh-engel, k., windt, j., frenzel, c., & hobson, a. (2013). measuring consciousness in dreams: the lucidity and consciousness in dreams scale. consciousness and cognition, 22(1), 8–21. https://doi.org/10.1016/j.concog.2012.11 .001 watson, d., clark, l. a., & tellegen, a. (1988). development and validation of brief measures of positive and negative affect: the panas scales. journal of personality and social psychology, 54(6), 1063–1070. https://doi.org/10.1037/00223514.54.6.1063 windt, j. m. (2010). the immersive spatiotemporal hallucination model of dreaming. phenomenology and the cognitive sciences, 9(2), 295–316. https://doi.org/10.1007/s11097-010-9163-1 windt, j. m., nielsen, t., & thompson, e. (2016). does consciousness disappear in dreamless sleep? trends in cognitive sciences, 20(12), 871–882. https://doi.org/10.1016/j.tics.2016.09.006 windt, j. m., & noreika, v. (2011). how to integrate dreaming into a general theory of consciousness—a critical review of existing positions and suggestions for future research. consciousness and cognition, 20(4), 1091–1107. https://doi.org/ 10.1016/j.concog.2010.09.010 windt, j. m., & voss, u. (2018). spontaneous thought, insight, and control in lucid dreams. in k. c. r. fox & k. christoff (eds.), the oxford handbook of spontaneous thought: mind-wandering, creativity, and dreaming (pp. 385–410). oxford university press. worsley, a. (1988). personal experiences in lucid dreaming. in j. gackenbach & s. laberge (eds.), conscious mind, sleeping brain: perspectives on lucid dreaming (pp. 321–341). springer new york. https://doi.org/10.1007/978-1-4757-04235_13 worsley, a. (1984). lucid dream definition. lucidity letter, 3(2), 11–12. yu, c., & shen, h. (2020). bizarreness of lucid and non-lucid dream: effects of metacognition. frontiers in psychology, 10. https://doi.org/10.3389/fpsyg.2019.02946 zadra, a. l., & pihl, r. o. (1997). lucid dreaming as a treatment for recurrent nightmares. psychotherapy and psychosomatics, 66(1), 50–55. https://doi.org/10.1159/000289106 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. mallett et al. (2021). exploring the range of reported dream lucidity. philosophy and the mind sciences, 2, 1. https://doi.org/10.33735/phimisci.2021.63 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1038/nn.3719 https://doi.org/10.1093/sleep/32.9.1191 https://doi.org/10.1016/j.concog.2012.11.001 https://doi.org/10.1016/j.concog.2012.11.001 https://doi.org/10.1037/0022-3514.54.6.1063 https://doi.org/10.1037/0022-3514.54.6.1063 https://doi.org/10.1007/s11097-010-9163-1 https://doi.org/10.1016/j.tics.2016.09.006 https://doi.org/10.1016/j.concog.2010.09.010 https://doi.org/10.1016/j.concog.2010.09.010 https://doi.org/10.1007/978-1-4757-0423-5_13 https://doi.org/10.1007/978-1-4757-0423-5_13 https://doi.org/10.3389/fpsyg.2019.02946 https://doi.org/10.1159/000289106 https://doi.org/10.33735/phimisci.2021.63 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction methods participants procedure and lucid dream induction methods measures analyses lucidity induction semi-lucid interrogation correlates of lucidity results lucidity induction interrogation of semi-lucidity correlates of lucidity discussion semi-lucid dreams methodological decisions in reporting induction success adherence to lucidity induction protocols lucidity and bizarreness positive impact of lucidity the continuity of consciousness conclusion is mental time travel real time travel? is mental time travel real time travel? michael barkasia (michael.barkasi@gmail.com) melanie g. rosenb (melanie.rosen@gmail.com) abstract episodic memory (memories of the personal past) and prospecting the future (anticipating events) are often described as mental time travel (mtt). while most use this description metaphorically, we argue that episodic memory may allow for mtt in at least some robust sense. while episodic memory experiences may not allow us to literally travel through time, they do afford genuine awareness of past-perceived events. this is in contrast to an alternative view on which episodic memory experiences present past-perceived events as mere intentional contents. hence, episodic memory is a way of coming into experiential contact with, or being again aware of, what happened in the past. we argue that episodic memory experiences depend on a causal-informational link with the past events being remembered, and that, assuming direct realism about episodic memory experiences, this link suffices for genuine awareness. since there is no such link in future prospection, a similar argument cannot be used to show that it also affords genuine awareness of future events. constructivist views of memory might challenge the idea of memory as genuine awareness of remembered events. we explain how our view is consistent with both constructivist and anti-causalist conceptions of memory. there is still room for an interpretation of episodic memory as enabling genuine awareness of past events, even if it involves reconstruction. keywords episodic memory ∙ mental time travel ∙ perceptual experience ∙ causal theory of memory ∙ objects of memory 1 in what sense is there mental time travel? going back to his book elements of episodic memory (1983), endel tulving describes episodic memory as enabling mental time travel. in a more recent article, he says “when one thinks today about what one did yesterday, time’s arrow is bent into a loop. the rememberer has mentally traveled back into her past” (2002, p. 2). this colorful way of describing episodic memory has been widely adopted by both scientists (e.g. beaty, seli, & schacter, 2019) and philosophers (e.g. michaelian, 2016b). tulving is well aware that mental time travel does not really bend time’s arrow: “an event happens […] memory traces are laid down […]. the memory auniversity of toronto mississauga baarhus university barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-4070-8196 https://orcid.org/0000-0002-7545-5558 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 2 traces […] are retrieved, and the person remembers the event. […] there is certainly no violation of any law of time” (tulving, 2002, p. 19). this suggests that talk of mental time travel (hereafter ‘mtt’) is just a catchy metaphor for an otherwise straightforward psychological mechanism.1 both scientists (suddendorf & corballis, 2007) and philosophers (byrne, 2010; debus, 2014; matthen, 2010) have taken it this way. while mtt certainly is not literal time travel, we shall argue here that it is more than mere metaphor. the neuro-cognitive system behind episodic memory is generally taken to facilitate not only past recall but also prospecting the future (beaty et al., 2019; buckner & carroll, 2007; schacter & addis, 2007; suddendorf & corballis, 2007; tulving, 2002), imagining counterfactual ways the past could have gone (de brigard, 2014), and other forms of episodic simulation (michaelian, 2016b ch. 6). thus, we should not talk about a dedicated episodic memory system, but a more general episodic cognition system (michaelian, 2016b, p. 105). focusing on past recall and future prospection, tulving (2002, p. 20) has bundled all the activities of this episodic cognition system under the mtt label, saying that “mental time travel involves awareness not only of what has been but also of what may come”. others use the term in the same way. for example, de brigard (2014, p. 174) says mtt refers “to our psychological ability to mentally travel back in time […] and to project ourselves onto the future”. suddendorf and corballis (2007, p. 299) say mtt refers “to the faculty that allows humans to mentally project themselves backwards in time to re-live, or forwards to pre-live, events”. michaelian (2016b, p. 98) says mtt “can be directed both at the personal past and the personal future”. these theorists use the term ‘mental time travel’ to refer to whatever people are doing as they recall the past, imagine the future, and engage in other forms of episodic cognition. there is nothing wrong with a catchy, metaphorical term like ‘mental time travel.’2 still, we will articulate an interesting and controversial way in which episodic memory actually affords something like mtt in a non-metaphorical sense. on this notion, an experience involves (something like) mtt if it not only presents some event outside the present as a merely intentional content but also affords its subject genuine awareness of that event. on this notion, the episodic cognition system would enable mtt if the “re-experiencing” (tulving, 2002, p. 6) of a past event, or “pre-experiencing” (michaelian, 2016b, p. 98) of a future event, involved genuine awareness of the recalled or prospected event itself. it will take some work to explain precisely what is meant by ‘genuine awareness’, but if past recall or future prospection involve genuine awareness of the recalled or prospected event itself, then the subject of the recall or prospection experience is in a special sort of contact with that event. the event is “brought into consciousness”, as we might 1throughout this paper we adopt the philosophical convention of using single quotes to mention a term. we use double quotes for direct quotations of other texts and as scare quotes. 2debus (2014, p. 335) suggests that the point of this terminology within psychology is to emphasize the underlying psychological and neurobiological similarities between past recall, future prospection, and other forms of episodic simulation. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 3 put it. so, while mtt (in this sense) does not involve the literal transposition of your physical body to spacetimes outside the present, it does involve bringing “into mind” events outside the present. with this substantive notion of mtt in hand, it becomes an open question whether either past recall or future prospection amount to mtt. it is helpful to distinguish between the episodic cognition system – the set of neural mechanisms and the computations they implement – and the abilities and mental states it enables. it is possible that some, but not all, of these states count as mtt in our substantive sense. thus, the question is not whether the episodic cognition system enables mtt, broadly speaking, but which, if any, of the states it enables count as mtt. we argue that the past recall enabled by the episodic cognition system is, in fact, mtt in our sense, but that the future prospection enabled by the system is not. specially, we shall argue that episodic memory experiences (recall experiences, for short) involve genuine awareness of the past-perceived events being remembered, while prospection experiences present imagined future events as mere intentional contents. the argument turns on the neurobiology of memory as involving memory traces which provide a causal-informational link to the past recalled events; the lack of such links in future prospection is what keeps it from also being mtt. so, the episodic cognition system enables mtt in one direction (past) but not the other (future), because only its past-directed states make contact with events outside the present. in the following, section 2 lays out three possible ways to understand the claim that episodic cognition enables mtt. these are uninteresting, as nearly everyone will accept these senses of mtt. section 3 articulates the notion of mtt as genuine awareness and explains how it involves bringing events outside the present into consciousness. after the main argument from causal-informational links in section 4, sections 5-7 consider to what extent the constructive nature of the episodic cognition system undercuts the case for past recall involving genuine awareness. 2 three preliminary notions of mtt applying standard concepts from the philosophy of mind, we get three preliminary notions of mtt: 1. mtt as representation of the past/future: the claim that the episodic cognition system enables mtt could be a claim about the representational content of the states we are in as we recall the past or prospect the future. according to this claim, the accuracy of your current recalled memory or future prospection is determined by comparing what it represents to some past or future episode. in this sense, episodic cognition would enable mtt if it enabled representing something as outside the present. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 4 2. mtt as phenomenal similarity with past/future experiences: the claim that the episodic cognition system enables mtt could be a claim about the phenomenology of the states we are in as we recall the past or prospect the future. according to this claim, what it is like (nagel, 1974) to remember some episode from the past or imagine oneself in the future is the same as (or matches in some key respects) what that episode was initially, or will be, like for you. in this sense, episodic cognition would enable mtt if it enabled repeating (or repeating in relevant respects) some past experience or preempting a future one.3 3. mtt as the lack of phenomenal presence: the claim that the episodic cognition system enables mtt could be a claim about the lack of phenomenal presence felt towards what is experienced as we recall the past and prospect the future. according to this claim, the phenomenology of these states is such that you experience the recalled or prospected events not as present here and now, but as past or future in another place. in this sense, episodic cognition would enable mental time travel if it enabled experience which introspectively seems to you to be experience of the past or future. it is widely accepted that the episodic cognition system enables mtt in sense (1), i.e. that it provides representations of the past and future (e.g. michaelian, 2016b). the mental states you are in as you prospect the future or remember the past can fail to match their target future or past events; so, they have veridicality conditions which involve the future or past and hence are representations of the future or past on most definitions of ‘representation’ (e.g. burge, 2010, p. 292). all this still holds even if it is not the function of the episodic cognition system to produce accurate representations (de brigard, 2014).4 similarly, most will accept that the episodic cognition system enables mtt in sense (2). recalling the past or imagining the future has phenomenal similarity with the recalled past or imagined future experience.5 this is presumably what tulving means when he says that episodic memory “allows people to consciously 3if you are a strong representationalist who thinks that representational content determines phenomenology, and the states we are in as we recall the past or prospect the future match (at least some of) the content of the remembered or prospected perceptual states, then (1) entails (2). similarly, on certain views about the relationship between phenomenology and representational content, (2) entails (1). but these are all substantive views and as a conceptual matter (1) and (2) are independent notions of mtt. 4a naïve realist like debus (2008), or those inclined to something like noë’s (2004) sensorimotor enactivism, may take issue with the claim that past recall and future prospection involve representation. but as has been discussed in the philosophy of perception literature (see siegel, 2010 ch. 2), there is no reason why a naïve realist cannot in principle accept that these experiences have content. more recently, sant’anna and michaelian (2019, p. 198), adapting an argument from travis (2004), argue that states of the episodic cognition system leave their contents indeterminate, and hence presumably lack representational content. if correct, this would mean these states are not representations at all. 5martin (2001, p. 273) might be an example of someone who denies this claim. matthen (2010) barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 5 re-experience past experiences” (tulving, 2002, p. 6). although the vividness with which people “re-experience” the past (or “pre-experience” the future) varies (beaty et al., 2019), it is generally accepted that past recall and future prospection involve phenomenal experience.6 this experience, although perhaps fainter, matches the target past or future experience to some extent. for example, what it is like for you as you recall eating breakfast is, at least in some respects, what it was like for you to eat breakfast (debus, 2014; matthen, 2010). what it is like to imagine the color green being similar but fainter than what it is like to see green (hume, 1777/1993, p. 10) is a reasonable analogy; the mental image is similar to the percept. it is fairly trivial that, at least most of the time, the episodic cognition system enables mtt in sense (3). past recall and future prospection lack phenomenal presence (martin, 2001, p. 273). we do not normally confuse our recalled episodic memories for perception of the here and now (tulving, 2002, p. 2) and likewise for our future prospections, although there have been cases of relaxed mindfulness being confused for hallucination in simple experimental conditions of sensory deprivation (foulkes & fleisher, 1975). past recall and future prospection do not, introspectively, seem to us to be experience of the here and now. we have, as tulving puts it, “autonoetic consciousness” (tulving, 1983): we are aware of our recall and prospection experiences as experiences of the past and future (perrin, 2016, p. 45). while autonoeticity does not entail a lack of phenomenal presence, it would be hard to explain if recall and prospection experiences had phenomenal presence. 3 mtt as genuine awareness the above three notions of mtt are all metaphorical. we shall argue, however, that the episodic cognition system enables mtt in recall but not prospection experiences in the following sense: 4. mtt as genuine awareness of the past: the episodic cognition system enables mtt in the sense that it enables experiences which afford genuine awareness of past events. explaining what is meant by ‘genuine awareness’ will take some work. we need to start with some preliminary ground clearing. as we shall explain, this sense of mtt as genuine awareness of the past, while not literal time travel, goes beyond mere metaphor. emphasizes the phenomenal differences between recalling an event and the original experience, but this is consistent with much similarity. 6this experience is often called ‘mental imagery’ (e.g. martin & deutscher, 1966; beaty et al., 2019), although we mostly avoid this term because it is naturally read as presupposing an indirect realist account of recall and prospection experiences. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 6 3.1 direct realism there is something it is like to recall an event. as is often said, this episodic recall involves “re-experiencing” the event or “re-living” the original experience. if this recall is mtt in sense (4), then this (re-)experience is an experience of the past event that is being recalled. experiencing past events themselves, as opposed to (say) mental images of those events, implies direct realism. direct realism is the view that we experience the mind-independent world itself, not intermediary mind-dependent objects like mental images or mental representations. in contrast, indirect realism, which involves something akin to mental images, does not allow for mtt in the sense being articulated here. there are two ways to be a direct realist in the philosophical literature on perception (crane, 2006). naïve realism (e.g. campbell, 2002; fish, 2009; martin, 2004), sometimes called relationalism, can be characterized several ways (see genone, 2016). put simply it holds that perceptual experiences are relations of awareness to distal sensory stimuli that are unmediated by representations. representationalism holds that while we experience distal sensory stimuli themselves, that experience is a form of, or grounded in, representations of those stimuli (e.g. dretske, 1995; burge, 2010; speaks, 2015). representationalism at first glance seems to not be direct realism. after all, representationalists either take experiences to be, as a matter of metaphysical nature, relations to mental representations (e.g. sant’anna, 2020), or simply take them to be mental representations (e.g. dretske, 2003). but within the philosophical literature on perception, many representationalists take themselves to be direct realists. dretske (2003, p. 73) says that “seeing a tree [for example] is not to be understood as awareness of some mental intermediary (an image, a sense-datum) having the properties the tree appears to have”, but rather as awareness of a tree. clark agrees: the kinds of mechanism i have sketched provide a means by which an agent may come to perceive the world. they do not, for example, provide a means by which an agent comes to perceive her own internal representations of the world. for what is perceived are not representations but the world. (clark, 2012, p. 767) what is going on? the key is to distinguish between two kinds of directness. russell (1912/1997), and later jackson (1977), famously describe the first kind (see also genone, 2016, pp. 2–3). on this notion of directness, you directly perceives a distal sensory stimulus if and only if you do not perceive it by experiencing something else, like a mental image or sense-datum. the second notion of directness refers to metaphysical nature. perception is direct if and only if it does not depend on some other state, like the tokening of a mental representation. representationalism – as understood by philosophers of perception like dretske (2003), burge (2010), clark (2012), and many others – makes perceptual experience direct in the first sense, but not the second. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 7 under this version of representationalism, we perceive a distal sensory stimulus when we are in a certain kind of mental state which represents that stimulus. but what we experience while in that state is not the state (i.e., the representation) itself. what we experience is what is represented by that state: the distal sensory stimulus. as crane (2006, p. 136) explains, the experience itself (i.e., the state of the subject) might be a representation, but what is experienced by the subject in that state is what is represented by it (see also bernecker, 2008 ch. 5-6). as dretske (2003) stresses, the distinction between representational vehicles and representational contents is important: the state itself is a representational vehicle, but that vehicle’s (wide) content is what is experienced. relating this to memory (michaelian, 2016b, pp. 62–65), while some memory theorists are naïve realists (e.g. debus, 2008), others are representationalists (e.g. martin, 2001; bernecker, 2008). memory representationalists hold that states of recall are representations (i.e., representational vehicles) of the remembered event, but allow that the (re-)experiences accompanying these states are experiences of the represented past event (i.e., of that vehicle’s content). naïve realists about memory deny that states of recall are representations, saying instead they are metaphysically direct relations of awareness to the past. for both views, during recall, what we experience are recalled events themselves, not representations of them or some other mental intermediary. a representationalist about episodic memory who wanted to emphasize the metaphysical nature of recall experiences might say that their view is a version of indirect realism. michaelian (2016b, p. 65) uses this terminology, calling memory representationalists like bernecker (2008) ‘moderate indirect realists,’7 although bernecker rejects this terminology. given that there are two notions of directness at play, one can use the terminology of ‘direct realism’ and ‘indirect realism’ with an eye towards either. we follow the philosophy of perception literature and call all these views ‘direct realist’, since we emphasize what recall experiences make us aware of. nothing stops representationalists from endorsing indirect realism in both senses. under this view, influentially defended by jackson (1977) and still found in much of the literature on episodic memory, what we experience when perceiving or remembering are representations. for example, sant’anna and michaelian (2019, p. 197) write “according to representationalism, when we engage in episodic hypothetical thought, the direct objects of our mental states are internal representations of events”. within visual experiences, these representations might be thought of as mental pictures, mental images, or percepts. we will call this version ‘indirect representationalism’ and use the term ‘direct representationalism’ for the version of representationalism described above (and endorsed by dretske, 2003; bernecker, 2008; clark, 2012 among others). 7elsewhere, sant’anna and michaelian (2019, p. 190) define ‘direct realism’ to be the view we are here calling naïve realism (aka relationalism), i.e. the view that experiences are relations of direct awareness to distal sensory stimuli (like ordinary objects). they also (2019, p. 197) refer to “indirect realism or representationalism”, equating the two. contrast this with genone (2016, p. 6), who says that to get naïve realism one needs to add further claims to direct realism (e.g., the claim that awareness does not depend on the tokening of a mental representation). barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 8 contemporary philosophers of perception have largely abandoned indirect representationalism because it was initially motivated by the need to explain hallucination and illusion, phenomena which can be handled by direct representationalism. as genone (2016, p. 6) notes, illusory and hallucinatory experiences can be accounted for “without invoking sense-data or other mediating mental objects. in this respect, representationalism is consistent with direct realism, and can be thought of as one way of developing its central claims.” indirect representationalism is not needed to account for such cases. consider the traditional problem of hallucination for direct realism.8 the problem supposes that a given successful perceptual experience could, counterfactually, have been a hallucination. since the experience is the same either way (so the problem assumes), you would be aware of the same thing in both cases. but the counterfactual scenario involves no distal object for you to be aware of, and since the cases are phenomenally indistinguishable, in the actual successful perceptual experience, you are not aware of the seen object itself either. a similar problem arises for direct realism about memory (sant’anna & michaelian, 2019). false memories or confabulations, like perceptual hallucinations, allow for a parallel track of reasoning against the conclusion that we are aware of remembered events themselves. you could reply that “the direct objects of memory, both in cases of successful remembering and in cases of unsuccessful remembering, are internal representations” (sant’anna & michaelian, 2019, p. 194).9 indirect representationalists about perception, like jackson (1977), would have likewise said that the “objects” of perceptual experience (what we experience) are internal representations. but as noted by dretske (2003), bernecker (2008), genone (2016) and others, direct representationalism is able to handle this problem while saving direct realism. in both the good and bad case (of either perception or memory), what we experience is indeed the same. in both hallucinating and perceiving, according to the direct representationalist, you are in a state which represents some distal object. what you experience is the same, but there is an important difference. what is represented in the good case is an actual perceived or remembered stimulus. in the bad case, there is nothing out in the world which is represented by the (hallucinatory or confabulatory) experience. imagine you snap a digital photo of your desk. counterfactually, if you did not have a desk like that, you could have instead made a digital image from scratch which matches the photo pixel-for-pixel. in both cases, what is represented is the same: a certain desk. but in the one case this representation is satisfied; that is, it refers to an actual item out in the world. in 8this brief discussion is merely intended to be illustrative. for a more exhaustive discussion, with an eye towards memory in particular, see (bernecker, 2008 ch. 5-6). 9this is not the view which sant’anna and michaelian endorse in their paper (2019). instead, they develop an innovative pragmatist approach to the objects of memory, prospection, and other counterfactual-imaginative experiences. they simply suggest this as the representationalist response to the problem of hallucination and confabulation. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 9 the other case the representation is unsatisfied or remains merely intentional; there is no actual item to which it refers. experiences, according to the direct representationalist, are the same: sometimes they are satisfied by, or refer to, actual things out in the world, other times not.10 a perceptual experience and a matching hallucination can, in one sense, have the same object, while in another sense having different objects. the object of the experience in both cases is what is represented, but only in the good case of perception or memory is that representational content satisfied by something out in the world. 3.2 recall as genuine awareness the upshot of this preliminary ground clearing is that our thesis should not be rejected out of hand simply because it presupposes direct realism. the claim that we are, in recall experiences, genuinely aware of the recalled events themselves only presupposes direct realism in a weak sense. it only presupposes that we do not experience recalled events in virtue of experiencing some mental intermediary. it does not presuppose whether or not our recall experiences are mediated, or enabled, by intermediary representational states. more importantly, both the naïve realist and the direct representationalist about recall experiences can endorse our thesis. by entailing direct realism, our thesis does not automatically become a version of naïve realism about memory experience (e.g., like debus, 2008). our thesis is also not merely a restatement of direct realism. for the moment, set aside naïve realism (about perception or memory). consider the difference between veridical perception and perceptual hallucinations, along with the direct 10we are setting aside a number of important complications, such as the particularity of experience, for simplicity of exposition. according to the dominant account in the literature, prominently developed by dretske (2003), burge (2005) and matthen (2005), experiences are a kind of indexical, or demonstrative, representation. for example, if you see a yellow pencil, you are in a state with the content x is yellow and a pencil, where the open slot x is filled in contextually by the actual object with which you are causally interacting through your visual system. when you successfully perceive a yellow pencil, this slot is filled by the actual pencil you are looking at, and hence the particular token of the visual state you are in comes to represent that particular pencil. when you merely hallucinate a yellow pencil, this slot remains empty, because there is no object with which you are interacting through your visual system. while this does mean that the representational content of the two tokens (the good perceptual token and the bad hallucinatory token) of the visual state differ in their overall representational content, the two tokens still share the (narrow) content fixed by the state type, the content that there is a yellow pencil. direct representationalists like dretske, burge, and matthen say this is enough to explain the phenomenal similarity between the good the bad cases, as the particular object itself represented by the experience does not affect the phenomenology. only the general or property-level content that is shared across experience tokens affects phenomeonology. schellenberg (2011) develops this approach using “gappy” fregean modes of presentation. sant’anna (2020) adapts schellenberg’s approach to the case of memory. there is much debate over whether an account like this can explain all aspects of perceptual experience and the related problems associated with hallucination and illusion. but it is widely thought by direct representationalists that some explanation, along the lines sketched here, provides the tools for saving direct realism from such problems. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 10 representationalist explanation of each just discussed above. if you visually hallucinate a pencil, then a pencil is what you experience while hallucinating. the direct representationalist says that your hallucinatory experience is a representation of a pencil. you experience a pencil because that is what is represented by the experience. but this content of the hallucinatory experience is merely intentional, i.e. unsatisfied. there is no particular pencil out in the world which is represented. now contrast what the direct representationalist says about hallucinating a pencil with what they say about seeing a pencil. if you have a successful visual experience of some actual pencil in your environment, your visual experience represents a pencil, just as the hallucination did. but what it represents is an actual object out in the world: the pencil that is stimulating your photoreceptors. thus, the content of this representation (the content of your perceptual experience) is not merely intentional (martin, 2001, p. 275). what you experience is an actual object out in the world. as we will say, your visual perceptual experience affords you genuine awareness of this pencil. so, even if recall experiences are experiences of the recalled events themselves (direct realism), there is still an open question of whether they are more like the case of perception or the case of perceptual hallucination. as you episodically recall some past-perceived event, is what you experience a mere intentional content, or do you instead actually experience the event? when we say that recall experiences afford genuine awareness of recalled events, we mean that they are experiences in the latter sense. our talk of ‘genuine awareness’ is meant to capture that recall experiences do not merely make subjects aware of recalled events as mere intentional contents. it may seem that this distinction applies to perception but not to recall experiences. at first glance, what seems to make the content of a perceptual hallucination merely intentional is that there is no distal sensory stimulus out in the world to be experienced. in contrast, what is recalled in successful recall experiences are actual past-perceived events. hence, if the recall experience is successful and represents some actual past-perceived event, would not the content of that experience necessarily be satisfied, and so not “merely intentional”? this move is too fast. consider how you can hallucinate real objects. for example, you can visually hallucinate your favorite pencil when it is nowhere near you to be seen. in this case, what you experience (your favorite pencil) is a merely intentional content, even though that content represents an actual thing out in the world (assuming you have a favorite pencil). what makes the difference between genuine awareness and experiencing a mere intentional content is a certain kind of contact between subject and thing experienced. a hallucination of your favorite pencil is an experience of a mere intentional content, and not that pencil itself, because the experience fails to actually put you in contact with it. genuine awareness of a distal sensory stimulus happens when perceptual experience “brings” that stimulus “into consciousness.”11 11debus (2014 footnote 12) draws a similar distinction. what we have called ‘genuine awareness’ barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 11 keep in mind that this notion of genuine awareness is perfectly compatible with direct representationalism. indeed, direct representationalists about perceptual experience see themselves precisely as giving an account of how the senses manage to bring into mind distal stimuli. for example, here is clark again discussing his predictive-processing representationalist view of perceptual experience:12 considered in this way, the mechanistic story on offer simply shows how beings like us are able to achieve genuine access to the causal structure of our environment. they show, as my colleague matt nudds once put it to me, how it is that we can be “open to the world”. […] equipped with brains like ours we become porous to the world. (clark, 2012, p. 767) for the direct representationalist, we bring distal stimuli “into mind”, or “into consciousness”, by representing them in the right way. a hallucinatory experience of (or rather: “as of”) your favorite pencil, although representing it, does not put you in the sort of representational contact necessary for genuine awareness. the question then becomes whether recall experiences (at least the successful ones not based on false memories) afford genuine awareness, or whether instead (like perceptual hallucinations) they involve experience of mere intentional contents. we argue in the next section that recall experiences (when successful) afford genuine awareness. so far, we have focused on direct representationalist accounts. as noted above, debus (2008) argues for a naïve realist (or as she calls it, relationalist) account of memory in which recall experiences are direct relations to the recalled event. recall, in this view, involves not only (as we are calling it) genuine awareness of some past event, but also that awareness itself has the metaphysical structure of a relation. episodically remembering a past event is just a way to be related to it. our thesis, that recall experiences involve genuine awareness, is entailed by, but is weaker than, this view. we are not arguing for naïve realism (i.e., relationalism) and having genuine awareness of past objects through memory does not require it. while we agree with debus that memory (when successful) allows us genuine awareness of past events, we leave open the metaphysical nature of this awareness. in our view, recall experiences could be representational states in which (following the direct representationalist account outlined just above) the subject experiences what is represented. the arguments we give in the next section for our claim are also different from those given by debus. debus (2008) gives (a) a positive argument which parallels campbell’s (2002) claim that naïve realism is necessary to see how perception (or, in debus’ case, memory) affords understandis what debus describes as factive experiential awareness, while what we have called ‘experience of a mere intentional content’ is what debus describes as experience “as of” an object. 12the naïve realist might object that, because of the similarities (on the direct representationalist account) between perception and the hallucination, the perceptual experiences do not really (on the direct representationalist account) put one in contact with distal stimuli (e.g., see fish, 2009, p. 14). we are not here to adjudicate this dispute. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 12 ing and knowledge of demonstrative claims, and (b) various replies to the typical objections made to naïve realism. our arguments, for the weaker claim that recall experience involves genuine awareness, will instead focus on the causal structure of memory. if we are right that successful recall experiences involve genuine awareness of past-perceived events, why does this amount to mtt in a way that goes beyond mere metaphor? the main idea is that genuine awareness of a past-perceived event would entail that that past event is “brought into consciousness” (as we might put it), or (alternatively put) that our state of recall extends back into the past to include that event. to borrow some evocative language from the naïve realists, when you make experiential contact with the world, the world itself “shape[s] the contours” of your experience (martin, 2004, p. 64); there is “an incursion or intrusion of the objective within the nonobjective – of ‘brute’ nonconscious matter within conscious life” (hellie, 2014, p. 247). as debus (2008, p. 418) says, your “point of view extends over” the part of the world of which you are conscious. this metaphorical talk of “bringing” a past event “into consciousness”, or a mental state “extending” over the past event can be cashed out in terms of constitution. it is often assumed within philosophy of mind that experiences are constituted (at least in part) by the things of which they make us aware.13 specifically: what it is like for you to have an experience just is (at least in part) what the experienced thing is like. for example, if you were asked what it is like to visually experience a pencil, all you could really do is describe the pencil you see. so, what an experience makes you aware of becomes a constitutive part of the experience. this term ‘constitutive part’ raises some subtle issues. as noted by fish (2009, pp. 10–11), the idea that what we experience is a constitutive part of our experiences has been endorsed by both direct representationalists like dretske (2003), tye (1995), and lycan (2001), as well as by naïve realists like fish himself (2009) and martin (2004). when naïve realists (including debus) make this claim, they are claiming that the experience is a state of a type which necessarily involves the object or event experienced. when direct representationalists make the claim, they of course need to allow that any given experience is of a type which can be instanced even without the object or event that is being represented.14 hence the constitution claim, on their view, should be taken as a claim about experience tokens, not types. say you successfully perceive, or successfully remember perceiving, some distal object or event. while the distal object or event is not necessary for having another experience of the same type, and hence is not a constituent part of the experience type, it is a constituent part of the particular token instance of that type which you enjoy as you perceive it. 13when you add in the qualifier ‘at least in part’ (as we have done here), the claim should be amenable even to qualia theorists who think there are some properties of experience not identifiable with properties of what is experienced. 14in contrast, since naïve realists are mostly all disjunctivists, the issue of hallucination doesn’t arise from them. they say that when i have an experience of an object, that experience type includes that particular object as a constituent part. since the experience is of a type that constitutively involves the object perceived, that experience type can’t be reproduced in a hallucination. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 13 to summarize: if recall experiences afford genuine awareness of past events instead of simply presenting those past events as mere intentional contents, then they put us in contact with the past itself. the recalled event itself literally shows up within our experience, as a constituent part. while this does not amount to literal travel through time, it certainly makes talk of mtt more than mere metaphor.15 4 does episodic cognition enable genuine awareness of the past? we noted above that direct realism about recall experiences actually leaves open two possibilities: successful recall experiences could afford genuine awareness of what is recalled, or instead their contents could be merely intentional. the aim of this section is to argue for the first possibility, assuming direct realism. if a naïve realist account of memory like debus’ is true, then successful recall experiences (i.e., ones which involve recalling actual past-perceived events) will necessarily involve genuine awareness of the recalled events. experiences with merely intentional content are only possible for direct representationalists because, for them, awareness is mediated by representations. it is possible to represent something that is not really there, or with which you are not really in contact, and so it is possible to have an experience that does not put you into contact with anything. since the naïve realist denies that awareness is mediated by representations, you are aware of something only if there exists some actual distal object or event of which you are aware. so, if a naïve realist account of (successful) recall experiences is correct, it follows immediately that episodic cognition enables genuine awareness of the past. hence, any arguments for a naïve realist account are arguments for the claim that episodic cognition enables genuine awareness of the past. debus (2008) is one of the few theorists who have defended a naïve realist account of recall experiences.16 she also argues (2014) that a similar account will not work for the experiences involved in future prospection. her reasoning is straightforward: since future events are not yet actual, it is not possible to be genuinely aware of them. a similar, parallel worry can be raised for naïve realism about episodic memory: you might think past events no longer exist, and if not, you could not be conscious of them in a recall experience (michaelian, 2016b; moran, 2019; sant’anna & michaelian, 2019). note that the same problem will arise for the proposal that recall experiences involve genuine awareness of the past, even if 15we want to thank a reviewer for pressing us to clarify our claim that recall experiences involve mtt in some non-metaphorical sense. as they helpfully summarized our view, “the subject just ‘reaches out’ to the past mentally, from the here and now in which she remains mentally anchored”. 16debus, in her (2014) account, does not notice that her account entails a sense in which episodic memory is more than mere metaphor. she takes the mtt talk to be purely metaphorical. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 14 it is developed within the direct representationalist framework. if past events do not exist, then we cannot be genuinely aware of them, as that awareness entails their constitutive involvement in the (at least token) experience. we want to set aside both naïve realism and this metaphysical problem. since naïve realism is highly controversial (and not necessary for our thesis), it is more interesting to ask whether arguments for our thesis can be given independently of arguments for naïve realism. the metaphysical problem rests on deep questions about whether past objects and events really exist, questions which are outside the scope of this paper.17 while a final, definitive analysis of the issue will, of course, require some discussion of these metaphysical issues, we are instead more interested in pressing the discussion from the standpoint of empirical psychology and neuroscience. the interesting question, as we see it, is whether there is a case to be made, on the basis of the psychological facts, for or against the claim that recall experiences involve genuine awareness of the past. at first blush, there is a case to be made for the claim that recall experiences involve genuine awareness of the past. to start, consider the cases of perception and perceptual hallucination and what makes perception (but not hallucination) a case of genuine awareness. roughly, perception and hallucination are distinguished by the involvement of causal-informational links in the former. the reason a hallucinatory experience of your favorite pencil is merely “as of” that pencil and fails to make you genuinely aware of it is that your sensory systems are not interacting with it. there is no concurrent causal connection or information-carrying link between the pencil and your (hallucinatory) experience as there is when you perceive it. in perception your sensory systems interact with your favorite pencil in a way that couples your experience to it (chisholm, 1957, p. 149): what it is like for you as you perceive it depends on what it is like at present, as you interact with it through your sensory systems. now, it is plausible that this sensory interaction is what distinguishes perceptual experiences with mere intentional content (i.e., hallucinations) from successful perceptual experiences that afford genuine awareness of the distal world. if a perceptual experience of an object is generated via sensory interaction with that object so that the experience is coupled to the object, then it yields genuine awareness.18 the recall experiences of episodic memory are not perceptual experiences, but they still satisfy this condition. they are (at least when successful) generated by an information-carrying sensory interaction with the recalled event via memory traces. for example, what it is like for you as you recall eating breakfast this morning depends on what that event was in fact like as you interacted with the meal through your sensory systems. 17michaelian (2016b, p. 63), sant’anna and michaelian (2019, p. 191), and moran (2019) say that the view that past or future objects really exist (eternalism) is highly unintuitive, but at least one author of this paper disagrees. 18the qualifier that the experience is of the object is important. we are not suggesting that a causal link or coupling is sufficient for an experience to be of an object, only that if an experience is of an object and is causally coupled to that object, then it affords genuine awareness. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 15 the causal theory of memory accepts as a basic premise that episodic memory involves this causal-informational link to the recalled events (e.g. martin & deutscher, 1966; bernecker, 2008). it also is supported by the neurobiology of memory (debus, 2008, p. 411). during an initial sensory interaction, reciprocating patterns of neural activity cycle through sensory cortex and other related neural circuits. these patterns of activity tune synaptic connections between neurons via synaptic plasticity. this tuning primes the circuit to repeat the patterns when and if they are later partially reformed by other neural input, whether that input is stimulus-driven or top-down from extraperceptual processing (feldman, 2012; jackson, 2013; zylberberg & strowbridge, 2017). this priming facilitates episodic memory: episodic memories are recalled when partial inputs prompt the completion of activity patterns from previous sensory interactions (brogaard & gatzia, 2017, p. 9). the distributed effect of the synaptic tuning from the original pattern (which primes the circuit to repeat it) is the memory trace which is so often discussed in theorizing about memory (liu, ramirez, redondo, & tonegawa, 2014, p. 59). we can summarize the argument for genuine awareness of the past in episodic memory as follows: p1. recall experiences of the past depend on the neural patterns reactivated in the recall. p2. these neural activity patterns depend on the synaptic tuning which primed the neural circuit to repeat them. p3. this synaptic tuning depends on the original pattern of activity which shaped it. p4. that original pattern of activity was due to a particular past sensory interaction with particular objects in a particular event. conclusion 1: recall experiences of the past depend on these particular objects and events and what those objects and events were like. p5. if a recall experience of the past is of a past-perceived event and causally depends on that event and what it was like, then that experience involves genuine awareness of the event itself.19 19it might be objected that just because what it is like to recall the past depends on a particular event does not mean that what it is like to recall the past just is what that event was like. this is correct. but p5 is not meant as a conceptual truth. p5 is motivated by noticing that it is the dependence of perceptual experiences on the perceived object itself, via a causal-informational link, which seems to make the difference between them and hallucination. so long as the dependence of a recall experience on its past target is sufficiently like the dependence of a perceptual experience on its target, we can infer that both recall experiences and perceptual experiences involve the same kind of awareness: genuine awareness. note that this is not a move from causal dependence to the claim that experience is a simple, nonrepresentational relation to its object. rather, the premise barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 16 p6. recall experiences of the past are of past-perceived events. conclusion 2: the recall experiences of the past that occur in episodic memory involve genuine awareness of the remembered events themselves. note that this argument runs independently of whether we assume direct representationalism or naïve realism. p1-4 capture the neurobiological facts underlying past recall, while p5 gives a sufficient condition for genuine awareness which both direct representationalists and naïve realists can accept. p6 is just direct realism about recall experiences. it is important to reiterate that p1-4 assume successful recall. many times, when we recall episodic memories, the resulting recall, or perhaps the whole memory trace itself, is distorted to the point of fabrication. similar to how we can hallucinate and seem to perceive what is not there, we can misremember and seem to remember what did not happen. in that case, it is clear that our defective recall experience of the past does not depend on the particular items being remembered. in these cases, the recall experience’s content is merely intentional and does not afford genuine awareness of any past event.20 presumably, in these cases the causalinformation link discussed in p1-4 will itself be defective. so nothing about this argument, or our position, entails that episodic recall is always accurate or always affords genuine awareness of the past. just as perceptual experiences, such as hallucinations, can fail to afford genuine awareness of distal sensory stimuli, episodic recall will likewise fail to afford genuine awareness of recalled events when things go wrong. as we noted above, we are restricting our claim to recall experiences, not all experiences generated by the episodic cognition system. this is because you cannot run a similar argument for future prospection. there is no analogous causalinformational link to the future (sant’anna & michaelian, 2019, p. 193). presumably, prospection experiences have only merely intentional content. they do not afford us genuine awareness of the future, i.e. they do not put us into experiential contact with the future. so, if this argument is successful, then there is an asymmetry between past recall and future prospection: one puts us in experiential contact with times outside the present while the other does not. debus (2014) notices this potential asymmetry, although she takes the further step of inferring that because past recall but not future prospection involves experiential contact with events outside the present, the two are different kinds of mental states. similarly, perrin (2016, pp. 50–51) lands on a direct representationalist account of recall experiences moves from the causal dependence of the memory on the remembered event and the memory being an experience of that event to the memory being a certain kind of experience: the kind that affords genuine awareness of its object. memory as genuine awareness of the past does not entail naïve realism (relationalism). 20at least, this is what a direct representationalist about memory would say. the naïve realist will have to give a disjunctive account which treats these unsuccessful cases of memory as fundamentally different from the successful cases. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 17 as involving (as we are calling it) genuine awareness of the past while prospection experiences involve awareness of the future only in the sense of having future events as mere intentional contents.21 from this asymmetry (as well as others), he argues that past recall and future prospection are facilitated by different cognitive processes. we do not think this inference follows. the above argument is not intended to support any claims about past recall and future prospection being different kinds of states or being facilitated by different kinds of cognitive processes. consider successful perception and hallucination. according to direct representationalism, the former involves genuine awareness while the latter only involves experiences with merely intentional contents. but according to direct representationalism, the two are not different kinds of states nor supported by different systems. since our argument and conclusion are consistent with direct representationalism about recall experiences, disjunctivism is not implied. it is worth noting that even normal perception involves some mtt, i.e. some genuine awareness of objects outside the present. both perception and memory are delayed in terms of the occurrence of the relevant event and the experience of that event. we never perceive, visually or otherwise, events contemporaneously as they happen, since processing takes time.22 perceptual experiences present the past – how far into the past depends on how spatially distal the object is from the perceiver and how long it takes to process the information received via the senses. the light from stars can take billions of years to reach us, and we perceive milliseconds into the past when we look at objects nearby. temporal delay is not simply an incidental feature of perception but also necessary for some attributes of our experience. for example, the short delay between a sound being received by the two ears when they are not equidistant from the source partly accounts for our ability to detect the direction from which the sound emanated (mather, 2016). if temporal delay is a constant and necessary feature of perception, the temporal delay of memory itself is not a distinguishing feature, although the cause of the delay is distinct, i.e. storing and retrieving for memory vs. the travel and processing of stimulus for perception.23 we have argued that there is a plausible case to be made for memory as genuine awareness of past events. however, issues for this view arise under particular theories of memory, especially those that take memory to involve reconstruction. despite this, we argue that there is still room for our thesis under these views. 21like perrin, martin (2001) lands on a direct representationalist account of recall experiences as involving genuine awareness of the past, although he does not give an argument like the one developed here. so, debus (2008), martin (2001), and perrin (2016) endorse accounts on which episodic recall experiences afford genuine awareness of the past. 22predictive processing adds further complexity to this account. we discuss this alternative in section 6. 23it is possible that neither memory nor perception should be viewed in the way proposed. however, this goes beyond the goals of this paper. we simply mean to argue that successful memory can be considered as affording genuine awareness to the extent that successful perception can. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 18 5 constructivist views of episodic memory modern theories claiming that memory is a reconstruction rather than a replaying of past events pose a prima facie challenge to our interpretation of episodic memory. according to some theorists, memory does not even require a causal link to past events. if correct, would past recall still count as genuine awareness of past events? we argue that even these theories of memory which appear most contradictory to our view are in fact consistent with it. what exactly is episodic memory? it can be contrasted with what is known as ‘procedural memory’ in psychology (anderson, 1976; tulving, 1983; winograd, 1972) or ‘know how’ in philosophy (ryle, 1945), which involves skills or abilities. memory for facts is known as ‘propositional’ or ‘semantic memory’ (quillian, 1968; tulving, 1972). procedural and semantic memory are, for a good reason, sidelined in this discussion since they do not give the rememberer access to the past qua their own experienced past, and are ahistorical. the context in which these memories were experienced is not maintained as part of the memory. if this context is carried over, then there is a separate, episodic memory. for example, remembering where you were when you learned about the toxicity of vitamin a in a polar bear liver involves both episodic and semantic memories. for this reason, we are concerned with recollective, episodic or autobiographical memory, which involves remembering personal events from a first person perspective (brewer, 1986). historically, normally working episodic memory has been described as giving us access to the past through the preservation of sense impressions (hume, 1739/2000) or internal representations that are later reactivated (martin & deutscher, 1966). however, the view that memory simply replays past experiences has been largely rejected. memory is not simply stored information which is then later retrieved (martin, 2001). rather, according to constructivists, memory involves constructing “plausible representations of past events” (robins, 2019, p. 2136) instead of simply replaying stored representations of those events (see michaelian, 2011; de brigard, 2014). this constructive nature is evinced by how memory often involves the entwining of multiple events and is “blended, not laid down independently once and for all, and […] reconstructed rather than reproduced” (sutton, 1998, p. 2). memory rebuilds past representations on each recall (michaelian, 2016a). forgotten sections are filled in with memories of separate events or confabulated, for example, by inference making or imagination. de brigard (2014) likewise argues that memory is a process in which mental representations are optimised under probabilistic constraints. this brings together both encoded data of previous experiences and, with the same optimisation algorithms, fills the gaps where previous experience has not been encoded with information about what probably happened. sometimes, this information is an optimal construction, but when there is deviation from optimality, this is what we consider to be a misrepresentation or false memory. sutton (2003) notes that despite this turn from defining memories as accurate representation, this reconstruction barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 19 process does not necessarily render memories false, and at the same time, it does not entail that no memories can be classified as false. rather, all memories, both veridical and false, are constructed, and assessing memories as the former or the latter requires greater nuance. constructivists point out that reconstruction is necessary because we never remember an event in its entirety, and yet our memory is not usually experienced as being gappy. one clear example of how memory can be both veridical and at the same time reconstructed differently from the initial experience is when the perspective switches, so that the event is viewed from a different angle (rice & rubin, 2009). this observer perspective is commonly from a bird’s eye view. the original field perspective is not represented accurately, which can be immediately determined since that particular angle of perception would be physically impossible, but the events portrayed in the memory are often sufficiently accurate to be considered a true memory. although this is only one example, under this view, reinterpretation is the norm, not the exception. some theorists even deny that the purpose of memory is to give us knowledge of past events. another alternative is that memory is about life coherence. it enables us to “produce a coherent life history” (schechtman, 1994, p. 13). in this view, memory is a narrative that provides the important purpose of informing us of who we are as individuals. this not only incidentally but necessarily involves biases and reconstructions. while the aforementioned views describe remembering as involving reinterpretation of encoded data or stored memory traces, which can be more or less accurate, others reject the view that such traces are a necessary element of veridical memory altogether. michaelian (2016a, 2016b) rejects “causalist” theories according to which a causal connection with the event remembered is essential for an experience to be classified as successful remembering. instead, three conditions need to be met: accuracy, reliability and internality. internality is what distinguishes remembering from relearning, in which information that has been forgotten is relearned from an external source. this is not to say that causal factors, retention and access of information are never the cause of any token memory, but rather they are not the criterion. michaelian (2016a, p. 10) says “simulationism is compatible with the view that remembering often does involve retention: the simulationist’s claim is that remembering does not necessarily involve retention, not that it necessarily does not involve retention”. we could, for example, have a genuine accurate memory that is generated by some other process, such as inference making, that lacks any direct causal memory trace. so long as the process is reliable, comes to an accurate interpretation of events and is internal, it is memory irrespective of the process of memory generation. here, we do not aim to espouse the various strengths and weaknesses of these competing views, but rather analyse whether our thesis is consistent with them. anti-causalist views will be the most challenging, however, we believe it can be done. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 20 6 the constructivist challenge to a genuine awareness account of memory constructivist and anti-causalist views threaten to undercut our central premise that there is causal or informational dependence between a recall experience and the remembered event (sant’anna & michaelian, 2019, pp. 195–196). the constructivist would suggest that any given memory depends on a blending of cognitive features rather than simply such informational dependence. in the more extreme case of the anti-causalist, the dependence on neural patterns that occurred during the original experience is removed altogether, rejecting p3 of our argument, the premise that synaptic tuning of the relevant memory depends on the original pattern of activity which shaped it. if recall experiences are blended, or worse, if they need not have causally originated at all in the remembered event, can they afford genuine awareness of the past? first, let us start with the more radical anti-causalist position. anti-causalists like michaelian do not claim that no recall experiences are grounded in a causalinformational link to the recalled event. their claim is just that such a link is not necessary for successful recall. hence, a conciliatory approach is available. in line with our argument from section 4, we allow that the content of causally-detached recall experiences, when they occur, is merely intentional, but still hold that successful recall experiences which do have a causal-informational link afford genuine awareness. one might worry that this conciliatory approach is unstable.24 if only some successful recall experiences involve genuine awareness of the past and others do not, what difference does genuine awareness make? would it not be a simpler theory to deny that any of the cases involve genuine awareness of the past? our response to this worry is that simple theories are only to be preferred when there is no evidence providing greater support for the more complicated theory. we argue that when taking into account the empirical evidence, it is not plausible to simply claim that genuine awareness plays no role whatsoever. in line with our main argument from section 4, even if anti-causalism is true, some recall experiences still do have causal-informational links to the recalled event and are still dependent on the remembered event, and this dependence makes a difference for token memories. specifically, this dependence puts us into contact with the recalled event, or “brings it into consciousness”. what about memories which retain some causal link to what is remembered, but still involve distortion and reconstruction? we agree with martin (2001, p. 282), who notes that “direct contact with the past is still consistent with the idea that there is much fabrication in recall”. in support of this view, we highlight the analogy between memory to perception. since normal, successful perception can itself be constructed or have a tenuous causal connection to the distal stimulus 24thank you to a reviewer for this and the following suggestions. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 21 but still afford genuine awareness, these features of memory should not be seen to disqualify it as affording genuine awareness. let us consider this comparison with perception in more detail. according to bayesian brain approaches, perception of the world arises from unconscious inference making (helmholtz, 1924/2005). this view rejects the assumption that perceptual content is extracted from sensory stimulation. rather, the content of our perceptual experience involves “hypotheses” about events and objects in the world (gregory, 1980). perception is a process of inference making based on priors25 under the bayesian framework, and these priors become more relied upon when there is more ambiguity in the environment, say, in poor lighting (kersten & yuille, 2003). an influential solution to how bayesian principles can be applied to experience is the predictive processing framework of perception, according to which perceptual content is a prediction of sensory input that is updated when it comes into conflict with that input. a key element of perception, on this view, is prediction error minimisation (friston, 2003; friston & stephan, 2007; rao & ballard, 1999) or free energy minimisation (friston, 2010), which involves the attempt to minimise the error between predictions about what should be perceived and sensory inputs (clark, 2013). crucially, some theorists believe this is all there is to perception (e.g. hohwy, 2013, 2016). metzinger (2009) similarly considers the content of perception to be a transparent unified world model of the present time. perception, on this view, involves a complex interweaving of “shortterm memory and working memory, with recurrent loops in neural networks, and with the binding of single events into larger temporal gestalts (often simply called the psychological moment)” (metzinger, 2009, p. 35). according to these views, perceptual experience is distorted and constructed in much the same way as constructivist views of episodic memory. memory and perception are both constructed in order to give us an informational link to the environment, and these can be assessed for accuracy. if this constructive aspect of perception does not preclude perception from affording genuine awareness (drayson, 2018), the same applies for episodic memory. predictive processing views of perception also invite a second look at anticausalist views of episodic memory. on closer analysis, the anti-causalist view of episodic memory is not so distinct from the interpretation of predictive processing according to which perceptual content is just prediction constrained by sensations (friston & stephan, 2007). this interpretation makes even successful perception itself a kind of “controlled hallucination” (clark, 2015, p. 14). by describing sensory input as a constraint to rather than a cause of perception, michaelian’s anticausalist description of memory is quite analogous. but if this merely constraintbased nature of perception does not preclude it from affording us genuine awareness of the environment, neither should it preclude successful recall from affording genuine awareness of the past. we can still retain the kind of connection with the real world required for genuine awareness via constraints imposed by the environ25the probability distribution based on previous input. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 22 ment. so long as the relevant experience, whether perceptual or recall, depends on or is linked to the object or event, it provides genuine awareness of that object or event, even if there is no causal link. this analogy between the anti-causalist view of memory and perception as “controlled hallucination” may not seem entirely apt. even if sensory input is merely a constraint on the perceptual predictions affording the content of our perceptual experiences, sensory input still provides a sort of causal connection to the world that the anti-causalist claims can be missing from memory. for example, the anti-causalist might have in mind the following sort of case. say the original memory trace laid down during your perception of an event is lost via forgetting, but later you see video that you took of the event. watching the video lays down new memory traces that are accurate of your original experience, and subsequent recall experiences of the event are reconstructed by activating these new memory traces. in reply, we want to suggest that there might be instances of perception which are analogous to even the most extreme causal-disconnection in the memory. as aforementioned, higher ambiguity in sensory input means priors become more relied upon than that sensory input. does perception still afford genuine awareness of a distal stimulus when it relies heavily, or even entirely, on priors? more often than not, we do a good job at correctly inferring, based on priors, the distal causes of our sensory input. we might still take a perceptual experience to involve genuine awareness of its object if, say, the priors did a good job of representing the world accurately through correct inference making in absence of the relevant stimulus. a good example of prediction outweighing sensory input in the case of ambiguity is the mask illusion, and relatedly, perception of faces. a widely accepted explanation as to why it is nearly impossible for neurotypical individuals26 to see the inside of a mask as hollow is because of our priors about seeing faces. whether it is convex or concave is underdetermined by the visual cues, but faces are almost never experienced as hollow. because of their salience and importance for processing, both concave and convex masks are interpreted as being convex (dima et al., 2009; gregory, 1980). it is well accepted that what we see is determined by both retinal and several extra-retinal factors, with memory, arguably, being the main contributing extra-retinal factor (albright, 2012). one could say that the perceived convexity of a mask or even a face is priors-driven rather than stimulus-driven since the distinction between convex and concave is ambiguous. one can then infer that perception of convex masks or normal faces is accurate perception in the absence of disambiguating stimuli. these perceptions are reliable, in michaelian’s criteria, simply because faces and masks are usually convex. however, one might argue that despite elements of perception being entirely priors driven, perception of, say, an entire visual scene or event must involve some direct causal link in order to be both veridical and reliable. even ambiguous sensory inputs provide some direct causal constraint. in contrast, memory of an entire 26some individuals with schizophrenia do not see the illusion. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 23 event might be reliable in some other way – for example, being told about something you did as a child and “remembering” it accurately based on this telling of the event plus some good imaginative work and inference making. however, the view that a causal link is not part of the criterion of memory does not exclude the possibility that this link is, in practice, necessary for most reliably accurate memories. martin (2001, p. 282) notes that the important sense of memory as genuine awareness would be “lost if episodic memory had no special status in an account of our epistemic links with the past”. nonetheless, if it turns out that causal connectedness, or at least some informational link, is not a requirement of all reliable, accurate episodic memories, we can fall back on the conciliatory approach outlined at the start of this section. we argue that there is a case for saying that even causally-disconnected successful memories afford genuine awareness of the past, but if this argument fails, the causally-connected ones at least still afford genuine awareness. 7 the function challenge part of our response above was to claim that accurate tracking of the environment may suffice for the kind of contact which makes the difference between the genuine awareness of perception and the mere intentional content of hallucination, even if this tracking is not achieved by robust causal links. another challenge to our view of memory is that the function of memory is not even to represent the past accurately. memory is a key aspect of the sense of self in theories of psychological continuity of personal identity (parfit, 1971) and is important for our life narrative or trajectory (schechtman, 2011). because of this, there are a variety of biases that affect what and how we remember events. this is particularly obvious regarding emotional salience, which can alter both the likelihood that an event will be remembered and the associations related to these memories (kensinger, 2009). we can then ask whether perception carries similar functions. it does seem that salient factors can alter perception in a variety of ways. reward and punishment, for example, can affect what is perceived in binocular rivalry experiments in which two different visual stimuli are presented to different eyes and perception shifts back and forth between the two. balcetis, dunning, & granot (2012) found that after teaching participants to associate values to specific letters and numbers, specifically a monetary prize as the reward for earning points, the stimulus with a higher point value becomes dominant, i.e. participants are more likely to see that stimulus first and for longer. this is just one way in which value biases perception similarly to how biases affect memory. there are many biases and potential functions to memory as there are to perception, but we see this as no reason to reject the view that most memories involve contact with recalled events. if under these views, there are some mental states that do not accurately represent past events but should still be classed as memories, then we again adopt the conciliatory approach; these token memories are not cases of genuine awareness of the past. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 24 8 conclusion we have argued that episodic memory enables genuine awareness of the past. memories are (at least sometimes) reactivated neural patterns that depend on and repeat an original pattern of activity that occurred during the initial experience of an item or event. because of this, episodic recall experience involves genuine awareness of the event that was originally experienced. we have argued that this view is still applicable under theories of memory that appear, at first glance, to be incompatible with it, namely constructivist and anti-causalist views. however, particular token memories under such views may involve no causal link with the past and thus may not afford genuine awareness of past events. acknowledgments we would like to thank the anonymous referees and editors of this journal for their helpful comments which much improved our paper. rosen received funding from carlsbergfondet (cf16-0804). references albright, t. d. (2012). on the perception of probable things: neural substrates of associative memory, imagery, and perception. neuron, 74, 227–245. https://doi.org/10.1016/j.neuron.2012.04.001 anderson, j. r. (1976). language, memory, and thought. hillsdale: lawrence erlbaum associates. balcetis, e., dunning, d., & granot, y. (2012). subjective value determines initial dominance in binocular rivalry. journal of experimental social psychology, 48(1), 122–129. https://doi.org/10.1016/j.jesp.2011.08.009 beaty, r. e., seli, p., & schacter, d. l. (2019). thinking about the past and future in daily life: an experience sampling study of individual differences in mental time travel. psychological research, 83(4), 805–816. https://doi.org/10.1007/s00426-018-1075-7 bernecker, s. (2008). the metaphysics of memory. new york: springer. brewer, w. f. (1986). what is autobiographical memory? in d. c. rubin (ed.), autobiographical memory (pp. 25–49). cambridge: cambridge university press. brogaard, b., & gatzia, d. e. (2017). unconscious imagination and the mental imagery debate. frontiers in psychology, 8(799), 1–14. https://doi.org/10.3389/fpsyg.2017.00799 buckner, r. l., & carroll, d. c. (2007). self-projection and the brain. trends in cognitive sciences, 11(2), 49–57. https://doi.org/10.1016/j.tics.2006.11.004 burge, t. (2005). disjunctivism and perceptual psychology. philosophical topics, 33(1), 1–78. https://doi.org/10.5840/philtopics20053311 burge, t. (2010). origins of objectivity. oxford: oxford university press. byrne, a. (2010). recollection, perception, imagination. philosophical studies, 148, 15–26. https://doi.org/10.1007/s11098-010-9508-1 campbell, j. (2002). reference and consciousness. oxford: oxford university press. chisholm, r. (1957). perceiving: a philosophical study. ithaca: cornell university press. clark, a. (2012). dreaming the whole cat: generative models, predictive processing, and the enactivist conception of perceptual experience. mind, 121(483), 753–771. https://doi.org/10.1093/mind/fzs106 clark, a. (2013). whatever next? predictive brains, situated agents, and the future of cognitive science. behavioral and brain sciences, 36(3), 181–253. https://doi.org/10.1017/s0140525x12000477 clark, a. (2015). surfing uncertainty: prediction, action, and the embodied mind. oxford: oxford university press. crane, t. (2006). is there a perceptual relation? in t. szabo gendler & j. hawthorne (eds.), perceptual experience (pp. 126–146). oxford: oxford university press. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neuron.2012.04.001 https://doi.org/10.1016/j.jesp.2011.08.009 https://doi.org/10.1007/s00426-018-1075-7 https://doi.org/10.3389/fpsyg.2017.00799 https://doi.org/10.1016/j.tics.2006.11.004 https://doi.org/10.5840/philtopics20053311 https://doi.org/10.1007/s11098-010-9508-1 https://doi.org/10.1093/mind/fzs106 https://doi.org/10.1017/s0140525x12000477 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 25 de brigard, f. (2014). is memory for remembering? recollection as a form of episodic hypothetical thinking. synthese, 191, 155–185. https://doi.org/10.1007/s11229-013-0247-7 debus, d. (2008). experiencing the past: a relational account of recollective memory. dialectica, 62(4), 405–432. https://doi.org/10.1111/j.1746-8361.2008.01165.x debus, d. (2014). “mental time travel”: remembering the past, imagining the future, and the particularity of events. review of philosophy and psychology, 5, 333–350. https://doi.org/10.1007/s13164-014-0182-7 dima, d., roiser, j. p., dietrich, d. e., bonnemann, c., lanfermann, h., emrich, h. m., & dillo, w. (2009). understanding why patients with schizophrenia do not perceive the hollow-mask illusion using dynamic causal modelling. neuroimage, 46(4), 1180–1186. https://doi.org/10.1016/j.neuroimage.2009.03.033 drayson, z. (2018). direct perception and the predictive mind. philosophical studies, 175, 3145–3164. https://doi.org/10.1007/s11098-017-0999-x dretske, f. (1995). naturalizing the mind. cambridge, ma: the mit press. dretske, f. (2003). experience as representation. philosophical issues, 13(1), 67–82. https://doi.org/10.1111/1533-6077.00005 feldman, d. e. (2012). the spike-timing dependence of plasticity. neuron, 75, 556–571. https://doi.org/10.1016/j.neuron.2012.08.001 fish, w. (2009). perception, hallucination, and illusion. oxford: oxford university press. foulkes, d., & fleisher, s. (1975). mental activity in relaxed wakefulness. psychology, 84(1), 66–75. https://doi.org/10.1037/h0076164 friston, k. j. (2003). learning and inference in the brain. neural networks, 16(9), 1325–1352. https://doi.org/10.1016/j.neunet.2003.06.005 friston, k. j. (2010). the free-energy principle: a unified brain theory? nature reviews neuroscience, 11(2), 127–138. https://doi.org/10.1038/nrn2787 friston, k. j., & stephan, k. e. (2007). free-energy and the brain. synthese, 159, 417–458. https://doi.org/10.1007/s11229-007-9237-y genone, j. (2016). recent work on naïve realism. american philosophical quarterly, 53(1), 1–26. retrieved from https://www.jstor.org/stable/44982080 gregory, r. l. (1980). perceptions as hypotheses. philosophical transactions of the royal society of london. b, biological sciences, 290(1038), 181–197. https://doi.org/10.1098/rstb.1980.0090 hellie, b. (2014). love in the time of cholera. in b. brogaard (ed.), does perception have content? (pp. 241–261). oxford: oxford university press. helmholtz, h. von. (2005). treatise on physiological optics (vol. 3; j. p. c. southhall, ed.). mineola: dover. (original work published 1924) hohwy, j. (2013). the predictive mind. oxford: oxford university press. hohwy, j. (2016). the self-evidencing brain. noûs, 50(2), 259–285. https://doi.org/10.1111/nous.12062 hume, d. (1993). an enquiry concerning human understanding (2nd ed.; e. steinberg, ed.). indianapolis: hackett publishing company. (original work published 1777) hume, d. (2000). a treatise of human nature (d. f. norton & m. j. norton, eds.). oxford: clarendon press. (original work published 1739) jackson, f. (1977). perception: a representative theory. cambridge: cambridge university press. jackson, m. b. (2013). recall of spatial patterns stored in a hippocampal slice by long-term potentiation. journal of neurophysiology, 110, 2511–2519. https://doi.org/10.1152/jn.00533.2013 kensinger, e. a. (2009). remembering the details: effects of emotion. emotion review, 1(2), 99–113. https://doi.org/10.1177/1754073908100432 kersten, d., & yuille, a. (2003). bayesian models of object perception. current opinion in neurobiology, 13(2), 150–158. https://doi.org/10.1016/s0959-4388(03)00042-4 liu, x., ramirez, s., redondo, r. l., & tonegawa, s. (2014). identification and manipulation of memory engram cells. cold spring harbor symposia on quantitative biology, 79, 59–65. https://doi.org/10.1101/sqb.2014.79.024901 lycan, w. g. (2001). the case for phenomenal externalism. philosophical perspectives, 15, 17–35. https://doi.org/10.1111/0029-4624.35.s15.2 martin, c. b., & deutscher, m. (1966). remembering. philosophical review, 75(2), 161–196. https://doi.org/10.2307/2183082 martin, m. g. f. (2001). out of the past: episodic recall as retained acquaintance. in c. hoerl & t. mccormack (eds.), time and memory (pp. 257–284). oxford: oxford university press. martin, m. g. f. (2004). the limits of self-awareness. philosophical studies, 120(1-3), 37–89. https://doi.org/10.1023/b:phil.0000033751.66949.97 barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11229-013-0247-7 https://doi.org/10.1111/j.1746-8361.2008.01165.x https://doi.org/10.1007/s13164-014-0182-7 https://doi.org/10.1016/j.neuroimage.2009.03.033 https://doi.org/10.1007/s11098-017-0999-x https://doi.org/10.1111/1533-6077.00005 https://doi.org/10.1016/j.neuron.2012.08.001 https://doi.org/10.1037/h0076164 https://doi.org/10.1016/j.neunet.2003.06.005 https://doi.org/10.1038/nrn2787 https://doi.org/10.1007/s11229-007-9237-y https://www.jstor.org/stable/44982080 https://doi.org/10.1098/rstb.1980.0090 https://doi.org/10.1111/nous.12062 https://doi.org/10.1152/jn.00533.2013 https://doi.org/10.1177/1754073908100432 https://doi.org/10.1016/s0959-4388(03)00042-4 https://doi.org/10.1101/sqb.2014.79.024901 https://doi.org/10.1111/0029-4624.35.s15.2 https://doi.org/10.2307/2183082 https://doi.org/10.1023/b:phil.0000033751.66949.97 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org michael barkasi and melanie g. rosen 26 mather, g. (2016). foundations of sensation and perception (3rd ed.). new york: psychology press. matthen, m. (2005). seeing, doing, and knowing: a philosophical theory of sense perception. oxford: oxford university press. matthen, m. (2010). is memory preservation? philosophical studies, 148, 3–14. https://doi.org/10.1007/s11098-010-9501-8 metzinger, t. (2009). the ego tunnel: the science of the mind and the myth of the self. new york: basic books. michaelian, k. (2011). generative memory. philosophical psychology, 24(3), 323–342. https://doi.org/10.1080/09515089.2011.559623 michaelian, k. (2016a). confabulating, misremembering, relearning: the simulation theory of memory and unsuccessful remembering. frontiers in psychology, 25(1857), 1–13. https://doi.org/10.3389/fpsyg.2016.01857 michaelian, k. (2016b). mental time travel: episodic memory and our knowledge of the personal past. cambridge, ma: the mit press. moran, a. (2019). naïve realism, seeing stars, and perceiving the past. pacific philosophical quarterly, 100(1), 202–232. https://doi.org/10.1111/papq.12238 nagel, t. (1974). what is it like to be a bat? philosophical review, 83, 435–450. https://doi.org/10.2307/2183914 noë, a. (2004). action in perception. cambridge, ma: the mit press. parfit, d. (1971). personal identity. the philosophical review, 80(1), 3–27. https://doi.org/10.2307/2184309 perrin, d. (2016). asymmetries in subjective time. in k. michaelian, s. b. klein, & k. k. szpunar (eds.), seeing the future: theoretical perspectives on future-oriented mental time travel (pp. 39–61). https://doi.org/10.1093/acprof:oso/9780190241537.003.0003 quillian, m. r. (1968). semantic memory. in m. minsky (ed.), semantic information processing (pp. 227–270). cambridge, ma: the mit press. rao, r. p. n., & ballard, d. h. (1999). predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive-field effects. nature neuroscience, 2(1), 79–87. https://doi.org/10.1038/4580 rice, h. j., & rubin, d. c. (2009). i can see it both ways: firstand third-person visual perspectives at retrieval. consciousness and cognition, 18(4), 877–890. https://doi.org/10.1016/j.concog.2009.07.004 robins, s. k. (2019). confabulation and constructive memory. synthese, 196, 2135–2151. https://doi.org/10.1007/s11229-017-1315-1 russell, b. (1997). the problems of philosophy. oxford: oxford university press. (original work published 1912) ryle, g. (1945). knowing how and knowing that: the presidential address. proceedings of the aristotelian society, 46, 1–16. https://doi.org/10.1093/aristotelian/46.1.1 sant’anna, a. (2020). the hybrid contents of memory. synthese, 197, 1263–1290. https://doi.org/10.1007/s11229-018-1753-4 sant’anna, a., & michaelian, k. (2019). thinking about events: a pragmatist account of the objects of episodic hypothetical thought. review of philosophy and psychology, 10, 187–217. https://doi.org/10.1007/s13164-018-0391-6 schacter, d. l., & addis, d. r. (2007). the cognitive neuroscience of constructive memory: remembering the past and imagining the future. philosophical transactions of the royal society b: biological sciences, 362(1481), 773–786. https://doi.org/10.1098/rstb.2007.2087 schechtman, m. (1994). the truth about memory. philosophical psychology, 7, 3–18. https://doi.org/10.1080/09515089408573107 schechtman, m. (2011). memory and identity. philosophical studies, 153(1), 65–79. https://doi.org/10.1007/s11098-010-9645-6 schellenberg, s. (2011). perceptual content defended. noûs, 45(4), 714–750. https://doi.org/10.1111/j.1468-0068.2010.00791.x siegel, s. (2010). the contents of visual experience. oxford: oxford university press. speaks, j. (2015). the phenomenal and the representational. oxford: oxford university press. suddendorf, t., & corballis, m. c. (2007). the evolution of foresight: what is mental time travel, and is it unique to humans? behavioral and brain sciences, 30, 299–351. https://doi.org/10.1017/s0140525x07001975 sutton, j. (1998). philosophy and memory traces: descartes to connectionism. cambridge: cambridge university press. sutton, j. (2003). constructive memory and distributed cognition: towards an interdisciplinary framework. in b. kokinov & w. hirst (eds.), constructive memory (pp. 290–303). sofia: new bulgarian university. travis, c. s. (2004). the silence of the senses. mind, 113(449), 57–94. https://doi.org/10.1093/mind/113.449.57 tulving, e. (1972). episodic and semantic memory. in e. tulving & w. donaldson (eds.), organization of memory (pp. 381–403). new york: academic press. tulving, e. (1983). elements of episodic memory. oxford: oxford university press. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11098-010-9501-8 https://doi.org/10.1080/09515089.2011.559623 https://doi.org/10.3389/fpsyg.2016.01857 https://doi.org/10.1111/papq.12238 https://doi.org/10.2307/2183914 https://doi.org/10.2307/2184309 https://doi.org/10.1093/acprof:oso/9780190241537.003.0003 https://doi.org/10.1038/4580 https://doi.org/10.1016/j.concog.2009.07.004 https://doi.org/10.1007/s11229-017-1315-1 https://doi.org/10.1093/aristotelian/46.1.1 https://doi.org/10.1007/s11229-018-1753-4 https://doi.org/10.1007/s13164-018-0391-6 https://doi.org/10.1098/rstb.2007.2087 https://doi.org/10.1080/09515089408573107 https://doi.org/10.1007/s11098-010-9645-6 https://doi.org/10.1111/j.1468-0068.2010.00791.x https://doi.org/10.1017/s0140525x07001975 https://doi.org/10.1093/mind/113.449.57 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org is mental time travel real time travel? 27 tulving, e. (2002). episodic memory: from mind to brain. annual review of psychology, 53, 1–25. https://doi.org/10.1146/annurev.psych.53.100901.135114 tye, m. (1995). ten problems of consciousness: a representational theory of the phenomenal mind. cambridge, ma: the mit press. winograd, t. (1972). understanding natural language. cognitive psychology, 3(1), 1–191. https://doi.org/10.1016/0010-0285(72)90002-3 zylberberg, j., & strowbridge, b. w. (2017). mechanisms of persistent activity in cortical circuits: possible neural substrates for working memory. annual review of neuroscience, 40, 603–627. https://doi.org/10.1146/annurev-neuro-070815-014006 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. barkasi, m., & rosen, m. g. (2020). is mental time travel real time travel? philosophy and the mind sciences, 1(1), 1. https://doi.org/10.33735/phimisci.2020.1.28 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1146/annurev.psych.53.100901.135114 https://doi.org/10.1016/0010-0285(72)90002-3 https://doi.org/10.1146/annurev-neuro-070815-014006 https://doi.org/10.33735/phimisci.2020.1.28 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org in what sense is there mental time travel? three preliminary notions of mtt mtt as genuine awareness direct realism recall as genuine awareness does episodic cognition enable genuine awareness of the past? constructivist views of episodic memory the constructivist challenge to a genuine awareness account of memory the function challenge conclusion dynamic attentional mechanisms of creative cognition. dynamic attentional mechanisms of creative cognition shadab tabatabaeiana(stabatabaeian@ucmerced.edu) carolyn dicey jenningsa(cjennings3@ucmerced.edu) abstract in popular imagination creativity requires us to surrender control. yet, attention is at the heart of control, and many studies show attention to play a key role in the creative process. this is partly due to the selective nature of attention—creative cognition consists of two phases, idea generation and idea evaluation, and selective processes are essential for both phases. here, we investigate attentional (i.e., selective) mechanisms underlying each phase, using the framework of two major attention taxonomies: top-down/bottom-up and internal/external attention. we argue that creative cognition is supported by a dynamic interplay between the typically opposing sides of each taxonomy. further, we argue that this dynamic relationship is reflected in interactions across three large-scale brain networks: the default mode (dmn), frontoparietal control (fpn), and salience (sn) networks. our review of the evidence suggests that creative cognition is best achieved through the flexible use of multiple forms of attention, rather than through reduced attention. we thus propose a two-dimensional space, including one dimension for top-down/bottom-up attention and another for internal/external attention, which can sufficiently capture the flexibility and diversity of attentional mechanisms underlying different stages and components of creative cognition. keywords creative cognition ∙ top-down attention ∙ bottom-up attention ∙ internal attention ∙ external attention 1 introduction the popular view of creativity is that it opposes rationality and control. sometimes this is put in terms of hemispheric dominance: those who are rational and controlled are called “left brained,” while those who are creative and impulsive are called “right brained.” other times it is put in terms of inhibition: creative people are “freer” or less inhibited than other people. scientific accounts of creativity sometimes lend credence to this popular view (see, e.g., kaufman et al., 2010, p. 217). we take a different tack: we see control as playing a key role in creative auniversity of california, merced. tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-6936-5575 https://orcid.org/0000-0002-2222-1096 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 2 cognition, which we explore through the mechanisms of attention. as we will argue, not only is attention an essential part of the creative process, the more varied and dynamic the uses of attention, the more creative the agent is able to be. our account draws on evidence from psychology and neuroscience to offer a perspective that is distinctive, while also building on and synthesizing other recent accounts. as is noted in most other accounts, creativity is a process that aims for novel ideas or products that are useful to some question or enterprise (runco & jaeger, 2012). this process can be divided into two phases: idea generation and idea evaluation. in the first phase multiple ideas are generated, whereas in the second phase the most appropriate ideas are selected and further elaborated (allen & thomas, 2011). what is often overlooked is that selection is at the core of both phases. in the first phase, selection is used to bring forward the appropriate elements, such as words, musical notes, or shapes, which can then be combined to generate ideas. in the second phase selection is used to evaluate the ideas, separating the good from the bad. further, this type of selection counts as attention. that is, while selection is possible without attention, task-related selection is only possible with attention. the type of selection required for both phases is task-related selection (albeit sometimes with a task that is widely drawn). thus, to understand creative cognition, researchers must explore the role of attention. in this paper we argue that creativity is increased in the dynamic use of multiple forms of attention, rather than in its absence. two major taxonomies are discussed: top-down versus bottom-up, which is based on the source of attention (connor et al., 2004; posner, 1980; treisman & gelade, 1980), and internal versus external, which is based on its target (chun et al., 2011). the studies we review yield inconsistent results regarding the contributions of each taxon to creative cognition. while some find reduced top-down attention to facilitate creative cognition, at least for some tasks (chrysikou et al., 2013; reverberi et al., 2005), others report that increased top-down attention is conducive to creativity (nusbaum & silvia, 2011; silvia, 2015). many studies assign a significant role to internal attention but diverge on possible contributions of external attention. our account helps to explain these seeming contradictions. here, we use behavioral and neuroimaging studies to argue against an either/or narrative when it comes to attention and creative cognition. instead, we argue that creative cognition benefits from a dynamic interplay between typically opposed attentional mechanisms—top-down and bottom-up as well as internal and external. that is, whereas bottom-up and top-down attention are typically seen as in conflict, as when a distracting sound grabs your attention away from your current task, in creative cognition they work together. similarly, whereas internal and external attention normally seem opposed to one another, as when rehearsing a conversation prevents one from noticing something nearby, creative cognition is improved by cross-engagement. as we will show, interplay between these forms of attention is reflected in dynamic interactions across three large-scale brain networks: the default mode (dmn), frontoparietal control (fpn), and salience (sn) networks. in what follows we explain the unique role of attention in creative cogtabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dynamic attentional mechanisms of creative cognition 3 nition and the contributions of each type of attention while supporting our claim that a dynamic relationship between multiple attentional mechanisms best supports creativity. consequently, we propose a two-dimensional space of attention that sufficiently captures the flexibility of creative cognition and unifies seemingly contradictory accounts of its underlying attentional mechanisms. along the way we will also provide background on the dual-process account of creative cognition, the major taxonomies of attention, and how these are related. 2 background and motivation major accounts of creative cognition depict it as a multi-phase phenomenon, including an explorative phase wherein many ideas are generated and an evaluative phase where ideas are further assessed and elaborated (allen & thomas, 2011; runco & basadur, 1993). for example, wallas (1926) proposed a four-stage model of creative problem solving: preparation, incubation, illumination, and verification. individuals confront the problem in the preparation phase, stop consciously thinking about the problem in the incubation phase, experience an “aha” moment when they find a solution in the illumination phase, and then actively test the solution in the final verification stage. wallas (1926) suggested that only the first and last stages require reasoning, whereas the middle stages instead depend on imagination. streamlining these ideas, guilford (1950, 1967) proposed a dichotomy of idea production strategies, known as divergent and convergent thinking. divergent thinking is the ability to generate many ideas and is considered most useful for exploration and solving open-ended problems. convergent thinking is thought to be more analytical and suited for problems with one correct answer. guilford (1967) claimed that divergent thinking plays a more significant role in creativity, but that convergent thinking also contributes. divergent thinking ability is often assessed through tasks such as the alternative uses test (aut, torrance, 1966), in which the participant is asked to come up with many alternative ways of using a simple object, such as a paperclip, whereas convergent thinking ability is commonly evaluated through the remote associates test (rat, mednick & mednick, 1967), in which participants are asked to identify a word related to three cue words, such as “cheese” for the cue words “cottage,” “swiss,” and “blue.” campbell (1960) applied principles from evolutionary theory to propose a different set of phases. he contended that creative thinking comprises two stages: random variation and selective retention. in biological evolution, variation is produced through random mutations and the recombination of old genes.1 new variations are not intentionally designed to enhance adaptation, but the gene com1recent work suggests this process is not truly random, but this does not significantly impact campbell’s theory (monroe et al., 2022). tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 4 binations conducive to species survival are more likely to pass on to future generations. relatedly, campbell described idea generation as a process that allows for explorations “beyond the limits of foresight or prescience” (campbell, 1960, p. 381). once the generated ideas fit some criteria—set by intrinsic or environmental constraints—they are selected and stored in memory to produce further variations (see also simonton, 1999). aligning with these multi-phase descriptions of creativity, neuroimaging studies have rendered supporting evidence for a “biphasic” or “dual-process” account of creative cognition (beaty et al., 2018; ellamil et al., 2012; liu et al., 2012; rosen et al., 2020). such studies concur that creativity starts with an initial stage of idea generation followed by the search for an optimal solution through idea evaluation. the overwhelming evidence from such studies is that there is no specialized brain area or mechanism dedicated to creative cognition. instead, creativity emerges from dynamic interactions between common cognitive processes, including executive function, memory, and attention (abraham, 2018). what is missing from the above accounts is a concrete description of the cognitive mechanisms underlying each phase of the creative process and how those mechanisms are implemented in the brain. in this paper we take some initial steps toward remedying this gap by focusing on the role of attentional mechanisms in creative cognition in both behavioral and neural terms. future work will be required to develop this account into a testable model. 3 the role of attention in creative cognition the ubiquity of attention in contemporary psychology and neuroscience is paired with a lack of agreement as to its definition. it is a running joke that james once stated: “everyone knows what attention is” (james, 1890, p. 403), while others assert that “no one knows what attention is” (hommel et al., 2019). perhaps the most agreed upon feature of attention is that it concerns selection: the prioritization of some mental events or neural processes over others. beyond that, there is considerable disagreement as to its unity and scope. yet, most would agree that attention plays a central role in cognition; it was, after all, at the heart of the so-called “cognitive revolution” (nobre & kastner, 2014). we agree with nobre and kastner that the “consensus definition” is something like “the prioritization of processing information relevant to current task goals” (nobre & kastner, 2014, p. 1214).2 because this prioritization can happen at different levels of analysis and in many different parts of the brain, we discuss below two specific taxonomies of attention as 2as we will see below, this prioritization can nonetheless sometimes be driven by salient stimuli. while this may seem to be contradictory, it is not as long as we see attention as a process driven by multiple sources, including task set and salience, with a resulting selection that may appear more task or more stimulus driven. selective processing driven purely by stimuli, without reference to the subject’s interests or goals may be better understood as filtering, rather than attention (see, e.g., jennings, 2012). tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dynamic attentional mechanisms of creative cognition 5 they relate to creative cognition: top-down versus bottom-up and internal versus external attention. 3.1 top-down versus bottom-up attention the difference between top-down and bottom-up attention is based on the source of input to the process of prioritization: top-down inputs include all those that are related to the current task (e.g., one’s motivation to focus on the task) whereas bottom-up inputs include all those that are unrelated to the current task (e.g., salient stimuli that might distract from the task). for example, buschman and miller (2007, p. 1860) write: “attention can be focused volitionally by ‘top-down’ signals derived from task demands and automatically by ‘bottom-up’ signals from salient stimuli.” top-down attention is thus sometimes referred to as “endogenous” or goal-directed, whereas bottom-up attention is referred to as “exogenous” or stimulus-driven (posner, 1980). in neural terms, top-down and bottom-up attention correspond with different systems of activation (see, e.g., jennings, 2020, for an overview). bottom-up input is feedforward starting from sensory cortices, whereas top-down input relies on feedback from higher-order neural areas (see, e.g., katsuki & constantinidis, 2014 for an overview). specifically, top-down attention operates through the frontoparietal network (fpn), which is responsible for executive control in the brain. the fpn primarily comprises the dorsolateral prefrontal cortex (dlpfc) and posterior parietal cortex (ppc) and contributes to executive functions such as task initiation, task maintenance, and task switching (cole & schneider, 2007; vincent et al., 2008). because it relies on feedback from higher-level neural areas, top-down attention operates about 50-60 milliseconds slower than bottom-up attention (baluch & itti, 2011; corbetta & shulman, 2002). while these forms of attention can be examined as separate processes (e.g., treisman & gelade, 1980), they can also be examined as interacting forces in overall resource distribution (e.g., connor et al., 2004). this is sometimes described through the language of a “priority map”: “correlates of bottom-up and top-down attention reveal coactivation of the same network of parietal and prefrontal cortical areas, giving credence to the idea of a priority map simultaneously modulated by both types of factors” (katsuki & constantinidis, 2014, p. 510). in what follows we assume that top-down and bottom-up input typically interact in order to guide attention, rather than thinking of them as entirely separate processes, with “topdown attention” reflecting greater priority given to top-down input, and “bottomup attention” reflecting greater priority given to bottom-up input. the role played by top-down and bottom-up attention in creative cognition has been a point of contention (dietrich & kanso, 2010). while some accounts claim that reduced top-down attention supports creative cognition, others claim the opposite: that increased top-down attention brings about creativity (see below). in this section we argue that these accounts are compatible when we consider the tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 6 different phases associated with creative cognition, managed by the agent through her control of attention. 3.1.1 reduced top-down attention as key to creative cognition on one account, reduced top-down attention is conducive to creative cognition (for reviews, see chrysikou, 2018; zabelina, 2018). according to this account, reduced top-down attention, marked by deactivations of dlpfc, facilitates spontaneous associations and the generation of novel ideas (carson et al., 2003; kasof, 1997). as evidence for this account, it is reported that spontaneous jazz improvisation is characterized by deactivation of areas of the prefrontal cortex (limb & braun, 2008). it is also reported that patients with damaged pfc exhibit better performance on creative tasks such as matchstick problems (reverberi et al., 2005) and drawing (shamay-tsoory et al., 2011). in addition, reducing prefrontal activity in healthy subjects—through noninvasive transcranial direct current stimulation—leads to better creative performance in tasks such as aut (chrysikou et al., 2013). similarly, lesions in pfc are associated with improved performance in creative tasks that require overcoming the constraining influence of recently activated knowledge (abraham, 2018). along those lines, children younger than five years old encounter less functional fixedness—the tendency to perceive an object only in terms of its most common use—which is consistent with reduced top-down attention given the late development of pfc in children (german & defeyter, 2000; gopnik et al., 2017). reduced top-down attention allows for more openness to bottom-up stimuli that might not be directly relevant to the task at hand. for example, studies report that creative individuals are more sensitive to peripheral auditory cues (ansburg & hill, 2003). increased sensitivity to bottom-up stimuli is itself hypothesized to help creativity in two ways: 1) these stimuli might lead someone to make new connections between remote concepts, and 2) these stimuli expand an individual’s repertoire of potential building blocks for future creative ideas (zabelina, 2018). 3.1.2 increased top-down attention as key to creative cognition in contrast, other research finds that increased top-down attention promotes creative cognition. top-down attention, for instance, plays an important role in guiding memory retrieval, as well as in identifying and deploying task strategies, all of which support creative cognition. further, top-down attention is necessary for inhibiting habitual responses that would otherwise impede creative performance (aziz‐zadeh et al., 2009; bengtsson et al., 2007; edl et al., 2014). evidence supporting this approach includes studies that report a positive relationship between creativity and inhibition. for example, edl et al. (2014) used a modified stroop task to measure cognitive control. the stroop task is a standard clinical and experimental measure of inhibitory control that asks participants to name the color a word is printed in, rather than the printed word, which requires tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dynamic attentional mechanisms of creative cognition 7 them to inhibit the automatic tendency to read the printed word (see, e.g., herd et al., 2006, p. 79). they found that creative individuals did not experience stroop interference, whereas the control group exhibited a robust interference effect.3 the researchers concluded that more creative individuals “show stronger cognitive control” and “more efficiently suppress competing dominant but irrelevant actions, processes, or mental activities” (edl et al., 2014, p. 41). this research was supported by a follow up study that found “inhibition significantly predicted creativity (i.e., divergent thinking ability)” (benedek et al., 2014, p. 79). in another study, bengtsson et al. (2007) reported increased activation of cognitive control areas (i.e., dlpfc) in the generation of musical structures during improvisation. they concluded that cognitive control is necessary to adapt the improvisation to an “overall goal of producing an aesthetically satisfactory end-result” (bengtsson et al., 2007, p. 837). 3.1.3 flexible interactions between top-down and bottom-up attention as key to creative cognition the accounts we reviewed above seem contradictory. some attribute creative cognition to reduced top-down attention, arguing that it facilitates unconventional, spontaneous thinking. others consider increased top-down attention as conducive to creative cognition, arguing that top-down control is necessary for goal-directedness and the inhibition of conventional ideas in favor of novel ones. both accounts are supported by numerous behavioral and neuroimaging studies (see above). to address these seemingly inconsistent findings, pinho et al. (2015) conducted an experiment using functional magnetic resonance imaging (fmri). they instructed professional pianists to perform improvisations on a piano keyboard using either specific keys (“key condition”) or specific emotional content (“emotion condition”). pinho et al. found that the key condition corresponded with higher activity in dlpfc and greater functional connectivity between dlpfc and premotor and parietal areas.4 in contrast, in the emotion condition dlpfc was less active overall but had greater functional connectivity with a neural network known as the default mode network (dmn; this network will be further explained below). thus, pinho and colleagues concluded that activation of dlpfc is not intrinsic to creative cognition: creative performance can be associated with both an increase and reduction of activity in the dlpfc (and other working memory regions) 3in this study, participants’ creativity was assessed as a function of creative demands of their undergraduate studies. participants in the experimental group were enrolled in media and design studies while the control group studied psychology, social, and educational sciences. 4functional connectivity refers to simultaneous patterns of activation between brain regions that are anatomically separated. such patterns are thought to reflect the level of communication between different brain regions (heuvel & pol, 2010). tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 8 within-person, depending on task prerequisites. the findings clearly indicate that creativity is the result of a dynamic interplay between several brain regions, networks, and systems, and that the patterns of brain activity during creative problem solving depend strongly on the employed problem solving strategies. (pinho et al., 2015, p. 3061) this study suggests that instead of focusing on a specific form of attention as the key to creativity, one should focus on the agent’s flexible use of multiple forms to support the task at hand. multiple meta-analyses support this approach to creative cognition, using a coordinate-based method to detect overlaps between brain areas reported by individual studies (boccia et al., 2015; gonen-yaacovi et al., 2013). these meta-analyses indicate that creative cognition is supported by distributed neural networks comprising brain areas associated with both top-down and bottom-up attention. for example, gonen-yaacovi et al. (2013) reported that many creative activities share common neural correlates, including areas associated with both top-down (e.g., lateral prefrontal cortex) and bottom-up attention (e.g., angular gyrus), regardless of domain (e.g., verbal, musical, visual). yet, each creative activity also has its own neural signature. for example, activation of motor planning areas is more strongly associated with musical improvisation tasks relative to verbal tasks (boccia et al., 2015). these results suggest that creative cognition is supported by a flexible interplay between top-down and bottom-up attention. this flexible interplay corresponds with the phases of creative cognition: reduced top-down and increased bottom-up attention benefits the idea generation phase of creative cognition, while increased top-down attention benefits the evaluation phase (beaty et al., 2016; ellamil et al., 2012; kleinmintz et al., 2019). a study conducted by liu et al. (2015) illustrates this flexible relationship. the experimenters asked expert and novice poets to improvise and then revise poems in an fmri scanning session. they reported deactivation and dissociation of fpn areas during the generation phase, but increased activation and collaboration with other areas during the evaluation phase.5 such findings indicate that rather than dependence on top-down or bottom-up attention, creative cognition is characterized by leveraging both attentional processes to generate, modify, and evaluate ideas and products (beaty et al., 2016). the flexible relationship between top-down and bottom-up attention is reflected in recent models of creative cognition. for example, the model of creativity and attention (moca) suggests that different configurations of top-down and bottomup attention might be at work depending on different components and stages of 5in particular, they found dissociations between areas of fpn (dlpfc and intraparietal sulcus) and the medial prefrontal cortex (mpfc) during idea generation, such that mpfc showed increased activation, while dlpfc was deactivated. thus, the generation phase is characterized by reduced top-down control but also a more spontaneous and explorative mode of thought. the activation pattern, however, was reconfigured during the evaluation phase, such that dlpfc coupled with mpfc and exhibited increased activation (liu et al., 2015). tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dynamic attentional mechanisms of creative cognition 9 creativity (vartanian, 2009; zabelina, 2018; zabelina et al., 2019; zabelina & robinson, 2010). notably, while a flexible relation between top-down and bottom-up attention is conducive to most creative activities, the extent of their contribution depends on the type, context, and constraints of each creative activity (chrysikou et al., 2014; pinho et al., 2015; zabelina & ganis, 2018). this flexible view offers a more holistic and consistent account of attentional mechanisms of creative cognition. however, even this flexible view leaves some significant components of creative cognition unexplained. as we will see in the following section, spontaneous self-generated thoughts, imagination, and mindwandering make integral contributions to idea generation. this internal and selfgenerated content does not exactly align with the standard view of bottom-up attention since its source is not the surrounding world.6 similarly, such content does not perfectly fit the top-down definition since it might occur involuntarily and in a task-irrelevant manner that can distract from the task-at-hand. thus, the top-down/bottom-up taxonomy falls short of fully explaining the role of mental contents in creative cognition. below, we review a different taxonomy of attention, internal vs. external, that does a better job of capturing such content. 3.2 internal versus external attention internal and external attention is a newer taxonomy coined by chun et al. (2011). this taxonomy was proposed partly to counter the view of attention as a unitary construct or mechanism, instead asserting that attention is “a characteristic and property of multiple perceptual and cognitive control mechanisms” (chun et al., 2011, p. 74). instead of focusing on the source of input, chun, golomb, and turkbrowne are concerned with the targets of attention. the targets of external attention are externally-generated (e.g., sensory stimuli), whereas the targets of internal attention are internally-generated (e.g., memories; chun et al., 2011, p. 77). imagine sitting in front of your computer at work: external attention would include focusing on the computer screen, carefully listening to a conversation in the hallway, or enjoying the taste of coffee while sitting at your desk, while internal attention would include focusing on your thoughts, thinking about a talk you attended, trying to remember the name of an author, or deciding where to go for lunch (chun et al., 2011). of course, there is substantial overlap between internally and externally-generated mental content—focusing on a talk might be external attention one day and internal attention the next. along those lines, subsequent work has found internal and external attention to share significant resources: “ external and internal attention mostly operates based on a common, central pool of resources” (verschooren et al., 2019, p. 478). 6some researchers have nonetheless suggested that bottom-up attention can include internal stimuli, such as hallucinations (see, e.g., hugdahl, 2009, p. 553: “a new view is presented which sees auditory hallucinations as bottom-up driven perceptual phenomena originating from the left temporal lobe that are out of control due to failure of top-down inhibition”). tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 10 external attention is mainly supported by visual, somatomotor, and dorsal attention (dan) brain networks. the somatomotor network is responsible for movements and touch, comprising primary motor, premotor, and primary sensory cortices. dan allows for deliberate attention to visual stimuli and spatial locations. in contrast, internal attention is supported by the “limbic” network and dmn. the limbic network, including the subgenual anterior cingulate cortex and the amygdala, involves the processing of emotions (zabelina & andrews-hanna, 2016). dmn mainly consists of medial prefrontal cortex, the posterior inferior parietal cortex, and the medial temporal lobe. together, dmn regions support memory retrieval, imagination, future planning, metacognition, mental inference, and self-reflection (buckner et al., 2008; raichle et al., 2001). both internal and external attention can couple with cognitive control areas to support goal-directed behavior (dixon et al., 2014; zabelina & andrews-hanna, 2016). recent work emphasizes the importance of internal attention for creative cognition based on supporting evidence from behavioral and neuroimaging studies (beaty et al., 2019; benedek, 2018; benedek & fink, 2019). below, we review this evidence and argue that internal attention is not sufficient for the fullest expressions of creative cognition. instead, creativity is best supported by a dynamic interplay between internal and external attention and, as we will argue, flexible cognitive control. 3.2.1 the role of internal attention in creative cognition mind-wandering is often characterized by disengagement from the immediate environment, thinking about what is unrelated to the here and now (schooler et al., 2011; smallwood et al., 2007). it is thought to result from internal attention but can be spontaneous or deliberate (zedelius & schooler, 2020). due to its disruptive effects on task performance, mind-wandering has been traditionally associated with adverse effects. studies show that frequent mind-wandering impairs performance on tasks that require sustained engagement with the external world, such as learning, reading, and driving (for a review, see mooneyham & schooler, 2013). despite its negative influence on task performance, mind-wandering has been repeatedly shown to be positively related to creative cognition (agnoli et al., 2018; baird et al., 2012; fox & christoff, 2018; gable et al., 2019; zedelius & schooler, 2020). one study reported that a considerable proportion of creative ideas occur during episodes of mind-wandering (gable et al., 2019). in this study, experimenters observed professional writers and physicists to examine the unfolding of creative ideas over two weeks. participants were asked what they were doing or thinking when a creative idea occurred and how creative they thought it was. results indicated that a significant proportion (20%) of reported creative ideas happened during episodes of mind-wandering. in addition, mind-wandering was particularly tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dynamic attentional mechanisms of creative cognition 11 conducive to creative ideas that occurred after the individual hit an impasse while working on a problem (gable et al., 2019).7 in another study, agnoli et al. (2018) asked participants to complete selfreport measures on their tendency to engage in deliberate and spontaneous mind-wandering as well as a creative achievement questionnaire. they also instructed participants to develop creative titles for well-known novels and movies (a divergent thinking task). results suggested that a higher tendency for deliberate mind-wandering was associated with more original ideas. during episodes of mind-wandering, mental content shifts toward unsolved problems, revisiting and exploring them with fewer constraints and from new angles that can lead to novel solutions (gable et al., 2019). internal attention has become a focal point in creativity research due to its contributions to memory retrieval, prospective planning, and mental simulations (benedek et al., 2016; benedek, 2018). furthermore, as we will show later, the coupling of internal attention with cognitive control mechanisms allows for the active generation, manipulation, and evaluation of ideas (beaty et al., 2019; dixon et al., 2014; zabelina & andrews-hanna, 2016). nonetheless, mounting evidence points to possible contributions to creative cognition from external attention. 3.2.2 the role of external attention in creative cognition working memory, and increased working memory capacity, is known to play an important role in creative cognition (see, e.g. gong et al., 2023), since it allows one to hold multiple ideas in a single space, examining ways in which they might be combined. yet, working memory can be extended externally, allowing us to use our environment in the process of creative cognition (see, e.g., maglio & kirsh, 1992). in that vein, we explore the contribution of external attention to creative cognition. a series of studies reports lower performance in different creative tasks when the external stimulus is masked (benedek et al., 2011, 2016).8 in a within-subject design, participants performed either a convergent or divergent creative task.9 in half of the trials, the task stimulus was masked after 500 milliseconds (long enough to clearly see the stimulus), while in the other half, the stimulus remained unmasked for the entire duration of task performance. results showed that performance in 7see zedelius & schooler (2020) for discussions about types of daydreaming and mind-wandering that lead to higher creativity and how that relationship is impacted by individual differences. 8the original design of this experiment did not concern the role of external attention, but the results indicate that having access to external stimuli contributes to higher creative performance. 9a divergent task might present participants with a four-letter word stimulus (e.g., “post”) and ask them to generate an original and meaningful sentence by using the stimulus letters as initial letters (e.g., “oldies sometimes provoke tears”). a convergent task, on the other hand, might ask participants to rearrange all four letters of the stimulus to find a new and meaningful word (i.e., “stop). arguably, the divergent task requires a higher degree of self-generated thought and recombination of mental representations compared to the convergent task (benedek et al., 2016). tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 12 both convergent and divergent tasks was significantly higher in the unmasked condition relative to the masked condition. a possible explanation for this finding is that in the unmasked condition, the stimulus can be retrieved externally at any time, freeing up working memory resources. these additional resources can then be used for further higher-order processing and manipulation of mental representations, allowing for greater creativity. the masked condition, in contrast, would be associated with a higher load on working memory, since the stimulus has to be maintained and processed in memory throughout the task (benedek et al., 2011, 2016). in our view, the masked condition approximates the absence of external attention, since in both cases the agent is missing access to the external stimulus. thus, this study demonstrates the ways in which external attention might support creative cognition. observations of real-life creative performance—and not just tasks confined to laboratory settings—also point to the importance of dynamic interactions between external and internal attention. for example, when working on non-trivial proofs and problems, mathematicians tend to not only focus on the internal contents of the mind, but they do so while constantly interacting with the environment by creating and manipulating diagrams and inscriptions using chalk and blackboard or pen and paper (johansen & misfeldt, 2020; marghetis et al., 2019). we see this facilitation as akin to an extension of working memory, enabling greater creativity. it is clear in the above cases that external attention is conducive to creative cognition. we view its role as supplementary to internal attention, while nonetheless making significant contributions to the creative process. it does this, in part, by using the body and surrounding environment as extended memory, allowing for greater allocation of working memory resources toward the generation and manipulation of novel mental representations (ballard et al., 1997; clark & chalmers, 1998; goldin-meadow et al., 2001). one might look at this through the lens of “material anchors,” often used to describe the role of external representations in cognition. the thought is that humans project abstract concepts to external representations in order to anchor (stabilize) their understanding of such concepts. anchoring can lighten the load of memory and facilitate further higher-order processing and inference making (hutchins, 2005; see also fauconnier & turner, 2002). mathematicians’ constant interactions with diagrams and notations while working on non-trivial problems reflect the significance of external representations in creative cognition. 3.3 combining top-down/bottom-up and internal/external attention for creative cognition how are the two taxonomies we reviewed here related to each other? one might be tempted to map internal attention to top-down and external attention to bottomup, since external stimuli are typically responsible for bottom-up attention. however, it is best to think of the two taxonomies as orthogonal. both internal and tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dynamic attentional mechanisms of creative cognition 13 external attention can exhibit properties of either bottom-up or top-down attention. for example, external attention can be either task-relevant (i.e. top-down) or stimulus-driven (i.e. bottom-up) so long as its target is external stimuli. similarly, internal attention can be either top-down or bottom-up. intrusive thoughts or spontaneous mind-wanderings during which an individual unintentionally thinks about events or memories disconnected from the present time or location would count as bottom-up attention, whereas intentional and directed mind-wandering— for instance to plan future events, or to remember someone’s address—would count as top-down attention (dixon et al., 2014). there is, thus, no one-to-one map between the bottom-up/top-down and internal/external taxonomies of attention. consequently, we suggest that the role of attention in creative cognition can be depicted as a two-dimensional space, with the target of attention (i.e., internal and external) as one dimension and the source of attention (i.e., top-down and bottom-up) as the second dimension. we argue that this two-dimensional space sufficiently accounts for the reported flexibility and diversity of attention across different stages, components, and constraints of creative endeavors (see figure ⁇). this two-dimensional space is continuous but includes four zones of attention: external and top-down, external and bottom-up, internal and bottom-up, internal and top-down. zone 1 (external and top-down) captures instances of creative cognition when individuals intentionally pay attention to the external world. for example, in a jazz quartet musicians actively listen to the music made by other group members while also monitoring any feedback from the audience (dixon et al., 2014). likewise, when working on a problem, mathematicians actively make inscriptions and interact with them (marghetis et al., 2019). artists, too, intentionally pay attention to the physical properties of various media—the canvas, the colors, the clay, etc.—to adapt their creative endeavor to the constraints of each medium. zone 2 (external and bottom-up) captures instances of creativity that are triggered in a bottom-up fashion; that is, by the salient physical properties of some object in the world. accounts of serendipitous innovations and leaky attention (zabelina et al., 2016) suggest the involvement of this attentional zone. zone 2 might contribute to creative cognition by adding randomness to the search process. randomness can reduce bias for prior knowledge and common ideas which can in turn increase the probability of exploring a wider area of the search space (kirsh, 2014). for example, during jazz improvisation, musicians might produce novel and surprising notes inspired by perceived bottom-up cues from the audience or notes generated by other group members (ramalho & ganascia, 1994). a famous case of this is the serendipitous discovery of penicillin by fleming that occurred by his noticing an unusual pattern of bacterial growth in a petri dish (gaynes, 2017). zone 3 (internal and bottom-up) captures spontaneous internal thoughts that might distract from the task-at-hand but contribute to an overall creative goal. an example is poincaré’s account of the sudden discovery of a solution to a non-trivial mathematical problem while stepping onto a bus during vacation (poincaré, 2015). tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 14 figure 1: this two-dimensional space depicts different configurations of forms of attention that support creative cognition across a range of constraints, components, and stages. the vertical axis represents the target of attention (i.e., internal and external) and the horizontal axis represents the source of attention (i.e., top-down and bottom-up). this space is continuous; however, to make interpretation easier, we divide it into four general zones as marked by dashed lines and numbers (1, 2, 3, 4). these four zones characterize creative endeavors that are supported respectively by external and top-down, external and bottom-up, internal and bottom-up, and internal and top-down forms of attention. the circle in the middle represents the idea that most creative endeavors, from start to end, rely on all four zones, for example, musical improvisation or mathematical and scientific discoveries. tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dynamic attentional mechanisms of creative cognition 15 likewise, kekulé’s account of his major discovery of the hexagonal structure of the benzene molecule indicates that the idea came to him while daydreaming (see rothenberg, 1995)—or what we refer to as episodes of spontaneous mindwandering. similar to zone 2, zone 3 might also contribute to creativity through increased randomness, and thus, a more explorative search. but unlike zone 2, the randomness in zone 3 might primarily originate from internal thoughts, not salient stimuli in the immediate surrounding world. zone 4 (internal and top-down) captures instances of creative cognition that rely on guided memory retrieval and maintenance of internal trains of thought through inhibition of irrelevant internal and external stimuli. lab-based creativity tests such as aut and rat mostly involve this attentional zone; such tasks require minimal interaction with the outside world and are characterized by goal-directed memory retrieval and mental simulations. while some creative tasks can be located in one of the four zones (e.g., aut), most creative endeavors require interactions between multiple attentional zones encompassing internal, external, bottom-up and top-down forms of attention. take our earlier example of the musical improvisation of a jazz quartet. at some stages of the process, internally-oriented attention might be enough to spontaneously generate musical ideas, which are then translated into motor activities resulting in external sounds. yet, at most stages the musicians will also need cognitive control to make dynamic shifts between internal and external attention in order to optimize the creative process. the cooperation and coordination of top-down and bottom-up, internal and external attention allows musicians to be flexible and adapt themselves to unexpected changes in the musical patterns. jazz musicians retrieve prior knowledge of music and generate new musical patterns while physically interacting with an instrument and monitoring other band members, the overall rhythmic patterns, and feedback from the audience (see dixon et al., 2014; loui & guetta, 2019; ramalho & ganascia, 1994). the cooperation and coordination across forms of attention is also evident when we zoom in on accounts of major discoveries by poincaré, fleming, and kekulé, mentioned above. at first glance, these discoveries might seem solely reliant on attentional zones 2 and 3. nevertheless, in all three scenarios, the scientists had already started familiarizing themselves with the constraints of each problem and its corresponding solution space. what’s more, these accounts indicate that after finding the solution, the scientists deliberately tested the solution to make sure of its viability (gaynes, 2017; poincaré, 2015; rothenberg, 1995). that is, the formation of the solution was both preceded and followed by attentional zones 1 and 4. in other words, the creative process, from start to end, includes dynamic configurations of internal, external, top-down, and bottom-up attention. such dynamic attentional mechanisms of creative cognition are also reflected in recent neuroimaging studies of creativity. we will review some of these studies in the next section. tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 16 3.3.1 dynamic attentional mechanisms of creative cognition as evidenced by neuroimaging studies as mentioned earlier, internal and external attention can both draw on cognitive control to fulfill goal-directed behavior. in addition, internal and external attention can communicate and interact with each other through fpn and sn (zabelina & andrews-hanna, 2016). sn mainly consists of the anterior insula and dorsal anterior cingulate cortex (menon & uddin, 2010; uddin, 2017). the core function of this network is to identify salient stimuli (internal or external), mark them for additional processing, and guide target brain regions to generate the appropriate behavioral response. neuroimaging studies report a positive coupling between areas related to internal attention (dmn) and those related to cognitive control (fpn) during creative performances (beaty et al., 2018; de pisapia et al., 2016; ellamil et al., 2012; shi et al., 2018). consistent reports of this positive coupling suggest that although internal attention supports creativity by facilitating memory retrieval and mental simulations, it is insufficient. any retrieved memory contents and mental representations will require further manipulation to satisfy task constraints and stay in line with the creative goal. such goal-directed behavior is best supported by cognitive control mechanisms and their realizers (i.e., fpn). further, goal-directed internal attention needs to be maintained and shielded from distracting stimuli (both internal and external) over time; this function, too, is effectively fulfilled by cognitive control. additionally, depending on the type and constraints of the creative task, individuals might need to actively interact with the surrounding world, get inspired by, and/or manipulate relevant external stimuli. thus, creative cognition seems to maximally benefit from dynamic interactions between internal and external attention as well as cognitive control, such as topdown attention. this is evident from a study during which fmri data were recorded from 163 participants while performing a divergent thinking task (i.e., aut) and a control task (asking participants to think of typical characteristics of everyday objects, beaty et al., 2018). the data were then used to identify a whole-brain network associated with high-creative ability, which indicated that highly creative individuals can simultaneously engage three large-scale brain networks—dmn, fpn, and sn (beaty et al., 2018). this is in notable contrast to the observed anti-correlation of dmn and fpn in other, non-creative tasks (see, e.g., fox et al., 2005). a similar pattern of simultaneous engagement of three large-scale brain networks is reported for instances of mind-wandering. christoff et al. (2009) conducted an fmri study to examine the neural mechanisms of mind-wandering. they used an experience sampling method to intermittently probe participants about their current mental state as they performed a simple go/no-go task. results indicated that dmn, fpn, and sn regions were all recruited during mind-wandering episodes (for similar results, see fox et al., 2015; fox & christoff, 2018). tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dynamic attentional mechanisms of creative cognition 17 together, the neuroimaging studies of mind-wandering and creative task performance attest to dynamic attentional mechanisms underlying creative cognition. beaty et al. (2018, p. 1090) explain this as follows: the default network contributes to the generation of ideas via flexible and spontaneous combinatory mechanisms involved in memory retrieval and mental simulation. the salience network, in turn, functions to identify candidate ideas—potentially useful information generated via the default network—and forward such information to frontoparietal executive systems for high-order processing (e.g., idea evaluation, elaboration, or revision). effective communication across multiple attentional forms is reflected in the coupling of dmn and fpn through sn, which upon detecting a salient event, facilitates rapid access to relevant memory, attention, and motor resources (menon & uddin, 2010). 4 discussion creative cognition is characterized as a biphasic phenomenon, typically moving from idea generation to idea evaluation. both phases require selection in one form or another, whether to direct the process of generation or to support the process of evaluation. thus, examining attentional mechanisms is an excellent way to gain insight into creative cognition. we did this through the lens of two major taxonomies: top-down versus bottom-up attention, and internal versus external attention. previous research focused on the presence or absence of attention, typically through a single form of attention. we discussed, for example, studies that report either reduced top-down attention or increased internal attention as conducive to creative cognition. however, we find that a dynamic relationship between different forms of attention is most beneficial for creative cognition. various configurations of top-down and bottom-up, internal and external attention can benefit creativity; some configurations are more useful during idea generation (zone 2 and zone 3), whereas others are more useful during idea evaluation (zone 1 and zone 4). the dynamic interplay between different forms of attention is reflected in the neural underpinnings of creative cognition, and especially the close interaction between dmn, fpn, and sn. dmn and fpn are typically anti-correlated, with dmn being more active during resting states and fpn more active during tasks that rely on interactions with external stimuli. during resting states, one tends to engage with internal contents (e.g., memories) without exerting top-down control. hence, fpn shows reduced activation, while dmn shows increased activation. in contrast, during tasks that rely on interactions with external stimuli, mind-wandering can distract and impair performance. top-down control is needed to keep one engaged tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 18 with the task and inhibit unwanted thoughts (such as memories). hence, in this scenario, fpn shows increased activation, while dmn shows reduced activation. creative cognition, however, requires that we think beyond the typical opposition between task engagement and resting state. creative cognition requires cognitive control mechanisms, such as top-down attention, to maintain, manipulate, and assess ideas that are generated according to the constraints of the creative task. yet, it also requires that new ideas are generated “bottom-up,” which can occur through both internal (e.g., memory retrieval) and external (e.g., sensing) mechanisms. this flexibility and cooperation—across dmn, fpn, and sensory cortices—is facilitated by sn through a moment-by-moment identification of salient events (internal or external) and allocation of relevant memory, attentional, and motor resources to elicit an appropriate behavioral response. this is in contrast with non-creative tasks, which can be harmed by insufficient exclusion of other brain networks and stimuli. the coupling between dmn and fpn has also been reported in phenomena often associated with enhanced creativity, including mind-wandering (see the previous section), altered states of consciousness induced by hallucinogen consumption (palhano-fontes et al., 2015), and meditation (hasenkamp & barsalou, 2012; josipovic et al., 2012). to further explain the contributions of dmn/fpn coupling to creative cognition, some draw an analogy to simulated annealing (carhart-harris & friston, 2019; girn et al., 2020; gopnik et al., 2017). annealing is a metallurgy term that refers to the process of repeatedly heating a metal and then slowly cooling it down, resulting in a more robust structure. similarly, in computer science, the optimal solution is reached when the system starts with a broader (i.e., high-temperature) search and gradually moves to a narrower (i.e., low-temperature) search. a heated system is more flexible in its movements across the possible solutions, which enables it to escape local minima and continue the search until the optimal solution is reached (kirkpatrick et al., 1983; metropolis et al., 1953). likewise, the initial enhanced coupling between dmn and sn reflects a heated, explorative, and flexible search suitable for idea generation in creative cognition. the subsequent coupling between dmn and fpn through sn reflects the gradual cooling of the system, which allows for assessment, revision, elaboration, and selection of candidate ideas. 5 summary and conclusion creative cognition is a complex phenomenon emerging from interactions across common cognitive processes, such as memory, executive functions, and attention. the present paper reviewed attentional mechanisms of creative cognition with an emphasis on two major taxonomies: top-down/ bottom-up and internal/ external attention. these taxonomies can be conceived as orthogonal: the division between top-down and bottom-up attention is based on the source of attentional tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dynamic attentional mechanisms of creative cognition 19 input, while the division between internal and external attention is based on the target of attention. we argued that engagement across the taxonomies best captures the complex and flexible nature of creative cognition. creativity does not solely depend on top-down or bottom-up processes; neither does it solely depend on internal or external attention. a dynamic interplay between these different forms of attention is required to best enable novel ideas' generation, evaluation, and elaboration. we support these claims with research showing that dynamic interaction between types of attention in each taxonomy is conducive to creativity, as well as supporting research on the neural basis of creative cognition. however, we also found that the extent of contribution from each type hinges on the context and constraints of the creative activity. thus, determining the most conducive contributions of each type will require further research. references abraham, a. (2018). the forest versus the trees: creativity, cognition and imagination. in r. jung & o. vartanian (eds.), the cambridge handbook of the neuroscience of creativity (pp. 195–210). cambridge university press. https://doi.org/ 10.1017/9781316556238.012 agnoli, s., vanucci, m., pelagatti, c., & corazza, g. e. (2018). exploring the link between mind wandering, mindfulness, and creativity: a multidimensional approach. creativity research journal, 30(1), 41–53. https://doi.org/10.1080/1040 0419.2018.1411423 allen, a. p., & thomas, k. e. (2011). a dual process account of creative thinking. creativity research journal, 23(2), 109–118. https://doi.org/10.1080/10400419.2011.571183 ansburg, p. i., & hill, k. (2003). creative and analytic thinkers differ in their use of attentional resources. personality and individual differences, 34(7), 1141–1152. https://doi.org/10.1016/s0191-8869(02)00104-6 aziz‐zadeh, l., kaplan, j. t., & iacoboni, m. (2009). “aha!”: the neural correlates of verbal insight solutions. human brain mapping, 30(3), 908–916. https://doi.org/10.1002/hbm.20554 baird, b., smallwood, j., mrazek, m. d., kam, j. w. y., franklin, m. s., & schooler, j. w. (2012). inspired by distraction: mind wandering facilitates creative incubation. psychological science, 23(10), 1117–1122. https://doi.org/10.1177/09567976 12446024 ballard, d. h., hayhoe, m. m., pook, p. k., & rao, r. p. n. (1997). deictic codes for the embodiment of cognition. behavioral and brain sciences, 20(4), 723–767. https://doi.org/10.1017/s0140525x97001611 baluch, f., & itti, l. (2011). mechanisms of top-down attention. trends in neurosciences, 34(4), 210–224. https://doi.org/10.1 016/j.tins.2011.02.003 beaty, r. e., benedek, m., silvia, p. j., & schacter, d. l. (2016). creative cognition and brain network dynamics. trends in cognitive sciences, 20(2), 87–95. https://doi.org/10.1016/j.tics.2015.10.004 beaty, r. e., kenett, y. n., christensen, a. p., rosenberg, m. d., benedek, m., chen, q., fink, a., qiu, j., kwapil, t. r., kane, m. j., & silvia, p. j. (2018). robust prediction of individual creative ability from brain functional connectivity. proceedings of the national academy of sciences, 115(5), 1087–1092. https://doi.org/10.1073/pnas.1713532115 beaty, r. e., seli, p., & schacter, d. l. (2019). network neuroscience of creative cognition: mapping cognitive mechanisms and individual differences in the creative brain. current opinion in behavioral sciences, 27, 22–30. https://doi.org/10.1 016/j.cobeha.2018.08.013 benedek, m. (2018). internally directed attention in creative cognition. in r. e. jung & o. vartanian (eds.), the cambridge handbook of the neuroscience of creativity (1st ed., pp. 180–194). cambridge university press. https://doi.org/10.1017/ 9781316556238.011 benedek, m., bergner, s., könen, t., fink, a., & neubauer, a. c. (2011). eeg alpha synchronization is related to top-down processing in convergent and divergent thinking. neuropsychologia, 49(12), 3505–3511. https://doi.org/10.1016/j.neur opsychologia.2011.09.004 benedek, m., & fink, a. (2019). toward a neurocognitive framework of creative cognition: the role of memory, attention, and cognitive control. current opinion in behavioral sciences, 27, 116–122. https://doi.org/10.1016/j.cobeha.2018.11.002 tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1017/9781316556238.012 https://doi.org/10.1017/9781316556238.012 https://doi.org/10.1080/10400419.2018.1411423 https://doi.org/10.1080/10400419.2018.1411423 https://doi.org/10.1080/10400419.2011.571183 https://doi.org/10.1016/s0191-8869(02)00104-6 https://doi.org/10.1002/hbm.20554 https://doi.org/10.1177/0956797612446024 https://doi.org/10.1177/0956797612446024 https://doi.org/10.1017/s0140525x97001611 https://doi.org/10.1016/j.tins.2011.02.003 https://doi.org/10.1016/j.tins.2011.02.003 https://doi.org/10.1016/j.tics.2015.10.004 https://doi.org/10.1073/pnas.1713532115 https://doi.org/10.1016/j.cobeha.2018.08.013 https://doi.org/10.1016/j.cobeha.2018.08.013 https://doi.org/10.1017/9781316556238.011 https://doi.org/10.1017/9781316556238.011 https://doi.org/10.1016/j.neuropsychologia.2011.09.004 https://doi.org/10.1016/j.neuropsychologia.2011.09.004 https://doi.org/10.1016/j.cobeha.2018.11.002 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 20 benedek, m., jauk, e., beaty, r. e., fink, a., koschutnig, k., & neubauer, a. c. (2016). brain mechanisms associated with internally directed attention and self-generated thought. scientific reports, 6(1), 22959. https://doi.org/10.1038/srep22959 benedek, m., jauk, e., sommer, m., arendasy, m., & neubauer, a. c. (2014). intelligence, creativity, and cognitive control: the common and differential involvement of executive functions in intelligence and creativity. intelligence, 46, 73–83. https://doi.org/10.1016/j.intell.2014.05.007 bengtsson, s. l., csıkszentmihalyi, m., & ullen, f. (2007). cortical regions involved in the generation of musical structures during improvisation in pianists. journal of cognitive neuroscience, 19(5), 830–842. https://doi.org/10.1162/jocn.2007. 19.5.830v boccia, m., piccardi, l., palermo, l., nori, r., & palmiero, m. (2015). where do bright ideas occur in our brain? meta-analytic evidence from neuroimaging studies of domain-specific creativity. frontiers in psychology, 6, 1–12. https://doi.org/10.3389/fpsyg.2015.01195 buckner, r. l., andrews-hanna, j. r., & schacter, d. l. (2008). the brain’s default network. annals of the new york academy of sciences, 1124(1), 1–38. https://doi.org/10.1196/annals.1440.011 buschman, t. j., & miller, e. k. (2007). top-down versus bottom-up control of attention in the prefrontal and posterior parietal cortices. science, 315(5820), 1860–1862. https://doi.org/10.1126/science.1138071 campbell, d. t. (1960). blind variation and selective retentions in creative thought as in other knowledge processes. psychological review, 67(6), 380–400. https://doi.org/10.1037/h0040373 carhart-harris, r. l., & friston, k. j. (2019). rebus and the anarchic brain: toward a unified model of the brain action of psychedelics. pharmacological reviews, 71(3), 316–344. https://doi.org/10.1124/pr.118.017160 carson, s. h., peterson, j. b., & higgins, d. m. (2003). decreased latent inhibition is associated with increased creative achievement in high-functioning individuals. journal of personality and social psychology, 85(3), 499–506. https://doi.org/10.1037/0022-3514.85.3.499 christoff, k., gordon, a. m., smallwood, j., smith, r., & schooler, j. w. (2009). experience sampling during fmri reveals default network and executive system contributions to mind wandering. proceedings of the national academy of sciences, 106(21), 8719–8724. https://doi.org/10.1073/pnas.0900234106 chrysikou, e. g. (2018). the costs and benefits of cognitive control for creativity. in r. jung & o. vartanian (eds.), the cambridge handbook of the neuroscience of creativity (pp. 195–210). cambridge university press. https://doi.org/10.1 017/9781316556238.018 chrysikou, e. g., hamilton, r. h., coslett, h. b., datta, a., bikson, m., & thompson-schill, s. l. (2013). noninvasive transcranial direct current stimulation over the left prefrontal cortex facilitates cognitive flexibility in tool use. cognitive neuroscience, 4(2), 81–89. https://doi.org/10.1080/17588928.2013.768221 chrysikou, e. g., weber, m. j., & thompson-schill, s. l. (2014). a matched filter hypothesis for cognitive control. neuropsychologia, 62, 341–355. https://doi.org/10.1016/j.neuropsychologia.2013.10.021 chun, m. m., golomb, j. d., & turk-browne, n. b. (2011). a taxonomy of external and internal attention. annual review of psychology, 62(1), 73–101. https://doi.org/10.1146/annurev.psych.093008.100427 clark, a., & chalmers, d. (1998). the extended mind. analysis, 58(1), 7–19. https://doi.org/10.1093/analys/58.1.7 cole, m. w., & schneider, w. (2007). the cognitive control network: integrated cortical regions with dissociable functions. neuroimage, 37(1), 343–360. https://doi.org/10.1016/j.neuroimage.2007.03.071 connor, c. e., egeth, h. e., & yantis, s. (2004). visual attention: bottom-up versus top-down. current biology, 14(19), 850–852. https://doi.org/10.1016/j.cub.2004.09.041 corbetta, m., & shulman, g. l. (2002). control of goal-directed and stimulus-driven attention in the brain. nature reviews neuroscience, 3(3), 201–215. https://doi.org/10.1038/nrn755 de pisapia, n., bacci, f., parrott, d., & melcher, d. (2016). brain networks for visual creativity: a functional connectivity study of planning a visual artwork. scientific reports, 6(1), 39185. https://doi.org/10.1038/srep39185 dietrich, a., & kanso, r. (2010). a review of eeg, erp, and neuroimaging studies of creativity and insight. psychological bulletin, 136(5), 822. https://doi.org/10.1037/a0019749 dixon, m. l., fox, k. c. r., & christoff, k. (2014). a framework for understanding the relationship between externally and internally directed cognition. neuropsychologia, 62, 321–330. https://doi.org/10.1016/j.neuropsychologia.2014.05.024 edl, s., benedek, m., papousek, i., weiss, e. m., & fink, a. (2014). creativity and the stroop interference effect. personality and individual differences, 69, 38–42. https://doi.org/10.1016/j.paid.2014.05.009 ellamil, m., dobson, c., beeman, m., & christoff, k. (2012). evaluative and generative modes of thought during the creative process. neuroimage, 59(2), 1783–1794. https://doi.org/10.1016/j.neuroimage.2011.08.008 fauconnier, g., & turner, m. (2002). the way we think: conceptual blending and the mind’s hidden complexities. basic books. https://psycnet.apa.org/record/2002-02521-000 tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1038/srep22959 https://doi.org/10.1016/j.intell.2014.05.007 https://doi.org/10.1162/jocn.2007.19.5.830v https://doi.org/10.1162/jocn.2007.19.5.830v https://doi.org/10.3389/fpsyg.2015.01195 https://doi.org/10.1196/annals.1440.011 https://doi.org/10.1126/science.1138071 https://doi.org/10.1037/h0040373 https://doi.org/10.1124/pr.118.017160 https://doi.org/10.1037/0022-3514.85.3.499 https://doi.org/10.1073/pnas.0900234106 https://doi.org/10.1017/9781316556238.018 https://doi.org/10.1017/9781316556238.018 https://doi.org/10.1080/17588928.2013.768221 https://doi.org/10.1016/j.neuropsychologia.2013.10.021 https://doi.org/10.1146/annurev.psych.093008.100427 https://doi.org/10.1093/analys/58.1.7 https://doi.org/10.1016/j.neuroimage.2007.03.071 https://doi.org/10.1016/j.cub.2004.09.041 https://doi.org/10.1038/nrn755 https://doi.org/10.1038/srep39185 https://doi.org/10.1037/a0019749 https://doi.org/10.1016/j.neuropsychologia.2014.05.024 https://doi.org/10.1016/j.paid.2014.05.009 https://doi.org/10.1016/j.neuroimage.2011.08.008 https://psycnet.apa.org/record/2002-02521-000 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dynamic attentional mechanisms of creative cognition 21 fox, k. c. r., & christoff, k. (2018). the oxford handbook of spontaneous thought: mind-wandering, creativity, and dreaming. oxford university press. https://doi.org/10.1093/oxfordhb/9780190464745.001.0001 fox, k. c. r., spreng, r. n., ellamil, m., andrews-hanna, j. r., & christoff, k. (2015). the wandering brain: meta-analysis of functional neuroimaging studies of mind-wandering and related spontaneous thought processes. neuroimage, 111, 611–621. https://doi.org/10.1016/j.neuroimage.2015.02.039 fox, m. d., snyder, a. z., vincent, j. l., corbetta, m., essen, d. c., & raichle, m. e. (2005). the human brain is intrinsically organized into dynamic, anticorrelated functional networks. proceedings of the national academy of sciences, 102(27), 9673–9678. https://doi.org/10.1073/pnas.0504136102 gable, s. l., hopper, e. a., & schooler, j. w. (2019). when the muses strike: creative ideas of physicists and writers routinely occur during mind wandering. psychological science, 30(3), 396–404. https://doi.org/10.1177/0956797618820626 gaynes, r. (2017). the discovery of penicillin—new insights after more than 75 years of clinical use. emerging infectious diseases, 23(5), 849–853. https://doi.org/10.3201/eid2305.161556 german, t. p., & defeyter, m. a. (2000). immunity to functional fixedness in young children. psychonomic bulletin & review, 7(4), 707–712. https://doi.org/10.3758/bf03213010 girn, m., mills, c., roseman, l., carhart-harris, r., & christoff, k. (2020). updating the dynamic framework of thought: creativity and psychedelics. neuroimage, 116726, 116726. https://doi.org/10.1016/j.neuroimage.2020.116726 goldin-meadow, s., nusbaum, h., kelly, s. d., & wagner, s. (2001). explaining math: gesturing lightens the load. psychological science, 12(6), 516–522. https://doi.org/10.1111/1467-9280.00395 gonen-yaacovi, g., souza, l. c., levy, r., urbanski, m., josse, g., & volle, e. (2013). rostral and caudal prefrontal contribution to creativity: a meta-analysis of functional imaging data. frontiers in human neuroscience, 7, 1–22. https://doi.org/10.3389/fnhum.2013.00465 gong, z., miao, k., liu, x., luo, m., yu, y., & chen, z. (2023). a positive association between working memory capacity and human creativity: a meta-analytic evidence. journal of intelligence, 11(1), 15. https://doi.org/10.3390/jintelligenc e11010015 gopnik, a., o’grady, s., lucas, c. g., griffiths, t. l., wente, a., bridgers, s., aboody, r., fung, h., & dahl, r. e. (2017). changes in cognitive flexibility and hypothesis search across human life history from childhood to adolescence to adulthood. proceedings of the national academy of sciences, 114(30), 7892–7899. https://doi.org/10.1073/pnas.1700811 114 guilford, j. p. (1950). creativity. american psychologist, 5(9), 444–454. https://doi.org/10.1037/h0063487 guilford, j. p. (1967). the nature of human intelligence. mcgraw-hill. https://psycnet.apa.org/record/1967-35015-000 hasenkamp, w., & barsalou, l. w. (2012). effects of meditation experience on functional connectivity of distributed brain networks. frontiers in human neuroscience, 6, 38. https://doi.org/10.3389/fnhum.2012.00038 herd, s. a., banich, m. t., & o’reilly, r. c. (2006). neural mechanisms of cognitive control: an integrative model of stroop task performance and fmri data. journal of cognitive neuroscience, 18(1), 22–32. https://doi.org/10.1162/0898929067 75250012 heuvel, van den, m. p., & pol, h. (2010). exploring the brain network: a review on resting-state fmri functional connectivity. european neuropsychopharmacology, 20(8), 519–534. https://doi.org/10.1016/j.euroneuro.2010.03.008 hommel, b., chapman, c. s., cisek, p., neyedli, h. f., song, j. h., & welsh, t. n. (2019). no one knows what attention is. attention, perception, & psychophysics, 81(7), 2288–2303. https://doi.org/10.3758/s13414-019-01846-w hugdahl, k. (2009). “hearing voices”: auditory hallucinations as failure of top‐down control of bottom‐up perceptual processes. scandinavian journal of psychology, 50(6), 553–560. https://doi.org/10.1111/j.1467-9450.2009.00775.x hutchins, e. (2005). material anchors for conceptual blends. journal of pragmatics, 37(10), 1555–1577. https://doi.org/10.1 016/j.pragma.2004.06.008 james, w. (1890). the principles of psychology (vol. 1). henry holt. https://doi.org/10.1037/10538-000 jennings, c. d. (2012). the subject of attention. synthese, 189, 535–554. https://doi.org/10.1007/s11229-012-0164-1 jennings, c. d. (2020). the attending mind. cambridge university press. https://doi.org/10.1017/9781108164238 johansen, m. w., & misfeldt, m. (2020). material representations in mathematical research practice. synthese, 197(9), 3721– 3741. https://doi.org/10.1007/s11229-018-02033-4 josipovic, z., dinstein, i., weber, j., & heeger, d. j. (2012). influence of meditation on anti-correlated networks in the brain. frontiers in human neuroscience, 5, 183. https://doi.org/10.3389/fnhum.2011.00183 kasof, j. (1997). creativity and breadth of attention. creativity research journal, 10(4), 303–315. https://doi.org/10.1207/s1 5326934crj1004_2 katsuki, f., & constantinidis, c. (2014). bottom-up and top-down attention: different processes and overlapping neural systems. the neuroscientist, 20(5), 509–521. https://doi.org/10.1177/1073858413514136 tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1093/oxfordhb/9780190464745.001.0001 https://doi.org/10.1016/j.neuroimage.2015.02.039 https://doi.org/10.1073/pnas.0504136102 https://doi.org/10.1177/0956797618820626 https://doi.org/10.3201/eid2305.161556 https://doi.org/10.3758/bf03213010 https://doi.org/10.1016/j.neuroimage.2020.116726 https://doi.org/10.1111/1467-9280.00395 https://doi.org/10.3389/fnhum.2013.00465 https://doi.org/10.3390/jintelligence11010015 https://doi.org/10.3390/jintelligence11010015 https://doi.org/10.1073/pnas.1700811114 https://doi.org/10.1073/pnas.1700811114 https://doi.org/10.1037/h0063487 https://psycnet.apa.org/record/1967-35015-000 https://doi.org/10.3389/fnhum.2012.00038 https://doi.org/10.1162/089892906775250012 https://doi.org/10.1162/089892906775250012 https://doi.org/10.1016/j.euroneuro.2010.03.008 https://doi.org/10.3758/s13414-019-01846-w https://doi.org/10.1111/j.1467-9450.2009.00775.x https://doi.org/10.1016/j.pragma.2004.06.008 https://doi.org/10.1016/j.pragma.2004.06.008 https://doi.org/10.1037/10538-000 https://doi.org/10.1007/s11229-012-0164-1 https://doi.org/10.1017/9781108164238 https://doi.org/10.1007/s11229-018-02033-4 https://doi.org/10.3389/fnhum.2011.00183 https://doi.org/10.1207/s15326934crj1004_2 https://doi.org/10.1207/s15326934crj1004_2 https://doi.org/10.1177/1073858413514136 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 22 kaufman, a. b., kornilov, s. a., bristol, a. s., tan, m., & grigorenko, e. l. (2010). the neurobiological foundation of creative cognition. in j. kaufman & r. sternberg (eds.), the cambridge handbook of creativity (pp. 216–232). cambridge university press. https://doi.org/10.1017/9781316979839 kirkpatrick, s., gelatt, c. d., jr, & vecchi, m. p. (1983). optimization by simulated annealing. science, 220(4598), 671–680. https://doi.org/10.1126/science.220.4598.671 kirsh, d. (2014). the importance of chance and interactivity in creativity. pragmatics & cognition, 22(1), 5–26. https://escholarship.org/uc/item/9zv112f5 kleinmintz, o. m., ivancovsky, t., & shamay-tsoory, s. g. (2019). the two-fold model of creativity: the neural underpinnings of the generation and evaluation of creative ideas. current opinion in behavioral sciences, 27, 131–138. https://doi.org/10.1016/j.cobeha.2018.11.004 limb, c. j., & braun, a. r. (2008). neural substrates of spontaneous musical performance: an fmri study of jazz improvisation. plos one, 3(2), 1679. https://doi.org/10.1371/journal.pone.0001679 liu, s., chow, h. m., xu, y., erkkinen, m. g., swett, k. e., eagle, m. w., rizik-baer, d. a., & braun, a. r. (2012). neural correlates of lyrical improvisation: an fmri study of freestyle rap. scientific reports, 2(1), 1–8. https://doi.org/10.103 8/srep00834 liu, s., erkkinen, m. g., healey, m. l., xu, y., swett, k. e., chow, h. m., & braun, a. r. (2015). brain activity and connectivity during poetry composition: toward a multidimensional model of the creative process. human brain mapping, 36, 3351–3372. https://doi.org/10.1002/hbm.22849 loui, p., & guetta, r. e. (2019). music and attention, executive function, and creativity. in m. h. thaut & d. a. hodges (eds.), the oxford handbook of music and the brain (pp. 263–284). oxford university press. https://doi.org/10.1093/ox fordhb/9780198804123.001.0001 maglio, p., & kirsh, d. (1992). some epistemic benefits of action: tetris, a case study. in j. kruschke (ed.), proceedings of the fourteenth annual conference of the cognitive science society (vol. 14, p. 224). lawrence erlbaum associates, inc. https://philpapers.org/archive/davseb.pdf marghetis, t., samson, k., & landy, d. (2019). the complex system of mathematical creativity: modularity, burstiness, and the network structure of how experts use inscriptions. in a. k. goel, c. m. seifert, & c. freksa (eds.), proceedings of the 41st annual conference of the cognitive science society (pp. 763–769). cognitive science society. https://cogsci.m indmodeling.org/2019/papers/0148/0148.pdf mednick, s. a., & mednick, m. t. (1967). examiner’s manual, remote associates test: college and adult forms 1 and 2. houghton mifflin. https://worldcat.org/en/title/2865561 menon, v., & uddin, l. q. (2010). saliency, switching, attention and control: a network model of insula function. brain structure & function, 214(5–6), 655–667. https://doi.org/10.1007/s00429-010-0262-0 metropolis, n., rosenbluth, a. w., rosenbluth, m. n., teller, a. h., & teller, e. (1953). equation of state calculations by fast computing machines. the journal of chemical physics, 21(6), 1087–1092. https://doi.org/10.1063/1.1699114 monroe, j., srikant, t., carbonell-bejerano, p., becker, c., lensink, m., exposito-alonso, m., & weigel, d. (2022). mutation bias reflects natural selection in arabidopsis thaliana. nature, 602, 101–105. https://doi.org/10.1038/s4158602104269-6 mooneyham, b. w., & schooler, j. w. (2013). the costs and benefits of mind-wandering: a review. canadian journal of experimental psychology/revue canadienne de psychologie expérimentale, 67(1), 11–18. https://doi.org/10.1037/a00315 69 nobre, a. c., & kastner, s. (2014). attention: time capsule 2013. in a. c. nobre & s. kastner (eds.), the oxford handbook of attention (pp. 1201–1222). oxford university press. https://doi.org/10.1093/oxfordhb/9780199675111.001.0001 nusbaum, e. c., & silvia, p. j. (2011). are intelligence and creativity really so different? fluid intelligence, executive processes, and strategy use in divergent thinking. intelligence, 39(1), 36–45. https://doi.org/10.1016/j.intell.2010.11.002 palhano-fontes, f., andrade, k. c., tofoli, l. f., santos, a. c., crippa, j. a. s., hallak, j. e. c., ribeiro, s., & araujo, d. b. de. (2015). the psychedelic state induced by ayahuasca modulates the activity and connectivity of the default mode network. plos one, 10(2), 0118143. https://doi.org/10.1371/journal.pone.0118143 pinho, a. l., ullén, f., castelo-branco, m., fransson, p., & manzano, ö. (2015). addressing a paradox: dual strategies for creative performance in introspective and extrospective networks. cerebral cortex, 26(7), 3052–3063. https://doi.org/ 10.1093/cercor/bhv130 poincaré, h. (2015). the foundations of science: science and hypothesis, the value of science, science and method (g. halstead, trans.). cambridge university press. https://doi.org/10.1017/cbo9781107252950 posner, m. i. (1980). orienting of attention. quarterly journal of experimental psychology, 32(1), 3–25. https://doi.org/10.1 080/00335558008248231 raichle, m. e., macleod, a. m., snyder, a. z., powers, w. j., gusnard, d. a., & shulman, g. l. (2001). a default mode of brain function. proceedings of the national academy of sciences, 98(2), 676–682. https://doi.org/10.1073/pnas.98.2.676 tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1017/9781316979839 https://doi.org/10.1126/science.220.4598.671 https://escholarship.org/uc/item/9zv112f5 https://doi.org/10.1016/j.cobeha.2018.11.004 https://doi.org/10.1371/journal.pone.0001679 https://doi.org/10.1038/srep00834 https://doi.org/10.1038/srep00834 https://doi.org/10.1002/hbm.22849 https://doi.org/10.1093/oxfordhb/9780198804123.001.0001 https://doi.org/10.1093/oxfordhb/9780198804123.001.0001 https://philpapers.org/archive/davseb.pdf https://cogsci.mindmodeling.org/2019/papers/0148/0148.pdf https://cogsci.mindmodeling.org/2019/papers/0148/0148.pdf https://worldcat.org/en/title/2865561 https://doi.org/10.1007/s00429-010-0262-0 https://doi.org/10.1063/1.1699114 https://doi.org/10.1038/s41586-021-04269-6 https://doi.org/10.1038/s41586-021-04269-6 https://doi.org/10.1037/a0031569 https://doi.org/10.1037/a0031569 https://doi.org/10.1093/oxfordhb/9780199675111.001.0001 https://doi.org/10.1016/j.intell.2010.11.002 https://doi.org/10.1371/journal.pone.0118143 https://doi.org/10.1093/cercor/bhv130 https://doi.org/10.1093/cercor/bhv130 https://doi.org/10.1017/cbo9781107252950 https://doi.org/10.1080/00335558008248231 https://doi.org/10.1080/00335558008248231 https://doi.org/10.1073/pnas.98.2.676 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dynamic attentional mechanisms of creative cognition 23 ramalho, g., & ganascia, j.-g. (1994). simulating creativity in jazz performance. proceedings of the twelfth aaai national conference on artificial intelligence, 108–113. https://doi.org/10.5555/2891730.2891747 reverberi, c., toraldo, a., d’agostini, s., & skrap, m. (2005). better without (lateral) frontal cortex? insight problems solved by frontal patients. brain, 128(12), 2882–2890. https://doi.org/10.1093/brain/awh577 rosen, d. s., oh, y., erickson, b., zhang, f., kim, y. e., & kounios, j. (2020). dual-process contributions to creativity in jazz improvisations: an spm-eeg study. neuroimage, 213, 116632. https://doi.org/10.1016/j.neuroimage.2020.116632 rothenberg, a. (1995). creative cognitive processes in kekulé’s discovery of the structure of the benzene molecule. the american journal of psychology, 108(3), 419–438. https://doi.org/10.2307/1422898 runco, m. a., & basadur, m. (1993). assessing ideational and evaluative skills and creative styles and attitudes. creativity and innovation management, 2(3), 166–173. https://doi.org/10.1111/j.1467-8691.1993.tb00088.x runco, m. a., & jaeger, g. j. (2012). the standard definition of creativity. creativity research journal, 24(1), 92–96. https://doi.org/10.1080/10400419.2012.650092 schooler, j. w., smallwood, j., christoff, k., handy, t. c., reichle, e. d., & sayette, m. a. (2011). meta-awareness, perceptual decoupling and the wandering mind. trends in cognitive sciences, 15(7), 319–326. https://doi.org/10.1016/j.tics.2011. 05.006 shamay-tsoory, s. g., adler, n., aharon-peretz, j., perry, d., & mayseless, n. (2011). the origins of originality: the neural bases of creative thinking and originality. neuropsychologia, 49(2), 178–185. https://doi.org/10.1016/j.neuropsycholog ia.2010.11.020 shi, l., sun, j., xia, y., ren, z., chen, q., wei, d., yang, w., & qiu, j. (2018). large-scale brain network connectivity underlying creativity in resting-state and task fmri: cooperation between default network and frontal-parietal network. biological psychology, 135, 102–111. https://doi.org/10.1016/j.biopsycho.2018.03.005 silvia, p. j. (2015). intelligence and creativity are pretty similar after all. educational psychology review, 27(4), 599–606. https://doi.org/10.1007/s10648-015-9299-1 simonton, d. k. (1999). creativity as blind variation and selective retention: is the creative process darwinian? psychological inquiry, 10(4), 309–328. https://doi.org/10.1207/s15327965pli1004_4 smallwood, j., mcspadden, m., & schooler, j. w. (2007). the lights are on but no one’s home: meta-awareness and the decoupling of attention when the mind wanders. psychonomic bulletin & review, 14(3), 527–533. https://doi.org/10.3 758/bf03194102 torrance, e. p. (1966). torrance tests of creative thinking. norms-technical manual. research edition. verbal tests, forms a and b. figural tests, forms a and b. personnel press. https://worldcat.org/en/title/714040431 treisman, a. m., & gelade, g. (1980). a feature-integration theory of attention. cognitive psychology, 12(1), 97–136. https://doi.org/10.1016/0010-0285(80)90005-5 uddin, l. q. (2017). salience network of the human brain. academic press. https://doi.org/10.1016/c2015-0-01862-7 vartanian, o. (2009). variable attention facilitates creative problem solving. psychology of aesthetics, creativity, and the arts, 3(1), 57–59. https://doi.org/10.1037/a0014781 verschooren, s., schindler, s., raedt, r., & pourtois, g. (2019). switching attention from internal to external information processing: a review of the literature and empirical support of the resource sharing account. psychonomic bulletin & review, 26(2), 468–490. https://doi.org/10.3758/s13423-019-01568-y vincent, j. l., kahn, i., snyder, a. z., raichle, m. e., & buckner, r. l. (2008). evidence for a frontoparietal control system revealed by intrinsic functional connectivity. journal of neurophysiology, 100(6), 3328–3342. https://doi.org/10.1152/ jn.90355.2008 wallas, g. (1926). the art of thought. harcourt, brace; company. https://psycnet.apa.org/record/1926-10372-000 zabelina, d. l. (2018). attention and creativity. in o. vartanian & r. e. jung (eds.), the cambridge handbook of the neuroscience of creativity (pp. 161–179). cambridge university press. https://doi.org/10.1017/9781316556238.010 zabelina, d. l., & andrews-hanna, j. r. (2016). dynamic network interactions supporting internally-oriented cognition. current opinion in neurobiology, 40, 86–93. https://doi.org/10.1016/j.conb.2016.06.014 zabelina, d. l., friedman, n. p., & andrews-hanna, j. (2019). unity and diversity of executive functions in creativity. consciousness and cognition, 68, 47–56. https://doi.org/10.1016/j.concog.2018.12.005 zabelina, d. l., & ganis, g. (2018). creativity and cognitive control: behavioral and erp evidence that divergent thinking, but not real-life creative achievement, relates to better cognitive control. neuropsychologia, 118, 20–28. https://doi.or g/10.1016/j.neuropsychologia.2018.02.014 zabelina, d. l., & robinson, m. d. (2010). creativity as flexible cognitive control. psychology of aesthetics, creativity, and the arts, 4(3), 136–143. https://doi.org/10.1037/a0017379 zabelina, d. l., saporta, a., & beeman, m. (2016). flexible or leaky attention in creative people? distinct patterns of attention for different types of creative thinking. memory & cognition, 44(3), 488–498. https://doi.org/10.3758/s13421015-0569-4 tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.5555/2891730.2891747 https://doi.org/10.1093/brain/awh577 https://doi.org/10.1016/j.neuroimage.2020.116632 https://doi.org/10.2307/1422898 https://doi.org/10.1111/j.1467-8691.1993.tb00088.x https://doi.org/10.1080/10400419.2012.650092 https://doi.org/10.1016/j.tics.2011.05.006 https://doi.org/10.1016/j.tics.2011.05.006 https://doi.org/10.1016/j.neuropsychologia.2010.11.020 https://doi.org/10.1016/j.neuropsychologia.2010.11.020 https://doi.org/10.1016/j.biopsycho.2018.03.005 https://doi.org/10.1007/s10648-015-9299-1 https://doi.org/10.1207/s15327965pli1004_4 https://doi.org/10.3758/bf03194102 https://doi.org/10.3758/bf03194102 https://worldcat.org/en/title/714040431 https://doi.org/10.1016/0010-0285(80)90005-5 https://doi.org/10.1016/c2015-0-01862-7 https://doi.org/10.1037/a0014781 https://doi.org/10.3758/s13423-019-01568-y https://doi.org/10.1152/jn.90355.2008 https://doi.org/10.1152/jn.90355.2008 https://psycnet.apa.org/record/1926-10372-000 https://doi.org/10.1017/9781316556238.010 https://doi.org/10.1016/j.conb.2016.06.014 https://doi.org/10.1016/j.concog.2018.12.005 https://doi.org/10.1016/j.neuropsychologia.2018.02.014 https://doi.org/10.1016/j.neuropsychologia.2018.02.014 https://doi.org/10.1037/a0017379 https://doi.org/10.3758/s13421-015-0569-4 https://doi.org/10.3758/s13421-015-0569-4 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org shadab tabatabaeian and carolyn jennings 24 zedelius, c. m., & schooler, j. w. (2020). capturing the dynamics of creative daydreaming. in d. d. preiss, d. cosmelli, & j. c. kaufman (eds.), creativity and the wandering mind (pp. 55–72). academic press. https://doi.org/10.1016/b978-012-816400-6.00003-1 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. tabatabaeian, s., & jennings, c. (2023). dynamic attentional mechanisms of creative cognition. philosophy and the mind sciences, 4, 20. https://doi.org/10.33735/phimisci.2023.10020 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/b978-0-12-816400-6.00003-1 https://doi.org/10.1016/b978-0-12-816400-6.00003-1 https://doi.org/10.33735/phimisci.2023.10020 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction background and motivation the role of attention in creative cognition top-down versus bottom-up attention reduced top-down attention as key to creative cognition increased top-down attention as key to creative cognition flexible interactions between top-down and bottom-up attention as key to creative cognition internal versus external attention the role of internal attention in creative cognition the role of external attention in creative cognition combining top-down/bottom-up and internal/external attention for creative cognition dynamic attentional mechanisms of creative cognition as evidenced by neuroimaging studies discussion summary and conclusion the replication crisis and philosophy. the replication crisis and philosophy wesley buckwaltera (wesleybuckwalter@gmail.com) abstract the replication crisis is perceived by many as one of the most significant threats to the reliability of research. though reporting of the crisis has emphasized social science, all signs indicate that it extends to many other fields. this paper investigates the possibility that the crisis and related challenges to conducting research also extend to philosophy. according to one possibility, philosophy inherits a crisis similar to the one in science because philosophers rely on unreplicated or unreplicable findings from science when conducting philosophical research. according to another possibility, the crisis likely extends to philosophy because philosophers engage in similar research practices and face similar structural issues when conducting research that have been implicated by the crisis in science. proposals for improving philosophical research are offered in light of these possibilities. keywords cognitive science ∙ replication crisis ∙ methods ∙ thought experiments ∙ incentives ∙ collaboration 1 introduction a substantial proportion of published scientific research fails to replicate and is likely unreliable (ebersole et al., 2016; klein et al., 2018; nosek & lakens, 2014; osf, 2015; stroebe, 2019). several high profile replication attempts have shown that, of the results selected for replication attempts by researchers on various occasions, anywhere from 64% of scientific findings in psychology (osf, 2015), 39% in experimental economics (camerer et al., 2016), 89% in preclinical cancer research (begley & ellis, 2012; nosek & errington, 2017) and 100% of studies involving structural brain-behavior correlations in neuroscience (boekel et al., 2015) could not be confirmed. though it is currently unclear to what degree these percentages actually represent replication rates in these fields, concerns over the reliability of and confidence in scientific results are widely shared among scientists, the media, and the general public (aschwanden, 2019; baker, 2016). this has led researchers to conclude that “whether ‘crisis’ is the appropriate term to describe the current adepartment of philosophy, institute for philosophy and public policy, george mason university buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-6222-5083 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 2 state or trajectory of science,” there is nonetheless, “substantial room for improvement with regard to research practices to maximize the efficiency of the research community’s use of the public’s financial investment in research” (munafò et al., 2017, p. 1). with every crisis, as they say, also comes opportunity. one positive aspect of the replication crisis is the opportunity to study, better understand, and improve upon research practices moving forward. these efforts have largely come from within the fields of psychology and cognitive science. several psychologists have seized this opportunity by attempting to isolate the causes of the replication crisis and advocate changes to increase reliability and efficiency of research in their field (ioannidis et al., 2014; munafò et al., 2017; romero, 2018; simmons et al., 2011). these efforts have inspired both methodological and social reforms in psychology, and according to one estimate of average replicability, may have begun to show measurable improvements to published research in social psychological science (schimmack, 2017). the opportunity to improve replicable science also extends to researchers working in areas typically thought to be outside of science. many of the questions raised by the replication crisis are conceptual or philosophical in nature (fidler & wilcox, 2018; romero, 2019). for example, what does it mean to “replicate” a finding, successfully, convincingly, conceptually, or otherwise (brandt et al., 2014; hüffmeier et al., 2016; machery, 2020)? what role do values play in promoting reproducible research, and particularly, values such as openness, trust, civility, and shame in science (fiske, 2016; levin & leonelli, 2017; wilholt, 2012)? and given all of this, what does it mean to make “progress” in science (vazire, 2018)? such questions overlap significantly with foundational research topics in philosophy and philosophers of science may be well positioned to contribute to them. the overlap between philosophy and science also suggests that philosophers can learn from what is happening in science. though news of the replication crisis has been dominated by social psychology, all signs indicate that it likely extends to several other fields. according to one hypothesis, while the same problems face many other disciplines, various aspects of the research culture in psychology made the problem easier to detect in that field (gelman, 2016). for example, it could be that statistical sophistication and transparency made it more likely that the problem would be recognized in psychology. this raises a troubling question: just how far does the replication crisis extend, and could it extend to very different fields across or even outside the sciences? the aim of this paper is to investigate the possibility that the replication crisis and related challenges to science extend to the field of philosophy. this investigation is motivated by the assumption that just as philosophers can contribute to improving replicable science from a discipline traditionally thought outside of science, so too can scientists contribute to our understanding of philosophical research methods from outside of philosophy. given the typical objects of philosophical inquiry, this opportunity should be especially welcome. philosophers buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 3 often study objects that are intangible, unobservable, or challenging to measure in various respects. this makes it more difficult to directly assess the reliability of research findings in philosophy than it can sometimes be in science. but the crisis in science may serve as a useful model to assess these matters indirectly. if there is significant overlap between fields in some important ways, then it may indicate that the crisis extends to philosophical research. if the crisis does extend to philosophical research, then this might also suggest that similar reforms advocated in science could improve the reliability of research in philosophy, too. for this reason, philosophers might look to the replication crisis not only as an object of research, but also for insights to better understand and improve upon their own activity in light of similarities between what they are doing and what is happening in science. here is how the paper will proceed. section 2 presents several potential indicators or warning signs of an impending crisis in philosophy that motivate improvements to philosophical methods. the next two sections investigate concrete ways that the replication crisis might extend to philosophical research. section 3 assesses the possibility that a crisis extends to philosophy because philosophers rely on unreplicated or unreplicable evidence when conducting philosophical research. section 4 assesses the possibility that the crisis extends to philosophy because philosophers engage in similar practices and face similar structural issues as those implicated by the crisis in science. section 5 proposes several changes to philosophical research practices modeled after some reforms that have improved the reliability of scientific research. 2 indications of crisis before hypothesizing about ways that the replication crisis might extend to philosophy, a more fundamental question arises: are there antecedent reasons to suspect that philosophy is in crisis? after all, it doesn’t make much sense to speak of a crisis when there is no cause for alarm and there do not seem to be many philosophical emergencies. then again, the current problems facing science make a lot more sense in hindsight. several contributing factors to the replication crisis in science that were widely accepted by researchers only a few short years ago are regarded as completely unacceptable today. but at the time, many scientists openly engaged in and encouraged them. in some literatures, such as ego depletion, for example, foundational theories are now nearly debunked despite widespread support and hundreds of positive results reported in their favor (hagger et al., 2016). this has led researchers to wonder, “if a large sample pre-registered study found absolutely nothing, how has the ego depletion effect been replicated and extended hundreds and hundreds of times?” and “[…] more sobering still: what other phenomena, which we now consider obviously real and true, will be revealed to be just as fragile?” (inzlicht, 2016). it is possible that we could also be asking the same questions of philosophy in the not so distant future. buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 4 several potential warning signs support this possibility. the following discussion raises several warning signs that collectively suggest that there is room for improvement in philosophical methods. the first warning sign stems from the fact that philosophical research is mired in persistent disagreement. though there is no established measure of disagreement across the field, it is difficult to deny that disagreement is an inescapable aspect of philosophical activity. if there is research that is advanced by a philosopher, then the chances are good that there is a philosopher who is equally well trained and informed that advances the opposite. quite often, many of these disagreements are intractable. they are rarely resolved through reflection, discussion, or additional information. there have been many explanations offered to explain persistent disagreement in philosophy (beebee, 2018; daly, 2017; macbride, 2014). however, one reason disagreement continues unresolved, it is sometimes suggested, is because philosophical researchers rely on different evidence (beebee, 2018). to a large extent, this evidence consists of intuitions from thought experiments or the judgments that individual philosophers make about real or imagined cases that they construct to support their theories (more below). in metaphysics, for example, researchers have observed several clashes among case judgements in foundational thought experiments: you might think that designed ernie in alfred mele’s ‘zygote argument’ is not morally responsible, and conclude that nonhistoricist compatibilism is false (mele, 2006, p. 189). or you might not (fischer, 2011, p. 271). you might come up with what you take to be a castiron case of the failure of the transitivity of causation (mcdermott, 1995, p. 524). someone else will inevitably disagree (lewis, 2000, p. 194). you might think you have described two different possible worlds that agree with respect to the distribution of matters of particular fact but disagree with respect to the laws, thus refuting the claim that the latter supervene on the former (carroll, 1884, pp. 57–68). or you might not (beebee, 2000, pp. 586–592). and so on. (beebee, 2018, p. 5) when this happens, as in metaphysics above, for example, researchers have further observed that philosophers are left with very little recourse over how to proceed: we have no such method for resolving metaphysicians’ clashes of intuitions. when i say that designed ernie is morally responsible for his crime or that in our transitivity case x really was a cause of z, and you disagree, we have no agreed way of reconciling our differences. there are no empirical facts about which you might correct me, in the light of which i would recognise my mistake and change my mind. i have not illegitimately held fixed some facts that i should not have held fixed, or vice versa – not by my lights, anyway. we might try to convince each other otherwise, of course, and sometimes one of us succeeds; but buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 5 often we both fail. often we both fail simply because each of us is holding fixed some element of our own background philosophical theory that the other rejects. but then our thought experiment serves only to provide an example of the different consequences of our respective theories; it cannot adjudicate between them. (beebee, 2018, pp. 5–6) one hypothesis for why intractable disagreement persists is because one philosopher is claiming to offer basic evidence that another philosopher cannot replicate, reproduce, or confirm. in this case, the relevant evidence is a judgment or intuitive reaction from a thought experiment when using the method of cases. if, try as they might, a philosopher cannot reproduce this evidence, while another insists that the evidence exists and supports their theory, then the pair are at an impasse. few would claim that researchers must always agree with one another in order for progress to occur. but failure to converge on even basic evidence in core thought experiments when practicing the case method is a sign that philosophical methods are in crisis, especially when there is no way to adjudicate judgments between researchers. such an impasse leads to several questions reminiscent of those raised in the early days of the scientific crisis. if your peers conduct research that generates results diametrically opposed to your own, and you have every reason to believe that these peers are just as good at conducting research as you are, then it is reasonable to begin questioning the research process. if disagreement persists because two researchers are drawing on different evidence when evaluating thought experiments, then this reasonably calls for a careful investigation into the nature of that evidence. if upon investigation, there is no way to adjudicate the conflicting evidence generated between researchers, then this reasonably calls into question the source of that evidence and points to weaknesses in the research method used to generate it. a second warning sign stems from the questions of whether philosophy produces a distinctive body of knowledge or makes appreciable progress. many philosophers are skeptical that it does. according to thomas reid, philosophy only manages to “cast a ‘darkness visible’ on the human faculties, and to disturb the peace and security enjoyed by happier people” (reid, 1764/1997). william lycan writes that “philosophical consensus is far more the result of zeitgeist, fad, fashion, and careerism than of accumulation of probative argument” (lycan, 2013, pp. 116–117). still other researchers have claimed either that it is “not clear that the philosophical enterprise has served as a source of knowledge” (kornblith, 2013, p. 260), or worse, that “there is no information and there are no facts to be learned besides information and facts about what certain people think” in philosophy (van inwagen, 2015, p. 11). when evaluating progress in the field, philosophers often compare the state of philosophical research to that of the natural sciences. for example, it has been suggested that “since science took its modern form in the seventeenth century, it has been one long success story” and that “philosophy compares badly with science on this score” (papineau, 2017). it is also often suggested that science makes buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 6 progress because science is self-correcting. the basic idea is that while scientists might make just as many mistakes as other researchers do in the short term, science somehow corrects for this in such a way that promotes truth or instrumental advances in the long term. however, the replication crisis has recently challenged whether science makes significantly more progress than other fields. for example, it is reasonable to question the degree to which science self-corrects for certain types of errors (romero, 2019). and without successful replication in science, it has even been suggested that, “perpetuated and unchallenged fallacies may comprise the majority of the circulating evidence” (ioannidis, 2012, p. 645). this sounds a lot like what some skeptics have said about philosophy. these considerations offer a tentative though suggestive warning that philosophy faces a crisis similar in kind to that of the natural sciences and calls for the improvement of philosophical methods. it could be not that it was wrong to question progress in philosophy, but that it was wrong not to be equally suspicious of scientific progress. the lack of error correction mechanisms needed for the satisfactory promotion of true theories or accumulation of knowledge might well be shared by both fields. in science, the lack of error correction has been tied to concerns over replication and reliability of evidence. thus, it is reasonable to investigate whether the same features might account for errors that also limit progress or knowledge creation in philosophy. a third warning sign stems from the growing body of research in cognitive science suggesting that many popular research programs in theoretical philosophy have turned out to be false starts, in some cases because case judgments are not shared or are misunderstood (buckwalter, 2014; buckwalter & turri, 2019; colaco et al., 2014; kneer & machery, 2019; machery, 2017; rose et al., 2014, 2017, 2019; turri et al., 2015; turri, 2017). this research demonstrates how theorists can be led astray by confounded thought experiments or idiosyncratic understandings of concepts and terminology. in many instances, judgments about cases are either not shared by other researchers or the general public, or have replicated in these ways but are overstated, misdiagnosed, or explained by extraneous factors that were not of philosophical interest. two recent developments exemplify these concerns. in epistemology, for example, philosophers have constructed famous pairs of cases manipulating certain variables such as stakes and have claimed that different judgments about them motivate the theory that the word “knows” is a contextually sensitive expression (derose, 1992, 2009). in fact, it is often claimed that the effect is so strong that it can reverse the truth of knowledge sentences. when researchers subjected these thought experiments to controlled testing, however, they found little evidence for the predicted effect anywhere in the world (rose et al., 2019). in a large-scale cross-cultural replication attempt involving forty-five hundred participants in over a dozen countries, researchers failed to replicate the effect at sixteen out of nineteen international research sites. what effects researchers were able to detect were small and amounted to a three percent difference in judgments across conditions. buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 7 regardless of whether the effect in question is ultimately detectable in some form or other (turri et al., 2016; turri, 2017), this state of affairs suggests that there is room for improvement in investigating it. in other subfields, such as normative ethics, researchers have partially replicated foundational intuitions and judgments about cases, but in ways that question the development of the resulting philosophical research programs. in one series of experiments, for example, researchers could not detect many of the everyday judgments said to underlie the philosophical puzzle of resultant moral luck (kneer & machery, 2019). according to philosophers, this puzzle arises because we are sometimes inclined to judge others for the unlucky consequences of actions even though the results are beyond their ability to fully control (williams, 1981). when subjecting multiple moral luck cases to controlled experimental testing, however, researchers discovered little evidence for the claim that wrongness, blame, or permissibility judgments differed considerably as a result of lucky or unlucky outcomes. what evidence researchers did find for this effect appeared to be artifacts of the testing situation or perhaps attributable to cognitive biases, such as hindsight bias. at the same time, researchers did find evidence for a substantial effect on other judgments in moral luck cases, such as punishment judgments. if these findings are correct, then they suggest that either moral luck is not a genuine ethical puzzle that arises for many people upon reflection or that while there may be a glimmer of a puzzle here, it has been misunderstood or mischaracterized for several decades. whether or not these observations are indicative of a crisis, they indicate room for improvement and warrant further exploration of the possibility. there is persistent philosophical disagreement between researchers that may be perpetuated by differences in case judgements. longstanding debates about the nature of philosophical progress may point to the need for new methods or procedures to limit the accumulation of errors, as the replication crisis suggests is needed in science. lastly, replication failure in philosophy may be quite literal. to the extent that case judgments motivating leading theories are not widely shared or unreliably generated, philosophy may share key features of unreplicable science. 3 philosophy in the age of replication crisis science given that there is initial theoretical and practical motivation to investigate the replication crisis in philosophy, what forms could it take? one of the most straightforward ways that the replication crisis extends to philosophy lies in the fact that much research in philosophy relies on unreplicated or unreplicable science. contrary to how it may sometimes appear, a non-trivial proportion of philosophical research draws on scientific evidence. philosophers regularly rely on new, interesting, or surprising findings across the sciences to motivate, challenge, or support buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 8 philosophical theorizing. they also often rely on classic or foundational research in science to do these things, which until recently at least, was considered settled science. however, both kinds of findings are frequently subject to failed replication. thus, to the extent that philosophers draw on scientific findings in their theorizing, and those findings are implicated by the replication crisis, the replication crisis straightforwardly extends to philosophy. to what extent does philosophical research rely on evidence from science? because philosophy is a broad discipline, the answer to this question is bound to vary widely based on the area and research question. some areas may rely on scientific evidence more than others do or prioritize analytic aspects of various phenomena. other areas of philosophy, however, rely quite heavily on scientific findings and methods. some of these areas are necessarily interdisciplinary. philosophy of science, philosophy of cognitive science, moral psychology, applied social and political philosophy, feminist philosophy, and applied ethics all heavily appeal to developments in various social and natural sciences. this is to be expected of many areas of philosophy that are more likely to address practical philosophical questions that arise in social or political life. such appeals are not only restricted to areas of philosophy that study applied questions. areas of philosophy traditionally associated with or dominated by armchair methods have also dramatically shifted to embrace empirical methods. one particularly striking example of the shift from a priori to empirical methods in a relatively short amount of time has been documented in philosophy of mind (knobe, 2015). to document this shift, researchers analysed the 397 highest cited articles in the same set of philosophy journals that were published either in the mid-to-late twentieth century (1960–1999), or those published in the early twenty-first century (2009–2013). what researchers found was that there was a change in predominate methods. papers in the twentieth century were dominated by a priori armchair methods (62.4%). conversely however, the majority of twentieth century papers relied on empirical research generated in science (61.8%), or contributed new experimental research to the research record (26.8%). very few papers in the latter period relied on purely a priori methods (11.5%). researchers also documented a surprising shift in topics of philosophical research during these periods. the earlier period heavily focused on research pertaining to the metaphysics of mind, such the mind-body problem or the nature of content. however, the later period focused on interdisciplinary topics in cognitive science, such as perception, theory of mind, or cognition. these findings suggest a transformation has occurred in analytic philosophy of mind, prioritizing interdisciplinary topics and an engagement with evidence from science. appeals to scientific evidence may also extends to several other core areas of philosophy. often, it is not so much debated whether philosophy relies heavily on science, but whether this development is good or bad. in moral psychology, for example, researchers have argued that popular forms of argument in the field such as debunking arguments “tend to rely on a problematic scientism, privilegbuckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 9 ing scientific causal explanation of targeted ethical or meta-ethical beliefs while ignoring or downplaying important philosophical alternatives” (fitzpatrick, 2018, p. 234). in epistemology, researchers have claimed that “science not only helps us to address the philosophical questions that we had before we became acquainted with scientific advances; it helps us to revise the philosophical questions we ask in light of the better understanding of various phenomena that science provides” (kornblith, 2018, p. 146). in other areas, researchers lament that philosophers do not rely on science enough. in metaphysics for instance, researchers have argued that analytic metaphysics should be discontinued because “no alternative kind of metaphysics can be regarded as a legitimate part of our collective attempt to model the structure of objective reality” than one that is “radically naturalistic” (ladyman et al., 2014, p. 1). these observations suggest that scientific evidence is relevant to where philosophy has been, or where it may soon be headed. given that philosophical research relies on scientific evidence generated during a replication crisis, to what extent are the findings that philosophers rely on implicated by that crisis? at present writing we do not know the answer to this question. we do not know this because we do not know the true replication rate in science. successful replication is both bound to fluctuate between scientific fields and vary by individual research questions. relatively few scientific findings have been subject to registered replication attempts. to further complicate matters, even if findings do replicate, we do not know their true effect size, whether they have been correctly interpreted, or whether they extend from the lab to meaningful contexts in everyday life. nonetheless, several high-profile scientific findings of deep philosophical interest have played a significant role in both philosophy and the replication crisis. the stanford prison experiment, a study frequently discussed in ethics and regarded as a powerful source of evidence for situationism has been called “a lie” (blum, 2018) involving a “biased and incomplete collection of data” (le texier, 2019). the effect of disgust induction on moral judgment, a finding heavily invoked in the metaethics and philosophy of emotion, has been overstated, has failed to replicate, and may be accounted for by publication bias (ghelfi, 2020; johnson et al., 2016; landy & goodwin, 2015). relevant to the egoism and altruism debate, the claim that deliberate perspective taking of needy others increases empathetic concern has been heavily challenged (mcauliffe et al., 2020), as has bystander apathy, relevant to the bystander effect (philpot et al., 2020). it is unlikely that stereotype threat and intelligence mindsets, frequent topics in feminist philosophy and social political philosophy, can explain performance outcomes outside of lab contexts after controlling for publication bias (bahník & vranka, 2017; finnigan & corker, 2016; flore & wicherts, 2015; shewach et al., 2019). frequent objects of theorizing in philosophy of mind and philosophy of cognitive science, such as ego depletion (hagger et al., 2016) or backfire effect (wood & porter, 2019) fail to replicate. an extensive meta-analysis appears to undermine previous claims that manipulating free will beliefs is associated with anti-social behavior, such as cheating (genschow buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 wesley buckwalter 10 et al., 2022). the implicit association test, a keystone of philosophical theorizing across ethics, metaphysics, and philosophy of mind faces a series of challenges regarding construct validity (schimmack, 2019) or usefulness (buckwalter, 2019; forscher et al., 2019; machery, 2022). and the finding that intuitions about intentional action are influenced by implicit bias fails to replicate (klein et al., 2018). these challenges deeply question the starting assumptions of large bodies of literature in philosophy and require significant attention to address. a final question that remains is to consider how the effects of the scientific crisis might translate to fields typically thought to be outside of science, such as philosophy, when drawing on scientific evidence. one reasonable hypothesis is that the effects will largely be the same between fields. researchers across these fields ultimately rely on the same body of unreliable evidence to support their theories and will suffer equally when they turn out to be based on unreliable evidence. another hypothesis, though, is that the effects in philosophy will be worse. there are some important field-specific differences that might magnify the impact of the replication crisis in philosophy. for instance, philosophical research often tends to prioritize extended theorizing. this extended theorizing abstracts away from the specific findings reported to explore the possible philosophical implications or applications of findings. if this is the case, then unreplicable science might license more discussion and speculation in philosophy than it does in science. extended speculation could magnify the problem by increasing the reach of a single unreplicable result considerably further than data driven science typically allows and by giving a veneer of scientific authority to claims that have not been tested or shown. another difference is that many scientific results are not settled and that philosophers have less expertise in evaluating this than professional scientists do. it can be incredibly difficult to accurately assess what a paper has shown, even for those with extensive hands on experience conducting experiments in that research area. but scientists often have better access to experiences, types of background information, and networks between scientists than others do. these things all provide information that often doesn’t make it into published papers and that can help contextualize results. without this access and experience, even the most conscientious researchers are more likely to misinterpret contributions or overlook important red flags. for these reasons, a little bit of unreliable scientific evidence might translate into large effects in philosophy. large bodies of literature meticulously investigating the possible philosophical implications of unreplicable scientific findings could well be without foundation. 4 replication and the method of cases is the replication crisis only a concern for empirical approaches to conducting philosophical research? it is unsurprising that the replication crisis in science would extend to fields outside of science that sometimes rely on the same evidence from science. thus, it might be thought that limiting the role or influence buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 11 of evidence from unreplicable science would prevent the challenge from spreading beyond science. but what of conceptual, analytic, or broadly speaking nonscientific methods in philosophy, could the replication crisis extend to philosophical research produced in these ways regardless of formal contact with scientific evidence? given the natures of the typical objects of philosophical inquiry, it is difficult to assess this possibility directly. however, it is possible to assess it indirectly, by examining whether the research practices implicated in the scientific crisis are also present in more analytic philosophical methods. this approach is motivated in part by bertrand russell, who argued that philosophy should draw its inspiration from science. by this, however, russell does not necessarily mean that philosophy should simply inherit the results of science as discussed above, but rather, that it should emulate certain features about the way he perceived science to be optimally conducted: much philosophy inspired by science has gone astray through preoccupation with the results momentarily supposed to have been achieved. it is not results, but methods that can be transferred with profit from the sphere of the special sciences to the sphere of philosophy. what i wish to bring to your notice is the possibility and importance of applying to philosophical problems certain broad principles of method which have been found successful in the study of scientific questions. (russell, 2008 [1917], 98-99) the sort of methods that russell has in mind all generally pertain to a kind of scientific spirit or temperament that prioritizes the use of logic, a dispassionate search for truth, patience, and modesty in the research process. applying these principles to philosophical research, russell writes, “is to ensure a progress in method whose importance it would be almost impossible to exaggerate” (ibid., 113-114). applying russell’s insight in the present case, the thought is that we might be able to continue to improve philosophical research by emulating recent improvements in scientific method. or put another way, we might be able to isolate weaknesses in conducting philosophical research and improve upon them by observing common points of overlap with methods in science associated with the replication crisis. though philosophers might utilize many methods in the course of conducting philosophical research, one natural point of comparison between philosophical and scientific research practices involves the use of the method of cases. the method of cases is a distinctive method in philosophy (machery, 2017; strevens, 2019). when applying this method, philosophers plan, construct, and evaluate cases to assess philosophical claims about philosophical phenomena, concepts, common sense, or natural language. many features of such cases are objects of lengthy philosophical research. in general, however, since the use of cases shares several common features with controlled experiments, it is no surprise that cases are often referred to as “thought experiments”. there are many basic similarities between thought experiments and controlled experiments. when conducting a thought experiment, a philosopher designs a buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 12 vignette involving a real or imaginative situation. the vignette is designed to test a specific hypothesis, for example, that knowledge is compatible with luck, that justice is repaying debts, that scientific progress can occur without justification, and so on. to test hypotheses about these research questions, philosophers often manipulate factors, for example, by varying whether a protagonist in a vignette acquires information by luck. philosophers then collect evidence in the form of judgments about the case. if the philosopher is inclined to judge of this situation that, for example, the protagonist has knowledge despite the lucky circumstance, then this is taken as data supporting the original hypothesis that knowledge is compatible with luck. if not, then this is taken as evidence that the hypothesis should be rejected or refined. this is similar in form to what some scientific researchers do when they conduct controlled experiments, particularly in social psychology. often, psychologists develop narrative cover stories to use in experimental materials that closely resemble the thought experiments that philosophers use in their papers. the cover stories typically describe ordinary situations that isolate and manipulate variables of psychological interest. these variables frequently overlap with many of the same areas that philosophers study, such as belief, knowledge, morality, intention, or punishment. sometimes, experimentalists even adopt the exact text of thought experiments originally introduced by philosophers, such as moral dilemmas like trolly cases. of course, psychologists typically recruit more participants to evaluate cover stories than philosophers do. psychologists also typically subject those judgments to some kind of statistical analysis beyond armchair reflection. interestingly, some ways that psychologists have done these things has probably exacerbated the crisis. but in any event, and in this corner of social psychology, at least, the differences to thought experiments in philosophy seems to be a matter of detail rather than kind. and while differences can be substantial when comparing particular research programs in philosophy and psychology, the overlap in the case method motivates the search for additional similarities. the following is an investigation of many more similarities between research in philosophy using the method of cases and research in social psychology utilizing case-based experiments. the investigation reveals that many of the similarities between these things are problematic. in its use and practice in philosophy, the method of cases embodies several key factors implicated in the replication crisis in science. reviewed in what follows are some general similarities between thought experiments in philosophy and controlled experiments in science at the forefront of methodological discussions in science. sample size. one key factor perpetuating the replication crisis is low sample size. it is well known that studies with small sample sizes often have low statistical power, which increases the likelihood that measurable effects are not true effects (button et al., 2013). it is also well known that studies with smaller sample sizes tend to show larger effect sizes and display greater heterogeneity between studies on the same research question (inthout et al., 2015). some of the ways that sample buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 13 size has been shown to relate to low replication rates in science are mathematical, while others are observational and may vary across research fields. in both ways, small studies are more susceptible to replication failure, false positives, and inflated effect sizes than larger studies. they also tend to exacerbate existing biases that are present when researchers conduct their research. philosophers rely on extremely small samples. the method of cases is usually conducted by individual researchers or a small group of collaborators. for these results to be published, the judgments from cases that researchers report must only be shared (or at least not rejected) by a small group, typically consisting of a handful of editors and peer reviewers. with judgments that result from samples this small, however, we simply do not know if they are shared among researchers. if other researchers do share the relevant judgments, we do not know to what strength or degree. to the degree case judgments are reported and validated in this way, the greater the likelihood that the results are false positives, overstated, or do not reflect true discoveries about philosophical phenomena. the clear conclusion to draw from this is that both fields should assess the samples they are drawing from in conjunction with the inferences they are making about the effects they claim to demonstrate. reporting and publication bias. a second factor perpetuating the replication crisis involves biased reporting and the publication of research findings. publication bias, often referred to as the “file drawer problem” occurs when the result of the experiment influences the likelihood that the study will be reported or published (rosenthal, 1979; scargle, 2000). according to this problem, we cannot tell within any given research area how many studies have been conducted but not reported. the reason we cannot tell this is because there is a strong preference to only report and publish positive findings. because we have differential access to positive and negative findings, researchers cannot fully assess the strength of the existing evidence for a research claim. this is troubling and contributes to the replication crisis because a shockingly small number of unpublished results can greatly increase the risk that positive findings are spurious (rosenthal, 1979). all indications suggest that philosophical research is extremely susceptible to publication bias. these biases are likely on both the individual and institutional levels. when philosophers use the method of cases in their research, they almost always only report positive evidence. in philosophy, positive evidence consists of judgments that support a desired theoretical application such as motivating a theory or constituting a counterexample. this means that we do not know how many variations of those cases the philosophers tried before getting the desired judgment. of course, philosophy is different than science in that it sometimes only takes one successful case to prove the point, whereas science is typically interested in establishing central tendencies. even in these circumstances, however, reporting only successes and burying failures obscures the full picture. when this happens, for example, the research community cannot assess what seemingly innocuous changes were made to cases get the desired result and evaluate the philosophical buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 14 significance of this. for instance, does the intuition only arise in cases that involve extreme affect, alternate universes, or outer space? it could be valuable to consider this. neither can we assess how many other philosophers working in the area also tried and failed to construct a case to reach that judgment. if only positive evidence is reported, then we cannot fully evaluate how strong the evidence is for claims made across research areas and potentially valuable information is lost. when it comes to disagreements about famous thought experiments such as twin earth cases, for example, philosophers have observed that publications have become “intramural sports among believers” and that “those who do not share the intuition are simply not invited to the games” (cummins, 1998, p. 116). it is also less common, by comparison, that cases and case judgments not supporting desired theoretical outcomes are published. like science, philosophical publishing also prioritizes positive evidence. when using the method of cases in philosophy, this amounts to reaching a desired judgment that either motivates philosophical theories or serves as a counterexample to them. for this reason, there are strong institutional biases against publishing negative findings from the method of cases that do not support one’s position. to the extent that publishing dissenting case judgments is discouraged, one of two bad things is likely to happen. researchers can self-select out of research areas where they are not able to replicate the same evidence everyone else is drawing on when generating theories, which further contributes to publication bias. or alternatively, researchers can use the method of cases to generate new judgments about unrelated cases that are more conducive to their own preferred theories about the same research question, in which case the cycle is bound to repeat itself. lack of replication. a third factor perpetuating the replication crisis is that relatively few scientific findings are subject to replication labor. replicable science is a cornerstone of reliable science. one way to ensure that science is replicable and reliable is to subject results to systematic replication attempts. such replication efforts help to rule out systematic error in the research record and improve our confidence in scientific publications. the lack of systematic efforts to do this is arguably the largest factor in explaining why the replication crisis continued as long as it did and was one of the first steps in assessing the damage. like the recent picture in science, there have not been enough coordinated replication efforts of case judgments in philosophy. some replication attempts have recently been carried out cross culturally and this number is growing (cova et al., 2019; machery et al., 2017; rose et al., 2019). for many foundational cases in philosophy, however, we know far too little about the reliability of published findings. we do not know to what degree other philosophers share the same judgments about them, the extent to which these judgments are made by researchers in other fields more broadly, or beyond. as in science, understanding these things is key to understanding and improving the evidence used in philosophical research. some aspects of the case method in philosophy may approximate replication but are insufficient. for example, it might be thought that dialectical exchanges buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 15 between philosophers in published journal articles can sometimes play this role. philosophy is often conducted through call and response. while this is true, responses are more likely to involve theoretical implications of case judgments or add new case judgements to the research record rather than dispute judgments in particular thought experiments. the reason for this, as noted above is that clashes of intuitions often result in dialectical stalemates. it might also be thought that replication occurs in the classroom, when exposing students to the case method. while it is true that students are exposed to cases in the course of their philosophical education, we do not fully understand what effect this has and whether case judgements are shared. results are not shared publicly with the research community, which limits the positive effects of this practice with respect to replication. insufficient training. a fourth factor perpetuating the replication crisis is insufficient scientific training. many have suggested that problems in psychology stem from the fact that students are not well trained in basic concepts and procedures inherent to the method. in psychology, the lack of training involves basic training in statistics, and particularly, things like understanding p-values, effect sizes, and statistical power (button et al., 2013; cohen, 1962; greenland et al., 2016; sedlmeier & gigerenzer, 1989). lack of concern for statistical power and sample size has long been noted as a “remarkable phenomenon” that has been neglected in many “statistics textbooks used in the graduate training of the investigators” (cohen, 1962, p. 145). while metaphilosophy is a growing area of philosophical research in which the method of cases is often discussed, it is striking that researchers receive very little practical training in using the method of cases. particularly, philosophers are not directly trained in how to construct thought experiments and there are few norms for how to evaluate them. this is especially apparent given that thought experiments are often published despite basic weaknesses in their construction and interpretation. for example, they are often long, complicated, and feature strange situations that might well be responsible for judgments orthogonal to the philosophical factors of interest. cases are also typically published without adequate controls to help isolate the variables of interest. conclusions are drawn on the basis of a single case, rather than multiple cases that vary in topic or other incidental details. judgments are also often based on a single question, without considering closely related variables or alternative ways to phrase that question. in some cases, researchers even go so far as to name protagonists in thought experiments intended to produce positive or negative verdicts “mr. havit” and “mr. nogot”, respectively (lehrer, 1965). these things increase the risk that cases and judgements made about them are not reliable or more widely generalizable. research would be improved by more formal training in designing and evaluating thought experiments. experimenter effects. a fifth factor perpetuating the replication crisis involves the ability of the experimenter to influence the outcome of an experiment. researchers must make many choices when designing, conducting, and analyzing experiments. sometimes these choices are innocuous and arbitrary. other times, buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 16 freedom or flexibility can be good, insofar as it promotes creativity and exploration. such choices can profoundly affect the outcome of experimental research (landy et al., 2020). when left unchecked or improperly acknowledged, however, researcher degrees of freedom can also introduce errors into the collection or analysis of data that biases the results towards a desired hypothesis (strickland & suben, 2012; wicherts et al., 2016). such degrees of freedom involve aspects of the experimental design, such as wording or measurement choices, data collection and analysis practices, such as insufficient blinding of participants or experimenters, and the manner that results are reported, such as presenting exploratory analyses as confirmatory. some of these choices increase the likelihood of false positives and decrease replicability. philosophical research is unparalleled in the freedom it affords. as in science, freedom in the selection and investigation of research projects can be beneficial. similarly, however, unchecked degrees of freedom can also increase the prevalence of questionable research practices and unreliable evidence. for example, researchers are free to utilize cases of any length, complexity, or topic, as well as employ any words, phrases, or characters they see fit in designing thought experiments in the method of cases. because there are few norms when it comes to conducting the method of cases, researchers are also free to present and discuss judgments about the cases in ways that might influence or alter their assessment. these things increase the likelihood that researchers will construct cases to reach a desired conclusion that may not reflect the way the world is organized and that different researchers might reach different findings when considering the same cases. one representative example involves case judgments in support of leading theories in epistemology. researchers have shown that subtle confounds in wording and probing in foundational cases can explain judgments about the word “knows” (turri 2017) and that understanding the mechanisms that cause them can undermine aspects of their theoretical significance (buckwalter, 2021). in these studies, researchers examined classic pairs of cases used to motivate epistemic contextualism. for example, consider the following low stakes case: keith and his wife jane are driving home from work on friday afternoon. they just received a large check from a client, which keith plans to deposit in their bank account. it is not important for him to deposit the check before monday: they definitely do have enough money in their account for all their checks to clear. as they drive past the bank, they see that the lines inside are very long. keith says, ‘i hate waiting in line. i’ll just come back tomorrow morning instead.’ jane responds calmly, ‘this is really not important, but lots of banks are closed on saturdays. do you know that our bank is open tomorrow?’ keith answers, ‘it was two saturdays ago that i went to our bank, and it was open. so, yes, i do know that our bank is open tomorrow.’ (turri, 2017, p. 144) buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 17 and the following high stakes case: keith and his wife jane are driving home from work on friday afternoon. they just received a large check from a client, which keith plans to deposit in their bank account. it is very important for him to deposit the check before monday: otherwise they won’t have enough money in their account for all their checks to clear. as they drive past the bank, they see that the lines inside are very long. keith says, ‘i hate waiting in line. i’ll just come back tomorrow morning instead.’ jane responds anxiously, ‘this is really very important, and lots of banks are closed on saturdays. do you know that our bank is open tomorrow?’ keith answers, ‘it was two saturdays ago that i went to our bank, and it was open. so, no, i don’t know that our bank is open tomorrow.’ (ibid.) it is often claimed that cases with this basic structure motivate epistemic contextualism because they demonstrate the effect of contextual standards on judgments about the truth of knowledge sentences (derose, 2011). more specifically, they are often taken to show that when a protagonist says “i do know” in low stakes cases and “i don’t know” in high stakes case both statements seem true. if the cases are otherwise identical, then it must be shifting epistemic standards that explain these judgments. what researchers noticed about these thought experiments, however, was that the cases were not otherwise identical. the cases manipulated more than just contextual standards. they also manipulated whether self-attribution or self-denial of knowledge occurred, i.e., whether or not a protagonist says, “i know” or “i don’t know”. the presence of this additional variable creates a confound that threatens to undermine the explanation of the case judgment. if it is the way the knowledge statement is phrased, rather than the contextual standard being manipulated between cases, then this challenges the evidence for the effect of contextual standards. when subjected to controlled testing, this is exactly what researchers found (turri, 2017, experiment 2). the difference in the way knowledge statements were phrased produced the agreement patterns predicted by contextualism, even when stakes do not vary. in short, judgments were being caused by seemingly innocuous or incidental features introduced into thought experiments by researchers rather than factors of theoretical significance. hidden moderators. a fifth factor perpetuating the replication crisis involves unaccounted for features of an experiment that contribute to its result. more specifically, the term “hidden moderators” is often used to describe unobserved or unmeasured factors that influence effects. out there in the world, of course, there are a lot of potential moderators that could impact experimental findings as a result of the context in which any one experiment was conducted. this kind of contextual sensitivity might involve, for example, features of the participants, such as culture or native language, the times or settings in which the experiment took place, or design features of the experiment itself, such as details of stimuli or probing method. buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 18 insufficient sensitivity to these potential moderators increases the likelihood that a given finding is overstated or is not generalizable to new experimental contexts. it is also possible, in some cases, that these effects explain why findings are either not replicable or might not replicate on some experimental occasions (van bavel et al., 2016). the possibility of hidden moderators may constitute one of the most significant challenges to the method of cases. the reason for this is simple. we do not know nearly enough about the individual features of a thought experiment that are responsible for the judgments that we make about it. quite often, it is simply taken for granted that we do. almost always, the research assumption is that whatever judgments thought experiments elicit are caused by what the researcher intended the thought experiment to assess. if a thought experiment is constructed to assess the relationship between knowledge and luck, for example, it is assumed that whatever judgment it elicits is about luck and the direct effect that luck has on knowledge judgments. but this is little more than an assumption. it is only an assumption because the pathways or mechanisms that generate philosophical judgments are often complex, interact with several variables or combinations of variables, and are not fully accessible by introspection. as in science, this raises the possibility that judgments about cases are partially explained by previously unaccounted for factors that alter or constrain their theoretical significance. some research in experimental cognitive science and experimental philosophy suggests that this may be common (buckwalter, 2021; buckwalter & turri, 2015; chituc et al., 2016; kneer & machery, 2019; rose et al., 2017; turri, 2017). we often misdiagnose the reasons for judgments or misrepresent their causal structure when utilizing the method of cases. one representative example comes from research in metaphysics and action theory concerning intuitive free will judgments. case judgments have played a significant role in the free will literature, with some philosophers claiming that compatibilism is natural (fischer & ravizza, 1998; nahmias, 2014), while others claiming incompatibilism is natural (pereboom, 2014; strawson, 1986). most of the evidence for natural compatibilism or incompatiblism comes from claims about free will and responsibility judgments made in deterministic thought experiments. consider, for example, the following case: imagine that in the next century we discover all the laws of nature, and we build a supercomputer which can deduce from these laws of nature and from the current state of everything in the world exactly what will be happening in the world at any future time. it can look at everything about the way the world is and predict everything about how it will be with 100% accuracy. suppose that such a supercomputer existed, and it looks at the state of the universe at a certain time on march 25th, 2150 a.d., twenty years before jeremy hall is born. the computer then deduces from this information and the laws of nature that jeremy will definitely rob fidelity bank at 6:00 pm on january 26th, 2195. as buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 19 always, the supercomputer’s prediction is correct; jeremy robs fidelity bank at 6:00 pm on january 26th, 2195. (nahmias et al., 2005, p. 559) given that the thought experiment stipulates that determinism is true, if we intuitively judge that jeremy acted of his own free will or that jeremy was morally responsible after considering it, then this demonstrates that determinism is naturally compatible with free will or moral responsibility. however it is possible to question whether judgments about free will or responsibility support natural compatibilism simply because they follow from reading a case involving determinism. for case judgments to provide evidence that either compatibilism or incompatiblism is natural, it is not enough, it might be thought, to simply make judgments about freedom or responsibility after reading deterministic cases. it is essential that such judgments are also properly responsive to the deterministic nature of the scenarios. if judgments are made without such responsiveness, then it is unclear what evidence they contribute to the debate about whether compatibilism or incompatiblism is natural. this is precisely what researchers have shown. more specifically, researchers have shown that some judgments in classic free will cases used to motivate compatibilism between determinism, freedom, and moral responsibility are made without tracking the deterministic features of these scenarios (nadelhoffer et al., 2020; rose et al., 2017). instead, researchers have shown that many of the free will and responsibility judgments taken as evidence for natural compatibilism can be explained by commitments to indeterminism about human actions and decision making. despite what the scenario might say above for instance, people judge that there is a slight chance that jeremy would not rob the bank as the computer predicted he would. even though jeremy actually did what the computer predicted he would do, the thinking seems to go, it was possible for jeremy to do something else instead. in short, indeterministic commitments were hidden or unmeasured when processing deterministic features of the cases, and the unmeasured effect of these things may explain case judgments. this threatens to undermine their evidential role in theorising about natural compatibilism and raises larger questions about the use of the case method without controlling for such effects. a related but distinct body of examples also suggests that judgments from the method of case may be moderated by demographics, order effects, framing effects, and situational cues. for example, researchers have shown well-replicated effects of culture or heritable personality traits on philosophical case judgments in philosophy of language and action theory (beebe & undercoffer, 2016; feltz & cokely, 2012; machery et al., 2004). researchers have shown that presentation, framing, and order effects persist in foundational case judgments in ethics, such as trolley problem and asian disease scenarios, with or without professional philosophical training (liao et al., 2012; schwitzgebel & cushman, 2015; wiegmann et al., 2012). these things are important to consider and can be difficult to detect when conducting thought experiments. buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 20 one of the most comprehensive challenges to the method of cases to date is due to edouard machery (machery, 2017). in this book length assessment to the case method, machery candidly assesses the reliability of effects of culture, gender, age, personality, order, and framing on case judgments. while some effects are more reliable than others, the result of this examination overall is that “nearly all the cases that have been examined are influenced either by demographic or by presentation variables, and this influence is frequently large,” (ibid. p. 88). machery goes on to conclude that, “variation and instability (due to presentation variables such as different framings or different orders of presentation) are thus both substantial and widespread,” (ibid). according to a separate analysis of these differences, more than 90 studies to date involving over 75,000 participants in experimental cognitive science have reported demographic variation in judgments to philosophical cases (stich & machery, 2022). while there is currently some debate regarding the frequency or strength to which philosophical intuitions vary by demographic groups (knobe, 2019), the possibility that demographic or presentation effects impact judgments when using the method of cases cannot be ignored. this must be considered and measured responsibly as a matter of course. incentive structure. a seventh factor perpetuating the replication crisis involves the way in which conducting inquiry in science is socially organized. researchers in science work within a highly institutionalized system of academic credit and reward (may, 2021). rewards in science such as placement and academic promotion require publishing. to a large extent, publishing requires positive findings. in turn, the pressure to generate positive findings increases the likelihood that the findings that are reported are not reliable or replicable. this situation has led researchers to conclude that “the most powerful incentives in contemporary science actively encourage, reward and propagate poor research methods” and that this process drives the “natural selection of bad science” (smaldino & mcelreath, 2016, p. 2). science, of course, is not the only field susceptible to research incentives. seemingly all academic researchers are incentivised toward maximizing credit and reward in their respective fields. however, the effects of this may be especially pronounced for philosophy and the method of cases. philosophy also requires publishing for placement and career advancement. and one of the main ways that philosophers build careers in philosophy is by generating examples that question or motivate theories. to generate these things, philosophers often rely on judgments from the method of cases. thus, they are strongly incentivised to create thought experiments that yield the desired judgments. but unlike controlled experiments, there is a relative lack of procedures, norms, or oversight in the use of thought experiments. moreover, solitary inquiry and single-authored publications have traditionally been the norm in philosophy. co-authored publications are sometimes even discouraged or viewed as less valuable contributions in philosophy, which is virtually the opposite of science. one consequence of these practices for the method is that individual researchers often end up being the sole buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 21 participants of the thought experiments they’ve created for the purposes of supporting or challenging the theories they wish to support or challenge and the only data generated are their own best judgments about whether they’ve been successful. this encourages the use of poorly designed thought experiments to produce the desired judgments. in summary, there are several similarities between the use of controlled experiments in social science and thought experiments in philosophy that overlap significantly with the causes of the replication crisis in science. both controlled experiments and thought experiments have utilized small samples that increase the likelihood that observed effects are false, overstated, or amplify existing biases in research practices. lack of replication and biased reporting perpetuate the use of unreliable evidence and limit our understanding of the total evidence. experimenter effects brought on by rampant researcher degrees of freedom and hidden moderators abound. lastly, researchers in both fields receive insufficient training in central aspects of the methods and are strongly incentivised to embrace methods that produce desired findings. these similarities between core aspects of the method of cases and methods in social psychological science are suggestive of a similar crisis in philosophy. indeed, perhaps one way to view the factors that caused the replication crisis in science is that they were improvements to the philosophical method of cases that did not go far enough. these improvements also made visible the problems in the way the method of cases had traditionally been practiced. 5 proposals for reliable philosophical research given that there are some initial warning signs in philosophy, it is reasonable to begin investigating the possibility that the replication crisis and related challenges to the reliability or credibility of science might extend to philosophy. one possibility is that this happens because philosophy relies on evidence generated by replication-crisis-era science. a second possibility is that this happens because some philosophical research methods overlap with structural factors that give rise to unreliable research. given these possibilities, it is reasonable to explore ways that the costs of the replication crisis might be avoided in philosophy by seizing the opportunity to improve philosophical research. fortunately, several proposals have been made for improving the replicability of social psychological science (munafò et al., 2017). given overlap between fields when it comes to evidence or methods, such proposals might also improve the reliability and efficiency of philosophical research. and while there is a lengthy debate in metaphilosophy about the products of the method of cases and the degree to which they constitute good evidence, the likelihood that they can is increased by adopting the following reforms. one proposal is to increase sample size through collaboration, group inquiry, and ideally, inquiry led by diverse groups. diversity and inclusion are widely disbuckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 22 cussed factors for improving professional philosophy. there is agreement that diversity and inclusion are not only paramount for just and productive inquiry, but also that these things can decrease the effect of individual biases within research communities (longino, 1990). prior research has also suggested that collaboration among diverse groups is valued in lay populations when choosing to engage with and study philosophical questions (buckwalter & turri, 2016). the present discussion contributes a more specific way that diversity and inclusion are paramount. by including a wider range of researchers into the research process, it increases the chances that case judgments are repeatable and reliable. when the group members have different disciplinary backgrounds, this can also have a training effect and balance out the lack of scientific literacy. a second proposal is to increase training in both scientific literacy and in using the method of cases. “as the devil can quote scripture”, some researchers have noted, “so the philosopher can quote science” (russell, 2008 [1917], pp. 43–44). if a large percentage of evidence in philosophy is empirical in nature, then philosophers require the ability to assess it. some efforts to improve training in some subfields, such as in moral psychology, for instance are underway but have not yet been widely adopted (machery & doris, 2017). doing so lowers the likelihood that philosophers or even entire subfields are duped by misinterpreted or questionable evidence. improving scientific literacy among philosophers is especially important when training and mentoring students, who might otherwise be encouraged to spend years of their lives on a thesis that is fundamentally motivated by results of no scientific merit. likewise, additional training is required for conducting the method of cases. this should involve basic training in the construction of cases, use of controls and multiple cover stories, and the unbiased presentation and assessment of cases (buckwalter, in press). a third proposal is to encourage reporting and publication of negative findings relevant to philosophical research. within philosophy, what “negative findings” amount to might take many forms. one idea is to incentivise the publication of papers that grapple with failed arguments for prized claims. one classic example of this comes from the paper “why is belief involuntary” by jonathan bennett. the paper begins as follows: this paper will present a negative result an account failure to explain why belief is involuntary. when i announced my question a year or so ahead of time, i had a vague idea it might be answered, but i cannot make it work out. necessity, this time, has not given birth to invention. (bennett, 1990, p. 87) the idea here is that exploring why our arguments for conclusions we like fail can be beneficial to research, especially for claims in philosophy that are currently popular, such as the claim that belief is involuntary. a different conception of publishing negative results involves changing incentives to encourage more reporting of unsuccessful uses of the case method. if a philosopher has a different intuition or diagnosis of the exact same thought exbuckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 23 periment, they should be encouraged to publish this finding. but more than just registering disagreement with case judgments from cases already in the published literature, philosophers might also be incentivised to report instances in which they could not build satisfactory cases to yield the case judgments that would generate a problem for a theory that they do not like or support one that they do. for instance, perhaps the researcher can only get the case to work in certain hypothetical situations that are not of central philosophical concern. sharing this information provides valuable insights into the process of case construction. for example, this might reveal confounding variables that would have otherwise impacted case judgments and distract us from the philosophical research question being investigated. one way that scientists have helped encourage reporting and transparency is through the use of preregistration and it is interesting to explore how this practice might potentially be used in other fields. preregistration involves specifying a research plan in a registry in advance of conducting a study (nosek brian et al., 2018). of course, journal articles in philosophy often do not clearly separate methods from results sections and philosophical research does not always proceed by testing hypotheses with empirical studies. nonetheless, preregistration may still benefit researchers by improving theoretical aspects of research projects (sarafoglou et al., 2022). for example, one thing preregistration forces researchers to do is to define primary and any secondary research questions and to clearly articulate research hypotheses. this practice can be incredibly valuable when structuring and honing research projects and to execute those research projects more efficiently. preregistration also forces researchers to specify an analysis plan, which in philosophy, might encourage more careful reflection about methods and evidence. for example, this might involve determining ahead of time whether evidence will come from social observation, personal experience, logical inference, historical analysis, or the method of cases, as well as encourage further reflection about the strengths and weaknesses of these sources for answering the central research question that has been identified. these may be things that early career researchers in particular could benefit from, especially if they involve input from peers at an early stage as to whether a research question, argument, or vignette is well specified. lastly, judgments from broader samples of researchers or the community should also be encouraged. thankfully, philosophy journals have indicated some willingness to publish replication attempts of foundational case judgments and the reinterpretation of past findings (kim & yuan, 2015; sytsma & livengood, 2011; turri, 2014). likewise, we might encourage additional review articles that document the strength of existing evidence in a research area. more broadly, a systematic repository of philosophically relevant replication attempts across the sciences would be extremely valuable in raising awareness about the reliability of scientific findings for philosophical theorizing. a fourth proposal for improving philosophical research is to limit the influence of experimenter effects and investigate potential moderators for philosophbuckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 24 ical judgments when practicing the method of cases. to do so requires a better scientific understanding of the psychological processes and mechanisms that underly case judgements. this necessity was partially foreseen by david hume, who wrote that “the only expedient, from which we can hope for success in our philosophical researches” is to study “human nature itself; which being once masters of, we may every where else hope for an easy victory” (hume, 1978, pp. i:6–8). in the same passage, hume tells us that “there is no question of importance, whose decision is not comprised in the science of man; and there is none, which can be decided with any certainty, before we become acquainted with that science” (ibid). applying hume’s insight to the present discussions, scientific research into the reasons philosophers make the judgments that they do goes a long way to improving our understanding of philosophical activity and research outcomes. one way to interpret hume’s vision is to fully embrace the methods of science when utilizing the case method in philosophy by conducting controlled experiments. there are some initial indications that doing so objectively improves the likelihood that the method produces reliable philosophical evidence. for example, one recent audit found that nearly three out of four studies in the field of experimental philosophy selected for attempted replication were successful, suggesting a relatively high replication rate compared to other areas of social science (cova et al., 2021). a separate audit found that statistical irregularities in experimental philosophy may be lower than those in other fields (colombo et al., 2018). of course, just because a judgment is shared does not necessarily mean that it is philosophically significant. but these audits do suggest that adopting experimental techniques with the reforms above in mind can be an effective way to improve the reliability of philosophical evidence. another way to interpret hume is to emphasize the important advantages of interdisciplinary research that includes both strong conceptual and experimental foundations. this suggests a fifth proposal, which is to encourage interdisciplinary research with these components. for example, the topic of mind wandering and the phenomenon of streams of consciousness is of both deep philosophical and scientific interest. in recent years, however, it has become clear that an efficient way to make progress understanding this phenomenon involves a combination of conceptual clarity in philosophy in defining central terms, as well as rigorous data collection concerning the neural correlates, ordinary concepts, experiences, and phenomenology of mind wandering (irving & glasser, 2020; mills et al., 2018). to cite another example, foundational questions in philosophy of mind involve the relationships between memory, self-knowledge, and group minds. recent research suggests that combining conceptual and experimental research methods on autobiographical recall clarifies the ways that collaborative processes impact the quality of certain memories (selwood et al., 2020). it is reasonable to suppose that grounding research in strong conceptual frameworks as well as experimental reforms motivated by crisis science may increase the reliability of research. future buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 25 research might profitably explore the relationship between replication rates and this kind of interdisciplinary inquiry. in the spirit of interdisciplinarity, a final question remains as to the scope of the argument and the impact of the crisis. philosophy is not the only field outside of what is typically considered science that either relies on scientific evidence or is subject to the structural pressures of conducting research implicated by the replication crisis. further study may reveal that the present discussion is but one case study in documenting a problem that generalizes from science to several other disciplines across the arts and humanities. researchers in these fields may also begin to examine the impact of the scientific crisis on their scholarly activity to improve research. acknowledgments thanks to carolyn buckwalter, david liggins, fraser macbride, john turri, and jennifer windt for comments on previous drafts of this manuscript. references aschwanden, c. (2019). we’re all “p-hacking” now: an insiders’ term for scientific malpractice has worked its way into pop culture. is that a good thing? www.wired.com/story/were-all-p-hacking-now/ bahník, š., & vranka, m. a. (2017). growth mindset is not associated with scholastic aptitude in a large sample of university applicants. personality and individual differences, 117, 139–143. https://doi.org/10.1016/j.paid.2017.05.046 baker, m. (2016). 1,500 scientists lift the lid on reproducibility. nature, 533, 452–454. https://doi.org/10.1038/533452a beebe, j. r., & undercoffer, r. (2016). individual and cross-cultural differences in semantic intuitions: new experimental findings. journal of cognition and culture, 16(3-4), 322–357. https://doi.org/10.1163/15685373-12342182 beebee, h. (2000). the non-governing conception of laws of nature. philosophy and phenomenological research, 61(3), 571– 594. https://doi.org/10.2307/2653613 beebee, h. (2018). philosophical scepticism and the aims of philosophy. proceedings of the aristotelian society, 118(1), 1–24. https://doi.org/10.1093/arisoc/aox017 begley, c. g., & ellis, l. m. (2012). raise standards for preclinical cancer research. nature, 483(7391), 531–533. https://doi.org/10.1038/483531a bennett, j. (1990). why is belief involuntary? analysis, 50(2), 87–107. https://doi.org/10.2307/3328852 blum, b. (2018). the lifespan of a lie. https://gen.medium.com/the-lifespan-of-a-lie-d869212b1f62 boekel, w., wagenmakers, e. j., belay, l., verhagen, j., brown, s., & forstmann, b. u. (2015). a purely confirmatory replication study of structural brain-behavior correlations. cortex, 66, 115–133. https://doi.org/10.1016/j.cortex.2014.11.019 brandt, m. j., ijzerman, h., dijksterhuis, a., farach, f. j., geller, j., giner-sorolla, r., grange, j. a., perugini, m., spies, j. r., & veer, a. van ’t. (2014). the replication recipe: what makes for a convincing replication? journal of experimental social psychology, 50, 217–224. https://doi.org/10.1016/j.jesp.2013.10.005 buckwalter, w. (2014). factive verbs and protagonist projection. episteme, 11(4), 391–409. https://doi.org/10.1017/epi.2014.22 buckwalter, w. (2019). implicit attitudes and the ability argument. philosophical studies, 176(11), 2961–2990. https://doi.org/10.1007/s11098-018-1159-7 buckwalter, w. (2021). error possibility, contextualism, and bias. synthese, 198, 2413–2426. https://doi.org/10.1007/s11229019-02221-w buckwalter, w. (in press). a guide to thought experiments in epistemology. in e. sosa, m. steup, j. turri, & b. roeber (eds.), contemporary debates in epistemology, 3rd edition. wiley-blackwell. buckwalter, w., & turri, j. (2015). inability and obligation in moral judgment. plos one, 10(8), 212–222. https://doi.org/10.1371/journal.pone.0136589 buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.wired.com/story/were-all-p-hacking-now/ https://doi.org/10.1016/j.paid.2017.05.046 https://doi.org/10.1038/533452a https://doi.org/10.1163/15685373-12342182 https://doi.org/10.2307/2653613 https://doi.org/10.1093/arisoc/aox017 https://doi.org/10.1038/483531a https://doi.org/10.2307/3328852 https://gen.medium.com/the-lifespan-of-a-lie-d869212b1f62 https://doi.org/10.1016/j.cortex.2014.11.019 https://doi.org/10.1016/j.jesp.2013.10.005 https://doi.org/10.1017/epi.2014.22 https://doi.org/10.1007/s11098-018-1159-7 https://doi.org/10.1007/s11229-019-02221-w https://doi.org/10.1007/s11229-019-02221-w https://doi.org/10.1371/journal.pone.0136589 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 26 buckwalter, w., & turri, j. (2016). perceived weaknesses of philosophical inquiry: a comparison to psychology. philosophia, 44(1), 33–52. https://doi.org/10.1007/s11406-015-9680-9 buckwalter, w., & turri, j. (2019). moderate scientism in philosophy. in j. de ridder, r. peels, & r. van woudenberg (eds.), scientism: prospects and problems (pp. 280–300). oxford university press. button, k. s., ioannidis, j. p. a., mokrysz, c., nosek, b. a., flint, j., robinson, e. s. j., & munafò, m. r. (2013). power failure: why small sample size undermines the reliability of neuroscience. nature reviews neuroscience, 14(5), 365– 376. https://doi.org/10.1038/nrn3475 camerer, c. f., dreber, a., forsell, e., ho, t.-h., huber, j., johannesson, m., kirchler, m., almenberg, j., altmejd, a., chan, t., heikensten, e., holzmeister, f., imai, t., isaksson, s., nave, g., pfeiffer, t., razen, m., & wu, h. (2016). evaluating replicability of laboratory experiments in economics. science, 25;351(6280), 1433–1436. https://doi.org/10.1126/science.aaf0918 carroll, j. w. (1884). laws of nature. cambridge university press. chituc, v., henne, p., sinnott-armstrong, w., & de brigard, f. (2016). blame, not ability, impacts moral “ought” judgments for impossible actions: toward an empirical refutation of “ought” implies “can”. cognition, 150, 20–25. https://doi.org/10.1016/j.cognition.2016.01.013 cohen, j. (1962). the statistical power of abnormal-social psychological research: a review. the journal of abnormal and social psychology, 65(3), 145–153. https://doi.org/10.1037/h0045186 colaco, d., buckwalter, w., stich, s., & machery, e. (2014). epistemic intuitions in fake-barn thought experiments. episteme, 11(2), 199–212. https://doi.org/10.1017/epi.2014.7 colombo, m., duev, g., nuijten, m. b., & sprenger, j. (2018). statistical reporting inconsistencies in experimental philosophy. plos one, 13(4), e0194360. https://doi.org/10.1371/journal.pone.0194360 cova, f., olivola, c. y., machery, e., stich, s., rose, d., alai, m., angelucci, a., berniūnas, r., buchtel, e. e., chatterjee, a., cheon, h., cho, i.-r., cohnitz, d., dranseika, v., lagos, á. e., ghadakpour, l., grinberg, m., hannikainen, i., hashimoto, t., … zhu, j. (2019). de pulchritudine non est disputandum? a cross-cultural investigation of the alleged intersubjective validity of aesthetic judgment. mind & language, 34(3), 317–338. https://doi.org/10.1111/mila.12210 cova, f., strickland, b., abatista, a., allard, a., andow, j., attie, m., beebe, j., berniūnas, r., boudesseul, j., colombo, m., cushman, f., diaz, r., n’djaye nikolai van dongen, n., dranseika, v., earp, b. d., torres, a. g., hannikainen, i., hernández-conde, j. v., hu, w., … zhou, x. (2021). estimating the reproducibility of experimental philosophy. review of philosophy and psychology, 12(1), 9–44. https://doi.org/10.1007/s13164-018-0400-9 cummins, r. (1998). reflections on reflective equilibrium. in m. depaul & w. ramsey (eds.), rethinking intuition: the psychology of intuition and its role in philosophical inquiry (pp. 113–127). rowman; littlefield. daly, c. (2017). ii—persistent philosophical disagreement. proceedings of the aristotelian society, 117(1), 23–40. https://doi.org/10.1093/arisoc/aow020 derose, k. (1992). contextualism and knowledge attributions. philosophy and phenomenological research, 52(4), 913–929. https://doi.org/10.2307/2107917 derose, k. (2009). the case for contextualism. oxford university press. derose, k. (2011). contextualism, contrastivism, and x-phi surveys. philosophical studies, 156(1), 81–110. https://doi.org/10.1007/s11098-011-9799-x ebersole, c. r., atherton, o. e., belanger, a. l., skulborstad, h. m., allen, j. m., banks, j. b., baranski, e., bernstein, m. j., bonfiglio, d. b. v., boucher, l., brown, e. r., budiman, n. i., cairo, a. h., capaldi, c. a., chartier, c. r., chung, j. m., cicero, d. c., coleman, j. a., conway, j. g., … nosek, b. a. (2016). many labs 3: evaluating participant pool quality across the academic semester via replication. journal of experimental social psychology, 67, 68–82. https://doi.org/10.1016/j.jesp.2015.10.012 feltz, a., & cokely, e. t. (2012). the philosophical personality argument. philosophical studies, 161(2), 227–246. https://doi.org/10.1007/s11098-011-9731-4 fidler, f., & wilcox, j. (2018). reproducibility of scientific results. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (winter 2018). metaphysics research lab, stanford university. https://plato.stanford.edu/archives/win2018/entries/ scientific-reproducibility/ finnigan, k. m., & corker, k. s. (2016). do performance avoidance goals moderate the effect of different types of stereotype threat on women’s math performance? journal of research in personality, 63, 36–43. https://doi.org/10.1016/j.jrp.2016.05.009 fischer, j. m. (2011). the zygote argument remixed. analysis, 71(2), 267–272. https://doi.org/10.1093/analys/anr008 fischer, j. m., & ravizza, m. (1998). responsibility and control: an essay on moral responsibility. cambridge university press. fiske, t., susan. (2016). a call to change science’s culture of shaming. aps observer, 29(9). buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11406-015-9680-9 https://doi.org/10.1038/nrn3475 https://doi.org/10.1126/science.aaf0918 https://doi.org/10.1016/j.cognition.2016.01.013 https://doi.org/10.1037/h0045186 https://doi.org/10.1017/epi.2014.7 https://doi.org/10.1371/journal.pone.0194360 https://doi.org/10.1111/mila.12210 https://doi.org/10.1007/s13164-018-0400-9 https://doi.org/10.1093/arisoc/aow020 https://doi.org/10.2307/2107917 https://doi.org/10.1007/s11098-011-9799-x https://doi.org/10.1016/j.jesp.2015.10.012 https://doi.org/10.1007/s11098-011-9731-4 https://plato.stanford.edu/archives/win2018/entries/scientific-reproducibility/ https://plato.stanford.edu/archives/win2018/entries/scientific-reproducibility/ https://doi.org/10.1016/j.jrp.2016.05.009 https://doi.org/10.1093/analys/anr008 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 27 fitzpatrick, w. (2018). cognitive science and moral philosophy: challenging scientistic overreach. in j. de ridder, r. peels, & r. van woudenberg (eds.), scientism: prospects and problems (pp. 233–257). oxford university press. https://doi.org/10.1093/oso/9780190462758.003.0011 flore, p. c., & wicherts, j. m. (2015). does stereotype threat influence performance of girls in stereotyped domains? a meta-analysis. journal of school psychology, 53(1), 25–44. https://doi.org/10.1016/j.jsp.2014.10.002 forscher, p. s., lai, c. k., axt, j. r., ebersole, c. r., herman, m., devine, p. g., & nosek, b. a. (2019). a meta-analysis of procedures to change implicit measures. journal of personality and social psychology, 117(3), 522–559. https://doi.org/10.1037/pspa0000160 gelman, a. (2016). why does the replication crisis seem worse in psychology? slate. https://slate.com/technology/2016/10/w hy-the-replication-crisis-seems-worse-in-psychology.html genschow, o., cracco, e., schneider, j., protzko, j., wisniewski, d., brass, m., & schooler, j. w. (2022). manipulating belief in free will and its downstream consequences: a meta-analysis. personality and social psychology review, 10888683221087527. https://doi.org/10.1177/10888683221087527 ghelfi, c. d. c., eric. (2020). reexamining the effect of gustatory disgust on moral judgment: a multi-lab direct replication of eskine, kacinik, and prinz (2011). advances in methods and practices in psychological science, 3–23. https://doi.org/10.1177/2515245919881152 greenland, s., senn, s. j., rothman, k. j., carlin, j. b., poole, c., goodman, s. n., & altman, d. g. (2016). statistical tests, p values, confidence intervals, and power: a guide to misinterpretations. european journal of epidemiology, 31(4), 337–350. https://doi.org/10.1007/s10654-016-0149-3 hagger, m. s., chatzisarantis, n. l. d., alberts, h., anggono, c. o., batailler, c., birt, a. r., brand, r., brandt, m. j., brewer, g., bruyneel, s., calvillo, d. p., campbell, w. k., cannon, p. r., carlucci, m., carruth, n. p., cheung, t., crowell, a., ridder, d. t. d. d., dewitte, s., … zwienenberg, m. (2016). a multilab preregistered replication of the ego-depletion effect. perspectives on psychological science, 11(4), 546–573. https://doi.org/10.1177/1745691616652873 hüffmeier, j., mazei, j., & schultze, t. (2016). reconceptualizing replication as a sequence of different studies: a replication typology. journal of experimental social psychology, 66, 81–92. https://doi.org/10.1016/j.jesp.2015.09.009 hume, d. (1978). a treatise of human nature. oxford university press. inthout, j., ioannidis, j. p., borm, g. f., & goeman, j. j. (2015). small studies are more heterogeneous than large ones: a meta-meta-analysis. journal of clinical epidemiology, 68(8), 860–869. https://doi.org/10.1016/j.jclinepi.2015.03.017 inzlicht, m. (2016). reckoning with the past. http://michaelinzlicht.com/getting-better/2016/2/29/reckoning-with-the-past ioannidis, j. p. a. (2012). why science is not necessarily self-correcting. perspectives on psychological science, 7(6), 645–654. https://doi.org/10.1177/1745691612464056 ioannidis, j. p., munafo, m. r., fusar-poli, p., nosek, b. a., & david, s. p. (2014). publication and other reporting biases in cognitive sciences: detection, prevalence, and prevention. trends in cognitive sciences, 18(5), 235–241. https://doi.org/10.1016/j.tics.2014.02.010 irving, z. c., & glasser, a. (2020). mind-wandering: a philosophical guide. philosophy compass, 15(1), e12644. https://doi.org/10.1111/phc3.12644 johnson, d. j., wortman, j., cheung, f., hein, m., lucas, r. e., donnellan, m. b., ebersole, c. r., & narr, r. k. (2016). the effects of disgust on moral judgments: testing moderators. social psychological and personality science, 7(7), 640–647. https://doi.org/10.1177/1948550616654211 kim, m., & yuan, y. (2015). no cross-cultural differences in gettier car case intuition: a replication study of weinberg et al. 2001. episteme, 12(3), 355–361. https://doi.org/10.1017/epi.2015.17 klein, r. a., vianello, m., hasselman, f., adams, b. g., reginald b. adams, jr., alper, s., aveyard, m., axt, j. r., babalola, m. t., bahník, š., batra, r., berkics, m., bernstein, m. j., berry, d. r., bialobrzeska, o., binan, e. d., bocian, k., brandt, m. j., busching, r., … nosek, b. a. (2018). many labs 2: investigating variation in replicability across samples and settings. advances in methods and practices in psychological science, 1(4), 443–490. https://doi.org/10.1177/2515245918810225 kneer, m., & machery, e. (2019). no luck for moral luck. cognition, 182, 331–348. https://doi.org/10.1016/j.cognition.2018.09.003 knobe, j. (2015). philosophers are doing something different now: quantitative data. cognition, 135, 36–38. https://doi.org/10.1016/j.cognition.2014.11.011 knobe, j. (2019). philosophical intuitions are surprisingly robust across demographic differences. epistemology & philosophy of science, 56(2), 29–36. https://doi.org/10.5840/eps201956225 kornblith, h. (2013). is philosophical knowledge possible? in d. e. machuca (ed.), disagreement and skepticism (pp. 260– 276). routledge. kornblith, h. (2018). philosophy, science, and common sense. in j. de ridder, r. peels, & r. van woudenberg (eds.), scientism: prospects and problems (pp. 127–148). oxford university press. https://doi.org/10.1093/oso/9780190462758.003.0006 buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1093/oso/9780190462758.003.0011 https://doi.org/10.1016/j.jsp.2014.10.002 https://doi.org/10.1037/pspa0000160 https://slate.com/technology/2016/10/why-the-replication-crisis-seems-worse-in-psychology.html https://slate.com/technology/2016/10/why-the-replication-crisis-seems-worse-in-psychology.html https://doi.org/10.1177/10888683221087527 https://doi.org/10.1177/2515245919881152 https://doi.org/10.1007/s10654-016-0149-3 https://doi.org/10.1177/1745691616652873 https://doi.org/10.1016/j.jesp.2015.09.009 https://doi.org/10.1016/j.jclinepi.2015.03.017 http://michaelinzlicht.com/getting-better/2016/2/29/reckoning-with-the-past https://doi.org/10.1177/1745691612464056 https://doi.org/10.1016/j.tics.2014.02.010 https://doi.org/10.1111/phc3.12644 https://doi.org/10.1177/1948550616654211 https://doi.org/10.1017/epi.2015.17 https://doi.org/10.1177/2515245918810225 https://doi.org/10.1016/j.cognition.2018.09.003 https://doi.org/10.1016/j.cognition.2014.11.011 https://doi.org/10.5840/eps201956225 https://doi.org/10.1093/oso/9780190462758.003.0006 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 28 ladyman, j., collier, j. g., ross, d., & spurrett, d. (2014). every thing must go: metaphysics naturalized. oxford university press. landy, j. f., & goodwin, g. p. (2015). does incidental disgust amplify moral judgment? a meta-analytic review of experimental evidence. perspectives on psychological science, 10(4), 518–536. https://doi.org/10.1177/1745691615583128 landy, j. f., jia, m. l., ding, i. l., viganola, d., tierney, w., dreber, a., johannesson, m., pfeiffer, t., ebersole, c. r., gronau, q. f., ly, a., bergh, d. van den, marsman, m., derks, k., wagenmakers, e. j., proctor, a., bartels, d. m., bauman, c. w., brady, w. j., … uhlmann, e. l. (2020). crowdsourcing hypothesis tests: making transparent how design choices shape research results. psychological bulletin, 146(5), 451–479. https://doi.org/10.1037/bul0000220 le texier, t. (2019). debunking the stanford prison experiment. american psychologist, 74(7), 823–839. https://doi.org/10.1037/amp0000401 lehrer, k. (1965). knowledge, truth and evidence. analysis, 25(5), 168–175. https://doi.org/10.2307/3326431 levin, n., & leonelli, s. (2017). how does one “open” science? questions of value in biological research. science, technology, & human values, 42(2), 280–305. https://doi.org/10.1177/0162243916672071 lewis, d. (2000). causation as influence. the journal of philosophy, 97(4), 182–197. https://doi.org/10.2307/2678389 liao, s. m., wiegmann, a., alexander, j., & vong, g. (2012). putting the trolley in order: experimental philosophy and the loop case. philosophical psychology, 25(5), 661–671. https://doi.org/10.1080/09515089.2011.627536 longino, h. e. (1990). science as social knowledge: values and objectivity in scientific inquiry. princeton university press. lycan, w. g. (2013). on two main themes in gutting’s what philosophers know. book symposium, southern journal of philosophy, 51(1), 112–120. https://doi.org/10.1111/sjp.12003 macbride, f. (2014). analytic philosophy and its synoptic commission: towards the epistemic end of days. royal institute of philosophy supplement, 74, 221–236. https://doi.org/10.1017/s1358246114000095 machery, e. (2017). philosophy within its proper bounds. oxford university press. machery, e. (2020). what is a replication? philosophy of science, 87(4), 545–567. https://doi.org/10.1086/709701 machery, e. (2022). anomalies in implicit attitudes research. wiley interdisciplinary reviews: cognitive science, 13(1), e1569. https://doi.org/10.1002/wcs.1569 machery, e., & doris, j. m. (2017). an open letter to our students: doing interdisciplinary moral psychology. in b. g. voyer & t. tarantola (eds.), moral psychology: a multidisciplinary guide (pp. 119–143). springer international publishing. https://doi.org/10.1007/978-3-319-61849-4_7 machery, e., mallon, r., nichols, s., & stich, s. p. (2004). semantics, cross-cultural style. cognition, 92(3), 1–12. https://doi.org/10.1016/j.cognition.2003.10.003 machery, e., stich, s., rose, d., chatterjee, a., karasawa, k., struchiner, n., sirker, s., usui, n., & hashimoto, t. (2017). gettier across cultures. noûs, 51(3), 645–664. https://doi.org/10.1111/nous.12110 may, j. (2021). bias in science: natural and social. synthese, 199(1), 3345–3366. https://doi.org/10.1007/s11229-020-02937-0 mcauliffe, w. h. b., carter, e. c., berhane, j., snihur, a. c., & mccullough, m. e. (2020). is empathy the default response to suffering? a meta-analytic evaluation of perspective taking’s effect on empathic concern. personality and social psychology review, 24(2), 141–162. https://doi.org/10.1177/1088868319887599 mcdermott, m. (1995). redundant causation. the british journal for the philosophy of science, 46(4), 523–544. https://doi.org/10.1093/bjps/46.4.523 mele, a. (2006). free will and luck. oxford university press. mills, c., raffaelli, q., irving, z. c., stan, d., & christoff, k. (2018). is an off-task mind a freely-moving mind? examining the relationship between different dimensions of thought. consciousness and cognition, 58, 20–33. https://doi.org/10 .1016/j.concog.2017.10.003 munafò, m. r., nosek, b. a., bishop, d. v. m., button, k. s., chambers, c. d., percie du sert, n., simonsohn, u., wagenmakers, e.-j., ware, j. j., & ioannidis, j. p. a. (2017). a manifesto for reproducible science. nature human behaviour, 1(1), 0021. https://doi.org/10.1038/s41562-016-0021 nadelhoffer, t., rose, d., buckwalter, w., & nichols, s. (2020). natural compatibilism, indeterminism, and intrusive metaphysics. cognitive science, 44(e12873). https://doi.org/10.1111/cogs.12873 nahmias, e. (2014). is free will an illusion? confronting challenges from the modern mind sciences. in w. sinnottarmstrong (ed.), moral psychology, vol. 4: freedom and responsibility (pp. 2–56). mit press. nahmias, e. a., morris, s. g., nadelhoffer, t., & turner, j. (2005). surveying freedom: folk intuitions about free will and moral responsibility. philosophical psychology, 18(5), 561–584. https://doi.org/10.1080/09515080500264180 nosek, b. a., & errington, t. m. (2017). making sense of replications. elife, 6, e23383. https://doi.org/10.7554/elife.23383 nosek, b. a., & lakens, d. (2014). registered reports: a method to increase the credibility of published results. social psychology, 45(3), 137–141. https://doi.org/10.1027/1864-9335/a000192 buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1177/1745691615583128 https://doi.org/10.1037/bul0000220 https://doi.org/10.1037/amp0000401 https://doi.org/10.2307/3326431 https://doi.org/10.1177/0162243916672071 https://doi.org/10.2307/2678389 https://doi.org/10.1080/09515089.2011.627536 https://doi.org/10.1111/sjp.12003 https://doi.org/10.1017/s1358246114000095 https://doi.org/10.1086/709701 https://doi.org/10.1002/wcs.1569 https://doi.org/10.1007/978-3-319-61849-4_7 https://doi.org/10.1016/j.cognition.2003.10.003 https://doi.org/10.1111/nous.12110 https://doi.org/10.1007/s11229-020-02937-0 https://doi.org/10.1177/1088868319887599 https://doi.org/10.1093/bjps/46.4.523 https://doi.org/10.1016/j.concog.2017.10.003 https://doi.org/10.1016/j.concog.2017.10.003 https://doi.org/10.1038/s41562-016-0021 https://doi.org/10.1111/cogs.12873 https://doi.org/10.1080/09515080500264180 https://doi.org/10.7554/elife.23383 https://doi.org/10.1027/1864-9335/a000192 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the replication crisis and philosophy 29 nosek brian, a., ebersole charles, r., dehaven alexander, c., & mellor david, t. (2018). the preregistration revolution. proceedings of the national academy of sciences, 115(11), 2600–2606. https://doi.org/10.1073/pnas.1708274114 osf. (2015). estimating the reproducibility of psychological science. 349(6251), aac4716. https://doi.org/10.1126/science.aac4716 papineau, d. (2017). is philosophy simply harder than science? the times literary supplement, december 2, 2019. pereboom, d. (2014). free will, agency, and meaning in life. oxford university press. philpot, r., liebst, l., levine, m., bernasco, w., & lindegaard, m. r. (2020). would i be helped? cross-national cctv footage shows that intervention is the norm in public conflicts. american psychologist, 75(1), 66–75. https://doi.org/10.1037/amp0000469 reid, t. (1764/1997). an inquiry into the human mind on the principles of common sense. pennsylvania state university press. romero, f. (2018). who should do replication labor? advances in methods and practices in psychological science, 1(4), 516– 537. https://doi.org/10.1177/2515245918803619 romero, f. (2019). philosophy of science and the replicability crisis. philosophy compass, 14:e12633. https://doi.org/10.1111/phc3.12633 rose, d., buckwalter, w., & nichols, s. (2017). neuroscientific prediction and the intrusion of intuitive metaphysics. cognitive science, 41, 482–502. https://doi.org/10.1111/cogs.12310 rose, d., buckwalter, w., & turri, j. (2014). when words speak louder than actions: delusion, belief, and the power of assertion. australasian journal of philosophy, 92(4), 683–700. https://doi.org/10.1080/00048402.2014.909859 rose, d., machery, e., stich, s., alai, m., angelucci, a., berniūnas, r., buchtel, e. e., chatterjee, a., cheon, h., cho, i.r., cohnitz, d., cova, f., dranseika, v., lagos, á. e., ghadakpour, l., grinberg, m., hannikainen, i., hashimoto, t., horowitz, a., … zhu, j. (2019). nothing at stake in knowledge. noûs, 53(1), 224–247. https://doi.org/10.1111/nous.12211 rosenthal, r. (1979). the file drawer problem and tolerance for null results. psychological bulletin, 86(3), 638–641. https://doi.org/10.1037/0033-2909.86.3.638 russell, b. (2008 [1917]). mysticism and logic and other essays. project gutenberg. https://www.gutenberg.org/ebooks/25447 sarafoglou, a., kovacs, m., bakos, b., wagenmakers, e.-j., & aczel, b. (2022). a survey on how preregistration affects the research workflow: better science but more work. royal society open science, 9(211997). https://doi.org/10.1098/rsos.211997 scargle, d., jeffrey. (2000). publication bias: the “file-drawer” problem in scientific inference. journal of scientific exploration, 14(1), 91–106. schimmack, u. (2017). preliminary 2017 replicability rankings of 104 psychology journals: vols. october 8, 2019 (october 8, 2019). https://replicationindex.com/2017/10/24/preliminary-2017-replicability-rankings-of-104-psychologyjournals/ schimmack, u. (2019). the implicit association test: a method in search of a construct. perspectives on psychological science, 1745691619863798. https://doi.org/10.1177/1745691619863798 schwitzgebel, e., & cushman, f. (2015). philosophers’ biased judgments persist despite training, expertise and reflection. cognition, 141, 127–137. https://doi.org/10.1016/j.cognition.2015.04.015 sedlmeier, p., & gigerenzer, g. (1989). do studies of statistical power have an effect on the power of studies? psychological bulletin, 105(2), 309–316. https://doi.org/10.1037/0033-2909.105.2.309 selwood, a., harris, c. b., barnier, a. j., & sutton, j. (2020). effects of collaboration on the qualities of autobiographical recall in strangers, friends, and siblings: both remembering partner and communication processes matter. memory, 28(3), 399–416. https://doi.org/10.1080/09658211.2020.1727521 shewach, o. r., sackett, p. r., & quint, s. (2019). stereotype threat effects in settings with features likely versus unlikely in operational test settings: a meta-analysis. journal of applied psychology, 104(12), 1514–1534. https://doi.org/10.1037/ apl0000420 simmons, j. p., nelson, l. d., & simonsohn, u. (2011). false-positive psychology: undisclosed flexibility in data collection and analysis allows presenting anything as significant. psychological science, 22(11), 1359–1366. https://doi.org/10.1 177/0956797611417632 smaldino, p. e., & mcelreath, r. (2016). the natural selection of bad science. royal society open science, 3(160384). https://doi.org/10.1098/rsos.160384 stich, s. p., & machery, e. (2022). demographic differences in philosophical intuition: a reply to joshua knobe. review of philosophy and psychology. https://doi.org/10.1007/s13164-021-00609-7 strawson, g. (1986). freedom and belief. oxford university press. strevens, m. (2019). thinking off your feet: how empirical psychology vindicates armchair philosophy. harvard university press. buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1073/pnas.1708274114 https://doi.org/10.1126/science.aac4716 https://doi.org/10.1037/amp0000469 https://doi.org/10.1177/2515245918803619 https://doi.org/10.1111/phc3.12633 https://doi.org/10.1111/cogs.12310 https://doi.org/10.1080/00048402.2014.909859 https://doi.org/10.1111/nous.12211 https://doi.org/10.1037/0033-2909.86.3.638 https://www.gutenberg.org/ebooks/25447 https://doi.org/10.1098/rsos.211997 https://replicationindex.com/2017/10/24/preliminary-2017-replicability-rankings-of-104-psychology-journals/ https://replicationindex.com/2017/10/24/preliminary-2017-replicability-rankings-of-104-psychology-journals/ https://doi.org/10.1177/1745691619863798 https://doi.org/10.1016/j.cognition.2015.04.015 https://doi.org/10.1037/0033-2909.105.2.309 https://doi.org/10.1080/09658211.2020.1727521 https://doi.org/10.1037/apl0000420 https://doi.org/10.1037/apl0000420 https://doi.org/10.1177/0956797611417632 https://doi.org/10.1177/0956797611417632 https://doi.org/10.1098/rsos.160384 https://doi.org/10.1007/s13164-021-00609-7 https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wesley buckwalter 30 strickland, b., & suben, a. (2012). experimenter philosophy: the problem of experimenter bias in experimental philosophy. review of philosophy and psychology, 3(3), 457–467. https://doi.org/10.1007/s13164-012-0100-9 stroebe, w. (2019). what can we learn from many labs replications? basic and applied social psychology, 41(2), 91–103. https://doi.org/10.1080/01973533.2019.1577736 sytsma, j., & livengood, j. (2011). a new perspective concerning experiments on semantic intuitions. australasian journal of philosophy, 89(2), 315–332. https://doi.org/10.1080/00048401003639832 turri, j. (2014). the problem of esee knowledge. ergo, 1(4), 101–127. https://doi.org/10.3998/ergo.12405314.0001.004 turri, j. (2017). epistemic contextualism: an idle hypothesis. australasian journal of philosophy, 95(1), 141–156. https://doi.org/10.1080/00048402.2016.1153684 turri, j., buckwalter, w., & blouw, p. (2015). knowledge and luck. psychonomic bulletin and review, 22, 378–390. https://doi.org/10.3758/s13423-014-0683-5 turri, j., buckwalter, w., & rose, d. (2016). actionability judgments cause knowledge judgments. thought, 5(3), 212–222. https://doi.org/10.1002/tht3.213 van bavel, j. j., mende-siedlecki, p., brady, w. j., & reinero, d. a. (2016). contextual sensitivity in scientific reproducibility. proceedings of the national academy of sciences of the united states of america, 113(23), 6454–6459. https://doi.org/10.1073/pnas.1521897113 van inwagen, p. (2015). metaphysics. westview press. vazire, s. (2018). implications of the credibility revolution for productivity, creativity, and progress. perspectives on psychological science, 13(4), 411–417. https://doi.org/10.1177/1745691617751884 wicherts, j. m., veldkamp, c. l. s., augusteijn, h. e. m., bakker, m., aert, r. c. m. van, & assen, m. a. l. m. van. (2016). degrees of freedom in planning, running, analyzing, and reporting psychological studies: a checklist to avoid phacking. frontiers in psychology, 7, 1832–1832. https://doi.org/10.3389/fpsyg.2016.01832 wiegmann, a., okan, y., & nagel, j. (2012). order effects in moral judgment. philosophical psychology, 25(6), 813–836. https://doi.org/10.1080/09515089.2011.631995 wilholt, t. (2012). epistemic trust in science. the british journal for the philosophy of science, 64(2), 233–253. https://doi.org/10.1093/bjps/axs007 williams, b. a. o. (1981). moral luck. cambridge university press. wood, t., & porter, e. (2019). the elusive backfire effect: mass attitudes’ steadfast factual adherence. political behavior, 41(1), 135–163. https://doi.org/10.1007/s11109-018-9443-y open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. buckwalter, w. (2022). the replication crisis and philosophy. philosophy and the mind sciences, 3, 16. https://doi.org/10.33735/phimisci.2022.9193 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s13164-012-0100-9 https://doi.org/10.1080/01973533.2019.1577736 https://doi.org/10.1080/00048401003639832 https://doi.org/10.3998/ergo.12405314.0001.004 https://doi.org/10.1080/00048402.2016.1153684 https://doi.org/10.3758/s13423-014-0683-5 https://doi.org/10.1002/tht3.213 https://doi.org/10.1073/pnas.1521897113 https://doi.org/10.1177/1745691617751884 https://doi.org/10.3389/fpsyg.2016.01832 https://doi.org/10.1080/09515089.2011.631995 https://doi.org/10.1093/bjps/axs007 https://doi.org/10.1007/s11109-018-9443-y https://doi.org/10.33735/phimisci.2022.9193 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction indications of crisis philosophy in the age of replication crisis science replication and the method of cases proposals for reliable philosophical research the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. the neural correlates of consciousness under the free energy principle from computational correlates to computational explanation wanja wiesea (wanja.wiese@rub.de) karl j. fristonb (k.friston@ucl.ac.uk) abstract how can the free energy principle contribute to research on neural correlates of consciousness, and to the scientific study of consciousness more generally? under the free energy principle, neural correlates should be defined in terms of neural dynamics, not neural states, and should be complemented by research on computational correlates of consciousness – defined in terms of probabilities encoded by neural states. we argue that these restrictions brighten the prospects of a computational explanation of consciousness, by addressing two central problems. the first is to account for consciousness in the absence of sensory stimulation and behaviour. the second is to allow for the possibility of systems that implement computations associated with consciousness, without being conscious, which requires differentiating between computational systems that merely simulate conscious beings and computational systems that are conscious in and of themselves. given the notion of computation entailed by the free energy principle, we derive constraints on the ascription of consciousness in controversial cases (e.g., in the absence of sensory stimulation and behaviour). we show that this also has implications for what it means to be, as opposed to merely simulate a conscious system. keywords active inference ∙ computational correlates of consciousness ∙ computational explanation ∙ consciousness ∙ free energy principle ∙ islands of awareness ∙ minimal unifying model ∙ neural correlates of consciousness this article is part of a special issue on “the neural correlates of consciousness,” edited by sascha benjamin fink and ying-tung lin. ainstitute of philosophy ii, ruhr-universität bochum bwellcome centre for human neuroimaging, ucl wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-3338-7398 https://orcid.org/0000-0001-7984-8909 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 2 1 introduction the free energy principle (fep, friston, 2010) provides an information-theoretic analysis of the concept of existence of self-organising systems (hohwy, 2020). it entails that self-organising systems at non-equilibrium steady state1 bound the entropy of their sensory signals by minimising – on average – a quantity known as variational free energy. as such, fep is not a theory of consciousness. consequently, it is not obvious that fep has any relevance to research on neural correlates of consciousness (nccs, see chalmers, 2000). furthermore, fep is a computational principle;2 hence, it may seem even less relevant to investigating the neural structures associated with consciousness. here, we will argue that fep provides constraints on computational correlates of consciousness (cccs, see atkinson et al., 2000; cleeremans, 2005; mathis & mozer, 1996; reggia et al., 2016, 2019). cccs specify computational properties instantiated by nccs, and thereby provide some insights as to why a particular neural structure or type of activity is associated with consciousness. more specifically, we will argue that fep supports the following three observations with respect to the debate on nccs and cccs. (i) according to fep, nccs must be defined in terms of neural dynamics, not neural states. (ii) according to fep, there is a relevant distinction to be made between the probabilities of neural states (or of trajectories) and the probabilities encoded by neural states.3 these distinct families of probability distributions can be regarded as points on statistical manifolds, with corresponding information geometries. in friston, wiese, et al. (2020), the information geometries – associated with the probabilities of states and with probabilities encoded by states – are called intrinsic and extrinsic information geometries, respectively. in line with this distinction, neural dynamics (with nccs as a special case) pertain to movements on the intrinsic statistical manifold, whereas neural computations (with cccs as a special case) pertain to movements on the extrinsic manifold. furthermore, the computations ascribed to movements on the extrinsic manifold can be regarded as inferences or, more poetically, selfevidencing (hohwy, 2016).4 (iii) some candidates for cccs that have been pro1a system is at non-equilibrium with its environment if it is in exchange with its environment. it is in steady state if it has characteristic features that remain invariant during this exchange, i.e., if it continues to exist. 2technically, the fep is a variational principle of stationary action. the “action” in question here is a path integral of a functional of probabilistic beliefs encoded by the internal states of a system. as such, fep becomes a computational principle; in that the probabilistic beliefs in question are about something (i.e., the external states of a system). for a discussion of the relationship between fep, computation, and representation, see wiese & friston (2021). 3a general, lucid analysis of the two kinds of information processing associated with probabilities of states (representational vehicles) and with probabilities represented by states (probabilities in represented content), respectively, is provided by sprevak (2020). for a related, excellent discussion with respect to fep, see kiefer (2020). 4this is because every point on the extrinsic manifold is equipped with a free energy functional, namely a function of the probabilistic belief corresponding to that point. crucially, this functional wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 3 posed in the literature are – from the point of view of fep – just mathematical descriptions of nccs. (iv) if a goal of research on correlates of consciousness is to infer the presence or absence of consciousness in controversial cases, then one should regard these correlates as necessary (not sufficient) conditions for consciousness. in particular, this holds for computational principles entailed by fep. a key contribution to the debate that follows from these observations is that computational explanations of consciousness (or “explanatory correlates of consciousness,” seth, 2009) must be specified with respect to the extrinsic information geometry. that is, a computational explanation of consciousness requires more than a formal description of neural dynamics. it must also specify computations instantiated by these dynamics. moreover, fep has implications for the very idea of a computational explanation of mental phenomena. fep not only specifies computations that have to be performed by self-organising systems at non-equilibrium steady state, but also entails that descriptions of these computations are equivalent to descriptions of physical systems interacting with their environment. therefore, it provides additional constraints on what it means to be, as opposed to merely simulate a member of a certain class of computational systems.5 this enables a nontrivial role for computational correlates in explanations of consciousness, without having to accept what david chalmers (2011) calls the thesis of computational sufficiency (according to which the right computational structure is sufficient for mind and consciousness). in particular, our account is compatible with the possibility of synthetic phenomenology, but does not entail that every computer that performs the right computations is conscious. although fep itself does not provide an explanation of consciousness, it provides a quantitative framework within which models and measures of consciousness can be developed. more specifically, process theories conforming to fep, such as active inference (friston et al., 2017), can facilitate formulating and testing hypotheses about necessary computational mechanisms that may underpin different cognitive capacities, or different measures of consciousness. because of the broad applicability of the framework (for a recent overview, see costa, parr, et al., 2020), completely different cognitive or formal properties associated with consciousness can be modelled and analysed. since all models that conform to fep describe the phenomena that are modelled as ways of minimising free energy, fep provides a unifying perspective (hohwy & seth, 2020) from which it becomes tractable to determine what necessary computational mechanisms the different phenomena have in common (wiese, 2018). (function of a function) is an upper bound on surprisal. equivalently, the negative free energy constitutes a lower bound on model evidence (a.k.a. marginal likelihood) in bayesian statistics. 5this means fep not only provides the foundation of a framework for consciousness, but also building blocks for a theory of consciousness. this contrasts with the proposal in hohwy & seth (2020), according to which predictive processing (pp) only provides a framework for research on consciousness (and nccs), without itself being a theory of consciousness. (although hohwy and seth suggest that “aspects of pp may themselves coalesce into a theory of consciousness in its own right,” hohwy & seth, 2020, p. 24.) wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 4 the rest of this paper is structured as follows. in section 2, we review the canonical notion of an ncc, as defined by chalmers (2000). we highlight three challenges for research on neural correlates and suggest that these challenges can be overcome by research on computational correlates of consciousness (cccs). in section 3, we discuss research on cccs (section 3.1) and argue that some computational principles associated with consciousness (cleeremans, 2005) are already entailed by fep (or by fep and further, general assumptions; section 3.2). this is one reason why cccs should be regarded as necessary conditions for consciousness (not sufficient conditions; section 3.3). furthermore, we take up a special case of a challenge identified in section 2, viz. the challenge of “islands of awareness” (bayne et al., 2020). we argue that fep furnishes a constraint on the ascription of consciousness in such unusual cases: according to fep, such systems must minimise the statistical complexity of their internal model (section 3.4). since empirical evidence suggests that activity in conscious systems has high dynamical complexity, we briefly discuss how these notions relate to one another and suggest that it should be possible to define dynamical complexity in terms of a system’s extrinsic information geometry (section 3.5). the benefit of such a definition is that it would provide not just a fundamental theoretical motivation for associating consciousness with dynamical complexity, but also the basis for a computational explanation of consciousness. the explanatory value of fep is a matter of controversy. hence, in section 4, we explain what contribution fep would make to a computational explanation of consciousness. finally, in section 5, we discuss its potential role in developing a minimal unifying model of consciousness, following a suggestion made in wiese (2020). 2 from neural correlates to computational correlates of consciousness according to the canonical notion of an ncc, as defined by chalmers (2000), “[a]n ncc is a minimal neural system n such that there is a mapping from states of n to states of consciousness, where a given state of n is sufficient under conditions c, for the corresponding state of consciousness” (chalmers, 2000, p. 31). a classic example of an ncc is provided by crick & koch (1990), who hypothesized that synchronised oscillations (in a range around 40–70 hz) in parts of visual cortex may underpin visual consciousness. the neural system n would be visual cortex (or parts of it). states of – or, rather, activity in – visual cortex corresponding to visual experiences would be synchronised oscillations within a certain frequency. note that the definition cited above is a general definition for neural correlates of states of consciousness. that is, it can be applied to neural differences between consciousness and the absence of consciousness, as well as to neural differences between background states, such as wakefulness, dreaming, or hypnosis, or between particular conscious experiences, such as smelling coffee, hearing a voice, wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 5 seeing a face, or seeing a house. in addition to this general notion, chalmers (2000) proposes a more specific definition for neural correlates of individual contents of consciousness. in what follows, we shall focus on the general definition, because some parts of our discussion mainly apply to neural correlates of global states of consciousness (e.g., being conscious vs. unconscious, being awake vs. dreaming, etc.). we entertain three challenges for chalmers’ general definition of an ncc. the aim of this discussion is not to argue that the definition is problematic.6 instead, we wish to highlight the virtues of computational correlates of consciousness (cccs, cleeremans, 2005; reggia et al., 2016). cccs, as understood here, are not defined in terms of neural structures, but in terms of computational properties that can be instantiated by different substrates. cccs are thus more general and promise to be more explanatory than nccs. however, they also pose challenges of their own. 2.1 challenge 1: global dynamics the definition of an ncc cited above is meant to pick out core correlates of consciousness. an implicit hypothesis is that the difference between neural activity that is associated with consciousness, and neural activity that is not, can be specified in terms of a difference in the neural systems involved, as well as by differences in the type of activity (“states”) instantiated by these systems. this hypothesis is challenged by theoretical considerations, according to which states cannot be mapped to conscious experiences, but only trajectories (fekete & edelman, 2011); furthermore, it is in tension with considerations that privilege global (as opposed to local) processes (mashour et al., 2020; tononi et al., 2016).7 recent empirical approaches support such theoretical considerations, by highlighting global dynamics that are associated with consciousness. examples include differences in functional connectivity (huang et al., 2020; luppi et al., 2019) or in the power spectral density of resting eeg (colombo et al., 2019; he et al., 2010). a further example is given by research on differences between global neural dynamics and state transitions during wakefulness and non-rem sleep (stevner et al., 2019), or global dynamics during wakefulness and different types of unrespon6some of these challenges are explicitly mentioned in chalmers (2000) already; for discussions of chalmers’ definition, and ncc research more generally, see noë & thompson (2004); block (2005); miller (2007); tononi & koch (2008); hohwy (2009); seth (2009); aru et al. (2012); tsuchiya et al. (2015); koch et al. (2016); fink (2016); mashour (2018). 7chalmers’ definition of an ncc is compatible with this possibility: it may turn out that the minimal neural system that is sufficient for states of consciousness is the entire brain. however, the possibility also illustrates that nccs, defined in terms of neural systems, may be less informative if consciousness mainly depends on how neural activity evolves over a certain period of time, and depends less on where in the brain the neural activity takes place. the virtue of computational correlates is that they can be informative regardless of whether they describe properties of the entire brain or of small regions in the brain. wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 6 siveness, such as anaesthesia, unresponsive wakefulness syndrome, or minimally conscious states (demertzi et al., 2019). computational correlates are not defined in terms of neural structures, and hence are neutral with respect to the question whether conscious experiences correlate with global neural activity or not. furthermore, they are compatible with the possibility that neural state transitions are more important than the states themselves. hence, computational correlates promise progress on the challenge of accounting for global dynamics associated with consciousness. 2.2 challenge 2: non-arbitrary mappings a further challenge consists in determining why a particular neural structure n is an ncc of consciousness (see hohwy & seth, 2020). chalmers’ (2000) definition only requires that there be a mapping between neural states and states of consciousness. this mapping can, as chalmers himself points out, be “seemingly arbitrary” (chalmers, 2000, p. 23). the mapping can be arbitrary in the sense that there seems to be no epistemically necessary connection between certain neural states and states of consciousness; however, such mappings will be non-arbitrary in the sense that the connection must be general and lawful: states of an ncc, as defined by chalmers, must be sufficient for states of consciousness, given certain general conditions. finding arbitrary mappings between neural states and conscious experiences can help illuminate how consciousness arises in human beings, but it will not lead to a deep understanding of consciousness. chalmers (2000, p. 23) suggests that a non-arbitrary mapping could be established if not only neural states themselves are mapped to states of consciousness, but if also relations between neural states are mapped to relations between conscious states (a recent approach pursuing this idea using category theory is presented in northoff et al., 2019). drawing on such a mapping, one can extrapolate from existing nccs to predict how certain changes in neural activity will change the correlated conscious experience (a similar idea, though not using the notion of an ncc, is explored in churchland, 2005). cccs provide another way of increasing explanatory and predictive power. if it is shown that neural activity associated with consciousness implements certain computations, then this provides an explanation of why that activity is associated with consciousness. in principle, this can then be used to predict the presence (or absence) of consciousness in novel (or controversial) cases. this brings us to the final challenge we will consider here. 2.3 challenge 3: unusual conditions a problem for research on nccs consists in specifying the conditions under which a correlation between neural and conscious states exists. chalmers’ (2000) definition circumvents this problem to some extent, in that his definition only requires wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 7 that the states of an ncc be sufficient for conscious states. it does not require that they be necessary. if, by contrast, a perfect positive correlation between states of an ncc and states of consciousness were required, then states of an ncc would also have to be necessary. however, even a (mere) sufficiency requirement is challenged by unusual cases. perhaps a certain type of neural activity goes along with consciousness in ordinary cases, but does not when brain functioning is affected by lesions? chalmers anticipates this challenge and restricts the required mapping between neural states and conscious states (i) to ordinary functioning brains in ordinary environments, (ii) with unusual inputs, and (iii) under local brain stimulation. although this restriction creates a safe methodological footing for research on nccs, it also comes with a limitation. ideally, we would like to go beyond usual cases, and make inferences about unusual cases. for instance, we would like to know whether an unresponsive patient is non-conscious, or whether they are just unable to report their conscious experience. failing to find neural activity that is sufficient for consciousness cannot rule out that the patient is conscious. this would be different if an ncc provided a necessary condition for consciousness. an extreme case of an unusual condition is exemplified by the possibility of consciousness in the complete absence of both sensory stimulation and observable behaviour. bayne et al. (2020) call such cases “islands of awareness” (ioa). the authors define an ioa as a “conscious stream (or system) whose contents are not shaped by sensory input from either the external world or the body and which cannot be expressed via motor output” (bayne et al., 2020, p. 7). cases such as dreaming or locked-in syndrome provide an approximation to ioas (but not complete ioas), in which some connections to the environment (via sensory input or motor output) remain. genuine ioas could, as the authors point out, be found in ex cranio brains, hemispherotomy, or cerebral organoids (bayne et al., 2020, pp. 6–11). an ex cranio brain is a brain that has been extracted from the cranium and is kept alive outside of the body (post-mortem). an example can be found in a study by vrselja et al. (2019), who extracted pig brains and observed “spontaneous synaptic activity” (vrselja et al., 2019, p. 336) in them. in hemispherotomy (de ribaupierre & delalande, 2008), a cortical hemisphere is more or less disconnected from the rest of the brain, leaving only vascular connectivity intact. a cerebral organoid is a neural structure grown in the lab (lancaster et al., 2013). how would one determine whether conscious experience is sustained in such isolated systems? nccs can provide some guideline, at least if activity that has been found to be sufficient for consciousness in ordinary cases is measured in ioas, as well. however, it would be desirable to have additional criteria for the ascription of consciousness, ideally ones that are also necessary. bayne et al. (2020) suggest using measures of dynamical complexity, which have been found to be reliable proxies for consciousness (and its absence, respectively), across a variety of conditions (demertzi et al., 2019; li & mashour, 2019; schartner et al., 2015; schartner et al., 2017; sitt et al., 2014). wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 8 for this reason, we shall take a closer look at the concept of dynamical complexity in the following section. in particular, we will discuss whether it can be regarded as a computational correlate of consciousness. we will also revisit the challenge of ioas, as this challenge is especially helpful for discussing the relevance of fep to research on nccs and cccs: an application of fep to ioas seems to suggest that ioas must minimise complexity, whereas empirical research suggests that conscious systems (and hence also ioas) display high complexity. we shall argue that this apparent tension is resolved upon a closer look; in fact, free energy minimising systems are ipso facto capable of producing activity with dynamical complexity – and certain kinds of systems are stipulatively defined by their complex dynamics.8 3 computational correlates of consciousness and the free energy principle a computational correlate of consciousness (ccc) is a set of computational properties that are associated with consciousness. such properties are specified by computational models. below we will argue that cccs should be construed as necessary conditions for consciousness. this constitutes a major departure from the concept of an nccs, which is usually regarded as a (minimally) sufficient condition for consciousness. as we point out below (in section 3.2), this conforms to how the notion of a ccc is construed in the literature. furthermore, necessary conditions for consciousness have some advantages that merely sufficient conditions do not have (in particular, they can be used to infer the absence of consciousness, viz., if a necessary condition is not fulfilled in a given case, see fink, 2016). although characterising cccs in terms of “computational properties associated with consciousness” is relatively general and vague, it already shows that the first two challenges identified above can be overcome by cccs. a ccc can be specified by a computational model of global neural dynamics, thereby meeting the first challenge. in addition, cccs are non-arbitrary in the sense that they describe computational functions that are realised by neural activity, which can help explain cognitive capacities associated with consciousness (see section 5 below for more details). the second challenge can therefore also be met by cccs.9 in order to enrich the general characterisation of a ccc just given, we first review an influential paper by cleeremans (2005) on cccs. we then connect some of the ideas presented in that paper to fep and explain how the notion of a ccc would be conceived from the point of view of fep. 8there can still be large differences in the level of complexity displayed by different systems. that is, differences in the level of complexity may still account for differences in consciousness. 9of course, cccs may still seem arbitrary in the sense that they do not directly address the hard problem (chalmers, 1995). a complete computational explanation of consciousness, based on cccs, must therefore also address the meta-problem (chalmers, 2018). wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 9 3.1 cccs and computational principles cleeremans (2005) takes the “contrastive approach to consciousness” as a starting point and argues that the study of the correlates of consciousness must go beyond differences in neural activity, by including differences in computations (and behaviour) that are associated with consciousness (see cleeremans, 2005, p. 84). following mathis & mozer (1996), he refers to these computations as the computational correlates of consciousness. based on this idea, cleeremans suggests two relatively general “computational principles” associated with consciousness. the first principle, “quality of representation,” is defined in terms of a representation’s “stability in time, strength, and distinctiveness” (cleeremans, 2005, pp. 91–92). the stability of a representation corresponds to the amount of time for which its contents are available. the strength of a representation is a property of its vehicles and corresponds to the number of “processing units” of which it consists, as well as of their activation strength (cleeremans, 2005, p. 92). distinctiveness is primarily a content property, namely its specificity. according to cleeremans’ second principle, consciousness involves meta-representations. why are these two computational principles important? one reason is that they refer to properties that can help explain how presumed functions of consciousness are realised. the functions mentioned by cleeremans (2005) include: “flexible, adaptive control over action,” simulating possible actions and their consequences, and “error-correcting functions” (cleeremans, 2005, p. 85). high-quality representations, as characterised by the first principle, are required for control over action and planning, and error correction requires meta-representation (because representations of error are meta-representational, see shea, 2012b). the benefit of these general computational principles is that they make relatively little assumptions about consciousness but can still provide some guidance for specific computational models of consciousness. furthermore, they can suggest why a given type of activity correlates with consciousness: if the neural populations involved can be interpreted as a representation having the properties described by the two principles, and if this helps explain a given type of behaviour (a behavioural correlate of consciousness), then we understand (to some extent), why the measured neural activity is associated with consciousness. the generality of these computational principles also has a disadvantage: the properties described by them may be necessary for many types of conscious experience, but probably not sufficient. in particular, stability, strength, and metarepresentation may be individually necessary, but it is unclear whether any combination of (individually necessary) cccs would be sufficient for consciousness. in other words, a complete explanation of consciousness would require more (cleeremans et al., 2020). at the same time, it could even be questioned whether the properties are necessary. in order to address this question, we show to what extent, and under what assumptions, some computational principles can be derived from first principles. wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 10 3.2 deriving computational principles from first principles here, we shall succinctly summarise the gist of results presented in more detail elsewhere (see especially friston, 2019b; friston, wiese, et al., 2020). this will show that some properties highlighted by the two computational principles given above are already instantiated by a very general class of systems, comprising single-cell organisms, but also more complex systems, such as human beings. this underscores that the computational principles reviewed above are not sufficient for consciousness (unless one is willing to ascribe consciousness to cellular organisms and similarly simple types of system). a general (coarse-grained) form in which a random dynamical system can be described is in terms of langevin dynamics (ao, 2008; seifert, 2012; sekimoto, 1998): ̇𝑥(𝜏 ) = 𝑓 (𝑥, 𝜏 ) + 𝜔. here, the state of the system 𝑥(𝜏 ) at time 𝜏 is constituted by slowly-changing macroscopic variables, which are grounded in microscopic variables with faster dynamics. this is why the result is a stochastic differential equation, comprising not only the state-dependent flow 𝑓, but also a stochastic term 𝜔 (which is a gaussian stochastic term with mean equal to zero and covariance 2γ – so γ is the amplitude of the random fluctuation). the fundamental assumption we shall make here is that things that persist must have measurable properties that remain invariant over a certain amount of time. formally, this means that the system’s long-term behaviour can be described with reference to a global random attractor (or pullback attractor). put differently, there is a (random) set of states to which the system will converge. since the attractor is a random set, it can be described in terms of a density, its non-equilibrium steadystate (ness) density. the ness density is a solution to the fokker-planck equation (for details, see friston, 2019b): ̇𝑝(𝑥, 𝜏 ) = ∇ ⋅ (γ∇ − 𝑓 )𝑝(𝑥, 𝜏 ) the ness density will be a density 𝑝(𝑥, 𝜏 ) that does not change over time, i.e. ̇𝑝(𝑥, 𝜏 ) = ∇ ⋅ (γ∇ − 𝑓 )𝑝(𝑥, 𝜏 ) = 0. a solution to the fokker-planck equation satisfying ̇𝑝(𝑥, 𝜏 ) = 0 is 𝑝(𝑥, 𝜏 ) = exp(−ℑ(𝑥, 𝜏 )), where ℑ(𝑥) is self-information (or surprisal): ℑ(𝑥) = − ln 𝑝(𝑥). wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 11 this means we can express the flow 𝑓 in terms of surprisal (and hence, in terms of the ness density): 𝑓 (𝑥) = (𝑄(𝑥) − γ) ⋅ ∇ℑ(𝑥) = (γ − 𝑄(𝑥)) ⋅ ∇ ln 𝑝(𝑥). here, 𝑄 denotes solenoidal flow. furthermore, following a variational principle of least action, the ensuing dynamics are a gradient flow on surprisal. in other words – on average – the dynamics minimise self-information (surprisal) ℑ(𝑥) (friston, 2019b, pp. 12–13). to recap, anything that exists will have characteristic features that do not change over time. formally, we express this by assuming that the system is a random dynamical system in non-equilibrium steady state (ness). this means that the system is in exchange with its environment and that its dynamics can be described in terms of its ness probability density, i.e., a probability distribution that does not change during the time over which the system is said to exist. such a system has a random attractor, which can be thought of in two ways: first, it can be considered as the trajectory of systemic states as they evolve over time. here, the key aspect is that the system will, after sufficient time, revisit particular regions of its state space, i.e., those regions that constitute the attracting set. the second interpretation considers the random attractor as subtending a probability density 𝑝 over the states that the system will be found in, when sampled at random. the density dynamics are described by the fokker-planck equation. in nonequilibrium steady state, this density does not change with time. the solution (i.e., setting the equation to zero) shows that there is a lawful relationship between the flow of states 𝑓 and the probability density 𝑝. a further assumption we shall make is that the system possesses a markov blanket. a markov blanket can be conceived as a boundary between a system and its environment. it is defined as that set of states that statistically separate a system’s internal states from external states: the system’s internal states 𝜇 are conditionally independent of external states 𝜂, given the markov blanket 𝑏. here, the markov blanket 𝑏 = {𝑠, 𝑎} comprises both sensory states 𝑠 (which are not directly influenced by internal states) and active states 𝑎 (which are not directly influenced by external states). note that internal, external, and blanket states jointly constitute the systemic states 𝑥 = {𝜂, 𝑠, 𝑎, 𝜇}. the term “markov blanket” was originally coined by pearl (1988), in the context of bayesian networks, i.e., directed acyclic graphs. however, the notion can also be applied to directed cyclic graphs (i.e., dependency networks) and undirected graphs (i.e., markov random fields). for examples of applications of the notion, see clark (2017), palacios et al. (2017), parr et al. (2020), pellet & elisseeff (2008).10 10in fact, there is an ongoing discussion about how the markov blanket construct, as used in friston (2013), friston (2019b), friston, wiese, et al. (2020), and related works, differs from the original definition presented in pearl (1988). the original definition only considers synchronic properties, wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 12 here, what matters is that the existence of a markov blanket has subtle, but profound implications. the fact that internal states are conditionally independent of external states, given blanket states, allows us to disentangle the system’s dynamics. in other words, we can write down separate equations of motion for internal states and active states – that only depend on blanket states: 𝑓𝛼(𝜋) = (𝑄𝛼𝛼 − γ𝛼𝛼)∇𝛼ℑ(𝜋) 𝛼 = {𝑎, 𝜇} (autonomous states) 𝜋 = {𝑠, 𝛼} (particular states) here, 𝑓𝛼(𝜋) is the flow of internal and active states 𝛼, to which we refer as autonomous states. note that the flow only depends on particular (non-external) states 𝜋 = {𝑠, 𝛼}. this in turn enables us to express the dynamics of internal states in terms of a probability distribution over external states – encoded by internal states. that is, for any blanket state 𝑏, we can map internal states 𝜇 to a probability distribution 𝑞𝜇(𝜂) over external states, given blanket states. strictly speaking, we can only map expected internal states, because there need not be a unique internal state, associated with a given blanket state. in practice, one must therefore consider averages of internal states (such as average activity of neural populations), which approximate their expected value. if the probability distribution 𝑞𝜇(𝜂) is sufficiently similar to the actual conditional distribution 𝑝(𝜂|𝜋) over external states (where 𝑝 is the ness density), then we can approximate the flow of autonomous states using 𝑞𝜇(𝜂). equivalently, we can say that the gradient flow of autonomous states can be considered as a gradient flow on variational free energy 𝐹 : 𝑓𝛼(𝜋) ≈ (𝑄𝛼𝛼 − γ𝛼𝛼)∇𝛼𝐹 (𝜋), with 𝐹 (𝜋) ≜ 𝐸𝑞[ℑ(𝜂, 𝜋)] − 𝐻[𝑞𝜇(𝜂)] = ℑ(𝜋) + 𝐷[𝑞𝜇(𝜂)||𝑝(𝜂|𝜋)] = 𝐸𝑞[ℑ(𝜋|𝜂)] + 𝐷[𝑞𝜇(𝜂)||𝑝(𝜂)] ≥ ℑ(𝜋). according to this way of expressing 𝑓𝛼(𝜋), the system’s autonomous states 𝛼 must change in such a way that the variational free energy 𝐹 (𝜋) is minimised – where variational free energy is a functional of the probability distribution 𝑞𝜇(𝜂) encoded by internal states. hence, as internal states change, 𝑞𝜇(𝜂) changes, as well; more specifically, it changes in a way that makes it more similar to 𝑝(𝜂|𝜋) (as measured by the kullback-leibler divergence 𝐷[𝑞𝜇(𝜂)||𝑝(𝜂|𝜋)]). changes of internal states therefore go along with changes in the probability distribution encoded by internal states. we can thus map a given internal state whereas its application in the context of fep involves diachronic properties (i.e., it is meant to formalise dependencies in an action-perception loop). for discussion, see rosas et al. (2020); see also friston, costa, et al. (2020); friston et al. (2021). wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 13 to a point on a statistical manifold, and associate changes of internal states with movement on the statistical manifold. this manifold has an information geometry, which we shall call the extrinsic information geometry (because it is the geometry of probabilistic beliefs about external states). the extrinsic information geometry is thus associated with the probability distributions encoded by internal states. now note that the ness density also specifies a probability of internal states. we can regard it as a point on another statistical manifold, which has a geometry we shall call the intrinsic11 information geometry. if we parametrise the density over internal states with time, we can associate the progression of time with a movement on the system’s intrinsic manifold. however, since the ness is constant, movement on the intrinsic manifold will eventually stop (once the system has reached non-equilibrium steady state). in other words, if – as physicists like to say – we “prepare” the internal states in some initial configuration, and then “push” the probabilistic configuration forwards in time, the internal states pass through a series of probabilistic configurations until they reach the point on the intrinsic statistical manifold corresponding to the ness. furthermore, if we assume some of the internal states are exchangeable (e.g., one neuron in a neuronal population can be treated as the same as another from the same population), we equip internal states with a statistical mechanics or thermodynamics; i.e., the initial configuration will, over time, converge to the ness and thermodynamic free energy will fall progressively. in short, the intrinsic geometry underwrites the dynamical complexity and thermodynamics of neuronal populations. conversely, on the extrinsic manifold, the internal states minimise variational free energy, when conditioned on the blanket states. in other words, whenever the blanket states change, the expected internal state changes and there is movement on the extrinsic manifold that – by construction – can be cast as bayesian belief updating. this is because the (expected) internal states encode probabilistic (i.e., bayesian) beliefs about external states. the blanket states here can be construed as the sensory impressions of external states on the markov blanket that contains the internal states. in short, it may be more useful to consider the system from the point of view of the extrinsic information geometry (probabilities encoded by internal states) than from the point of view of the intrinsic information geometry (probabilities of internal states). we can now relate some of the computational principles of conscious processing, as suggested by cleeremans (2005), to the probability density dynamics just sketched. in particular, we shall consider the “strength” of neural representations (the number of “processing units” that constitute the representational vehicle) and meta-representations. firstly, recall that only expected internal states can be said 11it is not just intrinsic in the sense that it is about internal states, it is also intrinsic in the sense that it does not depend on the manifold associated with the probabilities encoded by internal states. the extrinsic information geometry, by contrast, presupposes the existence of the manifold associated with probabilities of internal states. wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 14 to encode a probability distribution over external states. if we want to interpret actual neural activity as encoding a probability distribution, we therefore have to consider population averages. any neural representation will therefore have high “strength”, from the point of view of fep. secondly, note that – under gaussian assumptions about the encoded probabilistic beliefs – minimising variational free energy becomes prediction error minimisation (friston & kiebel, 2009). prediction error signals are meta-representational (shea, 2012b), so the feature of metarepresentation is easily fulfilled by many systems.12 we draw the following lessons from these observations. (i) the activity of a large class of systems conforms to at least some of the computational principles that are identified as computational correlates of consciousness in cleeremans (2005). in particular, this is true for many non-conscious systems (see also reggia et al., 2016, p. 111). (ii) this suggests, more generally, that computational correlates should not be regarded as minimally sufficient conditions for consciousness (in contrast to chalmers’ nccs, as presented in his 2000). in particular, computational correlates need not be associated with the difference between conscious and unconscious processing (pace cleeremans, 2005). (iii) instead, we should expect computational correlates to specify necessary conditions for consciousness.13 for instance, reggia et al. (2019) present a model of working memory and propose that cccs are identified by the “underlying computational mechanisms [that] are critically needed to realize such a model and distinguish it from other contemporary neural networks in general” (reggia et al., 2019, pp. 265–266). this is a useful feature, as not fulfilling a necessary condition can license an inference to the absence of consciousness (fink, 2016). furthermore, existing evidence for the presence of consciousness in controversial cases can be strengthened, if necessary conditions for consciousness are fulfilled. 3.3 computational correlates as necessary, sufficient, or necessary and sufficient conditions for consciousness? we noted, almost in passing, that some of the principles identified as computational principles by cleeremans should be regarded as merely necessary conditions for 12in fact, one could argue that all systems that exist – in the above sense – are meta-representational on this definition. this follows because the gradients of variational free energy can always be expressed as a form of prediction error. one could object that this only entails (meta)representations in an instrumentalist sense, i.e., we can describe internal states of the system as if they were representations (ramstead et al., 2020). in wiese & friston (2021), we argue that at least living organisms that embody a hierarchical generative model are representational systems in a realist sense. however, we are aware that this is a controversial claim (see bruineberg et al., 2018; downey, 2018; hutto, 2018; kirchhoff & robertson, 2018; van es, 2019). 13some of them may be trivial (e.g., single-cell organisms may fulfill them). a challenge is to find computational correlates that specify non-trivial necessary conditions. one strategy is to focus on computational properties associated with functions of consciousness (see cleeremans, 2005, and section 5). wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 15 consciousness. even if one agrees with this remark (since, otherwise, one would have to accept that many very simple systems are conscious), one could argue that computational correlates should be sufficient conditions for consciousness. that is, cleeremans’ principle should then only be regarded as part of a bundle of computational principles that (together) yield sufficient conditions for consciousness. in particular, according to this line of reasoning, the search for computational correlates of consciousness should consist in searching for necessary and sufficient conditions for consciousness. we disagree. we argue that even finding merely necessary conditions for consciousness can be useful and important in the science of consciousness. first of all, necessary conditions for consciousness can complement “minimally sufficient” conditions for consciousness. that is, the search for necessary conditions need not replace the search for minimally sufficient conditions. secondly, whether sufficient or necessary conditions are more useful depends on the domain of application. for instance, if one already knows that a creature is conscious (e.g., a human being with a normally functioning brain in alert wakefulness), minimally sufficient conditions for consciousness in this type of creature will be most informative. however, if it is unclear whether a creature (or machine) is conscious or not, determining whether necessary conditions for consciousness are fulfilled can be immensely informative. if a necessary conditions is not fulfilled, the question (“is it conscious or not?”) can even be settled.14 thirdly, necessary conditions may provide unification. for instance, if cccs are investigated by modelling different cognitive capacities that are associated with consciousness, necessary computational mechanisms of these different capacities can reveal what they have in common. (and it is to be expected that they have something in common if they are all enabled, or facilitated, by the same phenomenon, i.e., consciousness. see birch, 2020, pp. 8–9.) we shall elaborate on this point in section 5 below. 3.4 the challenge of islands of awareness we can now return to a special case of challenge 3 (identified in section 2 above), viz. the possibility of what bayne et al. (2020) call “islands of awareness” (ioa): conscious experiences that are instantiated in the absence of sensory stimulation and observable behaviour. most measures of consciousness rely on some kind of neural response to sensory stimulation, which is precluded in the case of ioas. from the point of view of fep, the challenge may seem even harder. fep has 14in a related vein, fink (2016) discusses four problems with chalmers’ definition of an ncc, one of which is that sufficient conditions do not enable one to rule out that a system is conscious. fink argues: “in order to overcome some of the problems of the chalmers-ncc, it seems that we have to introduce necessity somewhere” (fink, 2016, p. 9). fink suggests identifying a feature bundle 𝔽 that is shared by different neural state tokens that are (each) sufficient for the same phenomenal state type. crucially, 𝔽 must be a necessary condition for consciousness. if the feature bundle consists of neurofunctional and neurocomputational features, then it may also be applicable to many non-human animals, or even machines (fink, 2016, p. 11). wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 16 been developed for systems that sense and act, not for systems that do not receive sensory input and are incapable of producing motor output. in response to this challenge, we will propose to treat ioas in analogy to brains in altered states of consciousness, such as hallucination or sleep and dreaming. as previously argued (friston, wiese, et al., 2020; hobson & friston, 2012; hobson & friston, 2014), neuronal dynamics during sleep minimise the complexity of the brain’s generative model (i.e., by removing redundant model parameters). this can be seen by noting that variational free energy can be expressed as follows: 𝐹 (𝜋) = 𝐸𝑞[ℑ(𝜋|𝜂)] + 𝐷[𝑞𝜇(𝜂)||𝑝(𝜂)]. here, 𝐷[𝑞𝜇(𝜂)||𝑝(𝜂)] describes the statistical complexity of the internal model (“how much do i have to change my mind to accommodate current changes in sensory signals?”), whereas −𝐸𝑞[ℑ(𝜋|𝜂)] describes accuracy (so minimising free energy entails minimising the difference between accuracy and complexity). note that accuracy depends on the particular states 𝜋 (comprising blanket states and internal states). during a period of (partial) disconnection from blanket states, such as dreaming, the variational free energy gradient in the following equation will mainly be driven by the complexity part (which does not depend on blanket states): 𝑓𝛼(𝜋) ≈ (𝑄𝛼𝛼 − γ𝛼𝛼)∇𝛼𝐹 (𝜋). this entails that neural processes can perform the computations required to minimise free energy, even when the coupling with the environment is temporarily suspended. using sleep and dreaming as a model, fep can be applied to ioas and yields the same result: isolated systems will minimise free energy by reducing the complexity of the generative model. this creates a seeming tension between fep and the suggestion by bayne et al. (2020) that high complexity could serve as a proxy for consciousness in ioas. measures used to infer the presence of consciousness in altered states (rohaut et al., 2017; sarasso et al., 2014; sarasso et al., 2015) are measures of entropic complexity such as lempel-ziv (lz) complexity (lempel & ziv, 1976) that provide an upper bound on algorithmic complexity (ruffini, 2017) – where the algorithmic complexity of a signal is defined as the length of the shortest algorithm that produces the signal. this is exactly the same quantity that underwrites variational free energy. in fact, variational free energy minimization in machine learning was predicated on minimizing algorithmic (computational) complexity (mackay, 1995; wallace & dowe, 1999), which is the basis of universal computation (hutter, 2005). consciousness researchers assume complexity has to be large and yet the imperatives for universal computation, in general – and the free energy principle, in particular – say exactly the opposite. this apparent paradox can be resolved easily by noting that free energy is a bound on marginal likelihood (a.k.a. model evidence) in the same way that the wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 17 lz complexity is an upper bound on algorithmic complexity. the logarithm of evidence can always be expressed as accuracy minus complexity. in brief, if consciousness entails belief updating (technically, movement on a statistical manifold) then the imperative for this movement is to minimise the difference between accuracy and complexity. on this view, complexity provides an incomplete metric for any ccc because it fails to account for accuracy, i.e., the marginal likelihood of the sensorium. in the rhetoric of, e.g., integrated information theory (tononi et al., 2016) it is the “matching” that matters not the complexity (where matching is scored by model evidence). in a deeply structured world, an accurate account of the sensorium can only be supplied by a complex (bayesian) belief about the causes of sensations. this means that complexity is always chasing accuracy, so that it will look as if complexity is being maximized. technically, however, it is the difference between complexity and accuracy that is being minimized. sleep and related ioa present an interesting scenario. because there is a transient suspension of exchange with the sensorium (or other parts of the brain) it would appear that there is no accuracy (because there are no inputs to provide an accurate account of). however, these ioa are themselves accountable to a larger context; namely, all the inputs that have been experienced – and will be experienced. on this view, it is then no surprise to see that a good model of a complex world will show a high degree of algorithmic or statistical complexity, even when observed in temporary isolation. this perspective provides fundamental constraints on (the utility of) complexity measures of consciousness. high complexity (as measured by lz complexity) is necessary for consciousness in a complex world, but it is not sufficient. under fep, the minimisation of free energy constitutes a necessary condition for consciousness. this second condition entails the first condition when, and only when, a high degree of complexity is necessary to explain the sensorium. hence, evidence for the presence of consciousness in ioas can be garnered from at least two sources: first, from measurements of apparent algorithmic complexity (such as lz complexity). secondly, from models of ioas that show such systems minimise free energy when their insular nature disappears. 3.5 dynamical complexity as a ccc? the proposed solution to the problem posed by potential ioas is slightly inelegant. while it is pleasing that it does not constitute a problem for fep per se, it may still seem that fep can only make a tiny contribution to understanding such cases (or to “measuring consciousness”). this is because measures of dynamical complexity (such as lz complexity of neural time series) may seem to provide much stronger evidence than any criteria derived from fep: existing empirical results already suggest that dynamical complexity is a correlate of consciousness (demertzi et al., 2019; li & mashour, 2019; schartner et al., 2015; schartner et al., 2017; sitt et wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 18 al., 2014). hence, it is not unreasonable to extrapolate from this to unusual cases (bayne et al., 2020), perhaps even without requiring any additional evidence. let us take a step back and ask: “what is dynamical complexity?” – rather than: “how can it be measured?” in dynamical systems theory, three types of complex behaviour can be distinguished: chaotic itinerancy, heteroclinic cycling, and the switching under multistability (friston, 2019a). all three types of behaviour involve meandering movements between regions that are revisited time and time again, but always following slightly different, unpredictable trajectories. a difference between these types of complex behaviour rests on how they are brought about. either the system switches between basins of attraction (due to noise, as in multistability, or due to the proximity of basins of attraction, as in chaotic itinerancy), or the system’s state space contains a circle of connected saddles with both attracting and repelling manifolds. fep provides an insight into the conditions under which stochastic dynamical systems display behaviour that is characteristic of dynamical complexity (i.e., behaviour such as chaotic itineracy, heteroclinic cycling, or multi-stability and switching). as we saw above, any system with a markov blanket is trying to minimise its self-information, i.e. surprisal. one can show that this brings about a tendency to self-organised criticality and dynamical complexity (friston, 2019a; friston et al., 2012). at the same time, there is a relation between minimising surprisal of blanket states and minimising statistical (and algorithmic) complexity15: the very process by which a system minimises surprisal of blanket states can, equivalently, be described as the process of minimising variational free energy with respect to the internally encoded recognition density. recall that variational free energy can be described as statistical complexity minus accuracy. hence, if a system minimises surprisal of blanket states by minimising variational free energy, it will also tend to keep statistical complexity within bounds. developing a measure of dynamical complexity within the framework of fep, i.e., a measure that invokes a system’s extrinsic information geometry, would help determine under what conditions systems (that minimise variational free energy) display high dynamical complexity, and under what conditions they display low dynamical complexity. furthermore, existing measures of dynamical complexity cannot be said to measure computational correlates of consciousness, from the 15the relation between surprisal minimisation and minimising algorithmic complexity (kolmogorov complexity) is not straightforward. heuristically, it can be shown that maximising the posterior probability of external states, given blanket states, 𝑝(𝜂|𝑏), is equivalent to minimising the algorithmic complexity of blanket states (wallace & dowe, 1999). since minimising variational free energy entails that the recognition density 𝑞𝜇(𝜂) approximates the posterior 𝑝(𝜂|𝑏), there is at least an indirect link to algorithmic complexity. furthermore, in aixi, a formalism for artificial general intelligence, a reinforcement learning agent optimises its hypotheses about the environment by minimising their algorithmic complexity (hutter, 2000, 2005). since active inference (which involves minimising expected free energy) subsumes reinforcement learning as a special case (costa, sajid, et al., 2020), it should in principle be possible to describe an aixi agent as an agent that minimises (expected) variational free energy, under a particular set of priors – this remains an important task for future work. wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 19 point of view of fep. this is because the computations performed by a system that minimises variational free energy are inferences, which correspond to movements on the system’s extrinsic statistical manifold. existing measures of dynamical complexity, however, do not measure complexity in terms of probabilities encoded by neural activity, but in terms of statistical properties of neural activity, that is, properties of the intrinsic statistical manifold. hence, dynamical complexity, as measured by existing approaches, should be associated with nccs (at least from the point of view of fep). interestingly, fep suggests an alternative. an fep-inspired measure of dynamical complexity would most naturally be based on the information length of neurally encoded probability distributions (i.e., bayesian beliefs): instead of considering the statistical properties of a neural time series as such, one would have to consider the probability distributions encoded by this activity – or, equivalently, trajectories on the extrinsic statistical manifold. the information length of a trajectory then simply corresponds to the distance travelled on the manifold (which is measured at any point on the path by the fisher information metric). even if a measure based on information length should turn out to be equivalent to existing measures of dynamical complexity, it would be advantageous, because it could unify existing approaches under a single overarching principle (i.e., the fep). note that, although many existing measures of dynamical complexity are based on lz complexity (demertzi et al., 2019; li & mashour, 2019; sitt et al., 2014), some also use amplitude coalition entropy and synchrony coalition entropy (schartner et al., 2015; schartner et al., 2017), or the spectral exponent of resting eeg (colombo et al., 2019) as a proxy. in fact, there is no obviously “correct” way of measuring dynamical complexity. to the extent that proposed measures of dynamical complexity have empirical validity (as measure of consciousness), they may be regarded as empirically adequate. but it would still be desirable to provide a fundamental justification because this could further support their application to controversial cases. in addition, it could help resolve an uncertainty about whether existing measures of dynamical complexity also track algorithmic complexity. it is well-known that lz complexity, for instance, constitutes an upper bound on algorithmic complexity. however, this entertains the possibility that conscious activity maximises lz complexity (and related measures of dynamical complexity), but minimises algorithmic complexity (this hypothesis has been stated by ruffini, 2017.). finally, developing a measure of dynamical complexity within fep could help to provide a computational explanation of consciousness (or “explanatory correlates” of consciousness, see seth, 2009; seth & edelman, 2009). wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 20 4 from computational correlates to a computational explanation of consciousness measuring neural activity in conscious creatures yields insights into correlates of consciousness. here, we shall probe the idea that cccs, construed as necessary conditions for consciousness, may actually yield a computational explanation of consciousness. do computational explanations of consciousness imply that implementing the right computations is sufficient for being conscious?16 or is it possible that some simulations of conscious systems perform all the computations performed by actual conscious systems, without being conscious? here, we will argue that implementing the right computations is not sufficient for instantiating consciousness. but what would then count as a computational explanation of consciousness? we shall argue that the computations must be implemented by the right kind of system, and that fep puts constraints on what the right kind of system is. to illustrate, consider the following example.17 a digital assistant (say, a sufficiently sophisticated successor of siri or alexa) may be able to pass a version of the turing test, if it implements the right computations. but implementing these computations will not ensure its continued existence. in particular, it may be able to provide sensible answers to any questions it is asked – but it would also continue to exist if it were to respond with nonsensical answers. such a system computes and may represent the world, but it is not committed to the existence of anything that corresponds to the way it represents the world (smith, 2019). this is different for systems like us. our continued existence is systematically related to the way we register (and interact with) the world. similarly, a digital computer may implement all computations underpinning conscious experience. but it must also (have the potential to18) successfully interact with the world by virtue of these computations, in order to be conscious. we elaborate on this point below (section 4.2). before that, we discuss a more fundamental problem associated with computational explanations of consciousness (section 4.1) 4.1 the scope of computational explanations computational explanations seek to explain a target phenomenon characteristic of a system (e.g., a particular type of behaviour) in terms of computations performed by that system. if no computational properties are sufficient for consciousness, then it seems there is something that is left out by computational explanations.19 16this is a special case of what chalmers (2011) calls the thesis of computational sufficiency. 17this example was brought up in a discussion with karen konkoly’s reading group; we are grateful for this suggestion. 18this clause is necessary to accommodate the possibility of islands of awareness. 19note that this is not simply the explanatory gap (levine, 1983). an explanatory gap arises if there is no necessary epistemic connection between the explanans and the explanandum. the absence wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 21 let us consider this problem in a bit more detail. a notorious challenge for computational explanations is that they depend on an account of physical computation, i.e., an account of what it means for a concrete, physical system to perform a given computation (milkowski, 2013; piccinini, 2015). such an account should rule out trivial implementations (sprevak, 2019). for instance, if a computational explanation of consciousness specifies internal state transitions that are also instantiated by a rock or a piece of swiss cheese, then it is doubtful whether describing a system as computing can explain why it is conscious. in other words, if a computational explanation focuses on abstract computational properties (disregarding how they are implemented in the system), then it must be complemented by an account of physical implementation, showing that the same computational properties cannot be instantiated by systems we would not regard as computational. there is, however, no consensus on whether any account of physical computation can avoid trivial implementations (sprevak, 2019), nor on whether computational properties are intrinsic properties of physical systems (dewhurst, 2018; fresco, 2015). for this reason, schweizer (2019) argues that a computational explanation needs to specify more than just internal state transitions that account for abstractly described input-output patterns. rather, a computational explanation should specify inputs and outputs relative to a particular system. for instance, a computational explanation of mental arithmetic, as performed by actual human beings, would have to specify in which sensory modality and in which format the input is presented to the subject (e.g., whether in terms of verbal questions), as well as how the output is produced (verbally, by way of written response, etc.; see schweizer, 2019, p. 294). internal state transitions posited by a computational explanation must then not only be mapped to internal states of the system in question, but these internal states must also be causally connected to the sensory input and behaviour figuring in the description of the input-output patterns that are to be explained. for instance, in a computational explanation of mental arithmetic, as performed by a human being, one would ideally specify neuronal states corresponding to the postulated computational steps, i.e., concrete internal state transitions, which are causally implicated in bringing about behaviour associated with the capacity to perform mental arithmetic (schweizer, 2019, p. 303). this account of computational explanation is applicable to behavioural explananda. it is not obvious how to apply it to the problem of consciousness.20 of a necessary epistemic connection is compatible with the presence of a necessary ontological connection. for instance, if pain experiences are identical with the firing of certain neural populations, then there is a necessary ontological connection, but it may still be reasonable to ask why this type of neural activity is identical with pain experiences. by contrast, if all computations performed by a conscious being can be performed by an unconscious computer, then there is not just an epistemic gap, but also an ontological gap between computational properties and consciousness. 20schweizer makes this point: “the only non-abstract effects that instantiated formalisms are required to preserve are defined in terms of their input/output profiles, and thus internal experiwiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 22 indeed, it seems it poses a dilemma for computational explanations of consciousness. one option is that the computational explanation targets behaviour that is associated with consciousness. but this would always beg the question whether the explanation really explains consciousness – even assuming there are types of behaviour that cannot be performed without consciousness (by human beings at least). on the face of it, such an explanation would only explain behaviour. the other option is that a computational explanation of consciousness targets internal processes in a conscious system (during conscious processing). in fact, given that the first option seemingly only explains behaviour, it seems this is what a computational explanation must account for. however, if one believes that implementing the right kind of computational system is insufficient for instantiating consciousness, then there will be non-conscious systems that still display all the computational properties instantiated by conscious beings. the properties that account for a system’s being conscious must therefore also comprise non-computational properties. can fep bring anything more to the table? fep is not an account of physical computation. however, it can provide constraints that may complement existing accounts of computation. in other words, if the fep applies to a given physical process, then that process has to possess certain properties. an account of physical computation specifies what it means to perform a computation (such as variational free-energy minimisation). furthermore, it is likely that some computing devices will never be conscious, regardless of which computations they perform (e.g., a desktop pc). such systems might simulate consciousness, but will never be conscious. 4.2 distinguishing simulation from instantiation a fundamental constraint provided by fep is that it only applies to active systems that engage with their environment (but see the remarks about islands of awareness in section 3 above). a computer simulation of a system that minimises variational free energy may perform computations that are also instantiated by a conscious agent that interacts with its environment. but if these computations are realised by a piece of hardware that cannot be regarded as an agent (e.g., they cannot move or change external states), then they can at most lead to a simulation of consciousness (islands of awareness would constitute a limit case, i.e., they must at least have the potential to interact with the environment, if the causal connection to the environment via sensory and motor systems is restored; see footnote 26 below). let us unpack this idea. a system in ness with a markov blanket comprises internal, external, and blanket states. as shown in section 3, such a system can also be described as performing certain computations (inferences), by virtue of causal relations between ences, qua actual events, are in principle omitted” (schweizer, 2002, p. 144; see also 2019, footnote 11). wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 23 internal, external, and blanket states. this is different for a mere simulation.21 to be sure, a simulating system can perform the same computations, and it will perform them because of causal relations between its parts. however, it will not perform them because of causal relations between its internal, external, and blanket states. the simulation may instantiate a virtual machine, which may have internal, external, and blanket states, such that its internal states encode a probability distribution over external states, given blanket states. but the internal states of the virtual machine will not be the internal states of the physical machine on which the virtual machine is running. in particular, consider a computer simulation on a von neumann computer that stores the values of a system’s states in its memory registers. there must be some (indirect) causal links between these memory states, but the interaction between them will always be mediated by the cpu. the causal flow in the virtual machine is thus different from the causal flow in the physical machine. put differently, not all virtual machines that perform approximate bayesian inference by encoding a probability distribution over external states, given blanket states, are realised by physical machines with the same markov blanket partition. if we keep this in mind, we can maintain a distinction between simulating and instantiating consciousness, but still retain the hypothesis that the right computational properties are sufficient for consciousness (chalmers, 2011), if they are instantiated by the right kind of system.22 moreover, the “right kind of system” is not simply defined in terms of its physical properties (i.e., it is not about siliconvs. carbon-based systems). rather, the right kind of system is one in which the internal, external, and blanket states that figure in a computational description of the system are also physical states of the system, such that the system’s internal dynamics can equivalently be described (a) with respect to the system’s ness density, or (b) with respect to the probability density (i.e., bayesian beliefs) encoded by internal states.23 one might worry that this constraint still does not suffice to distinguish simulation from duplication. a computer simulating a system that performs variational inference over external states must have physical states encoding the probability distributions over external states that figure in those inferences. define these physical states as internal. there will then be a unique markov blanket for these internal states. hence, the physical system’s internal states will be numerically identical with the internal states that encode the probability distributions figuring in the computational system implemented by the physical system. 21technically, in dynamical systems theory, this kind of “simulation” corresponds to a skew-product or master-slave system, in which the influence of external states on internal states (mediated vicariously by blanket states) was not complemented by an influence of internal states on external states. in other words, the circular causality – that renders the system autonomous – is precluded. 22in other words, our account is compatible with a slightly modified version of the thesis of computational sufficiency. however, it is also compatible with the possibility that computational explanations cannot explain all aspects of consciousness. 23a related argument is presented in kiefer (2020). wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 24 this consideration overlooks that a system’s markov blanket must be specified with respect to the system’s dynamics. these dynamics depend on the structure of the system. for instance, the dynamics of a human organism are quite different from the dynamics of a digital computer, even if both have internal states that encode the same probability distributions.24 moreover, a computer can simulate a vast number of different types of system (with different markov blankets), but the computer’s markov blanket will not change when it simulates different systems (or at least it will not change in such a way as to match the markov blanket of the simulated system). this means the computer’s continued existence does not depend on the computations it performs in simulating another system. conversely, digital computers with different causal topologies (and different markov blankets) may be able to simulate the same virtual machine (gamez, 2014, p. 180). 5 computational explanations of consciousness and minimal unifying models so far, we have argued that fep can provide the basis for a computational explanation of consciousness. a computational explanation of consciousness must specify computations that are sufficient for at least some features of consciousness, under the assumption that the computations are performed by the right kind of system. if the computations can be described as inferences over external states, then fep tells us what the right kind of system is: a physical system with internal states that encode the probability distributions figuring in the computational explanation, such that the physical system’s internal dynamics can equivalently be described a) with respect to the system’s ness density, or b) with respect to the probability distribution encoded by internal states. if correct, this only shows that fep can provide a framework for computational explanations of consciousness (similarly to the proposal by hohwy & seth, 2020). it does not directly provide computational explanations. how should one expect to get to an explanation (or a theory) of consciousness, starting from a framework that itself is not a theory of consciousness (because it applies to basically everything)? the account on offer here is that a computational explanation of consciousness should explain teleological functions of consciousness, i.e., cognitive capacities associated with consciousness (see also cleeremans, 2005). explanations of cognitive capacities cannot be provided by fep itself, but by process theories such as active inference (which are theories of how minimising variational free energy is realised). the fep can fulfil a unificatory role: using fep as a framework, one can 24a special case of the possibility described in the paragraph above is a computer simulation of an island of awareness. in a mere simulation, blanket states are virtual, not just (temporally) disconnected from internal states; a real artificial island of awareness would have to be able to continue to function (i.e., to continue minimising variational free energy) if it were connected to physical sensors and actuators – not just virtual ones. wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 25 formulate and test hypotheses about necessary computational mechanisms that may underpin different cognitive capacities,25 or different measures of consciousness. this could also be used as a defense against the following charge: operationalisations of consciousness employed by consciousness researchers are so heterogeneous that they do not serve to unequivocally pick out a unique phenomenon (irvine, 2013, 2017). hence, it has been argued that the concept of consciousness should be eliminated from scientific discourse (irvine, 2012; irvine & sprevak, 2020). as a unifying framework, fep (and process theories such as active inference) could enable insights into what the phenomena picked out by different measures of consciousness have in common (see also wiese, 2018). we shall unpack this idea in what follows. wiese (2020) argues that theoretical work in consciousness research should not focus on creating new theories of consciousness, but on creating a minimal unifying model, which embodies assumptions that existing theories have in common, and are thus more likely to be true than any entire theory. in doing this, only necessary conditions for consciousness are picked out (instead of sufficient conditions) – this is the main sense in which such a model will be minimal (see also metzinger, 2020). this dovetails with the results presented here, in that cccs should, from the point of view of fep, be regarded only as necessary properties of consciousness, not as (minimally) sufficient properties (as suggested by chalmers’ ncc notion). minimising variational free energy could count as a minimal unifying model: variational free energy must be minimised by every self-organising system that persists, and hence also by any conscious system. minimising variational free energy is thus a necessary condition, trivially presupposed by any theory of consciousness that does not explicitly reject fep. unfortunately, this necessary condition is not really informative, because it is trivial. however, recall that slightly more informative constraints directly follow from fep. cccs should be specified in terms of probabilities encoded by internal states, not in terms of probabilities of internal states. conscious (and unconscious) systems must minimise the difference between accuracy and complexity – and systems with (transiently absent) connection to sensory inputs and motor outputs will tend to minimise complexity. apart from these implications, more informative minimal unifying models can be developed under the fep by taking results from consciousness research into account. for instance, a promising approach is to focus on cognitive capacities associated with consciousness, such as trace conditioning, complex types of learning (birch, 2020; ginsburg & jablonka, 2019; kanai et al., 2019), or perhaps working memory (reggia et al., 2019). it is unlikely that any particular cognitive capacity by itself is sufficient for consciousness (birch, 2020; wiese, 2020). instead, the 25a related idea, put forward by ginsburg & jablonka (2019), is that there is a single learning capacitiy that underpins various characteristic features of consciousness. wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 26 hypothesis is that consciousness enables a number of different cognitive abilities. furthermore, if consciousness is a natural kind (shea, 2012a; shea & bayne, 2010), then it is to be expected that there is a single underlying mechanism for these cognitive abilities. whether consciousness is indeed a natural kind is an empirical question: finding a positive correlation between different cognitive capacities associated with consciousness would support the assumption. fep-based models can contribute to this approach from a theoretical perspective. for instance, active inference models (friston et al., 2017) could establish that minimising (expected) variational free energy in deep models is sufficient for a broad range of cognitive capacities associated with consciousness (at least if certain further, to-be-specified conditions are fulfilled; see wiese, 2018). more specifically, assume there is a cluster of consciousness-linked cognitive capacities 𝑐1, 𝑐2, 𝑐3, … strong support for the natural-kind hypothesis would be provided if a minimal model of 𝑐1 were ipso facto a minimal model of 𝑐2 (and of 𝑐3, …). this is unlikely, but it may turn out that minimal models of 𝑐1, 𝑐2, 𝑐3, …, respectively, share at least some core structural features (wiese, 2020) that play a central role in accounting for these different cognitive abilities. a model with these core structural features would then point to a common mechanism underlying these capacities. the computational description afforded by such a unifying model would provide an essential part of a computational explanation of consciousness. the explanatory “heavy lifting” would not be done by fep itself, but by specific models of the respective cognitive capacities (schlicht & dołega, 2021). however, finding common core structural features of these models (fink et al., 2021; song, 2021) will be facilitated by (or will even be conditional on) expressing the models within a single formal framework, as provided by fep (and active inference). 6 conclusion how can research on neural correlates of consciousness (nccs) contribute to an explanation of consciousness, and what role could the free energy principle (fep) possibly play in this endeavour? we have suggested that research on neural correlates should be complemented by research on computational correlates of consciousness (cccs), in order to yield a computational explanation of consciousness. moreover, according to fep, nccs must be defined in terms of neural dynamics, not neural states, and cccs should be defined in terms of probabilities encoded by neural states. consequently, a computational explanation of consciousness requires more than a formal description of neural dynamics. it must also specify computations implemented by these dynamics. a key contribution by fep is that it provides additional constraints on what it means to be, as opposed to merely simulate a conscious system. put differently, the possibility of a computational explanation of consciousness is compatible with the assumption that some systems instantiate all computational properties associated with consciousness, without being conscious. wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 27 there are many open problems and questions we have barely mentioned, or even outright ignored, such as the hard problem (chalmers, 1995), the meta problem (chalmers, 2018), or the computational explanatory gap (reggia et al., 2017; reggia et al., 2014). a computational explanation of consciousness can only be achieved if these problems are addressed as well. furthermore, the possibility of a computational explanation of consciousness also raises the possibility of machine consciousness (reggia, 2013), as well as ethical concerns that should be addressed (agarwal & edelman, 2020; metzinger, 2013, 2017, 2021). however, our goal has been more modest. all we hope to have shown is that a computational explanation of consciousness is possible, and that the free energy principle can make a significant contribution to this project. acknowledgments we are extremely grateful to all attendants of a virtual theoretical neurobiology meeting at the wellcome centre for human neuroimaging, at which an early version of this paper was discussed. ww is grateful to joe dewhurst and mark sprevak for correspondence on issues related to ideas presented in this paper, as well as to audiences at the tokyo consciousness club, the active inference lab, the mathematical consciousness science seminar, and karen konkoly’s reading group. we also thank two anonymous reviewers and ying-tung lin and sascha fink for their critical and constructive comments. this research was supported by a wellcome trust principal research fellowship (kf; ref: 088130/z/09/z). references agarwal, a., & edelman, s. (2020). functionally effective conscious ai without suffering. journal of artificial intelligence and consciousness, 7(01), 39–50. https://doi.org/10.1142/s2705078520300030 ao, p. (2008). emerging of stochastic dynamical equalities and steady state thermodynamics. communications in theoretical physics, 49(5), 1073–1090. https://doi.org/10.1088/0253-6102/49/5/01 aru, j., bachmann, t., singer, w., & melloni, l. (2012). distilling the neural correlates of consciousness. neuroscience & biobehavioral reviews, 36(2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.12.003 atkinson, a. p., thomas, m. s. c., & cleeremans, a. (2000). consciousness: mapping the theoretical landscape. trends in cognitive sciences, 4(10), 372–382. https://doi.org/10.1016/s1364-6613(00)01533-3 bayne, t., seth, a. k., & massimini, m. (2020). are there islands of awareness? trends in neurosciences, 43(1), 6–16. https://doi.org/10.1016/j.tins.2019.11.003 birch, j. (2020). the search for invertebrate consciousness. noûs, 1–21. https://doi.org/10.1111/nous.12351 block, n. (2005). two neural correlates of consciousness. trends in cognitive sciences, 9(2), 46‐52. https://doi.org/10.1016/j.tics.2004.12.006 bruineberg, j., kiverstein, j., & rietveld, e. (2018). the anticipating brain is not a scientist: the free-energy principle from an ecological-enactive perspective. synthese, 195, 2417–2444. https://doi.org/10.1007/s11229-016-1239-1 chalmers, d. j. (1995). facing up to the problem of consciousness. journal of consciousness studies, 2(3), 200–219. https://www.ingentaconnect.com/content/imp/jcs/1995/00000002/00000003/653 chalmers, d. j. (2000). what is a neural correlate of consciousness? in t. metzinger (ed.), neural correlates of consciousness: empirical and conceptual questions (pp. 17–40). mit press. chalmers, d. j. (2011). a computational foundation for the study of cognition. journal of cognitive science, 12, 323–357. chalmers, d. j. (2018). the meta-problem of consciousness. journal of consciousness studies, 25(9-10), 6–61. https://www.ingentaconnect.com/content/imp/jcs/2018/00000025/f0020009/art00001 churchland, p. (2005). chimerical colors: some phenomenological predictions from cognitive neuroscience. philosophical psychology, 18(5), 527–560. https://doi.org/10.1080/09515080500264115 wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1142/s2705078520300030%20 https://doi.org/10.1088/0253-6102/49/5/01 https://doi.org/10.1016/j.neubiorev.2011.12.003 https://doi.org/10.1016/s1364-6613(00)01533-3 https://doi.org/10.1016/j.tins.2019.11.003 https://doi.org/10.1111/nous.12351 https://doi.org/10.1016/j.tics.2004.12.006 https://doi.org/10.1007/s11229-016-1239-1 https://www.ingentaconnect.com/content/imp/jcs/1995/00000002/00000003/653 https://www.ingentaconnect.com/content/imp/jcs/2018/00000025/f0020009/art00001 https://doi.org/10.1080/09515080500264115 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 28 clark, a. (2017). how to knit your own markov blanket. in t. k. metzinger & w. wiese (eds.), philosophy and predictive processing. mind group. https://doi.org/10.15502/9783958573031 cleeremans, a. (2005). computational correlates of consciousness. progress in brain research, 150, 81–98. https://doi.org/10.1016/s0079-6123(05)50007-4 cleeremans, a., achoui, d., beauny, a., keuninckx, l., martin, j.-r., muñoz-moldes, s., vuillaume, l., & heering, a. de. (2020). learning to be conscious. trends in cognitive sciences, 24(2), 112–123. https://doi.org/10.1016/j.tics.2019.11.011 colombo, m. a., napolitani, m., boly, m., gosseries, o., casarotto, s., rosanova, m., brichant, j. f., boveroux, p., rex, s., laureys, s., massimini, m., chieregato, a., & sarasso, s. (2019). the spectral exponent of the resting eeg indexes the presence of consciousness during unresponsiveness induced by propofol, xenon, and ketamine. neuroimage, 189, 631–644. https://doi.org/10.1016/j.neuroimage.2019.01.024 costa, l. d., parr, t., sajid, n., veselic, s., neacsu, v., & friston, k. (2020). active inference on discrete state-spaces: a synthesis. https://arxiv.org/abs/2001.07203 costa, l. d., sajid, n., parr, t., friston, k., & smith, r. (2020). the relationship between dynamic programming and active inference: the discrete, finite-horizon case. https://arxiv.org/abs/2009.08111 crick, f., & koch, c. (1990). towards a neurobiological theory of consciousness. seminars in the neurosciences, 2, 263–275. https://doi.org/https://resolver.caltech.edu/caltechauthors:20130816-103136937 de ribaupierre, s., & delalande, o. (2008). hemispherotomy and other disconnective techniques. neurosurgical focus, 25(3), e14. https://doi.org/10.3171/foc/2008/25/9/e14 demertzi, a., tagliazucchi, e., dehaene, s., deco, g., barttfeld, p., raimondo, f., martial, c., fernández-espejo, d., rohaut, b., voss, h., & others. (2019). human consciousness is supported by dynamic complex patterns of brain signal coordination. science advances, 5(2), eaat7603. https://doi.org/10.1126/sciadv.aat7603 dewhurst, j. (2018). computing mechanisms without proper functions. minds and machines, 28(3), 569–588. https://doi.org/10.1007/s11023-018-9474-5 downey, a. (2018). predictive processing and the representation wars: a victory for the eliminativist (via fictionalism). synthese, 195(12), 5115–5139. https://doi.org/10.1007/s11229-017-1442-8 fekete, t., & edelman, s. (2011). towards a computational theory of experience. consciousness and cognition, 20(3), 807–827. https://doi.org/10.1016/j.concog.2011.02.010 fink, s. b. (2016). a deeper look at the “neural correlate of consciousness.” frontiers in psychology, 7. https://doi.org/10.3389/fpsyg.2016.01044 fink, s. b., lyre, h., & kob, l. (2021). a structural constraint on neural correlates of consciousness. philosophy and the mind sciences, 2, 7. https://doi.org/10.33735/phimisci.2021.79 fresco, n. (2015). objective computation versus subjective computation. erkenntnis, 80(5), 1031–1053. https://doi.org/10.1007/s10670-014-9696-8 friston, j. k. (2019a). complexity and computation in the brain. the knowns and the known unknowns. in w. singer, t. j. sejnowski, & p. rakic (eds.), the neocortex (pp. 269–291). mit press. friston, k. (2010). the free-energy principle: a unified brain theory? nature reviews neuroscience, 11(2), 127–138. https://doi.org/10.1038/nrn2787 friston, k. (2013). life as we know it. journal of the royal society interface, 10(86). https://doi.org/10.1098/rsif.2013.0475 friston, k. (2019b). a free energy principle for a particular physics. https://arxiv.org/abs/1906.10184 friston, k., breakspear, m., & deco, g. (2012). perception and self-organized instability. frontiers in computational neuroscience, 6, 44. https://doi.org/10.3389/fncom.2012.00044 friston, k., costa, l. d., & parr, t. (2020). some interesting observations on the free energy principle. https://arxiv.org/abs/2002.04501 friston, k., fitzgerald, t., rigoli, f., schwartenbeck, p., & pezzulo, g. (2017). active inference: a process theory. neural computation, 29(1), 1–49. https://doi.org/10.1162/neco_a_00912 friston, k. j., fagerholm, e. d., zarghami, t. s., parr, t., hipólito, i., magrou, l., & razi, a. (2021). parcels and particles: markov blankets in the brain. in network neuroscience (no. 1; vol. 5, pp. 211–251). https://doi.org/10.1162/netn_a_00175 friston, k. j., wiese, w., & hobson, j. a. (2020). sentience and the origins of consciousness: from cartesian duality to markovian monism. entropy, 22(5), 516. https://doi.org/10.3390/e22050516 friston, k., & kiebel, s. (2009). predictive coding under the free-energy principle. philosophical transactions of the royal society b: biological sciences, 364(1521), 1211–1221. https://doi.org/10.1098/rstb.2008.0300 gamez, d. (2014). can we prove that there are computational correlates of consciousness in the brain? journal of cognitive science, 15(2), 149–186. https://doi.org/10.17791/jcs.2014.15.2.149 wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.15502/9783958573031 https://doi.org/10.1016/s0079-6123(05)50007-4 https://doi.org/10.1016/j.tics.2019.11.011 https://doi.org/10.1016/j.neuroimage.2019.01.024 https://arxiv.org/abs/2001.07203 https://arxiv.org/abs/2009.08111 https://resolver.caltech.edu/caltechauthors:20130816-103136937 https://doi.org/10.3171/foc/2008/25/9/e14 https://doi.org/10.1126/sciadv.aat7603 https://doi.org/10.1007/s11023-018-9474-5 https://doi.org/10.1007/s11229-017-1442-8 https://doi.org/10.1016/j.concog.2011.02.010 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.33735/phimisci.2021.79 https://doi.org/10.1007/s10670-014-9696-8%20 https://doi.org/10.1038/nrn2787 https://doi.org/10.1098/rsif.2013.0475 https://arxiv.org/abs/1906.10184 https://doi.org/10.3389/fncom.2012.00044 https://arxiv.org/abs/2002.04501 https://doi.org/10.1162/neco_a_00912 https://doi.org/10.1162/netn_a_00175 https://doi.org/10.3390/e22050516 https://doi.org/10.1098/rstb.2008.0300 https://doi.org/10.17791/jcs.2014.15.2.149 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 29 ginsburg, s., & jablonka, e. (2019). the evolution of the sensitive soul: learning and the origins of consciousness. mit press. he, b. j., zempel, j. m., snyder, a. z., & raichle, m. e. (2010). the temporal structures and functional significance of scale-free brain activity. neuron, 66(3), 353–369. https://doi.org/10.1016/j.neuron.2010.04.020 hobson, j. a., & friston, k. j. (2012). waking and dreaming consciousness: neurobiological and functional considerations. progress in neurobiology, 98(1), 82–98. https://doi.org/10.1016/j.pneurobio.2012.05.003 hobson, j. a., & friston, k. j. (2014). consciousness, dreams, and inference: the cartesian theatre revisited. journal of consciousness studies, 21(1-2), 6–32. https://www.ingentaconnect.com/content/imp/jcs/2014/00000021/f0020001/art00001 hohwy, j. (2009). the neural correlates of consciousness: new experimental approaches needed? consciousness and cognition, 18(2), 428–438. https://doi.org/10.1016/j.concog.2009.02.006 hohwy, j. (2016). the self-evidencing brain. noûs, 50(2), 259–285. https://doi.org/10.1111/nous.12062 hohwy, j. (2020). self-supervision, normativity and the free energy principle. synthese, 1–25. https://doi.org/10.1007/s11229-020-02622-2 hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 huang, z., zhang, j., wu, j., mashour, g. a., & hudetz, a. g. (2020). temporal circuit of macroscale dynamic brain activity supports human consciousness. science advances, 6(11). https://doi.org/10.1126/sciadv.aaz0087 hutter, m. (2000). a theory of universal artificial intelligence based on algorithmic complexity. corr, cs.ai/0004001. https://arxiv.org/abs/cs/0004001 hutter, m. (2005). universal artificial intelligence: sequential decisions based on algorithmic probability. springer science & business media. hutto, d. d. (2018). getting into predictive processing’s great guessing game: bootstrap heaven or hell? synthese, 195(6), 2445–2458. https://doi.org/10.1007/s11229-017-1385-0 irvine, e. (2012). consciousness as a scientific concept: a philosophy of science perspective. springer science & business media. irvine, e. (2013). measures of consciousness. philosophy compass, 8(3), 285–297. https://doi.org/10.1111/phc3.12016 irvine, e. (2017). explaining what? topoi, 36(1), 95–106. https://doi.org/10.1007/s11245-014-9273-4 irvine, e., & sprevak, m. (2020). eliminativism about consciousness. in u. kriegel (ed.), the oxford handbook of the philosophy of consciousness (pp. 348–370). oxford university press. kanai, r., chang, a., yu, y., magrans de abril, i., biehl, m., & guttenberg, n. (2019). information generation as a functional basis of consciousness. neuroscience of consciousness, 2019(1). https://doi.org/10.1093/nc/niz016 kiefer, a. b. (2020). psychophysical identity and free energy. journal of the royal society interface, 17(169), 20200370. https://doi.org/10.1098/rsif.2020.0370 kirchhoff, m. d., & robertson, i. (2018). enactivism and predictive processing: a non-representational view. philosophical explorations, 21(2), 264–281. https://doi.org/10.1080/13869795.2018.1477983 koch, c., massimini, m., boly, m., & tononi, g. (2016). neural correlates of consciousness: progress and problems. nature reviews neuroscience, 17(5), 307–321. https://doi.org/10.1038/nrn.2016.22 lancaster, m. a., renner, m., martin, c.-a., wenzel, d., bicknell, l. s., hurles, m. e., homfray, t., penninger, j. m., jackson, a. p., & knoblich, j. a. (2013). cerebral organoids model human brain development and microcephaly. nature, 501(7467), 373–379. https://doi.org/10.1038/nature12517 lempel, a., & ziv, j. (1976). on the complexity of finite sequences. ieee transactions on information theory, 22(1), 75–81. levine, j. (1983). materialism and qualia: the explanatory gap. pacific philosophical quarterly, 64(9), 354–361. li, d., & mashour, g. a. (2019). cortical dynamics during psychedelic and anesthetized states induced by ketamine. neuroimage, 196, 32–40. https://doi.org/https://doi.org/10.1016/j.neuroimage.2019.03.076 luppi, a. i., craig, m. m., pappas, i., finoia, p., williams, g. b., allanson, j., pickard, j. d., owen, a. m., naci, l., menon, d. k., & stamatakis, e. a. (2019). consciousness-specific dynamic interactions of brain integration and functional diversity. nature communications, 10(1), 4616. https://doi.org/10.1038/s41467-019-12658-9 mackay, d. j. (1995). free energy minimisation algorithm for decoding and cryptanalysis. electronics letters, 31(6), 446–447. https://doi.org/10.1049/el:19950331 mashour, g. a. (2018). the controversial correlates of consciousness. science, 360(6388), 493–494. https://doi.org/10.1126/science.aat5616 mashour, g. a., roelfsema, p., changeux, j. p., & dehaene, s. (2020). conscious processing and the global neuronal workspace hypothesis. neuron, 105(5), 776–798. https://doi.org/10.1016/j.neuron.2020.01.026 wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neuron.2010.04.020 https://doi.org/10.1016/j.pneurobio.2012.05.003 https://www.ingentaconnect.com/content/imp/jcs/2014/00000021/f0020001/art00001 https://doi.org/10.1016/j.concog.2009.02.006 https://doi.org/10.1111/nous.12062 https://doi.org/10.1007/s11229-020-02622-2 https://doi.org/10.33735/phimisci.2020.ii.64 https://doi.org/10.1126/sciadv.aaz0087 https://arxiv.org/abs/cs/0004001 https://doi.org/10.1007/s11229-017-1385-0 https://doi.org/10.1111/phc3.12016 https://doi.org/10.1007/s11245-014-9273-4 https://doi.org/10.1093/nc/niz016 https://doi.org/10.1098/rsif.2020.0370 https://doi.org/10.1080/13869795.2018.1477983 https://doi.org/10.1038/nrn.2016.22 https://doi.org/10.1038/nature12517 https://doi.org/10.1016/j.neuroimage.2019.03.076 https://doi.org/10.1038/s41467-019-12658-9 https://doi.org/10.1049/el:19950331 https://doi.org/10.1126/science.aat5616 https://doi.org/10.1016/j.neuron.2020.01.026 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 30 mathis, d., & mozer, m. c. (1996). conscious and unconscious perception: a computational theory. in g. cottrell (ed.), proceedings of the eighteenth annual conference of the cognitive science society (pp. 324–328). erlbaum. metzinger, t. (2013). two principles for robot ethics. in e. hilgendorf & j.-p. günther (eds.), robotik und gesetzgebung. beiträge der tagung vom 7. bis 9. mai 2012 in bielefeld. (pp. 263–302). nomos. metzinger, t. (2017). suffering, the cognitive scotoma. in k. almqvist & a. haag (eds.), the return of consciousness. a new science on old questions (pp. 237–262). axel and margaret axson johnson foundation. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 metzinger, t. (2021). artificial suffering: an argument for a global moratorium on synthetic phenomenology. journal of artificial intelligence and consciousness, 1–24. https://doi.org/10.1142/s270507852150003x milkowski, m. (2013). explaining the computational mind. mit press. miller, s. m. (2007). on the correlation/constitution distinction problem (and other hard problems) in the scientific study of consciousness. acta neuropsychiatrica, 19(3), 159–176. https://doi.org/10.1111/j.1601-5215.2007.00207.x noë, a., & thompson, e. (2004). are there neural correlates of consciousness? journal of consciousness studies, 11(1), 3–28. https://www.ingentaconnect.com/content/imp/jcs/2004/00000011/00000001/1400 northoff, g., tsuchiya, n., & saigo, h. (2019). mathematics and the brain: a category theoretical approach to go beyond the neural correlates of consciousness. entropy, 21(12), 1234. https://doi.org/10.3390/e21121234 palacios, e. r., razi, a., parr, t., kirchhoff, m., & friston, k. (2017). biological self-organisation and markov blankets. biorxiv. https://doi.org/10.1101/227181 parr, t., da costa, l., & friston, k. (2020). markov blankets, information geometry and stochastic thermodynamics. philosophical transactions of the royal society a: mathematical, physical and engineering sciences, 378(2164), 20190159. https://doi.org/10.1098/rsta.2019.0159 pearl, j. (1988). probabilistic reasoning in intelligent systems: networks of plausible inference. morgan kaufmann publishers. pellet, j. p., & elisseeff, a. (2008). using markov blankets for causal structure learning. journal of machine learning research, 9, 1295–1342. http://jmlr.org/papers/v9/pellet08a.html piccinini, g. (2015). physical computation: a mechanistic account. oxford university press. ramstead, m. j. d., friston, k. j., & hipólito, i. (2020). is the free-energy principle a formal theory of semantics? from variational density dynamics to neural and phenotypic representations. entropy, 22(8), 889. https://doi.org/10.3390/e22080889 reggia, j. a. (2013). the rise of machine consciousness: studying consciousness with computational models. neural networks, 44, 112–131. https://doi.org/10.1016/j.neunet.2013.03.011 reggia, j. a., katz, g., & davis, g. p. (2019). modeling working memory to identify computational correlates of consciousness. open philosophy, 2, 252–269. https://doi.org/10.1515/opphil-2019-0022 reggia, j. a., katz, g., & huang, d.-w. (2016). what are the computational correlates of consciousness? biologically inspired cognitive architectures, 17, 101–113. https://doi.org/10.1016/j.bica.2016.07.009 reggia, j. a., monner, d., & sylvester, j. (2014). the computational explanatory gap. journal of consciousness studies, 21(9-10), 153–178. https://www.ingentaconnect.com/content/imp/jcs/2014/00000021/f0020009/art00007 reggia, j., huang, d.-w., & katz, g. (2017). exploring the computational explanatory gap. philosophies, 2(1), 5. https://doi.org/10.3390/philosophies2010005 rohaut, b., raimondo, f., galanaud, d., valente, m., sitt, j. d., & naccache, l. (2017). probing consciousness in a sensory-disconnected paralyzed patient. brain injury, 31(10), 1398–1403. https://doi.org/10.1080/02699052.2017.1327673 rosas, f. e., mediano, p. a. m., biehl, m., chandaria, s., & polani, d. (2020). causal blankets: theory and algorithmic framework. https://arxiv.org/abs/2008.12568 ruffini, g. (2017). an algorithmic information theory of consciousness. neuroscience of consciousness, 3(1). https://doi.org/10.1093/nc/nix019 sarasso, s., boly, m., napolitani, m., gosseries, o., charland-verville, v., casarotto, s., rosanova, m., casali, a. g., brichant, j.-f., boveroux, p., & others. (2015). consciousness and complexity during unresponsiveness induced by propofol, xenon, and ketamine. current biology, 25(23), 3099–3105. https://doi.org/10.1016/j.cub.2015.10.014 sarasso, s., rosanova, m., casali, a. g., casarotto, s., fecchio, m., boly, m., gosseries, o., tononi, g., laureys, s., & massimini, m. (2014). quantifying cortical eeg responses to tms in (un)consciousness. clinical eeg and neuroscience, 45(1), 40–49. https://doi.org/10.1177/1550059413513723 schartner, m. m., pigorini, a., gibbs, s. a., arnulfo, g., sarasso, s., barnett, l., nobili, l., massimini, m., seth, a. k., & barrett, a. b. (2017). global and local complexity of intracranial eeg decreases during nrem sleep. neuroscience of consciousness, 2017(1), niw022. https://doi.org/10.1093/nc/niw022 wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://doi.org/10.1142/s270507852150003x https://doi.org/10.1111/j.1601-5215.2007.00207.x https://www.ingentaconnect.com/content/imp/jcs/2004/00000011/00000001/1400 https://doi.org/10.3390/e21121234 https://doi.org/10.1101/227181 https://doi.org/10.1098/rsta.2019.0159 http://jmlr.org/papers/v9/pellet08a.html https://doi.org/10.3390/e22080889 https://doi.org/10.1016/j.neunet.2013.03.011 https://doi.org/10.1515/opphil-2019-0022 https://doi.org/10.1016/j.bica.2016.07.009 https://www.ingentaconnect.com/content/imp/jcs/2014/00000021/f0020009/art00007 https://doi.org/10.3390/philosophies2010005 https://doi.org/10.1080/02699052.2017.1327673 https://arxiv.org/abs/2008.12568 https://doi.org/10.1093/nc/nix019 https://doi.org/10.1016/j.cub.2015.10.014 https://doi.org/10.1177/1550059413513723 https://doi.org/10.1093/nc/niw022 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the neural correlates of consciousness under the free energy principle 31 schartner, m., seth, a., noirhomme, q., boly, m., bruno, m.-a., laureys, s., & barrett, a. (2015). complexity of multi-dimensional spontaneous eeg decreases during propofol induced general anaesthesia. plos one, 10(8), e0133532. https://doi.org/10.1371/journal.pone.0133532 schlicht, t., & dołega, k. (2021). you can’t always get what you want: predictive processing and consciousness. philosophy and the mind sciences, 2, 8. https://doi.org/10.33735/phimisci.2021.80 schweizer, p. (2002). consciousness and computation. minds and machines, 12(1), 143–144. https://doi.org/10.1023/a:1013741414324 schweizer, p. (2019). triviality arguments reconsidered. minds and machines, 29, 287–308. https://doi.org/10.1007/s11023-019-09501-x seifert, u. (2012). stochastic thermodynamics, fluctuation theorems and molecular machines. reports on progress in physics, 75(12), 126001. https://doi.org/10.1088/0034-4885/75/12/126001 sekimoto, k. (1998). langevin equation and thermodynamics. progress of theoretical physics supplement, 130, 17–27. https://doi.org/10.1143/ptps.130.17 seth, a. k. (2009). explanatory correlates of consciousness: theoretical and computational challenges. cognitive computation, 1(1), 50–63. https://doi.org/10.1007/s12559-009-9007-x seth, a. k., & edelman, g. m. (2009). consciousness and complexity. in r. a. meyers (ed.), springer encyclopedia of complexity and systems science (pp. 1424–1443). springer. shea, n. (2012a). methodological encounters with the phenomenal kind. philosophy and phenomenological research, 84(2), 307–344. https://doi.org/10.1111/j.1933-1592.2010.00483.x shea, n. (2012b). reward prediction error signals are meta-representational. noûs, 48(2), 314–341. https://doi.org/10.1111/j.1468-0068.2012.00863.x shea, n., & bayne, t. (2010). the vegetative state and the science of consciousness. the british journal for the philosophy of science, 61(3), 459–484. https://doi.org/10.1093/bjps/axp046 sitt, j. d., king, j. r., el karoui, i., rohaut, b., faugeras, f., gramfort, a., cohen, l., sigman, m., dehaene, s., & naccache, l. (2014). large scale screening of neural signatures of consciousness in patients in a vegetative or minimally conscious state. brain, 137(pt 8), 2258–2270. https://doi.org/10.1093/brain/awu141 smith, b. c. (2019). the promise of artificial intelligence: reckoning and judgment. mit press. song, c. (2021). structural basis of consciousness. philosophy and the mind sciences, 2, 6. https://doi.org/10.33735/phimisci.2021.75 sprevak, m. (2019). triviality arguments about computational implementation. in m. sprevak & m. colombo (eds.), routledge handbook of the computational mind (pp. 175–191). routledge. sprevak, m. (2020). two kinds of information processing in cognition. review of philosophy and psychology, 11(3), 591–611. https://doi.org/10.1007/s13164-019-00438-9 stevner, a., vidaurre, d., cabral, j., rapuano, k., nielsen, s. f. v., tagliazucchi, e., laufs, h., vuust, p., deco, g., woolrich, m., & others. (2019). discovery of key whole-brain transitions and dynamics during human wakefulness and non-rem sleep. nature communications, 10(1), 1–14. https://doi.org/10.1038/s41467-019-08934-3 tononi, g., boly, m., massimini, m., & koch, c. (2016). integrated information theory: from consciousness to its physical substrate. nature reviews neuroscience, 17(7), 450–461. https://doi.org/10.1038/nrn.2016.44 tononi, g., & koch, c. (2008). the neural correlates of consciousness: an update. annals of the new york academy of sciences, 1124, 239–261. https://doi.org/10.1196/annals.1440.004 tsuchiya, n., wilke, m., frässle, s., & lamme, v. a. f. (2015). no-report paradigms: extracting the true neural correlates of consciousness. trends in cognitive sciences, 19(12), 757–770. https://doi.org/https://doi.org/10.1016/j.tics.2015.10.002 van es, t. (2019). minimizing prediction errors in predictive processing: from inconsistency to non-representationalism. phenomenology and the cognitive sciences, 1–21. https://doi.org/10.1007/s11097-019-09649-y vrselja, z., daniele, s. g., silbereis, j., talpo, f., morozov, y. m., sousa, a. m., tanaka, b. s., skarica, m., pletikos, m., kaur, n., zhuang, z. w., liu, z., alkawadri, r., sinusas, a. j., latham, s. r., waxman, s. g., & sestan, n. (2019). restoration of brain circulation and cellular functions hours post-mortem. nature, 568(7752), 336. https://doi.org/10.1038/s41586-019-1099-1 wallace, c. s., & dowe, d. l. (1999). minimum message length and kolmogorov complexity. the computer journal, 42(4), 270–283. https://doi.org/10.1093/comjnl/42.4.270 wiese, w. (2020). the science of consciousness does not need another theory, it needs a minimal unifying model. neuroscience of consciousness, 2020(1). https://doi.org/10.1093/nc/niaa013 wiese, w. (2018). toward a mature science of consciousness. frontiers in psychology, 9, 693. https://doi.org/10.3389/fpsyg.2018.00693 wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1371/journal.pone.0133532 https://doi.org/10.33735/phimisci.2021.80 https://doi.org/10.1023/a:1013741414324 https://doi.org/10.1007/s11023-019-09501-x https://doi.org/10.1088/0034-4885/75/12/126001 https://doi.org/10.1143/ptps.130.17 https://doi.org/10.1007/s12559-009-9007-x https://doi.org/10.1111/j.1933-1592.2010.00483.x https://doi.org/10.1111/j.1468-0068.2012.00863.x https://doi.org/10.1093/bjps/axp046 https://doi.org/10.1093/brain/awu141 https://doi.org/10.33735/phimisci.2021.75 https://doi.org/10.1007/s13164-019-00438-9 https://doi.org/10.1038/s41467-019-08934-3 https://doi.org/10.1038/nrn.2016.44 https://doi.org/10.1196/annals.1440.004 https://doi.org/10.1016/j.tics.2015.10.002 https://doi.org/10.1007/s11097-019-09649-y https://doi.org/10.1038/s41586-019-1099-1 https://doi.org/10.1093/comjnl/42.4.270 https://doi.org/10.1093/nc/niaa013 https://doi.org/10.3389/fpsyg.2018.00693 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org wanja wiese and karl j. friston 32 wiese, w., & friston, k. j. (2021). examining the continuity between life and mind: is there a continuity between autopoietic intentionality and representationality? philosophies, 6(11), 18. https://doi.org/10.3390/philosophies6010018 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. wiese, w., & friston, k. j. (2021). the neural correlates of consciousness under the free energy principle: from computational correlates to computational explanation. philosophy and the mind sciences, 2, 9. https://doi.org/10.33735/phimisci.2021.81 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.3390/philosophies6010018 https://doi.org/10.33735/phimisci.2021.81 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction from neural correlates to computational correlates of consciousness challenge 1: global dynamics challenge 2: non-arbitrary mappings challenge 3: unusual conditions computational correlates of consciousness and the free energy principle cccs and computational principles deriving computational principles from first principles computational correlates as necessary, sufficient, or necessary and sufficient conditions for consciousness? the challenge of islands of awareness dynamical complexity as a ccc? from computational correlates to a computational explanation of consciousness the scope of computational explanations distinguishing simulation from instantiation computational explanations of consciousness and minimal unifying models conclusion understanding as a bottleneck for the data-driven approach to psychiatric science. understanding as a bottleneck for the data-driven approach to psychiatric science barnaby crooka (barnaby.crook@uni-bayreuth.de) abstract the data-driven approach to psychiatric science leverages large volumes of patient data to construct machine learning models with the goal of optimizing clinical decision making. advocates claim that this methodology is well-placed to deliver transformative improvements to psychiatric science. i argue that talk of a data-driven revolution in psychiatry is premature. transformative improvements, cashed out in terms of better patient outcomes, cannot be achieved without addressing patient understanding. that is, how patients understand their own mental illnesses. i conceptualize understanding as the possession of adaptive mental constructs through which experience is mediated. i suggest that this notion of understanding serves as a bottleneck which any prospective approach to psychiatry must address to be efficacious. subsequently i argue that, though the data-driven approach is undoubtedly powerful, it does not have a straightforward means of unblocking the bottleneck of understanding. i suggest that the data-driven approach must be supplemented with significant theoretical progress if it is to transform psychiatry. keywords big data ∙ machine learning ∙ mental illness ∙ psychiatry ∙ recovery ∙ understanding this article is part of a special issue on “models and mechanisms in philosophy of psychiatry,” edited by lena kästner and henrik walter. heavenly hurt, it gives us – we can find no scar, but internal difference – where the meanings, are – emily dickinson (1861/1998), there’s a certain slant of light auniversity of bayreuth, department of philosophy. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 2 1 introduction mental illness is one of the greatest sources of human suffering worldwide (insel et al., 2015; vigo et al., 2016). governments, research institutions, and health organizations are leading calls for improved mental healthcare to ameliorate the devastating costs that mental illness imposes on individuals and societies (health & care, 2021; trust, 2020; world health organization, 2019). despite frequent claims that current approaches to tackling mental illness are inadequate, theorists are divided about how progress should be made (bickman, 2020; cuthbert & insel, 2013; hengartner & lehmann, 2017). recently, methods utilizing big data and artificial intelligence (ai) have been gaining increased attention. proponents of this data-driven approach are hopeful that sophisticated machine learning models will “redesign the current landscape of mental illness” (fernandes et al., 2017, p. 5) through “tailored interventions for better outcomes” (fernandes et al., 2017, p. 3). in this paper i critically assess the claim that the data-driven approach, as currently conceived, can bring about a transformative improvement to psychiatric science. i argue that, due to the central role patient understanding plays in recovery from mental illness, talk of an ai revolution in psychiatry is premature. my work follows other authors who have engaged with the philosophical issues raised by the idea of ai-driven psychiatry. in an article structured as a backand-forth debate, brown and colleagues (2021) discuss the possibility of future ai replacing human psychiatrists altogether. in a paper focusing on schizophrenia, starke and colleagues (2021) discuss the ethical principles that must be upheld if machine learning is to be introduced into clinical practice. in a response to that article, gauld and colleagues (2021) analyze the kind of change that the age of ai will presage for psychiatry and the training of psychiatrists, a topic also addressed by mccoy and colleagues (2020). finally, horn and weisz (2020) tackle the question of whether ai can improve psychotherapy research. my argument covers similar ground to all of the above work. the novelty of my approach consists in the key role it affords patient understanding and how thoroughly it engages with the evidence on what makes interventions effective. together, these ideas have important implications for how successful we can expect the data-driven approach to be. many practitioners and theorists worry that psychiatry is a science in crisis, with traditional frameworks insufficient to address the complexity of mental illness (bickman, 2020; mcgorry & nelson, 2019). advocates of the data-driven approach are optimistic that their framework is the first to be capable of tackling that complexity in a scientifically rigorous way (durstewitz et al., 2019; dwyer et al., 2018; rutledge et al., 2019). i begin the paper by describing this approach. methodologically, the data-driven approach eschews traditional hypothesis-driven inferential statistics in favor of the automated discovery of predictive patterns in large volumes of patient data. philosophically, i note that advocates of this approach espouse theory neutrality, rejecting specific theories of mental illness on the grounds crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 3 that data are a more reliable and objective guide to clinical decision-making. i also characterize the data-driven approach as function-oriented, viewing psychiatry as an interdependent set of optimization problems. i argue that these commitments lead to problems for the approach once the role of patient understanding is carefully considered. to set up my argument, i begin by clarifying what would constitute a transformative improvement to the status quo. i suggest that a reasonable interpretation should focus on patient outcomes, as “the amelioration of […] mental distress and suffering” (thornton, 2020, p. 237) is the fundamental goal of psychiatry 1. given such a view, i argue that for the data-driven approach to be transformative, it must bring about radical improvements in patient outcomes. my argument for tempering expectations of the data-driven approach centers on patient understanding. integrating ideas about understanding from pragmatist philosophers and theorists of psychiatry (beck & haigh, 2014; dilthey, 1984/1977; wampold & imel, 2015; wilkenfeld, 2013), i conceptualize patient understanding as a set of adaptive mental constructs through which a subject’s experiences are mediated. conceptualized this way, a patient possessing understanding becomes a pre-requisite for lasting recovery from mental disorder. i suggest that this notion of understanding serves as a bottleneck for any approach to psychiatric science. without addressing patient understanding, outcomes cannot be radically improved. the above analysis leads us to a specific question: can the data-driven approach address the bottleneck of understanding? i examine three ways that it might. first, i assess what i call the indirect path, which consists of optimization of the clinical pipeline in terms of nosology, diagnosis, prognosis, and treatment assignment. second, i assess what i call the direct path. that is, the tailoring of treatments for improved outcomes. third, i assess whether the data-driven approach might deliver radically improved patient outcomes by discovering novel insights about mental disorders that can be used to provide adaptive explanations. i argue that there are serious problems with each of these possibilities and, therefore, that the datadriven approach has no clear route to providing transformative improvements to psychiatric science. the structure of the paper is as follows. in section 2 i present the data-driven approach. i discuss what motivates it, how it works, and its philosophical commitments. in section 3 i make the claim that patient understanding functions as a bottleneck for any approach to psychiatric science. to do so, i present my analysis of understanding as possession of adaptive mental constructs, show how this makes understanding critical for robust recovery from mental distress, and explain in what sense this makes it a bottleneck. in section 4 i assess the potential of the data-driven approach to address the challenge of the bottleneck of understanding, 1to be clear then, radically improved scientific understanding of mental disorders, if unaccompanied by improved outcomes, would not be considered transformative by this criterion. however, one could reasonably adopt another criterion on which it would. i thank an anonymous reviewer for pushing me to clarify this point. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 4 arguing that, at least as currently practiced, it is ill-suited to do so. in section 5 i address possible objections to the view presented. in particular, i defend my conceptualization of understanding and justify the applicability of the bottleneck metaphor. i then conclude by suggesting that the data-driven approach can best contribute to psychiatric science in collaboration with theoretical research. 2 the data-driven approach in this section i characterize the data-driven approach to psychiatry. i begin by describing the theoretical landscape from which the approach has emerged. i then provide an overview of the data-driven methodology, including an example (koutsouleris et al., 2021). i conclude the section by discussing the philosophical commitments embodied by the approach. 2.1 motivating the approach: psychiatry in crisis reading the theoretical and philosophical literature on mental illness, one often comes across the view that psychiatry is a discipline in crisis. for example, stoyanov and maes (2021, p. 1) claim that “psychiatry remains in a permanent state of crisis […] evidenced by the many different competing approaches and ways to understand mental and psychiatric disorders”. leonard bickman (2020, p. 803) states that mental healthcare currently suffers from “overall poor effectiveness” caused by “poor alignment between what the patient needs, and the treatment provided”. and allen frances (2009, p. 391) writes that “our understanding of psychopathology is fairly primitive” and “lacks the fundamental understanding of pathogenesis”. these pessimistic assessments locate the precise source of the problem differently, but they all share the opinion that the complexity of mental disorders plays a significant role in thwarting the success of psychiatric science 2. mental disorders are complex in multiple ways. first, they are difficult to explain, resisting “a simple analysis in terms of biological pathways, endophenotypes, and neural mechanisms” (borsboom et al., 2019, p. 91). second, mental disorders are massively multifactorial (boer et al., 2021). that is, there are many different factors, “sprinkled across multiple causal levels” (kendler & gyngell, 2020, p. 44), that may be relevant to any given presentation of a mental disorder. this relates to the first point. if we ought, as engel suggested, to incorporate “all the factors contributing to both illness and patienthood” (engel, 1977, p. 133) into our explanations of mental disorder, then these will be dense explanations indeed. however, identifying the relevant factors does not suffice to fully characterize mental disorders. this is due to a third kind of complexity: interaction among the factors. as borsboom and colleagues put it, “mental disorders likely involve feedback loops that cross all 2the term complexity is used to refer to many different properties, both formal and informal (ladyman et al., 2013). it is beyond the scope of this paper to weigh in on which formal criteria for complexity mental disorders satisfy. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 5 of the traditional divides between levels of explanation, none of which can claim the status of ‘basis’ for the others” (borsboom et al., 2019, pp. 10–11). in summary, biological entities, psychological constructs, and social factors interact in intricate and diverse ways to produce particular instantiations of psychopathology. the idea that psychiatry is in a crisis of complexity provides a strong rationale for adopting the data-driven approach. for example, rutledge and colleagues (2019, p. 152) state that “big data and machine learning are uniquely placed to address [the complexities of psychiatric disorders]”. similarly, british psychiatrists brown and story (2021, p. 131) argue that modern technology “enables the capturing of rich, longitudinal, multimodal data, the analysis of which promises vastly improved characterization of illnesses and their trajectories”. while historical approaches to psychiatry have often been stuck either focusing on a small subset of factors or falling short of scientific standards of evidence (ghaemi, 2007), the data-driven approach leverages the availability of data and computation to integrate diverse sources of information in a rigorous way. unlike traditional statistical methods that focus on detecting group differences, machine learning methods are designed to exploit large volumes of multivariate data to make individual predictions (bennett et al., 2019; gillan & whelan, 2017). this means that clinical decisions, such as treatment recommendations, can be tailored to sophisticated statistical representations of individual patients. for this reason, the data-driven approach is sometimes termed precision psychiatry (e.g., bzdok & meyer-lindenberg, 2018; fernandes et al., 2017). some proponents of the data-driven approach have been forthright in claiming that this approach will suffice to bring about a transformative improvement to mental healthcare. for example, gauld and colleagues (2021, p. 2519) proclaim that “ai will undeniably transform the future of psychiatry”. similarly, bickman (2020, p. 803) extols the “revolutionary possibilities of artificial intelligence for improving mental healthcare”. and fernandes and colleagues (fernandes et al., 2017, p. 1) state that the data-driven approach “promises to be even more transformative than in other fields of medicine”. this view is grounded in the notion that the data-driven approach will provide models that allow clinicians “to identify the right treatment for each patient, first time around” (gillan & whelan, 2017, p. 34). as stated in the introduction, i interpret a transformative improvement to require radically improved patient outcomes. in order to assess whether the data-driven approach can really achieve this, we will need to take a closer look at how it works. 2.2 the methodology of the data-driven approach i thus turn to the methodology of the data-driven approach. readers familiar with big-data and machine learning can skip to section 2.3 where i will discuss the philosophical commitments of the approach. the data-driven approach is defined by a methodological hallmark: the employment of automated computational procedures for discovering predictive patterns crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 6 in large volumes of data. such approaches have become increasingly attractive as the cost of collecting, storing, and manipulating data has decreased (chekroud et al., 2021). at a very high level, the methodology of the data-driven approach can be described as consisting of four steps. first, a problem is defined. second, data are acquired. third, a model is designed and trained. fourth, the model is deployed. we will take a look at each of these steps in turn. 2.2.1 problem definition the first step in the data-driven approach is to define a problem. machine learning is a technique well-suited to classification, prediction, and clustering problems. in classification, the task is choosing which of a pre-defined set of classes a particular data instance belongs to. in prediction, the task is estimating the future value of a chosen variable. in clustering, the task is organizing a dataset into self-similar groups. models developed according to the data-driven approach have many possible applications in psychiatry. these can be coarsely divided into four (bzdok & meyer-lindenberg, 2018; dwyer et al., 2018). first, classification models can be used for diagnosis. for example, a model might be constructed to determine whether subjects should be diagnosed with major depressive disorder (e.g., sharma & verbeke, 2020). however, applying machine learning to diagnosis need not be limited to a binary decision about a single condition. for example, in multiclass logistic regression a model will return a probability distribution over multiple possible diagnoses (elujide et al., 2021; qureshi et al., 2016). this makes the data-driven approach amenable to transdiagnostic studies (pelin et al., 2021). second, models employing clustering algorithms can be used to aid nosology, the organization of mental illnesses into a set of categories (zachar & kendler, 2017). within this paradigm, the objective of the model is to find a partitioning of the data that best satisfies a chosen set of statistical criteria. when models are used to find clusters within extant disorder categories, this is called sub-typing (feczko et al., 2019). a third application of the data-driven approach is prognosis, the prediction of how an individual’s condition will evolve over time. machine learning models can be used to predict clinically relevant events, such as symptom onset (koutsouleris et al., 2021). recurrent neural networks, an architecture particularly well-suited to exploiting temporal regularities, can be used to compute disease trajectories (shickel et al., 2018; suhara et al., 2017). finally, perhaps the most important target application of all is treatment assignment, using patient data to decide which therapeutic approach to pursue 3. studies have attempted to apply machine learning to predict responses to both pharmacological and psychotherapeutic treatments (chekroud et al., 2021; su et al., 2020). 3initial research in this domain often involves predicting response to an individual treatment as opposed to assignment among treatments. clearly, if reliable estimates of treatment response could be computed across many treatments, this would lead to a straightforward procedure for assigning treatment. namely, predicting the response to multiple treatments and selecting the best one (indeed, the problems are treated as synonymous in chekroud et al., 2021). crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 7 2.2.2 data acquisition machine learning models are typically data hungry, meaning that they only learn patterns that are robust and generalizable if supplied with large volumes of data. this makes data acquisition a crucial part of the approach (dwyer et al., 2018). the reason for these onerous data requirements is known as the curse of dimensionality, a term referring to the fact that, as more dimensions of variation are considered, the numerosity of the data required to maintain model performance increases exponentially (keogh & mueen, 2017). this means that collecting the quantities of data required to train massively multivariate models of mental illness requires large-scale coordination (bzdok & meyer-lindenberg, 2018). since variation at almost every conceivable level of organization is thought to be relevant to mental illness (kendler & gyngell, 2020), data can be collected in innumerable ways, from questionnaires and cognitive tasks, to biomarkers and neuroimaging, to social media and wearable electronics (balaskas et al., 2021; garcíagutiérrez et al., 2020). the challenges involved vary across modalities. for example, collecting and analyzing neuroimaging data is time-consuming and costly (najafpour et al., 2021). on the other hand, clinical notes in electronic health record (ehr) data are abundant and accessible, but unstructured (shickel et al., 2018). pre-processing diverse data modalities such that they can be integrated by a single model is a technical challenge. additionally, it is often unclear whether different operationalizations of psychological concepts actually measure the same latent construct (poldrack & yarkoni, 2016). given these difficulties, shickel and colleagues (2018, p. 24) describe developing a unified representation of such disparate sources of information as the “holy grail of clinical deep learning research”. 2.2.3 model design and training given a rich and voluminous dataset, the next step is to design a model. in the machine learning paradigm this consists of specifying the objective function, architecture, and learning rule of the model. the objective may be to predict remission rates, responsiveness to specific medications, or symptom response trajectories (bzdok & meyer-lindenberg, 2018; chekroud et al., 2021; rutledge et al., 2019). translating the natural language expression of the goal into a mathematically specified objective function is a crucial part of the design phase. the architecture of a model consists of its components and how they are organized. for example, a deep neural network consists of multiple layers of neuron-like computational units equipped with activation functions (durstewitz et al., 2019), while a decision tree consists of hierarchically organized decision nodes (feczko et al., 2019). machine learning researchers have developed a veritable zoo of model architectures with a variety of mathematical and representational properties (bronstein et al., 2021). in principle, architectures can be designed to exploit the kinds of regularities in the sampled data domain. in the case of high-dimensional, multimodal patient data, however, design principles guiding architectural choices are not yet well estabcrook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 8 lished (si et al., 2021), reflecting the lack of scientific consensus on the nature of mental illness. once an objective and architecture are chosen, training takes place. a machine learning model is usually initialized with random parameter values (chollet, 2021, p. 46). what the model outputs is a function of the input and these parameter values. in a deep neural network, for example, the parameter values are the connection weights between the neurons and each individual neuron’s bias. the goal of learning is to find a set of parameter values that optimizes the objective. for example, if the objective is to diagnose depression, the model might take neuroimaging and physiological data associated with an individual and output a binary classification reflecting the presence or absence of a diagnosis (arbabshirani et al., 2017; sharma & verbeke, 2020). the model’s output is then compared with the true label for the data point. in this case, the label would be provided by a clinician’s diagnostic judgement about the patient. having access to the ground truth label makes this an example of supervised learning. should the machine’s output be incorrect, an error signal, often called the loss, is computed. this error signal, in conjunction with a learning rule, like stochastic gradient descent, is used to update the parameter values 4. after many iterations of this process, a model converges on a stable set of parameter values that encode its best approximation to the true function relating the input data to the output labels. in our example, the parameters can be considered as encoding a hypothesized relationship between biological variables and depression. crucially, to avoid overfitting on the particularities of the data sample used for training, models must always be validated on novel datasets (bzdok & meyer-lindenberg, 2018; rutledge et al., 2019). the simple example above is the tip of the iceberg when it comes to applications of the data-driven approach. while a thorough review is out of scope for this paper, it is worth noting the broad categories of unsupervised and semi-supervised methods, which do not rely on labels (bengio et al., 2013; bronstein et al., 2021; domingos, 2012). in our depression diagnosis case, ground truth labels were provided by human judgement. in addition to being costly and time-consuming to generate, the use of human judgment in producing labels limits the potential value data-driven models can provide (bickman, 2020). durstewitz and colleagues (2019) point out that the value of data-driven diagnosis is upper-bounded by how well extant diagnostic labels capture the underlying nature of mental illnesses. for this reason, the data-driven approach “needs to go beyond the mere prediction of symptoms by current diagnostic schemes, but rather has to help refining our diagnoses” (durstewitz et al., 2019, p. 1591). in line with this idea, researchers have employed unsupervised machine learning methods to discover novel clusters of patients that can be used to predict outcomes, stratify treatments, and increase understanding (gould et al., 2014; miranda et al., 2021; pelin et al., 2021). the crucial point is that, 4in practice, one iteration of a training loop usually involves a batch of data instances, rather than just one. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 9 while supervised learning is limited by human experts’ ability to provide useful labels, the data-driven approach as a whole need not be. 2.2.4 validation and deployment once data have been collected and a model has been trained, the next steps are validation and deployment. as of yet, the majority of studies implementing the data-driven approach are proof-of-concept and most trained models have not been clinically deployed (cearns et al., 2019; chekroud et al., 2021). in an insightful review, cearns and colleagues (2019) provide a checklist that any machine learning model should comply with to demonstrate readiness for clinical application. models should show: 1) an improvement on whatever approach constitutes the current state-of-the-art for the clinical application in question, 2) validation with a large, external dataset, and 3) clear specification of the scope of the model (i.e., the population for which its validity has been rigorously demonstrated). once deployed, a machine learning model must be continually evaluated to ensure that its accuracy remains robust, that it is secure, and that it does not lead to unacceptable algorithmic bias. practically, deployment also requires coordination with clinicians to ensure that models are well-utilized (tonekaboni et al., 2019). as the data-driven approach matures, further concrete issues relating to clinical deployment are sure to emerge. 2.2.5 the data-driven approach in practice: an example an example will help to elucidate how the data-driven approach plays out in practice. in a large-scale study, koutsouleris and colleagues (2021) used multimodal machine learning to predict whether patients with clinical high-risk states and recent-onset depression would transition to psychosis. in an extensive data acquisition process spanning five countries and more than three years, koutsouleris and colleagues collected data from many domains including sociodemographic, questionnaires, self-reports, structured interviews, cognitive and behavioral tasks, neuroimaging, and genetics. this highlights the importance of collecting spatially and temporally distributed data for constructing generalizable models that will be robust to variation in social and cultural factors (bzdok & meyer-lindenberg, 2018; dwyer et al., 2018). to make use of their avalanche of information, koutsouleris and colleagues built a stacked model, a kind of architecture suitable for automatically learning how to best integrate disparate data sources 5. the researchers’ multimodal predictive model was able to outperform all unimodal competitors as well as human clinical raters in prognostic accuracy. in addition to the predictive value of their model, the authors were also able to extract clinical insights into psychosis by analyzing their findings. for example, they found high-risk patients predicted 5stacked models consist of multiple base models, each exploiting a particular data modality, and a meta-model which learns how to weight the predictions of the base models. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 10 not to transition to psychosis had increased temporo-occipital brain volume relative to healthy controls, suggesting a neural basis for mechanisms of resilience. further, they found certain neurocognitive factors, such as impaired facial affect recognition, were markers of poor psychosis outcomes in both clinical high-risk and recent-onset depression groups.. this highlights another important feature of the data-driven approach. investigating a transdiagnostic population, that is, incorporating data from groups which are typically investigated separately, allows the data-driven approach to find and exploit patterns which transcend the group differences captured by traditional inferential approaches (hengartner & lehmann, 2017; mcgorry & nelson, 2019). 2.3 the philosophical commitments of the data-driven approach now that we have covered the methodological basis of the data-driven approach, we can turn to its philosophy. i raise two points. first, the data-driven approach purports to be theory neutral, relying on automated statistical procedures rather than theoretical assumptions to cut through the complexity of mental illness. second, the data-driven approach is function-oriented, focusing on building models to optimize each step of the clinical decision-making pipeline. these features will turn out to be important for assessing whether the approach can transform psychiatry by radically improving patient outcomes. 2.3.1 theory neutrality different schools of thought have long disagreed about the nature of mental illnesses (kendler, 2016; shorter, 1997; tsou, 2021; see also dembic, 2023; leder & zawidzki, 2023, this volume). should we think of them as brain disorders defined by neural dysfunction, socially constructed labels for deviant behavior, irreducibly subjective pathologies of psychology, or an intricate mixture of these ingredients? the data-driven approach chooses to wash its hands of the issue entirely, preferring to remain theory neutral (paulus, 2015). unlike the attempt to develop mechanistic models of disease processes, the data-driven approach to psychiatry makes no strong assumptions about the nature or causal structure of mental illness (bennett et al., 2019). practitioners take this feature to be a strength of the approach. for example, chekroud and colleagues (2021, p. 154) claim that machine learning offers a set of “powerful hypothesis-free approaches” to predicting treatment outcomes. similarly, in a review of techniques applying deep learning to electronic health record data, shickel and colleagues (2018, p. 17) describe the philosophy as “letting the data speak for itself by discovering latent relationships and hierarchical concepts from the raw data, without any human supervision or prior bias”. from the data-driven perspective, then, theoretical assumptions about mental illness do more harm than good, and we would be better off relying on the presumed objectivity of models trained with automated procedures. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 11 it is important to note that the data-driven approach is not, and indeed no approach can be, truly atheoretical (joober & tabbane, 2019). in practice, even huge quantities of data are useless without significant constraints on learning (domingos, 2012). researchers developing machine learning models make many choices about which data to collect, which architecture to use, how to specify their objective, and so on (see section 2.2). each of these decisions biases a model towards learning some structures rather than others, a phenomenon often discussed in machine learning under the term inductive bias (battaglia et al., 2018). despite this caveat, the intention to remain as neutral as possible with respect to theory still constitutes a significant philosophical commitment of the data-driven approach. in the argument to come, i will suggest that this is a mistake. without theory-driven research to improve understanding of how treatments can improve outcomes, the data-driven approach has limited potential. 2.3.2 function orientation powerful methodological tools alter how domains of inquiry are perceived by their users. cichy and kaiser (2019, p. 312) express this point thusly, “once technological artefacts are commonly used, they are not mere tools to realize predefined scientific goals but begin to shape social reality in a way that affects the user’s desires and interests”. the data-driven approach is no exception. as dwyer and colleagues put it (2018, p. 95), “modern machine learning methods can contribute greatly to clinical psychology and psychiatry by changing the way that problems are considered [emphasis added]”. i introduce the term function orientation to describe the perspective on psychiatry that the data-driven approach imposes. the term function here should be interpreted in the mathematical sense of mappings from inputs to outputs. note that the function-oriented perspective is largely implicit in the way data-driven research is carried out, rather than being explicitly endorsed by its practitioners. the function-oriented perspective decomposes psychiatric science into a set of statistical problems, each addressable through the development of the right machine learning model. it formulates these problems in mathematical terms, with solutions consisting of trained models judged by accuracy metrics. the overall goal is to optimize the entire pipeline of clinical decision making, from diagnosis to treatment assignment. as dwyer and colleagues (2018, p. 94) put it, “the ultimate aim of translational machine learning is to generate procedures that would be beneficial for clients, general practitioners, and in specialized hospital settings to improve patient outcomes”. this is a noble aim. however, the overall effect of viewing psychiatry this way is that discrete clinical decisions are foregrounded, while anything not amenable to being quantified and incorporated into a statistical learning problem is neglected. for example, the role of the psychiatrist as an empathic listener and the idiosyncratic content of mental distress, resistant to concise mathematical representation, can easily be obscured in this formulation (horn & weisz, 2020). of course, these are relative effects. function orientation does not crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 12 make one blind to the interpersonal nature of psychiatry. however, to the extent that one develops expertise in a way of seeing that relies on noting and measuring formalizable properties, one naturally becomes less attuned, on average, to those properties which cannot be easily formalized. this is what cichy and kaiser (2019, p. 312) mean when they stress that “models are not neutral tools.” in section 4.1, i will argue that function orientation is a flawed perspective. without addressing the limitations of currently available treatments, optimizing clinical decision making cannot be transformative. 3 the bottleneck of understanding having characterized the data-driven approach, i now introduce the notion of understanding that i claim to be of critical importance to treating mental illness 6. while this is a general account of understanding, when it is applied to how patients understand their own illnesses it takes on a crucial role for psychiatric science. in particular, i argue that, for patients suffering from mental illnesses, developing this form of understanding is necessary for recovery. according to my argument, patient understanding serves as a bottleneck which any effective approach to psychiatric science must address to improve patient outcomes. 3.1 understanding as possession of adaptive mental constructs i propose that understanding can be usefully conceived of as a set of adaptive mental constructs through which experience is mediated. i derive this definition by integrating ideas from theorists of psychiatry and the pragmatist tradition in philosophy. the notion of understanding has a rich history in psychiatry. perhaps best known is karl jaspers’ account (1959/1997), on which understanding is a psychiatrist’s intuitive, empathic means of interpreting a patient’s mental world. however, while the interpersonal relationship between practitioner and patient is of immense importance to psychiatry, understanding can also be isolated and considered at an individual level. for example, consider german polymath wilhelm dilthey’s account (dilthey, 1984/1977). central to dilthey’s notion of understanding is what he calls the “acquired nexus of psychic life” (1984/1977, p. 42). this nexus includes our values, habits, ideas, and goals and influences “every single act of consciousness” (1984/1977, p. 59). in dilthey’s view, these mental constructs form a coherent whole that “determines the nature of our understanding of ourselves and of others” (1984/1977, p. 55). following dilthey, then, we will take 6note that i am not arguing that this is the correct or best way to conceptualize understanding, either across all contexts or even within the domain of psychiatry. there are other notions of understanding that may be valuable and important in different contexts. see section 5.1 for more on this point. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 13 the basis of understanding to consist of a set of mental constructs through which experience is mediated. however, we focus our attention specifically on patients’ understanding of their own inner lives, as this is what is of particular relevance to the goals of psychiatric science. a further property that plays a role in conceptualizations of understanding from both inside and outside of psychiatry is that of adaptivity. in general, something is adaptive if it enables goals to be achieved under a set of environmental constraints. within psychiatry, this notion appears in aaron beck’s cognitive model of mental illness (beck, 1985; beck & haigh, 2014). on this view, a patient’s psychological distress is a downstream effect of their maladaptive cognitive schemas, which are the “beliefs, expectancies, evaluations, and attributions” that “serve to order everyday experience” (beck & haigh, 2014, p. 12). note the similarity to dilthey’s acquired nexus, described above. under the cognitive model, interventions involve cognitive restructuring, which is the active modification of a patient’s maladaptive beliefs and schemas, to make them adaptive. however, one need not adopt the specific conceptual lexicon of the cognitive model in order to subscribe to the view that an adaptive notion of understanding is crucial to psychiatry. for example, bruce wampold defines adaptive explanations as those that “provide a means to overcome or cope with […] difficulties” (wampold & imel, 2015, p. 58). wampold and budge suggest that “patients typically present with a maladaptive explanation for their disorder” and “a primary therapeutic activity of the therapist is to provide an adaptive explanation” (wampold & budge, 2012, p. 612). wampold’s work on the common factors of efficacious psychotherapy shows that the delivery of such adaptive explanations is a necessary feature of all effective therapies (wampold, 2015; wampold & imel, 2015). 7 while this line of work focuses on adaptive explanations, from a patient’s perspective, accepting and internalizing such an explanation just is coming to understand their experience in an adaptive way. the idea that understanding is usefully conceptualized as adaptive, affording particular abilities or skills, also resonates with recent currents of thought in the pragmatist philosophy of understanding (grimm, 2019; hills, 2016; regt, 2019; wilkenfeld, 2013). consider daniel wilkenfeld’s (2013) account: understanding is at root the possession of the right sort of mental representations of that which is understood; […] a mental representation counts as being “of the right sort” in virtue of the fact that possession of it enables one to perform […] feats relevant in that context. (wilkenfeld, 2013, p. 1000) 7the idea that diverse methods of psychotherapy owe their efficacy to a set of shared common factors was first proposed by saul rosenzweig (rosenzweig, 1936) before being explored in greater depth by jerome frank (frank & frank, 1993). various common factors, including the therapeutic alliance, empathy, and shared expectations, have been proposed (wampold, 2015). while there is no consensus on a particular conceptual model, there is very strong evidence that the common factors account for most of the variance in clinical outcomes (peterson, 2019). crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 14 notice that this view equates understanding with being empowered to do certain things. namely, whatever is relevant to a particular context. wilkenfeld’s account is general but can be readily applied to psychiatric patients. the mental representations relevant to patients may involve aspects of self-understanding, life narratives, memories of and beliefs about aberrant experiences, and so on (glover, 2020,; mcconnell, 2020). the relevant feats which the afflicted hope to be enabled to perform may be things like engagement in autonomous, competent behavior, the pursuit of goals, and effective self-regulation, particularly in the face of stressors (ivanov & schwartz, 2021; jacob, 2015). so then, putting the ideas above together, patient understanding can be cast as the patient’s possession of a set of adaptive mental constructs. following beck and dilthey, i use the term mental constructs to refer to things like beliefs, habits, goals, expectancies, evaluations, values, and so on 8. following wampold and wilkenfeld, i use the term adaptive to mean enabling one to do relevant things in a given context. notice two things about this notion of understanding. first, it covers both intuitive and intellectual senses of the word. by intuitive, i mean related to early, subconscious aspects of information processing, like the biased emotional processing hypothesized to be important to depression (harmer et al., 2009). by intellectual, i mean involving articulable propositional content, like a belief that one is incapable (beck, 1985). second, the notion of understanding we are using is not factive. that is, it does not require conformation to any particular standard of truth (doyle et al., 2019). this is suitable for our context for two reasons. first, in the study of effective psychotherapy there is no empirical evidence that the truth value of the explanation is related to outcome 9 (frank & frank, 1993; wampold et al., 2007). second, for some of the mental constructs relevant to this notion of understanding, such as evaluations, it is not clear whether or how a truth value could be ascribed. i further clarify and defend this analysis of understanding in section 5.1. 3.2 understanding as a bottleneck to recovery above, we conceptualized understanding as possession of adaptive mental constructs. for those suffering from mental illness, gaining understanding might involve developing beliefs, expectancies, and values which enable the self-regulation of affective experience and behavior. i now make the claim that patient understanding is a bottleneck for robust recovery.10 8the argument does not turn on which particular mental constructs play the role of organizing ongoing experience. see section 5.1 for further elaboration on this point. 9of course, such evidence may be forthcoming. there are theoretical reasons to expect some relationship between truth and efficacy, all else being equal (see baker et al., 2008; laska et al., 2014, for discussion of this issue). 10i follow defenders of the recovery model in viewing recovery from mental illness as describing a process towards a goal of living a self-efficacious and satisfying life (anthony, 2000; jacob, 2015; ramon et al., 2007). crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 15 a bottleneck is a component within a broader system that limits the rate at which something is produced. for example, a shortage of semiconductor chips limits the rate at which cars can be manufactured. i suggest that patient understanding is a bottleneck for recovery in the following sense: the long-term rate at which a patient recovers is limited by the rate at which that patient develops understanding of their illness. in other words, the outcomes psychiatric science wants to bring about cannot be achieved without finding a way to reliably increase understanding. including the ‘long-term’ qualifier here captures the idea that short-term improvements in an outcome measure could reflect random fluctuations in environmental factors such as the number and intensity of stressors. such fluctuations are one of the reasons why recovery is typically a non-linear process that happens in “fits and starts” (jacob, 2015, p. 118). casting understanding as a bottleneck is a way of framing the idea that, given how we have defined understanding, it is essential for a robust recovery. if one lacks adaptive mental constructs, one cannot perform relevant feats in one’s own context. as previously stated, these feats may include setting and seeking personal goals, regulating affective experience, and engaging in a community. if one lacks the mental constructs that enable the performance of such feats, it is difficult to see how one could live a satisfying and self-efficacious life, conditions constitutive of recovery11. i see the value of the bottleneck framing as being threefold. first, the concept of a bottleneck implies embeddedness in a system with multiple components. in this case, the system consists of psychiatric science as a whole, including disease classification, treatment development, therapeutic best practice, and so on. this makes the metaphor apt for sharpening thinking about the relationship between understanding and other levels of organization which may be pertinent to improving outcomes. for example, variables specifiable at the genetic, behavioral, and social levels may each be critical intervention targets for specific instantiations of psychiatric disorder. that is, inducing the adaptive mental constructs that constitute understanding may, in some cases, be most effectively achieved through interventions that are primarily medicinal or behavioral, rather than psychological12. however, the bottleneck framing highlights that in order to engender a robust recovery, any intervention must induce understanding. beck (1985, pp. 333–334) makes a related point, noting that both psychotherapy and pharmacotherapy must alter defective cognitive content to be efficacious. second, framing the challenge of psychiatry as addressing the bottleneck of understanding foregrounds subjective experience and psychopathology. why is this a good thing? because psychopathology is the very subject matter of psychiatry (stanghellini & broome, 2014). neurobiolog11while i have framed this in terms of the recovery model (see fn.10), the argument also applies, albeit less strongly, to symptom-focused models of recovery. to see this, consider that some of the feats relevant to patients will be constitutive of their diagnoses. for example, a patient with generalized anxiety disorder may hope to concentrate effectively on tasks and avoid becoming irritable (both symptoms of the disorder in the dsm 5, american psychiatric association, 2022). 12i thank an anonymous reviewer for pressing me to make this point clear. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 16 ical, genetic, and sociocultural approaches are necessarily studying the correlates of what is, fundamentally, mental disorder (see thornton, 2020, p. 235, for a more detailed version of this argument). finally, focusing on a person-centered, intentional, mental ability like understanding when conceptualizing the goals of psychiatry may be of pragmatic value. both theory and empirical evidence suggest that a sense of agency and self-efficacy are important for recovery from mental illness (glover, 2020; mcconnell & snoek, 2018). framing the challenge of recovery as the development of understanding emphasizes the power of the human subject to actively negotiate an alleviation of their own suffering (anthony, 2000; coulombe et al., 2016). 4 the data-driven approach and the bottleneck of understanding if the view presented in section 3 is correct, a transformative improvement to psychiatric science must involve unblocking the bottleneck of understanding. in this section i argue that the data-driven approach lacks a clear means of achieving this. first, i argue that, because extant treatments lack specificity, optimizing their allocation will not address the bottleneck of understanding. then, i argue that using the data-driven approach to target patient outcomes directly is fraught with theoretical challenges. finally, i argue that there are conceptual difficulties plaguing the translation of data-driven insights into therapeutically valuable information that can aid patient understanding. 4.1 limits of a function-oriented approach recall that the data-driven philosophy being function-oriented refers to how it conceptualizes psychiatry as a collection of machine learning problems. as discussed in section 2.3.2, methodological tools act like filters, rendering particular aspects of phenomena salient. in this case, the data-driven approach focuses on clinical decision-making under the implicit assumption that optimizing these decisions will dramatically improve outcomes. since clinical decisions can only be made with respect to existing interventions, the improvements the data-driven approach hopes to bring about must be realized through negotiating a more effective allocation of interventions (bzdok & meyer-lindenberg, 2018, p. 225). i call this the indirect path to transforming patient outcomes, since it does not involve targeting patient understanding directly. the problem with the indirect path is the lack of specificity of extant treatments. a treatment would be specific if it had an active ingredient that rectified a particular dysfunction (wampold & imel, 2015, p. 60). unfortunately, there is little evidence that treatments from either of the two major modalities, psychotherapy or pharmacotherapy, operate through specific ingredients that directly tarcrook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 17 get pathology (ivanov & schwartz, 2021; middleton & moncrieff, 2019; sathyanarayana rao & andrade, 2016; wampold & imel, 2015).13 in the case of psychotherapy, wampold and imel summarize their comprehensive review of the subject thusly, “there is no compelling evidence that the specific ingredients of any particular psychotherapy or specific ingredients in general are critical to producing the benefits of psychotherapy” (wampold & imel, 2015, p. 253). when it comes to pharmacotherapies, middleton and moncrieff state that “there is no evidence that antidepressants work by correcting a chemical imbalance or other identifiable abnormality” (middleton & moncrieff, 2019, p. 52). while there are many theories attempting to get a firmer handle on therapeutic mechanisms of action (e.g., beck & haigh, 2014; harmer et al., 2017), these have not yet led to the development of specific treatments. this lack of specificity in treatments is concerning for the data-driven approach. if it were the case that extant treatments were specific but misallocated, correcting their allocation would be a clear path to radically improved outcomes. however, given the lack of evidence that extant treatments are specific (budd & hughes, 2009; wolpert et al., 2021), the effects of improved allocation may be small. the problems associated with mental disorders are, as leichsenring and colleagues (2022, p. 141) put it when summarizing their findings from a survey of meta-analyses across major mental disorders, “not sufficiently addressed by the available treatments”. by analogy, imagine a restaurant full of diners randomly assigned variously shaped spoons to cut thick steaks. an optimal reallocation of the spoons according to hand size would be unlikely to bring about a transformative improvement in the average carving ability of the patrons. the aim of data-driven psychiatry can be summarized as the prescription of the right treatment to the right person at the right time (trivedi, 2016). however, if a specific treatment has not been devised yet, no algorithm can recommend its prescription. 4.2 limits of a theory-neutral approach above we questioned the indirect path by which the data-driven approach could revolutionize patient outcomes. what about a direct path? the idea here would be that the data-driven approach could improve treatments by optimizing for patient understanding itself. in other words, machine learning could be used to tailor treatments to achieve better outcomes. research in this area is already underway (see chekroud et al., 2021; su et al., 2020, for reviews). in one study, researchers coded therapist utterances in cognitive behavioral therapy sessions and trained a model to determine the association between therapeutic content and clinical outcomes (ewbank et al., 2020). in this case, the quantity of utterances classed as cognitive and behavioral change methods in therapy sessions was found to correlate with better chances of symptom improvement. while such early results are valu13as wampold and imel state (2015, p. 47), the necessity of some form of exposure for treating specific phobias may be an exception to this rule. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 18 able, notice the theoretical challenges involved in pushing such research programs further. ewbank and colleagues used a cognitive behavioral therapy framework to define twenty-four classes of utterance (e.g., mood check, set goals, therapeutic empathy). the coding of these utterances runs into thorny theory-laden issues of variability and interpretation (horn & weisz, 2020). what are the right categories? how many categories can a single utterance belong to? should semantically distinct ways of expressing similar sentiments be considered therapeutically equivalent? far from the theory-neutral philosophy of the data-driven approach, applying the methodology to improving treatment demands introducing significant theoretical assumptions. this should not be read as a criticism of this research direction. rather, it is an attempt to show that for the data-driven approach to tackle understanding, interdisciplinarity and theory are required. the hypothesisfree number crunching espoused by some proponents of the data-driven approach is not a viable option. there is a different kind of problem for the strategy of developing pharmacological treatments to overcome the lack of specificity mentioned in section 4.1. in chekroud and colleagues’ review of data-driven approaches to predicting treatment outcomes (2021) they find that factors at the psychological and social levels “consistently offer more meaningful and generalizable predictions” (2021, p. 165) than those at the neurobiological and genetic levels. in other words, when it comes to treatment outcomes, most of the predictive signal is not to be found at a level that can be targeted directly by pharmacological interventions. this means that, without a well-validated mapping between the predictive psychological constructs and the neurobiological entities that implement them, pharmacological interventions cannot be targeted directly at understanding (ivanov & schwartz, 2021; poldrack & yarkoni, 2016). as ivanov and schwartz (2021, p. 5) say, “we are still very far indeed” from possessing such a mapping. this critique is not meant to trivialize the important work of using the data-driven approach to improve the allocation of extant treatments for conditions like depression (chekroud et al., 2016; nie et al., 2018). however, these success stories are of the modest, indirect variety. they do not give any reason to think that the data-driven approach can optimize novel pharmacological treatments to directly target psychopathology. 4.3 translation problems for data-driven insights another avenue by which the data-driven approach might transform mental healthcare is by extracting therapeutically actionable insights from trained models. indeed, a common assertion among proponents of the data-driven approach is that these methods might discover subtle and complex patterns which aid mechanistic understanding of mental disorders. for example, durstewitz and colleagues (2019, p. 1592) suggest that as well as using deep neural networks for prediction, we may be able “to employ them to also gain insight into physiological, computational, and cognitive mechanisms”. along the same lines, koppe and colleagues (2021, crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 19 p. 187) state that using visualization techniques on machine learning models “may help uncover interpretable multi-modal biomarkers of psychiatric disease”. one reason to think this might be so is the convergence logic that characterizes data-driven success stories in science. i introduce the term convergence logic to refer to the idea that, because accurate prediction in complex, high-dimensional domains is exceedingly difficult, any model that attains high accuracy under tight constraints must do so in virtue of learning robust patterns which capture core features of the data domain being modeled (schrimpf et al., 2020). this line of reasoning is behind the neuroscientific research program using deep neural networks to model the primate ventral visual stream (yamins & dicarlo, 2016). in that case, the conjecture is that in order to be able to accurately classify images, a neural network has to converge upon mechanisms that approximate its biological target14. the convergence logic is also manifest in the linguistic structure that is learned by neural network language models (manning et al., 2020). in order for a model to accurately predict language tokens, it has to internalize aspects of linguistic structure. moving back to psychiatry, if large computational models incorporating multiple modalities are able to generate accurate prognoses or reliably pick out novel disorder subtypes, one might think it must be because they are converging on robust structure in the data15. on this view, machine learning models might be the right tools to discover the kind of “fuzzy, cross-level mechanisms” that kendler suggests might be the “true nature of psychiatric disorders” (kendler, 2012, p. 18). how would such discovery relate to patient understanding? kendler and campbell (2014, p. 1) suggest that scientific knowledge might “expand the domain of the understandable” by providing models of psychopathological processes that render them intelligible (note that kendler and campbell are referring to jaspers’ clinician-centered notion of understanding here). as an example, the authors discuss the pathological misattribution of meaning that leads to delusional beliefs in schizophrenia, considered paradigmatically un-understandable by jaspers. modern neuroscience posits the aberrant behavior of dopaminergic neurons encoding motivational salience to explain how these beliefs are formed. kendler and campbell suggest that this form of dopamine system dysfunction makes sense of the psychopathological experiences. if the data-driven approach could provide scientific knowledge about a range of psychiatric disorders, those insights might be used to provide adaptive explanations to patients and induce the understanding required to radically improve outcomes16. of course, there may be fundamental limits to this enterprise in certain cases. for example, it is difficult to see how scientific understanding could translate into an understanding 14naturally, this holds only for a given level of abstraction. 15we are assuming here that rigorous cross-validation rules out models achieving accuracy by overfitting on spurious correlations. 16other authors have argued that bringing neuroscience into the clinic could have negative consequences, inducing fatalism by conveying a deterministic world, stripped of human agency (hyman & mcconnell, 2020). crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 20 of psychopathological experiences which lack internal coherence17. this is an avenue that future research must explore. however, even if the convergence logic holds and novel multilevel patterns which can be mapped onto features of mental disorder are discovered18, there are still important barriers for translating data-driven insights into improved clinical outcomes. the first is the issue of intelligibility. the flexibility and expressivity of the machine learning paradigm come with a cost. how these models transform input into output is, prima facie, unintelligible (zednik, 2021). in their review of machine learning applications in psychiatry, cearns and colleagues (2019, p. 9) claim forcefully that “knowing every single one of the hundreds of millions of parameter values in a given ml model would fail to provide even a spec of practically useful insight into the inner workings of a trained model”. however, this assessment is overly pessimistic. while it is clear that no human could infer meaning directly from hundreds of millions of parameter values, functional characterization of trained models at a higher-level of abstraction, such as layers or pathways in a neural network, may be possible. in the domain of psychiatry, this would require translating the components of trained machine learning models into psychiatrically meaningful terms (sheu, 2020). strategies for interpreting the internal structure of trained models are being developed (elhage et al., 2021; koppe et al., 2021; olah et al., 2018; sheu, 2020). researchers following such strategies, such as anthropic’s mechanistic interpretability approach (elhage et al., 2021), use a combination of mathematics and experimental techniques to probe machine learning models in an attempt to characterize their functional organization. while further research is clearly required, applying interpretability techniques to large predictive models is a promising option for discovering the multi-level structure of psychiatric disorders. unfortunately, even the discovery of novel insights into mental illness would not guarantee a transformative impact for the data-driven approach. this is due to what i call the translation problem. namely, that the understanding gleaned by psychiatrists and the scientific community still needs to be translated into therapeutic techniques that induce patient understanding. if a patient cannot internalize an explanation, then that explanation will not be adaptive for them. explanations invoking fuzzy and complex multi-level mechanisms, even if understood by psychiatrists, may not be communicable to the average patient. in order for these hard-won insights to be clinically relevant, they need to be translated into language that resonates with the “assumptive worlds” of specific patients (frank & frank, 1993, p. 30).19 supporting this point, research has shown that having 17i thank an anonymous reviewer for raising this point. 18there is not space here to comment on whether such patterns would necessarily constitute mechanisms. the argument does not turn on whether or not they would. 19wampold and colleagues flesh this notion out by suggesting that explanations “should be proximal to the client’s currently held explanation or expectation, and should not create dissonance with the attitudes and values of the client” (wampold et al., 2007, p. 125). this coheres with the crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 21 a shared rationale with a practitioner is important to patients and contributes to improved outcomes (cuevas et al., 2014; johansson & eklund, 2003; wampold & budge, 2012). further, non-specific components of psychiatric treatments, such as clinical communication and treatment framing, also play a significant role in patient outcomes (priebe et al., 2020). this suggests that how explanations are presented is important and that the inferential distance between practitioner and patient should not be stretched too far. these points highlight the difficulty of the translation problem. discovering the multi-level structure of psychiatric disorder is not enough. the data-driven approach still requires important theoretical work to be done before the bottleneck of understanding can be addressed. 4.4 implications overall, then, while the data-driven approach is a fascinating and powerful addition to the landscape of psychiatric science, it has no straightforward means of unblocking the bottleneck of understanding. as such, the suggestion that this approach will bring about a transformative improvement to patient outcomes is not warranted. this is not to claim that the data-driven approach has no role to play. though i have argued it will not be transformative, increasing the efficiency of the allocation of extant treatments is still an important goal. further, the data-driven analysis of pharmacological and psychotherapeutic interventions is a promising line of research from which important principles and insights regarding treatment strategies can be derived (ewbank et al., 2021). finally, discoveries about how fine-grained patterns of variation across multiple levels relate to psychopathology and the course of mental illness will surely contribute to how we understand and treat psychiatric disorder. to be clear then, the conclusion of the argument is that the contribution of the data-driven approach ought to be incremental and interdisciplinary, rather than discontinuous and technocratic. proponents take pains to acknowledge the technical, ethical, and practical challenges facing their endeavor (bzdok & meyer-lindenberg, 2018; cearns et al., 2019). however, the functionoriented and theory-neutral philosophy often obscures the deep conceptual challenges pervading a data-driven approach to psychiatric science. the problem of inducing understanding in patients suffering mental distress cannot be solved by data alone. 5 responses to objections i now anticipate and respond to possible objections to two of the core concepts invoked in my argument. point, explained in section 3.1, that adaptive explanations, and by extension understanding, need not be true. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 22 5.1 on understanding i cast understanding as the possession of adaptive mental constructs that enable a subject to perform feats relevant to them. however, critics might protest that this pragmatic notion has not been adequately defended. i will briefly address three potential worries. first, there are many competing conceptualizations of understanding out there, so why pick this one in particular? second, does this notion really capture the views of the scholars i cited? third, does this conceptualization really capture anything unified at all? on the first issue, as to why i choose this notion of understanding rather than another, i make a clarification rather than supplying an argument. i endorse a form of conceptual pluralism whereby concepts can be invoked or defined stipulatively for the sake of argument so long as this practice is not abused to willfully obfuscate. the concept i am invoking was inspired by notions of understanding from pragmatist philosophy and theoretical work in psychiatry. i take this to suffice to show that my usage is neither without precedent nor needlessly idiosyncratic. ultimately, for the purposes of this paper, whether understanding is the right term to label the concept i have described is less important than whether or not that concept plays the role i have suggested in limiting the potential of the data-driven approach to psychiatry. a more serious worry is that i have erroneously conflated distinct views and presented them under a common label. along these lines, one might worry that the acquired nexus invoked by wilhelm dilthey (1984/1977) in his discussion of understanding and the cognitive schemas central to aaron beck’s (beck & haigh, 2014) theorizing are rather distinct from the adaptive explanations that wampold (wampold & budge, 2012) describes. the former two notions involve subconscious aspects of mentality that organize ongoing perception and experience, whereas the latter notion seems to be talking about explicit and propositional explanations. are these authors not pointing to distinct concepts? i think a close reading shows that they each have something broader in mind. first, both dilthey and beck include propositional entities like beliefs when spelling out their ideas. they do not limit their concepts to the unconscious. second, wampold takes a very liberal stance on what constitutes an explanation. for example, he describes an inability to recognize one has a problem as a dysfunctional explanation, even though this would clearly not be an articulable proposition (wampold et al., 2007, p. 123). as such, despite their different framings, these authors are capturing a broadly co-extensive set of mental constructs. namely, those that play a critical role in structuring, organizing, and influencing ongoing experience. of course, even if i am right that understanding captures what dilthey, beck, and wampold have in mind, it may still be the case that it does not play the role i am claiming it does. a third objection that one might have along these lines is that this notion of understanding is both too broad and too vague to be useful. it is broad in the sense described above, covering both subconscious and conscious constructs. it is vague in the sense that i have not specified precisely what set of crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 23 things is supposed to be captured by mental constructs. in response to this worry, i stress that both the broadness of the notion and the vagueness with respect to what kind of entities are being referred to are features of the account, not bugs. much of my argument turns on the fact that we have an inadequate theory of how exactly these influences – beliefs, values, goals, expectations, and the like – give rise to psychopathological experience. if we knew which constructs were relevant, how they interact with one another, and how to specify and measure them, the prospects of the data-driven approach addressing the bottleneck of understanding would be significantly improved (see poldrack & yarkoni, 2016 for an exposition of the view that there is unlikely to be any single ‘correct’ way to carve psychology into a set of mental constructs). as such, i gladly acknowledge that my account is underspecified. in my view, this reflects an appropriate degree of epistemic uncertainty with respect to the details. 5.2 on the bottleneck metaphor one possible objection to casting understanding as a bottleneck is that the mental states that make up understanding do not bear the appropriate relation to the other components in the system to justify the metaphor. in particular, a proponent of the biological approach to psychiatry might argue that a suitable pharmacological intervention not only alters a patient’s brain states, but also alters their mental states. this interlocuter could suggest that treating the underlying neurobiology and the mental dysfunction as independent smacks of cartesian dualism. on this view, regardless of the messy details, as long as subjective psychopathology is implemented by neural wetware, it stands to reason that intervening on that wetware constitutes intervening on the patient’s relevant mental constructs. this threatens the applicability of the bottleneck metaphor, which implies that pharmacological interventions are causally upstream of the mental constructs that feature in understanding. it is certainly true that the complex relation holding between brain and mind makes psychiatry significantly trickier to reason about than other systems which experience bottlenecks, such as factory production lines. however, for the bottleneck metaphor to be useful, all that is required is a dissociation between the neural changes known to be induced by pharmacological agents and the development of understanding. notice that such a dissociation is a function of our state of knowledge, and thus not a strong claim about the relation between brain and mind. the utility of treating the cognitive and chemical as dissociable can be seen in the cognitive neuropsychological theory of how antidepressants work (harmer et al., 2017, 2009). on this view, the early biological effects of selective serotonin reuptake inhibitors lead to a positive shift in the processing of emotionally salient information, which then leads to improved mood. this case highlights how scholars still find it useful to consider the chemical and the cognitive as functionally distinct. of course, harmer and colleagues are well aware that the cognitive effects crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 24 they point to are implemented by further biological changes (such as increased neural plasticity). the point is that the precise details of these changes are not yet known and hence cannot be targeted directly by pharmacological intervention (see also section 4.2). further, it has been shown that no specific feature of brain biology is sufficient to demarcate healthy and pathological populations (first et al., 2018; holmes & patrick, 2018). this implies that the dissociation should hold for any pharmacological agent which targets a particular biomarker, at least to the extent that there is uncertainty regarding the agent’s downstream effects. of course, it is possible that future technology will enable fundamentally different kinds of neurobiological intervention. however, this does not threaten the bottleneck framing, which is specifically intended to characterize the challenge psychiatric science faces today, given our current epistemic and technological situation. in sum, i maintain it is reasonable to consider patient understanding a bottleneck to robust recovery. 6 conclusion the focus of this article is the question: does the data-driven approach have transformative potential for psychiatric science? in providing an answer to this question, i focused on a core limitation. namely, that patient understanding functions as a bottleneck for recovery, and the data-driven approach is ill-suited to address that bottleneck. if this view is correct, talk of an ai revolution in mental healthcare is premature. given the current state of psychiatric science, the efficient allocation of treatments, which the data-driven approach can deliver, is likely to yield only modest improvements to mental health outcomes. in order to unlock additional benefits, further knowledge of the specificities of psychopathology and how treatments can be tailored to induce the understanding necessary for recovery are needed. this may involve engaging with research on features of psychiatry that are difficult to formalize, such as empathy, non-verbal communication, and the framing of treatments (priebe et al., 2020), the role of placebo effects in treatment outcomes (enck & zipfel, 2019), and the practitioner-patient relationship (horn & weisz, 2020). as explained in section 4.2, this line of research involves many researcher degrees of freedom and cannot proceed in a theory-neutral way. while the data-driven approach can contribute to these endeavors, it will do so best in collaboration with a theory-driven attempt to understand mental disorders. acknowledgments work on this paper was funded by the volkswagen foundation grant ”explainable intelligent systems” (eis) (grant numbers 9b 830 and 98 509). the author would like to thank all the participants of the minds, models, and mechanisms workshop (funded by dfg grant number 446794119) for stimulating discussions of these issues as well as the anonymous reviewers for many useful suggestions. thanks also to theresa waclawek, chiara caporuscio, and timo speith for their comments on a draft. finally, thanks to all members of the explainable intelligent systems research group for helpful feedback on the initial ideas from which this article grew. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 25 references american psychiatric association. (2022). diagnostic and statistical manual of mental disorders. https://doi.org/10.1176/ap pi.books.9780890425787 anthony, w. a. (2000). a recovery-oriented service system: setting some system level standards. psychiatric rehabilitation journal, 24(2), 159–168. https://doi.org/10.1037/h0095104 arbabshirani, m. r., plis, s., sui, j., & calhoun, v. d. (2017). single subject prediction of brain disorders in neuroimaging: promises and pitfalls. neuroimage, 145(pt b), 137–165. https://doi.org/10.1016/j.neuroimage.2016.02.079 baker, t. b., mcfall, r. m., & shoham, v. (2008). current status and future prospects of clinical psychology. psychological science in the public interest, 9(2), 67–103. https://doi.org/10.1111/j.1539-6053.2009.01036.x balaskas, a., schueller, s. m., cox, a. l., & doherty, g. (2021). ecological momentary interventions for mental health: a scoping review. plos one, 16(3), e0248152. https://doi.org/10.1371/journal.pone.0248152 battaglia, p. w., hamrick, j. b., bapst, v., sanchez-gonzalez, a., zambaldi, v., malinowski, m., tacchetti, a., raposo, d., santoro, a., faulkner, r., gulcehre, c., song, f., ballard, a., gilmer, j., dahl, g., vaswani, a., allen, k., nash, c., langston, v., … pascanu, r. (2018). relational inductive biases, deep learning, and graph networks. arxiv:1806.01261 [cs, stat]. http://arxiv.org/abs/1806.01261 beck, a. t. (1985). cognitive therapy, behavior therapy, psychoanalysis, and pharmacotherapy. in m. j. mahoney & a. freeman (eds.), cognition and psychotherapy (pp. 325–347). springer us. https://doi.org/10.1007/978-1-4684-75623_14 beck, a. t., & haigh, e. a. p. (2014). advances in cognitive theory and therapy: the generic cognitive model. annual review of clinical psychology, 10, 1–24. https://doi.org/10.1146/annurev-clinpsy-032813-153734 bengio, y., courville, a., & vincent, p. (2013). representation learning: a review and new perspectives. ieee transactions on pattern analysis and machine intelligence, 35(8), 1798–1828. https://doi.org/10.1109/tpami.2013.50 bennett, d., silverstein, s. m., & niv, y. (2019). the two cultures of computational psychiatry. jama psychiatry, 76(6), 563–564. https://doi.org/10.1001/jamapsychiatry.2019.0231 bickman, l. (2020). improving mental health services: a 50-year journey from randomized experiments to artificial intelligence and precision mental health. administration and policy in mental health, 47(5), 795–843. https://doi.org/10.100 7/s10488-020-01065-8 boer, n. s. de, bruin, l. c. de, geurts, j. j. g., & glas, g. (2021). the network theory of psychiatric disorders: a critical assessment of the inclusion of environmental factors. frontiers in psychology, 12, 221. https://doi.org/10.3389/fpsyg. 2021.623970 borsboom, d., cramer, a. o. j., & kalis, a. (2019). brain disorders? not really: why network structures block reductionism in psychopathology research. behavioral and brain sciences, 42, e2. https://doi.org/10.1017/s0140525x17002266 bronstein, m. m., bruna, j., cohen, t., & veličković, p. (2021). geometric deep learning: grids, groups, graphs, geodesics, and gauges. arxiv:2104.13478 [cs, stat]. http://arxiv.org/abs/2104.13478 brown, c., story, g. w., mourão-miranda, j., & baker, j. t. (2021). will artificial intelligence eventually replace psychiatrists? the british journal of psychiatry, 218(3), 131–134. https://doi.org/10.1192/bjp.2019.245 budd, r., & hughes, i. (2009). the dodo bird verdict–controversial, inevitable and important: a commentary on 30 years of meta-analyses. clinical psychology & psychotherapy, 16(6), 510–522. https://doi.org/10.1002/cpp.648 bzdok, d., & meyer-lindenberg, a. (2018). machine learning for precision psychiatry: opportunities and challenges. biological psychiatry: cognitive neuroscience and neuroimaging, 3. https://doi.org/10.1016/j.bpsc.2017.11.007 cearns, m., hahn, t., & baune, b. t. (2019). recommendations and future directions for supervised machine learning in psychiatry. translational psychiatry, 9(1), 1–12. https://doi.org/10.1038/s41398-019-0607-2 chekroud, a. m., bondar, j., delgadillo, j., doherty, g., wasil, a., fokkema, m., cohen, z., belgrave, d., derubeis, r., iniesta, r., dwyer, d., & choi, k. (2021). the promise of machine learning in predicting treatment outcomes in psychiatry. world psychiatry: official journal of the world psychiatric association (wpa), 20(2), 154–170. https://doi.org/10.1002/ wps.20882 chekroud, a. m., zotti, r. j., shehzad, z., gueorguieva, r., johnson, m. k., trivedi, m. h., cannon, t. d., krystal, j. h., & corlett, p. r. (2016). cross-trial prediction of treatment outcome in depression: a machine learning approach. the lancet. psychiatry, 3(3), 243–250. https://doi.org/10.1016/s2215-0366(15)00471-x chollet, f. (2021). deep learning with python, second edition. simon; schuster. cichy, r. m., & kaiser, d. (2019). deep neural networks as scientific models. trends in cognitive sciences, 23(4), 305–317. https://doi.org/10.1016/j.tics.2019.01.009 coulombe, s., radziszewski, s., meunier, s., provencher, h., hudon, c., roberge, p., provencher, m. d., & houle, j. (2016). profiles of recovery from mood and anxiety disorders: a person-centered exploration of people’s engagement in self-management. frontiers in psychology, 7. https://doi.org/10.3389/fpsyg.2016.00584 crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1176/appi.books.9780890425787 https://doi.org/10.1176/appi.books.9780890425787 https://doi.org/10.1037/h0095104 https://doi.org/10.1016/j.neuroimage.2016.02.079 https://doi.org/10.1111/j.1539-6053.2009.01036.x https://doi.org/10.1371/journal.pone.0248152 http://arxiv.org/abs/1806.01261 https://doi.org/10.1007/978-1-4684-7562-3_14 https://doi.org/10.1007/978-1-4684-7562-3_14 https://doi.org/10.1146/annurev-clinpsy-032813-153734 https://doi.org/10.1109/tpami.2013.50 https://doi.org/10.1001/jamapsychiatry.2019.0231 https://doi.org/10.1007/s10488-020-01065-8 https://doi.org/10.1007/s10488-020-01065-8 https://doi.org/10.3389/fpsyg.2021.623970 https://doi.org/10.3389/fpsyg.2021.623970 https://doi.org/10.1017/s0140525x17002266 http://arxiv.org/abs/2104.13478 https://doi.org/10.1192/bjp.2019.245 https://doi.org/10.1002/cpp.648 https://doi.org/10.1016/j.bpsc.2017.11.007 https://doi.org/10.1038/s41398-019-0607-2 https://doi.org/10.1002/wps.20882 https://doi.org/10.1002/wps.20882 https://doi.org/10.1016/s2215-0366(15)00471-x https://doi.org/10.1016/j.tics.2019.01.009 https://doi.org/10.3389/fpsyg.2016.00584 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 26 cuevas, las c. d., peñate, w., & rivera, de l. (2014). to what extent is treatment adherence of psychiatric patients influenced by their participation in shared decision making? patient preference and adherence, 8, 1547–1553. https://doi.org/10.2 147/ppa.s73029 cuthbert, b. n., & insel, t. r. (2013). toward the future of psychiatric diagnosis: the seven pillars of rdoc. bmc medicine, 11(1), 126. https://doi.org/10.1186/1741-7015-11-126 dembic, s. (2023). mental disorder: an ability-based view. philosophy and the mind sciences, 4. https://doi.org/10.33735/p himisci.2023.9630 dilthey, w. (1977). ideas concerning a descriptive and analytic psychology (1894). in w. dilthey (ed.), descriptive psychology and historical understanding (pp. 21–120). springer netherlands. https://doi.org/10.1007/978-94-009-9658-8_2 (original work published 1984) domingos, p. (2012). a few useful things to know about machine learning. communications of the acm, 55(10), 78–87. https://doi.org/10.1145/2347736.2347755 doyle, y., egan, s., graham, n., & khalifa, k. (2019). non-factive understanding: a statement and defense. journal for general philosophy of science, 50(3), 345–365. https://doi.org/10.1007/s10838-019-09469-3 durstewitz, d., koppe, g., & meyer-lindenberg, a. (2019). deep neural networks in psychiatry. molecular psychiatry, 24(11), 1583–1598. https://doi.org/10.1038/s41380-019-0365-9 dwyer, d. b., falkai, p., & koutsouleris, n. (2018). machine learning approaches for clinical psychology and psychiatry. annual review of clinical psychology, 14, 91–118. https://doi.org/10.1146/annurev-clinpsy-032816-045037 elhage, n., nanda, n., olsson, c., henighan, t., joseph, n., mann, b., askell, a., bai, y., chen, a., conerly, t., dassarma, n., drain, d., ganguli, d., hatfield-dodds, z., hernandez, d., jones, a., kernion, j., lovitt, l., ndousse, k., … olah, c. (2021). a mathematical framework for transformer circuits. transformer circuits thread. https://transformercircuits.pub/2021/framework/index.html elujide, i., fashoto, s. g., fashoto, b., mbunge, e., folorunso, s. o., & olamijuwon, j. o. (2021). application of deep and machine learning techniques for multi-label classification performance on psychotic disorder diseases. informatics in medicine unlocked, 23, 100545. https://doi.org/10.1016/j.imu.2021.100545 enck, p., & zipfel, s. (2019). placebo effects in psychotherapy: a framework. frontiers in psychiatry, 10. https://doi.org/10 .3389/fpsyt.2019.00456 engel, g. l. (1977). the need for a new medical model: a challenge for biomedicine. science (new york, n.y.), 196(4286), 129–136. https://doi.org/10.1126/science.847460 ewbank, m. p., cummins, r., tablan, v., bateup, s., catarino, a., martin, a. j., & blackwell, a. d. (2020). quantifying the association between psychotherapy content and clinical outcomes using deep learning. jama psychiatry, 77(1), 35–43. https://doi.org/10.1001/jamapsychiatry.2019.2664 ewbank, m. p., cummins, r., tablan, v., catarino, a., buchholz, s., & blackwell, a. d. (2021). understanding the relationship between patient language and outcomes in internet-enabled cognitive behavioural therapy: a deep learning approach to automatic coding of session transcripts. psychotherapy research, 31(3), 300–312. https://doi.org/10.1080/10503307 .2020.1788740 feczko, e., miranda-dominguez, o., marr, m., graham, a. m., nigg, j. t., & fair, d. a. (2019). the heterogeneity problem: approaches to identify psychiatric subtypes. trends in cognitive sciences, 23(7), 584–601. https://doi.org/10.1016/j.ti cs.2019.03.009 fernandes, b. s., williams, l. m., steiner, j., leboyer, m., carvalho, a. f., & berk, m. (2017). the new field of “precision psychiatry.” bmc medicine, 15(1), 80. https://doi.org/10.1186/s12916-017-0849-x first, m. b., drevets, w. c., carter, c., dickstein, d. p., kasoff, l., kim, k. l., mcconathy, j., rauch, s., saad, z. s., savitz, j., seymour, k. e., sheline, y. i., & zubieta, j.-k. (2018). clinical applications of neuroimaging in psychiatric disorders. the american journal of psychiatry, 175(9), 915–916. https://doi.org/10.1176/appi.ajp.2018.1750701 frances, a. (2009). whither dsm-v? the british journal of psychiatry: the journal of mental science, 195(5), 391–392. https://doi.org/10.1192/bjp.bp.109.073932 frank, j. d., & frank, j. b. (1993). persuasion and healing: a comparative study of psychotherapy. jhu press. garcía-gutiérrez, m. s., navarrete, f., sala, f., gasparyan, a., austrich-olivares, a., & manzanares, j. (2020). biomarkers in psychiatry: concept, definition, types and relevance to the clinical reality. frontiers in psychiatry, 11, 432. https://doi.org/10.3389/fpsyt.2020.00432 gauld, c., micoulaud-franchi, j.-a., & dumas, g. (2021). comment on starke et al.: “computing schizophrenia: ethical challenges for machine learning in psychiatry”: from machine learning to student learning: pedagogical challenges for psychiatry. psychological medicine, 51(14), 2509–2511. https://doi.org/10.1017/s0033291720003906 ghaemi, s. n. (2007). existence and pluralism: the rediscovery of karl jaspers. psychopathology, 40(2), 75–82. https://doi.org/10.1159/000098487 crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.2147/ppa.s73029 https://doi.org/10.2147/ppa.s73029 https://doi.org/10.1186/1741-7015-11-126 https://doi.org/10.33735/phimisci.2023.9630 https://doi.org/10.33735/phimisci.2023.9630 https://doi.org/10.1007/978-94-009-9658-8_2 https://doi.org/10.1145/2347736.2347755 https://doi.org/10.1007/s10838-019-09469-3 https://doi.org/10.1038/s41380-019-0365-9 https://doi.org/10.1146/annurev-clinpsy-032816-045037 https://transformer-circuits.pub/2021/framework/index.html https://transformer-circuits.pub/2021/framework/index.html https://doi.org/10.1016/j.imu.2021.100545 https://doi.org/10.3389/fpsyt.2019.00456 https://doi.org/10.3389/fpsyt.2019.00456 https://doi.org/10.1126/science.847460 https://doi.org/10.1001/jamapsychiatry.2019.2664 https://doi.org/10.1080/10503307.2020.1788740 https://doi.org/10.1080/10503307.2020.1788740 https://doi.org/10.1016/j.tics.2019.03.009 https://doi.org/10.1016/j.tics.2019.03.009 https://doi.org/10.1186/s12916-017-0849-x https://doi.org/10.1176/appi.ajp.2018.1750701 https://doi.org/10.1192/bjp.bp.109.073932 https://doi.org/10.3389/fpsyt.2020.00432 https://doi.org/10.1017/s0033291720003906 https://doi.org/10.1159/000098487 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 27 gillan, c. m., & whelan, r. (2017). what big data can do for treatment in psychiatry. current opinion in behavioral sciences, 18, 34–42. https://doi.org/10.1016/j.cobeha.2017.07.003 glover, j. (2020). psychiatry, folk psychology, and the impact of neuroscience a response to steve hyman’s loebel lectures. in j. savulescu, r. roache, w. davies, & j. p. loebel (eds.), psychiatry reborn: biopsychosocial psychiatry in modern medicine (pp. 301–319). oxford university press. gould, i. c., shepherd, a. m., laurens, k. r., cairns, m. j., carr, v. j., & green, m. j. (2014). multivariate neuroanatomical classification of cognitive subtypes in schizophrenia: a support vector machine learning approach. neuroimage: clinical, 6, 229–236. https://doi.org/10.1016/j.nicl.2014.09.009 grimm, s. r. (2019). varieties of understanding. in s. r. grimm (ed.), varieties of understanding: new perspectives from philosophy, psychology, and theology (pp. 2–13). oxford university press. https://doi.org/10.1093/oso/9780190860974. 001.0001 harmer, c. j., duman, r. s., & cowen, p. j. (2017). how do antidepressants work? new perspectives for refining future treatment approaches. the lancet. psychiatry, 4(5), 409–418. https://doi.org/10.1016/s2215-0366(17)30015-9 harmer, c. j., goodwin, g. m., & cowen, p. j. (2009). why do antidepressants take so long to work? a cognitive neuropsychological model of antidepressant drug action. the british journal of psychiatry, 195(2), 102–108. https://doi.org/10.1192/bjp.bp.108.051193 health, d. of, & care, s. (2021). mental health recovery plan backed by £500 million. in gov.uk. https://www.gov.uk/gov ernment/news/mental-health-recovery-plan-backed-by-500-million hengartner, m. p., & lehmann, s. n. (2017). why psychiatric research must abandon traditional diagnostic classification and adopt a fully dimensional scope: two solutions to a persistent problem. frontiers in psychiatry, 8, 101. https://doi.org/10.3389/fpsyt.2017.00101 hills, a. (2016). understanding why. noûs, 50(4), 661–688. https://doi.org/10.1111/nous.12092 holmes, a. j., & patrick, l. m. (2018). the myth of optimality in clinical neuroscience. trends in cognitive sciences, 22(3), 241–257. https://doi.org/10.1016/j.tics.2017.12.006 horn, r. l., & weisz, j. r. (2020). can artificial intelligence improve psychotherapy research and practice. administration and policy in mental health, 47(5), 852–855. https://doi.org/10.1007/s10488-020-01056-9 hyman, s. e., & mcconnell, d. (2020). mental illness: the collision of meaning with mechanism. in psychiatry reborn: biopsychosocial psychiatry in modern medicine (pp. 263–289). oxford university press. https://oxfordmedicine.com/v iew/10.1093/med/9780198789697.001.0001/med-9780198789697 insel, t. r., collins, p. y., & hyman, s. e. (2015). darkness invisible: the hidden global costs of mental illness. foreign affairs, 94(1), 127–135. ivanov, i., & schwartz, j. m. (2021). why psychotropic drugs don’t cure mental illness—but should they? frontiers in psychiatry, 12. https://doi.org/10.3389/fpsyt.2021.579566 jacob, k. s. (2015). recovery model of mental illness: a complementary approach to psychiatric care. indian journal of psychological medicine, 37(2), 117–119. https://doi.org/10.4103/0253-7176.155605 jaspers, k. (1997). general psychopathology. jhu press. (original work published 1959) johansson, h., & eklund, m. (2003). patients’ opinion on what constitutes good psychiatric care. scandinavian journal of caring sciences, 17(4), 339–346. https://doi.org/10.1046/j.0283-9318.2003.00233.x joober, r., & tabbane, k. (2019). from the neo-kraepelinian framework to the new mechanical philosophy of psychiatry: regaining common sense. journal of psychiatry & neuroscience : jpn, 44(1), 3–7. https://doi.org/10.1503/jpn.180240 kendler, k. s. (2012). levels of explanation in psychiatric and substance use disorders: implications for the development of an etiologically based nosology. molecular psychiatry, 17(1), 11–21. https://doi.org/10.1038/mp.2011.70 kendler, k. s. (2016). the nature of psychiatric disorders. world psychiatry, 15(1), 5–12. https://doi.org/10.1002/wps.20292 kendler, k. s., & campbell, j. (2014). expanding the domain of the understandable in psychiatric illness: an updating of the jasperian framework of explanation and understanding. psychological medicine, 44(1), 1–7. https://doi.org/10.101 7/s0033291712003030 kendler, k. s., & gyngell, c. (2020). multilevel interactions and the dappled causal world of psychiatric disorders. in psychiatry reborn: biopsychosocial psychiatry in modern medicine (pp. 25–45). oxford university press. https://oxford medicine.com/view/10.1093/med/9780198789697.001.0001/med-9780198789697 keogh, e., & mueen, a. (2017). curse of dimensionality. in c. sammut & g. i. webb (eds.), encyclopedia of machine learning and data mining (pp. 314–315). springer us. https://doi.org/10.1007/978-1-4899-7687-1_192 koppe, g., meyer-lindenberg, a., & durstewitz, d. (2021). deep learning for small and big data in psychiatry. neuropsychopharmacology, 46(1), 176–190. https://doi.org/10.1038/s41386-020-0767-z koutsouleris, n., dwyer, d. b., degenhardt, f., maj, c., urquijo-castro, m. f., sanfelici, r., popovic, d., oeztuerk, o., haas, s. s., weiske, j., ruef, a., kambeitz-ilankovic, l., antonucci, l. a., neufang, s., schmidt-kraepelin, c., ruhrmann, s., penzel, n., kambeitz, j., haidl, t. k., … meisenzahl, e. (2021). multimodal machine learning workflows for prediction crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.cobeha.2017.07.003 https://doi.org/10.1016/j.nicl.2014.09.009 https://doi.org/10.1093/oso/9780190860974.001.0001 https://doi.org/10.1093/oso/9780190860974.001.0001 https://doi.org/10.1016/s2215-0366(17)30015-9 https://doi.org/10.1192/bjp.bp.108.051193 https://www.gov.uk/government/news/mental-health-recovery-plan-backed-by-500-million https://www.gov.uk/government/news/mental-health-recovery-plan-backed-by-500-million https://doi.org/10.3389/fpsyt.2017.00101 https://doi.org/10.1111/nous.12092 https://doi.org/10.1016/j.tics.2017.12.006 https://doi.org/10.1007/s10488-020-01056-9 https://oxfordmedicine.com/view/10.1093/med/9780198789697.001.0001/med-9780198789697 https://oxfordmedicine.com/view/10.1093/med/9780198789697.001.0001/med-9780198789697 https://doi.org/10.3389/fpsyt.2021.579566 https://doi.org/10.4103/0253-7176.155605 https://doi.org/10.1046/j.0283-9318.2003.00233.x https://doi.org/10.1503/jpn.180240 https://doi.org/10.1038/mp.2011.70 https://doi.org/10.1002/wps.20292 https://doi.org/10.1017/s0033291712003030 https://doi.org/10.1017/s0033291712003030 https://oxfordmedicine.com/view/10.1093/med/9780198789697.001.0001/med-9780198789697 https://oxfordmedicine.com/view/10.1093/med/9780198789697.001.0001/med-9780198789697 https://doi.org/10.1007/978-1-4899-7687-1_192 https://doi.org/10.1038/s41386-020-0767-z https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 28 of psychosis in patients with clinical high-risk syndromes and recent-onset depression. jama psychiatry, 78(2), 1–15. https://doi.org/10.1001/jamapsychiatry.2020.3604 ladyman, j., lambert, j., & wiesner, k. (2013). what is a complex system? european journal for philosophy of science, 3(1), 33–67. https://doi.org/10.1007/s13194-012-0056-8 laska, k. m., gurman, a. s., & wampold, b. e. (2014). expanding the lens of evidence-based practice in psychotherapy: a common factors perspective. psychotherapy, 51(4), 467–481. https://doi.org/10.1037/a0034332 leder, g., & zawidzki, t. (2023). the skill of mental health: towards a new theory of mental health and disorder. philosophy and the mind sciences, 4. https://doi.org/10.33735/phimisci.2023.9684 leichsenring, f., steinert, c., rabung, s., & ioannidis, j. p. a. (2022). the efficacy of psychotherapies and pharmacotherapies for mental disorders in adults: an umbrella review and meta-analytic evaluation of recent meta-analyses. world psychiatry, 21(1), 133–145. https://doi.org/10.1002/wps.20941 manning, c. d., clark, k., hewitt, j., khandelwal, u., & levy, o. (2020). emergent linguistic structure in artificial neural networks trained by self-supervision. proceedings of the national academy of sciences, 117(48), 30046–30054. https://doi.org/10.1073/pnas.1907367117 mcconnell, d. (2020). the proper place of subjectivity, meaning, and folk psychology in psychiatry. in psychiatry reborn: biopsychosocial psychiatry in modern medicine (pp. 290–303). oxford university press. https://oxfordmedicine.com/v iew/10.1093/med/9780198789697.001.0001/med-9780198789697 mcconnell, d., & snoek, a. (2018). the importance of self-narration in recovery from addiction. philosophy, psychiatry, and psychology, 25(3), 31–44. https://doi.org/10.1353/ppp.2018.0022 mccoy, l. g., nagaraj, s., morgado, f., harish, v., das, s., & celi, l. a. (2020). what do medical students actually need to know about artificial intelligence? npj digital medicine, 3(1), 1–3. https://doi.org/10.1038/s41746-020-0294-7 mcgorry, p. d., & nelson, b. (2019). transdiagnostic psychiatry: premature closure on a crucial pathway to clinical utility for psychiatric diagnosis. world psychiatry, 18(3), 359–360. https://doi.org/10.1002/wps.20679 middleton, h., & moncrieff, j. (2019). critical psychiatry: a brief overview. bjpsych advances, 25(1), 47–54. https://doi.org/10.1192/bja.2018.38 miranda, l., paul, r., pütz, b., koutsouleris, n., & müller-myhsok, b. (2021). systematic review of functional mri applications for psychiatric disease subtyping. frontiers in psychiatry, 12, 665536. https://doi.org/10.3389/fpsyt.2021.665536 najafpour, z., fatemi, a., goudarzi, z., goudarzi, r., shayanfard, k., & noorizadeh, f. (2021). cost-effectiveness of neuroimaging technologies in management of psychiatric and insomnia disorders: a meta-analysis and prospective cost analysis. journal of neuroradiology = journal de neuroradiologie, 48(5), 348–358. https://doi.org/10.1016/j.neurad.202 0.12.003 nie, z., vairavan, s., narayan, v. a., ye, j., & li, q. s. (2018). predictive modeling of treatment resistant depression using data from star*d and an independent clinical study. plos one, 13(6), e0197268. https://doi.org/10.1371/journal.pone .0197268 olah, c., satyanarayan, a., johnson, i., carter, s., schubert, l., ye, k., & mordvintsev, a. (2018). the building blocks of interpretability. distill, 3(3), 10.23915/distill.00010. https://doi.org/10.23915/distill.00010 paulus, m. p. (2015). pragmatism instead of mechanism: a call for impactful biological psychiatry. jama psychiatry, 72(7), 631–632. https://doi.org/10.1001/jamapsychiatry.2015.0497 pelin, h., ising, m., stein, f., meinert, s., meller, t., brosch, k., winter, n. r., krug, a., leenings, r., lemke, h., nenadić, i., heilmann-heimbach, s., forstner, a. j., nöthen, m. m., opel, n., repple, j., pfarr, j., ringwald, k., schmitt, s., … andlauer, t. f. m. (2021). identification of transdiagnostic psychiatric disorder subtypes using unsupervised learning. neuropsychopharmacology: official publication of the american college of neuropsychopharmacology, 46(11), 1895– 1905. https://doi.org/10.1038/s41386-021-01051-0 peterson, b. s. (2019). editorial: common factors in the art of healing. journal of child psychology and psychiatry, 60(9), 927–929. https://doi.org/10.1111/jcpp.13108 poldrack, r. a., & yarkoni, t. (2016). from brain maps to cognitive ontologies: informatics and the search for mental structure. annual review of psychology, 67(1), 587–612. https://doi.org/10.1146/annurev-psych-122414-033729 priebe, s., conneely, m., mccabe, r., & bird, v. (2020). what can clinicians do to improve outcomes across psychiatric treatments: a conceptual review of non-specific components. epidemiology and psychiatric sciences, 29. https://doi.org/10.1017/s2045796019000428 qureshi, m. n. i., min, b., jo, h. j., & lee, b. (2016). multiclass classification for the differential diagnosis on the adhd subtypes using recursive feature elimination and hierarchical extreme learning machine: structural mri study. plos one, 11(8), e0160697. https://doi.org/10.1371/journal.pone.0160697 ramon, s., healy, b., & renouf, n. (2007). recovery from mental illness as an emergent concept and practice in australia and the uk. international journal of social psychiatry, 53(2), 108–122. https://doi.org/10.1177/0020764006075018 crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1001/jamapsychiatry.2020.3604 https://doi.org/10.1007/s13194-012-0056-8 https://doi.org/10.1037/a0034332 https://doi.org/10.33735/phimisci.2023.9684 https://doi.org/10.1002/wps.20941 https://doi.org/10.1073/pnas.1907367117 https://oxfordmedicine.com/view/10.1093/med/9780198789697.001.0001/med-9780198789697 https://oxfordmedicine.com/view/10.1093/med/9780198789697.001.0001/med-9780198789697 https://doi.org/10.1353/ppp.2018.0022 https://doi.org/10.1038/s41746-020-0294-7 https://doi.org/10.1002/wps.20679 https://doi.org/10.1192/bja.2018.38 https://doi.org/10.3389/fpsyt.2021.665536 https://doi.org/10.1016/j.neurad.2020.12.003 https://doi.org/10.1016/j.neurad.2020.12.003 https://doi.org/10.1371/journal.pone.0197268 https://doi.org/10.1371/journal.pone.0197268 https://doi.org/10.23915/distill.00010 https://doi.org/10.1001/jamapsychiatry.2015.0497 https://doi.org/10.1038/s41386-021-01051-0 https://doi.org/10.1111/jcpp.13108 https://doi.org/10.1146/annurev-psych-122414-033729 https://doi.org/10.1017/s2045796019000428 https://doi.org/10.1371/journal.pone.0160697 https://doi.org/10.1177/0020764006075018 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org understanding as a bottleneck for the data-driven approach to psychiatric science 29 regt, h. w. de. (2019). from explanation to understanding: normativity lost? journal for general philosophy of science, 50(3), 327–343. https://doi.org/10.1007/s10838-019-09477-3 rosenzweig, s. (1936). some implicit common factors in diverse methods of psychotherapy. american journal of orthopsychiatry, 6, 412–415. https://doi.org/10.1111/j.1939-0025.1936.tb05248.x rutledge, r. b., chekroud, a. m., & huys, q. j. (2019). machine learning and big data in psychiatry: toward clinical applications. current opinion in neurobiology, 55, 152–159. https://doi.org/10.1016/j.conb.2019.02.006 sathyanarayana rao, t. s., & andrade, c. (2016). classification of psychotropic drugs: problems, solutions, and more problems. indian journal of psychiatry, 58(2), 111–113. https://doi.org/10.4103/0019-5545.183771 schrimpf, m., kubilius, j., lee, m. j., ratan murty, n. a., ajemian, r., & dicarlo, j. j. (2020). integrative benchmarking to advance neurally mechanistic models of human intelligence. neuron, 108(3), 413–423. https://doi.org/10.1016/j.neur on.2020.07.040 sharma, a., & verbeke, w. j. m. i. (2020). improving diagnosis of depression with xgboost machine learning model and a large biomarkers dutch dataset (n = 11,081). frontiers in big data, 3. https://doi.org/10.3389/fdata.2020.00015 sheu, y.-h. (2020). illuminating the black box: interpreting deep neural network models for psychiatric research. frontiers in psychiatry, 11, 551299. https://doi.org/10.3389/fpsyt.2020.551299 shickel, b., tighe, p. j., bihorac, a., & rashidi, p. (2018). deep ehr: a survey of recent advances in deep learning techniques for electronic health record (ehr) analysis. ieee journal of biomedical and health informatics, 22(5), 1589–1604. https://doi.org/10.1109/jbhi.2017.2767063 shorter, e. (1997). a history of psychiatry: from the era of the asylum to the age of prozac. john wiley & sons. si, y., du, j., li, z., jiang, x., miller, t., wang, f., jim zheng, w., & roberts, k. (2021). deep representation learning of patient data from electronic health records (ehr): a systematic review. journal of biomedical informatics, 115, 103671. https://doi.org/10.1016/j.jbi.2020.103671 stanghellini, g., & broome, m. r. (2014). psychopathology as the basic science of psychiatry. the british journal of psychiatry, 205(3), 169–170. https://doi.org/10.1192/bjp.bp.113.138974 starke, g., clercq, e. d., borgwardt, s., & elger, b. s. (2021). computing schizophrenia: ethical challenges for machine learning in psychiatry. psychological medicine, 51(15), 2515–2521. https://doi.org/10.1017/s0033291720001683 stoyanov, d., & maes, m. h. (2021). how to construct neuroscience-informed psychiatric classification? towards nomothetic networks psychiatry. world journal of psychiatry, 11(1), 1–12. https://doi.org/10.5498/wjp.v11.i1.1 su, c., xu, z., pathak, j., & wang, f. (2020). deep learning in mental health outcome research: a scoping review. translational psychiatry, 10(1), 1–26. https://doi.org/10.1038/s41398-020-0780-3 suhara, y., xu, y., & pentland, a. ’sandy’. (2017). deepmood: forecasting depressed mood based on self-reported histories via recurrent neural networks. proceedings of the 26th international conference on world wide web, 715–724. https://doi.org/10.1145/3038912.3052676 thornton, t. (2020). psychiatry’s inchoate wish for a paradigm shift and the biopsychosocial model of mental illness. in psychiatry reborn: biopsychosocial psychiatry in modern medicine (pp. 229–239). oxford university press. https://oxfordmedicine.com/view/10.1093/med/9780198789697.001.0001/med-9780198789697 tonekaboni, s., joshi, s., mccradden, m. d., & goldenberg, a. (2019). what clinicians want: contextualizing explainable machine learning for clinical end use. proceedings of the 4th machine learning for healthcare conference, 359–380. https://proceedings.mlr.press/v106/tonekaboni19a.html trivedi, m. h. (2016). right patient, right treatment, right time: biosignatures and precision medicine in depression. world psychiatry, 15(3), 237–238. https://doi.org/10.1002/wps.20371 trust, w. b. (2020). a unified vision for the future of mental health, addiction, and well-being in the united states. in well being trust. https://wellbeingtrust.org/press-releases/ceos-from-14-top-mental-health-organizations-join-together/ tsou, j. y. (2021). philosophy of psychiatry. elements in the philosophy of science. https://doi.org/10.1017/9781108588485 vigo, d., thornicroft, g., & atun, r. (2016). estimating the true global burden of mental illness. the lancet. psychiatry, 3(2), 171–178. https://doi.org/10.1016/s2215-0366(15)00505-2 wampold, b. e. (2015). how important are the common factors in psychotherapy? an update. world psychiatry, 14(3), 270–277. https://doi.org/10.1002/wps.20238 wampold, b. e., & budge, s. l. (2012). the 2011 leona tyler award address: the relationship—and its relationship to the common and specific factors of psychotherapy. the counseling psychologist, 40(4), 601–623. https://doi.org/10.1177/ 0011000011432709 wampold, b. e., & imel, z. e. (2015). the great psychotherapy debate: the evidence for what makes psychotherapy work, 2nd ed. routledge/taylor & francis group. wampold, b. e., imel, z. e., bhati, k. s., & johnson-jennings, m. d. (2007). insight as a common factor. in insight in psychotherapy (pp. 119–139). american psychological association. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s10838-019-09477-3 https://doi.org/10.1111/j.1939-0025.1936.tb05248.x https://doi.org/10.1016/j.conb.2019.02.006 https://doi.org/10.4103/0019-5545.183771 https://doi.org/10.1016/j.neuron.2020.07.040 https://doi.org/10.1016/j.neuron.2020.07.040 https://doi.org/10.3389/fdata.2020.00015 https://doi.org/10.3389/fpsyt.2020.551299 https://doi.org/10.1109/jbhi.2017.2767063 https://doi.org/10.1016/j.jbi.2020.103671 https://doi.org/10.1192/bjp.bp.113.138974 https://doi.org/10.1017/s0033291720001683 https://doi.org/10.5498/wjp.v11.i1.1 https://doi.org/10.1038/s41398-020-0780-3 https://doi.org/10.1145/3038912.3052676 https://oxfordmedicine.com/view/10.1093/med/9780198789697.001.0001/med-9780198789697 https://proceedings.mlr.press/v106/tonekaboni19a.html https://doi.org/10.1002/wps.20371 https://wellbeingtrust.org/press-releases/ceos-from-14-top-mental-health-organizations-join-together/ https://doi.org/10.1017/9781108588485 https://doi.org/10.1016/s2215-0366(15)00505-2 https://doi.org/10.1002/wps.20238 https://doi.org/10.1177/0011000011432709 https://doi.org/10.1177/0011000011432709 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org barnaby crook 30 wilkenfeld, d. a. (2013). understanding as representation manipulability. synthese, 190(6), 997–1016. https://doi.org/10.1 007/s11229-011-0055-x wolpert, m., pote, i., & sebastian, c. l. (2021). identifying and integrating active ingredients for mental health. the lancet psychiatry, 8(9), 741–743. https://doi.org/10.1016/s2215-0366(21)00283-2 world health organization. (2019). special initiative for mental health (2019–2023). https://www.who.int/publicationsdetail-redirect/special-initiative-for-mental-health-(2019-2023) yamins, d. l. k., & dicarlo, j. j. (2016). using goal-driven deep learning models to understand sensory cortex. nature neuroscience, 19(3), 356–365. https://doi.org/10.1038/nn.4244 zachar, p., & kendler, k. s. (2017). the philosophy of nosology. annual review of clinical psychology, 13, 49–71. https://doi.org/10.1146/annurev-clinpsy-032816-045020 zednik, c. (2021). solving the black box problem: a normative framework for explainable artificial intelligence. philosophy & technology, 34(2), 265–288. https://doi.org/10.1007/s13347-019-00382-7 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. crook, b. (2023). understanding as a bottleneck for the data-driven approach to psychiatric science. philosophy and the mind sciences, 4, 5. https://doi.org/10.33735/phimisci.2023.9658 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11229-011-0055-x https://doi.org/10.1007/s11229-011-0055-x https://doi.org/10.1016/s2215-0366(21)00283-2 https://www.who.int/publications-detail-redirect/special-initiative-for-mental-health-(2019-2023) https://www.who.int/publications-detail-redirect/special-initiative-for-mental-health-(2019-2023) https://doi.org/10.1038/nn.4244 https://doi.org/10.1146/annurev-clinpsy-032816-045020 https://doi.org/10.1007/s13347-019-00382-7 https://doi.org/10.33735/phimisci.2023.9658 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction the data-driven approach motivating the approach: psychiatry in crisis the methodology of the data-driven approach problem definition data acquisition model design and training validation and deployment the data-driven approach in practice: an example the philosophical commitments of the data-driven approach theory neutrality function orientation the bottleneck of understanding understanding as possession of adaptive mental constructs understanding as a bottleneck to recovery the data-driven approach and the bottleneck of understanding limits of a function-oriented approach limits of a theory-neutral approach translation problems for data-driven insights implications responses to objections on understanding on the bottleneck metaphor conclusion dimensions of animal wellbeing. dimensions of animal wellbeing leonard dunga(leonard.dung@ruhr-uni-bochum.de) abstract whether animals fare well or not is of ethical significance. for this reason, their capacity for wellbeing, i.e., how good or bad the lives of animals can go, is of ethical significance as well. i assume that the wellbeing of most animals is mainly determined by their phenomenally conscious experiences. if consciousness differences between species determine wellbeing differences, then the kinds of conscious experience species are capable of may entail that some species systematically (can) have higher or lower wellbeing than others. then, i argue that not all phenomenally conscious states contribute to wellbeing equally. i discuss which features of consciousness are constituents of wellbeing and which can, for ethical purposes, be ignored. in addition, i scrutinize how much different features of experience contribute to wellbeing and how their presence can be detected empirically. this way, this paper exemplifies a novel consciousness-centered approach for the empirical investigation of animal wellbeing. the strengths and weaknesses of this approach are analyzed. while subsequent research is needed to refine the framework, i already note some preliminary implications for animal ethics. keywords animal consciousness ∙ animal welfare science ∙ hedonism ∙ judgement bias ∙ animal ethics ∙ wellbeing capacity whether animals fare well or not is of ethical significance. this is why it is important to investigate animal wellbeing. in this paper, i will tackle an important problem for animal welfare assessments which has only recently entered the literature (browning, 2023; gaffney et al., 2023; višak, 2022): interspecies differences in the capacity for wellbeing. due to differences in other capacities, particularly in the kinds of conscious experience different animal species are capable of, some species may be capable of higher (or lower) levels of wellbeing than others. for instance, it has been argued that elephants can experience ptsd (hoffman, 2020). if this is true and some other animals lack this kind of experience, then the lowest possible wellbeing levels of elephants might be lower than that of some other species. given these assumptions and all other things being equal, we have stronger reasons to make sure that elephants live in good conditions than animals with a lower wellbeing capacity. ainstitute of philosophy ii, ruhr-universität bochum. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-4154-5560 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 2 the main goal of this paper is to develop a novel framework for the investigation of interspecies differences in wellbeing capacity and, by extension, interspecies comparisons of wellbeing. those comparisons are notoriously hard to make since the measures typically used in animal welfare science, e.g. preference tests (dawkins, 2021), appear insufficient. the preferences of an animal can tell us which of two states it prefers and thus, presumably, where its wellbeing level is higher. however, it is not clear how we can reliably determine that an animal has a level of wellbeing higher than another (if both are positive). even if the first animal has its highest possible wellbeing level, the other animal might have a higher wellbeing if it has a higher capacity for wellbeing. even if a bee is on its highest possible wellbeing level, the wellbeing of a chimp might nevertheless be higher if the latter has a higher wellbeing capacity. since we currently lack an accepted framework for tackling interspecies differences in wellbeing capacity, this is what this paper aims to provide. i presuppose that interspecies differences in wellbeing capacity, if they exist (višak, 2022), are primarily rooted in interspecies differences in conscious experience. this motivates a three-step research program. first, we need to develop a classificatory scheme which is detailed enough to describe all relevant aspects of the conscious experience of various animals and to make room for distinctions between different kinds of consciousness. thankfully, birch et al. (2020) have already made significant advances on this task. so, i will use their framework as a basis.1 second, it is necessary to investigate which features of consciousness are constituents of wellbeing and which can, for ethical purposes, be ignored. this way, we can arrive at a framework which describes the different dimensions which jointly constitute animal wellbeing. this paper contributes to this endeavor. third, subsequent empirical research efforts into animal consciousness should consider ethical reasons to prioritize features of animal consciousness which constitute wellbeing. once we have gained sufficient knowledge of the relevant features of animal consciousness, we are in a position to ascertain animal wellbeing in a way which respects the possibility of interspecies differences in wellbeing capacity and thus enables interspecies comparisons of wellbeing. this paper is structured as follows. in §1, i will introduce the terminology relevant to frame this discussion. i will elaborate on the concepts of wellbeing, consciousness and constituent of wellbeing, situate this discussion with respect to the most important philosophical theories of wellbeing and motivate the importance of gaining knowledge about interspecies differences in wellbeing capacity. §2 contains a summary of the framework for the description of consciousness developed by birch et al. (2020) which i subsequently rely on. §3 and §4 make up the core of the paper. in these sections, i examine which features of animal consciousness are relevant to wellbeing, why and how much. where not already done by birch et al., i suggest empirical tests for these features. in §5, i will discuss how the 1for a recent multi-dimensional framework of animal consciousness which builds upon birch et al., see dung & newen (2023). dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 3 research program i propose differs from an approach to interspecies comparisons of animal wellbeing which is rooted in the proprietary tools of animal welfare science, as developed by browning (2023), and where its respective strengths and weaknesses lie. §6 mentions some potential ethical implications of the framework and summarizes the paper. 1 animal wellbeing 1.1 wellbeing and consciousness this paper explores which aspects of consciousness contribute to wellbeing. my usage of the terms ‘wellbeing’ and ‘consciousness’ is in line with the philosophical literature (fletcher, 2015; van gulick, 2022). in this paper, i use ‘consciousness’ as a shorthand for ‘phenomenal consciousness’. a being possesses phenomenal consciousness if and only if it undergoes states which are phenomenally conscious. mental states are phenomenally conscious if and only if they involve a subjective experiential feeling like smelling coffee or seeing a red rose often do in humans. this pertains to how the state is experienced from the first-person perspective respectively “what it’s like” (nagel, 1974) to be in such a state. comparative cognition researchers often refer to the same kind of property (e.g., gibbons et al., 2022; mason & lavery, 2022; nieder et al., 2020; panksepp, 2010). wellbeing is what is intrinsically good for the very person whose wellbeing it is (crisp, 2021). in other words, wellbeing grounds the prudential value of a life. this value is intrinsic (or non-instrumental), i.e., the value of wellbeing is independent from its relation to other states it may bring about, and it is conceptually independent from whether a life is morally good. prudential value only concerns whether things are good for the person in question. importantly, negative wellbeing is what is intrinsically bad for the person who is subject to it. when philosophers discuss what wellbeing is, they are asking about the property (or the set of properties) which determines to what extent someone's life is intrinsically and non-morally good or bad, i.e., they are asking which property wellbeing consists in or is constituted by. happiness may be one example. i will call this property a “ground” or “constituent” of wellbeing. like wellbeing itself, constituents of wellbeing are intrinsically good or bad. this distinguishes them from the conditions for wellbeing (browning, 2020). conditions for wellbeing may be things like health or a comfortable shelter. conditions for wellbeing are only non-intrinsically valuable, for their value derives from the fact that they causally affect wellbeing in virtue of affecting the constituents of wellbeing. finally, an indicator of wellbeing or of consciousness is a property which can be used to detect the wellbeing level or conscious experiences of an animal. typically, these indicators are causally related to consciousness or wellbeing. for instance, it has been argued that a stress response, for example expressed in a higher heart rate dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 4 and the release of epinephrine and norepinephrine, indicates compromised welfare (duncan & fraser, 1997). 1.2 theories of wellbeing and broad hedonism in philosophy, there are three broad families of entrenched theories of wellbeing, i.e., views on the grounds of wellbeing. according to hedonistic theories, wellbeing is determined by the balance of conscious mental states with positive and with negative valence (bramble, 2016; crisp, 2006; moore, 2019). positively valenced experiences, like enjoyment of a fine meal, are pleasant and (at least defeasibly) motivate us to continuously seek them. negatively valenced experiences, like pain, are unpleasant and (at least defeasibly) motivate us to avoid them. in addition to its valence, the determinants of the contribution a mental state makes to wellbeing encompass its intensity and its duration. according to hedonism, a life is good when the positive experiences outweigh the negative ones and a life is bad – or “not worth living” – if the negative experiences outweigh the positive ones (as weighted by their intensity and duration). according to desire-fulfillment theories, our wellbeing is determined by the extent to which our desires (weighted by their strength) are fulfilled (bruckner, 2010; heathwood, 2016, 2019). importantly, the constituents of wellbeing are not the positive feelings that may arise from the satisfaction of preferences; otherwise, the theory would collapse into a version of hedonism. instead, the constituent is the experience-independent fact that certain desires have been fulfilled. the last candidates are objective list theories (fletcher, 2013; griffin, 1988; lin, 2014). according to such theories, whether a life is (prudentially) good is determined by whether it contains some subject-independent list of goods. for instance, a good life may require friendship, play, knowledge and pleasure (rice, 2013). objective list theorists hold that some catalog of features is necessary and sufficient for wellbeing irrespective of whether the subject desires them. in contrast to hedonistic views, these features are regarded as consisting not only in types of conscious experience. these three families are general theories of the constituents of wellbeing: wellbeing is claimed to be constituted by conscious experiences, the satisfaction of desires or the possession of the set of goods on the objective list, respectively. the preceding theories have been developed and refined in light of philosophical reflection on human wellbeing. moreover, animal welfare scientists have come up with other notions of wellbeing (or ‘welfare’). in the animal welfare literature, the three dominant families of theories focus on the affective state (equivalently to hedonism), the biological functioning and the natural behavior of animals, respectively (appleby & sandøe, 2002; green & mellor, 2011; veit & browning, 2020). in addition, dawkins’ (dawkins, 2017, 2021) influential approach emphasizes animal health and the satisfaction of their preferences. i will set aside this tradition of conceptualizing animal wellbeing here, since i am interested in wellbeing as a dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 5 normative concept.2 by contrast, animal welfare – as understood in animal welfare science – is a thick concept which mixes normative and descriptive elements (birch, 2022b; robbins et al., 2018). as a working hypothesis, i endorse a view of animal wellbeing according to which wellbeing of most animals is entirely constituted by conscious experience. i call this view ‘broad hedonism’ to distinguish it from ‘narrow’ hedonist views as presented earlier. according to narrow hedonism, wellbeing is not only exhausted by conscious experience, but by valenced conscious experience. however, i take it that it may be the case that conscious experiences which are not valenced intrinsically contribute to wellbeing. for example, chalmers (2022) is sympathetic to the idea that vulcans – beings who are conscious but do not possess any affects – have moral status, i.e., matter morally for their own sake. in a broad sense of ‘affective’, the class of affective states is coextensive with the class of valenced states. arguably, it would be monstrous to kill a vulcan, if nothing of moral significance is gained thereby. one possible explanation for why killing vulcans for fun would be morally abhorrent is that they do indeed have a sort of wellbeing which is not grounded in valence. in general, imagining a vulcan in various conditions makes the idea more appealing that non-valenced perceptual and cognitive experiences may intrinsically contribute to wellbeing.3 while non-valenced conscious states may be direct constituents of wellbeing, is may also be the case that non-valenced properties of conscious experience impact4 wellbeing in virtue of changing how valenced states are experienced. for instance, in beings who are self-conscious, a conscious feeling that it is oneself who is suffering may permeate pain experiences and thus modify the character of the pain, even if it does not change its intensity. perhaps this change in the character of pain consciousness also changes its contribution to wellbeing. both possibilities, influence on wellbeing via non-valenced conscious experiences or via not valence-related properties of conscious experience, will be discussed further in the next sections. for now, it is sufficient to say that these prima facie reasons lead me to remain neutral on narrow hedonism. nevertheless, i assume broad hedonism as a working hypothesis.5 hedonism about animal wellbe2of course, the view that ‘wellbeing’ is a normative concept is compatible with holding that the constituents of wellbeing can be investigated empirically. 3one may alternatively take this as an argument that non-conscious states partially constitute wellbeing. this view is consistent with broad hedonism, as long as the latter view is only restricted to non-human animals. nevertheless, an adherent to the idea that non-valenced conscious states partially constitute wellbeing is committed to the view that there are elements constituting vulcan wellbeing which are absent in zombies (i.e., beings which are functionally identical to humans but do not possess consciousness). 4i intend to use the term ‘impact’ in a broad sense which includes non-causal influence since many conscious experiences not (necessarily) cause wellbeing, but constitute it. 5for an argument that hedonism is the correct theory of animal wellbeing, see browning (2020). for an argument that hedonism is one among multiple valuable perspectives on animal wellbeing, see veit & browning (2020). dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 6 ing is more plausible than hedonism about human wellbeing since it is less clear that experience-independent goods like autonomy, authenticity, knowledge and accomplishment matter for their own sake to animals than that they matter to humans (browning, 2020, chapter 2). that being said, wilcox (2021) explicitly argues that an objective list theory is true of non-human animal wellbeing. hence, it is important to note that broad hedonism is not necessary to establish the significance of this paper’s project. even if wellbeing is not exhausted by conscious experience, it is almost unanimously accepted that conscious experience partially grounds wellbeing (birch, 2022a; cutter, 2017; kriegel, 2019).6 on any remotely plausible view of animal wellbeing, whether animals experience, e.g., pain or joy, has huge bearing on their wellbeing. hence, if broad hedonism turns out to be false, then a theory of an animal’s capacities for consciousness does not amount to a complete theory of its capacity for wellbeing. nevertheless, since a close link between consciousness and wellbeing is incontestable, an animal’s capacities for consciousness will deeply inform our assessment of its wellbeing. thus, an adequate assessment of animal wellbeing under the assumption of hedonism approximates the true account of animal wellbeing.7 1.3 wellbeing capacity and wellbeing variance as a further piece of conceptual clarification, i note that wellbeing can be ascribed to different time periods of an animal’s life and that two different notions of wellbeing – both derived from the fundamental one that we just elucidated – can be so ascribed. we can ascribe a level of wellbeing to an individual regarding any time period we wish since any moment can be good or bad for a particular individual. most frequently, we either refer to wellbeing at a specific moment or over an entire life. it is important to keep these two uses separate. furthermore, we usually refer to a creature’s total level of wellbeing which is determined by the combination of all the constituents that add to and subtract from its wellbeing. however, it is also possible to focus on some subset of constituents and ask whether they ground positive or negative wellbeing. for example, we might want to know whether an animal has a mild feeling of hunger which contributes negatively to its wellbeing, even if its total level of wellbeing is positive. again, conflating these uses may lead to confusion. finally, we need to distinguish between the actual wellbeing of animal species (and individual animals) and their capacity for wellbeing. the actual wellbeing of animals depends on contingent empirical facts regarding their conditions of living. for example, the average wellbeing of chickens may be very low, but it would improve if humans would stop exploiting them via factory farming. by contrast, 6for an opposing argument, see kammerer (2019, 2022). for counterarguments, consider dung (forthcoming, 2022a, 2022c). 7see fischer (2022) for an argument that the choice of theory of wellbeing has a limited, although real and sometimes important, influence on estimates of the capacity for wellbeing. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 7 a species’ capacity for wellbeing is independent of such factors. the capacity for wellbeing is how bad or good a creature’s life can go, i.e., the range between a creature’s possible maximum positive and its possible maximum negative wellbeing level (schukraft, 2020a).8 we say that a creature’s capacity for wellbeing is ‘higher’ than another’s if the difference between its maximum possible positive and its maximum possible negative wellbeing level is larger. for instance, it seems that the way humans treat other animals, e.g. farmed fish, presupposes that those animals have a lower capacity for wellbeing than humans.9 an examination of animal consciousness is important because it sheds light on animals’ capacities for conscious experience and thus their capacity for wellbeing. the actual wellbeing is neglected in this investigation since it is a function of a creature’s capacity for wellbeing and the contingent conditions for wellbeing it finds itself in. in practice, when we want to estimate the wellbeing of chickens in factory farming, we need to factor in both, their living conditions as well as their capacity for wellbeing which, i argue, should be assessed based on an investigation of their capacities for conscious experiences. the basic assumption here is that creatures with a higher capacity for wellbeing ground more intrinsic value and are, thus, (ceteris paribus) morally more relevant. note that this assumption does not presuppose any particularly contentious ethical view, such as utilitarianism. the view that there are pro tanto moral reasons to increase wellbeing, including animal wellbeing, which is all that is assumed here, is shared by a large majority of ethicists (beauchamp, 2019; pandit, 2021). the only difference is that non-utilitarians hold that other considerations can, in many cases, override reasons based on wellbeing. actual wellbeing, rather than wellbeing capacity, cannot be the crucial moral factor since our moral deliberation depends on how different actions would affect an animal’s wellbeing, and only indirectly on what their actual wellbeing is. however, the nexus between moral weight and capacity for wellbeing depends on the empirical assumption that animals with higher capacity for wellbeing tend to have a higher variance in their wellbeing levels, measured as average distance from the neutral wellbeing level, i.e., such animals tend to more frequently enter extremely bad or extremely good states (schukraft, 2020a). this is crucial because this wellbeing variance is what we actually care about: if an animals wellbeing deviates more often and more strongly from relatively neutral wellbeing levels, it grounds more positive and negative intrinsic value. thus, 8as pointed out by schukraft (2020a), practical considerations should guide the scope of the possibilities involved here. while the notion of possible wellbeing states should not collapse into a notion too close to designating only actually experienced states, we care about wellbeing states that animals in reasonably normal living conditions can achieve. which kinds of wellbeing states occur in remote, nomologically possible worlds – involving, for instance, pervasive genetic or technological manipulation – is irrelevant. 9crucially, positing differences in wellbeing capacity is not the only way to justify privileging human interests over animal interests. most importantly, one may also claim that there are differences in moral status across species (kagan, 2019). dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 8 while our investigation of animal consciousness chiefly targets their capacity for wellbeing, it also needs to be sensitive to clues to how much of this capacity is realized, i.e., how often very good or very bad states indeed occur in normal circumstances of an animals’ life. 1.4 wellbeing, trade-offs, and moral weight to complete this stage setting, i will zoom in on a vital, but inconspicuous practical question: why is knowledge about differences in wellbeing capacity between species important? in essence, knowing the capacity for wellbeing of different animals is important information to our decision-making. if an animal’s wellbeing range is higher, it is of higher importance that it finds itself in good conditions of living.10 however, it seems that knowledge of wellbeing capacities is not always necessary. this is, for instance, the case where it is sufficient to rely on precautionary principles to govern our conduct with respect to animals. for example, birch (2017) proposes that such a precautionary principle should determine which animals fall under the scope of animal protection legislation. according to this principle, all animals should be included within the scope of these laws if there is credible evidence that at least one indicator of valenced conscious experience is present within an animal of the same order. to apply this principle, no knowledge about the extent of an animal’s capacity for wellbeing is necessary. an animal’s capacity for wellbeing is relevant as soon as we deal with tradeoffs. these trade-offs may concern different animal species, for example when a zookeeper deliberates on whether he should devote a given amount of money or effort to improve the zoo’s aquarium or its lion cage, or choices between improving non-human lives and other causes, e.g., when philanthropists reflect on whether they should donate to animal charities or to charities aiming to alleviate human poverty. when we need to conduct trade-offs, we need to know how much to weigh the wellbeing of different animals which is determined by their capacity for wellbeing.11 furthermore, trade-offs between various kinds of animal intervention are usually extremely sensitive to assumptions about animals’ wellbeing. this is due to 10alternatively, a skeptic of the approach proposed here might question why we need the detour through consciousness to measure wellbeing. why do we not just ignore consciousness and directly test for wellbeing by means of the measures animal welfare science has developed (dawkins, 2021)? a reason is that the notion of animal welfare which is normatively relevant refers, perhaps among other things, to consciousness (birch, 2022b). §5 contains a deeper discussion comparing the approach to animal wellbeing proposed here and a strategy which relies more on the tools of animal welfare science. 11of course, capacity for wellbeing is not the only consideration in such decisions. for instance, if – let us assume – lions have a higher capacity for wellbeing it can nevertheless be preferable to improve fish lives, e.g. when many more fish than lion lives can be improved by a single intervention. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 9 the fact that improving wellbeing is typically the main goal of such interventions and that we are currently extremely uncertain about when animals are doing well. for instance, there exist a variety of different estimates of the quantity of lifetime wellbeing cows experience on average during conventional factory farming (browning, 2022). those estimates even differ in whether they assess their wellbeing as positive or negative overall.12 however, this has tremendous consequences. if cow wellbeing is strongly negative, this aspect of meat production is probably morally objectionable, perhaps even reprehensible. on the other hand, if cow wellbeing is mostly positive, we may even be obligated to maintain factory farming practices to continually produce these above-neutral lives. similarly, whether (assuming that we can integrate wellbeing assessments in a single numerical score) salmons’ capacity for wellbeing causes them to have a wellbeing score of -3 or -9 when subjected to established farming practices makes a huge difference for how much we should prioritize alleviating salmon suffering. in the -9 scenario, the suffering caused by farming salmons is three times as bad. the project of this paper is to relate animals’ capacities for consciousness and the variety of experiences that are open to them to their capacity for wellbeing. if hedonism is true, wellbeing is entirely constituted by conscious experience. if we want to figure out what wellbeing animal species (can) have, we therefore need to examine which kinds of conscious experiences they (can) have. to answer this question, i will introduce and adopt the framework for describing kinds of animal consciousness that was developed by birch et al. (2020). summarizing this framework is the task of the next section. 2 the dimensions framework of consciousness 2.1 the five dimensions the framework developed by birch et al. (2020) enables fine-grained descriptions of the conscious inner life of different animal species. i will summarize it now, since their account serves as a basis for our investigation of which features of consciousness constitute wellbeing. its key claim is that variations in the degree of consciousness of different species can be helpfully characterized along five different dimensions which all pick out features of an animal’s conscious inner life: perceptual richness characterizes the level of detail with which animals consciously perceive the world. it further decomposes into bandwidth (the amount of content experienced at any given time), acuity (the sensitivity to fine perceptual differences) and categorization power (the capacity to group perceived properties into more abstract categories). 12independently of these numerical estimates, it seems that many studies take a critical perspective on cow welfare in factory farming (e.g., mee & boyle, 2020; park et al., 2020). dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 10 evaluative richness is the dimension concerned with the richness of experience of valence specifically. again, animals differ in how much valenced content they can experience at any given time (bandwidth) and how fine-grained their evaluative appraisals of the world are (acuity). in addition, there are probably kinds of experience, e.g. grief or guilt, which not all species can experience. while integration at a time (synchronic unity) concerns the degree to which experience is unified at any single point in time, integration across time (diachronic unity) is about the degree of temporal unity, i.e., whether the world is experienced as a continuous stream rather than through temporally fragmented experiences. differences in synchronic unity seem to obtain, for example, in human split-brain patients or in animals like octopodes. in both, complex parts of the nervous system can act remarkably autonomously. asynchronic unity is correlated with capacities for episodic-like memory, future planning or with the temporal fineness of grain of perceptual discrimination. lastly, an organism possesses self-consciousness to the extent that it is consciously aware of itself as distinct from the external world. according to birch et al., distinguishing between experiences that represent external states and that represent internal bodily state in the service of movement may involve a minimal form of self-consciousness. having an awareness of one’s body as persisting object may be a more demanding form of self-consciousness, while an even more sophisticated capacity consists in attributing mental states to oneself, as distinguished from other subjects. since it posits different dimensions of consciousness, i will call this account the ‘dimensions framework’ in what follows. all those dimensions admit of degrees, e.g., evaluative experiences can be more or less rich and different experiences can be more or less integrated. as a working hypothesis, it is assumed that those five dimensions vary to some extent independently from each other while different aspects of the same dimension, i.e., different ways for perceptual experience to be rich, tend to co-occur. furthermore, the framework presupposes that there is no further dimension which is crucial for understanding the conscious inner life of an animal and conceptually distinct from the five dimensions mentioned. thus, the nature of an animal’s conscious experience can be sufficiently captured by situating it with respect to those five dimensions. 2.2 empirical tests in addition, the different dimensions can be probed independently in animal experiments. birch et al. review and propose several experiments which each promise to shed light on one of the five dimensions. let’s look at three of them. to ascertain perceptual richness, one has to test for conscious perception of various different stimuli in different sensory modalities and for the ability to discriminate between slightly different experiences. to achieve this, one needs means to detect when a stimulus is perceived consciously. one potential candidate is trace conditioning. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 11 trace conditioning differs from standard classical conditioning only in that the unconditioned and the conditioned stimuli are separated by a temporal interval. evidence by clark & squire (1998) suggests that trace conditioning in humans might require conscious awareness of the stimuli and the time gap separating them. if so, when a non-human animal is able to perform trace conditioning on some stimulus, this is an indication that the stimulus is perceived consciously.13 if an animal can do trace conditioning on a wide range of stimuli in different modalities, that arguably reveals some significant degree of perceptual richness. furthermore, to investigate the evaluative richness dimension, the dimensions framework recommends motivational trade-off paradigms. in such experiments, animals are forced to trade-off between different types of stimuli which either constitute rewards or punishments. for instance, in the experiment of millsopp & laming (2008), fish reduced their feeding attempts in a part of an aquarium where they received a shock. the more intensive the shocks were, the less feeding attempts were conducted by the fish. yet, when the fish were increasingly fooddeprived, the number and duration of feeding attempts increased. thus, fish seem to trade off their need for food with their aversion to noxious stimuli (i.e. electrical shocks).14 this requires processing and integrating (crossmodally) different kinds of information to enable actions which appropriately take into account one’s overall internal state. birch et al. point out that integration and flexible use of information is often considered a hallmark of consciousness. third, a famous paradigm relevant to self-consciousness, which birch et al. adduce, is the so-called mirror self-recognition test. it probes whether an animal is able to recognize a mark seen in a mirror as belonging to its own body. if an animal can recognize itself as itself in the mirror, then it needs to possess at least some form of self-consciousness, one might think. while the three mentioned experimental strategies belong to the ones especially emphasized by birch et al., they present a more encompassing overview of promising empirical tests which i cannot summarize here. in the next section, i will discuss different aspects of the dimensions framework in relation to wellbeing. the goal is to shed light on which dimensions – and which of their aspects – are relevant to an animal’s capacity for wellbeing and how much. 13for a more detailed examination of the relation between trace conditioning and animal consciousness, see droege et al. (2021) and mason & lavery (2022). for an analysis of the general rationale for treating trace conditioning as well as other features as evidence of consciousness, see dung (2022b). 14for discussion relevant to ethical issues of inflicting pain on or killing animals in the context of experiments on animal welfare and consciousness, see mazor et al. (2022) and webb et al. (2020). dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 12 3 dimensions of animal wellbeing i: perceptual and evaluative experience 3.1 how can there be differences in wellbeing capacity? in this section, we will use the dimensions framework to uncover the different dimensions along which animal wellbeing can be characterized. crucially, for this exercise to be valuable, an animal’s capacities for consciousness need to substantially contribute to its capacity for wellbeing. the dimensions of wellbeing may be interpreted in one of two ways: perhaps each dimension captures a different constituent of wellbeing, such that a single numerical wellbeing score can ultimately be computed by weighting the different dimensions based on their importance for wellbeing and adding them together.15 alternatively, wellbeing may be irreducibly pluralistic such that different dimensions of consciousness constitute different kinds of wellbeing which cannot be integrated into a one-dimensional wellbeing assessment. further developing an idea by schukraft (2020a), i posit that there are three ways for an animal (species) to possess a higher capacity for wellbeing (in virtue of its conscious experience):16 first, an animal may be able to attain more of the specific (token) experiences which partially constitute wellbeing. for instance, an animal may be able to have more subjective pain experiences per objective unit of time than others. second, an animal may be able to have kinds of experiences which are not accessible to others. for instance, some animals may have the capacity to experience crushing guilt while others are not able to have this feeling. third, some animals may experience a shared experiential type differently than others. for instance, it is conceivable that some animals have the capacity for stronger pain experiences than others. to illustrate this threefold distinction, suppose that, e.g., dogs have a higher capacity for wellbeing than bees. this may be the case either because they can have experiences bees cannot have, say joy, or because the (e.g.) joy dogs can feel when playing surpasses the joy that bees can experience or because dogs can have more experiences like joy in the same amount of (objective) time. given broad hedonism, animals which do not possess consciousness can’t have any wellbeing. yet, it may well be that even some conscious beings do not possess any wellbeing, for perhaps not all kinds of consciousness are constituents of well15to make this exercise more complicated, the contribution of one dimension may not be constant but depend on how an animal scores on that particular dimension or other dimensions, i.e., a dimension may require a high weighting in extreme cases (say, an extremely synchronically disunified consciousness) but a low weighting in ordinary cases. while the two interpretations of wellbeing dimensions seem to exhaust all genuine options, hybrids are also possible. some wellbeing dimensions might be incommensurable, while others can be placed on a single scale. 16this includes the capacity for negative wellbeing. therefore, a higher capacity for wellbeing may also be explained by a higher capacity to attain constituents of negative wellbeing (like pain). dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 13 being. my aim is to investigate which kinds of conscious experience do constitute wellbeing. in the following, we will scrutinize the five dimensions contained in the dimensions framework in order. usually, i will skip over the question how the relevant features of conscious experience can be detected empirically, since i am in broad agreement with the suggestions by birch et al. regarding operationalizations of consciousness dimensions. 3.2 perceptual consciousness let’s start with perceptual richness. as already discussed, vulcans elicit the intuition that valenced experience is not necessary for wellbeing. however, one may trace attributions of wellbeing to vulcans to their non-conscious states, as opposed to their non-evaluative conscious states. vulcans – according to chalmers’ (2022) thought experiment – can possess active lives replete with colorful and rich activities. perhaps this is what constitutes the intuition that they possess a wellbeing? to rule this out, we should look at another case inspired by chalmers (2022): suppose there are conscious beings who do not possess affective states and spend their entire time carefully perceiving their environment. they do not move and do not engage with what they perceive. would we say that the lives of these inert beings can go better or worse for them? i am torn on this question. i feel an intuitive pull to answer affirmatively but this intuition may be caused by the artificial assumptions build into the thought experiment. it is hard to imagine perceptual experience which is not accompanied by some form of valence, like the feeling elicited by finding something beautiful. if perceptual experience can constitute wellbeing, then it seems plausible that higher bandwidth (i.e., more perceptual content) increases the level of wellbeing. that being said, it seems that the wellbeing range of our inert observer would in any case be vastly narrower than the range available to typical human beings. even if mere perceptual experience constitutes wellbeing, its contribution is miniscule compared to the wellbeing grounded in positive or negative affect. thought experiments make this clear. if we have the choice between reducing pain or fear experiences in some animals, on the one hand, and improving the conditions of hypothetical inert beings whose life consists only in a sequence of perceptual experiences, on the other, we should almost always favor the former. for practical purposes, we can – save for instrumental reasons mentioned below – largely ignore the dimension of perceptual richness when considering trade-offs involving animal wellbeing. to summarize, perceptual richness is a possible ground of wellbeing, but lacks much practical significance. however, perceptual richness is instrumentally important. it correlates with forms of evaluative richness since it is a prerequisite for some of them. for instance, we mentioned the positive experience distinctive of appreciating beauty. having this kind of experience depends on the appropriate sensory organs and perceptual processes which fall into the domain of perceptual richness. furtherdung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 14 more, high-bandwidth and fine-grained evaluations of environmental features are only possible if those features can be distinguished perceptually. a good example is pain which – according to the standard view – has distinct and dissociable sensory and evaluative components. while the sensory component of pain can appear without the negative evaluation, as in pain asymbolia (grahek, 2001) and sometimes under morphine, the reverse seems impossible. pain needs to be experienced perceptually to be experienced with a negative valence. whenever evaluative experiences are experiences of sensory features, evaluative richness presupposes some degree of perceptual richness. hence, while perceptual richness may not constitute wellbeing, it causally affects constituents of wellbeing. for this reason, an investigation of animal wellbeing cannot leave out the perceptual richness dimension completely. 3.3 evaluative consciousness the next dimension is evaluative richness. it is obvious that evaluative consciousness is a constituent of wellbeing. since (as in birch et al., 2020) evaluative richness refers to the feeling of valence, not to any kind of sophisticated cognitive evaluation, it is plausible that many animals, perhaps even invertebrates, have evaluative experiences (crook, 2021; crump et al., 2022; galpayage dona et al., 2022; gibbons et al., 2022). things can be good or bad for animals because they feel good or bad to them. the more serious question is whether there are non-evaluative constituents of wellbeing, not whether there are evaluative ones. in addition, it seems clear that the contribution of evaluative experiences to wellbeing is extremely large and often outweighs all other considerations. however, one can still ask which aspects of evaluative experience are the ones that are relevant for wellbeing and how they are weighted. first, evaluative experiences differ in their intensity or strength. this aspect is not explicitly mentioned by birch et al., but obviously possesses extreme, or even paramount, importance.17 pain experiences, to reuse the example of a paradigmatic and ethically crucial evaluative experience, can not only be more or less rich, but also more or less strong.18 how strong they are – a mild pinch or a cluster headache – is central to their contribution to wellbeing. intuitively, it seems to me as if the intensity of emotions – consider guilt, sadness and joy – also varies partially independently of their richness. a numbing feeling of sadness does not need to be complex and multi-faceted to have a large negative impact on wellbeing, if it is sufficiently strong. 17this feature corresponds to what dung & newen (2023), in their multidimensional analysis of consciousness, call ‘evaluative intensity’. 18according to an influential view, two dimensions – valence and arousal – are the core building blocks of affective experience (mendl et al., 2010; panksepp, 2010). the valence dimension seems to correspond to the degree to which a single experience is unpleasant. thus, what i call ‘intensity’ is a (very important) component of valence. by contrast, the arousal dimension does not seem to contribute to wellbeing. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 15 based on conceiving of various scenarios and observing actual cases, e.g., patients with cluster headaches (may, 2005), i am inclined to hold that the intensity of a valenced experience is most important when it comes to its relation to wellbeing. nevertheless, there are other features of evaluative consciousness which also constitute wellbeing. for starters, different types of valenced experience may differentially impact wellbeing. for instance, perhaps fear has a higher (negative) impact on wellbeing than the experience of emotional stress, even if they share the same intensity, duration and richness. on average, i expect that the negative impact of negatively valenced experiences outweighs the impact of positively valenced experiences of the same intensity.19 thus, two properties of evaluative experience besides its richness, namely its type and its intensity, contribute to wellbeing. is richness important? to begin with, we have to distinguish different kinds of richness. evaluative experience of an animal can be asynchronically rich in the sense that the animal is able to experience multiple types of evaluative experiences, i.e., types of pain, pleasure, emotions etc., albeit not simultaneously. if different types of conscious experiences differentially impact wellbeing, this is relevant for wellbeing. however, we already mentioned this point. ‘richness’ in the sense which is relevant here is a synchronic notion. it can be subdivided into type and token richness. type richness consists in the fact that an animal can have evaluative experiences of many different types at once. by contrast, token richness consists in the fact that an animal’s evaluative experience is fine-grained and has high bandwidth, i.e., the animal is able to experience much and finely differentiated evaluative content, comprising many token experiences, at once. the rationale for why token richness intrinsically contributes to wellbeing is obvious. given that single evaluative experiences ground wellbeing, it stands to reason that a higher number of evaluative experiences will ground a higher amount of wellbeing. having high evaluative token-richness implies that an animal can experience many contents (or especially complex contents) at the same time, such that its total wellbeing is, ceteris paribus, likely able to reach more extreme values. the case for the impact of evaluative type richness on the capacity for wellbeing is less clear-cut. high evaluative type richness involves a complex and multifaceted affective life, including many kinds of bodily sensations and especially emotions. plausibly, some animals are not able to experience the feeling of (e.g.) love, guilt or embarrassment which makes their consciousness less evaluatively rich. it is an open question whether the capacity for the simultaneous presence of many experiential types increases an animal’s capacity for wellbeing when we control for the factors mentioned above that some experiences may in virtue of their kind have higher impact on wellbeing, that the amount of total evaluative content is likewise relevant for wellbeing capacity and that experiences differ in their intensity. if type richness is indeed relevant, cognitively sophisticated animals – like 19alternatively, one may opt for a conceptual scheme which ascribes a higher average intensity to negative experiences. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 16 mammals and birds – may have (in this respect) a higher capacity for wellbeing, since they seem to have an emotional repertoire which surpasses simpler animals. in total, we discovered that the main contribution to (the capacity for) wellbeing stems from the evaluative dimension. in particular, evaluative intensity and token richness are very likely grounds of wellbeing and have much impact on wellbeing capacity. in addition, the type of experienced evaluative content and the number of types that can coincide in a single total evaluative experience may also be grounds of wellbeing. if so, they likely have less impact on wellbeing capacity than evaluative intensity. importantly, these features can be examined empirically, although we are still in the early stages of learning to track such fine distinctions between conscious experiences. 3.4 empirically accessing evaluative consciousness most accessible to empirical testing is the question which kinds of evaluative experience animals possess. given that an animal possesses the requisite mechanisms for consciousness and that it can represent the evaluative content e (let’s say, the emotion fear), it is a reasonable assumption that the animal sometimes experiences e (fear) consciously. since there exist tractable empirical research programs on animal emotions (de vere & kuczaj, 2016; kremer et al., 2020), we can use knowledge on the distribution of consciousness to draw inferences on conscious affect in various species. nevertheless, we mustn’t neglect the fact that this inference is defeasible. while the capacity to occupy mental state m and the capacity to have conscious mental states jointly provide good evidence for the capacity to consciously experience m, they don’t imply the latter capacity. tests for the intensity of evaluative experience are harder to come by. in respect to pain, one may investigate the threshold that – when crossed by some noxious stimulus – causes pain reactions. given the assumption that pain thresholds correlate with thresholds for conscious pain experience and that the correlation between noxious stimulation and actual pain is approximately linear, a lower pain threshold indicates that an animal experiences more intensive pain when subject to noxious stimulation. however, both assumptions are controversial and the second assumption likely breaks down when we are confronted with cases of drastic interspecies differences in the intensity of conscious experience. furthermore, one may estimate the strength of experiences via their effect on animals’ preferences. for example, present the animal with the choice between two rooms: in the first, there is an object it extremely likes but the room is cold. the second has no special features. observing changes in preference while continually varying the room temperature sheds light on the animals’ experience of temperature. by frequently using this method with objects whose attraction to the animal we have measured antecedently, we can thereby gain insight into when the temperature experience starts to assume a negative valence and how strong it is for different values of objective temperature. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 17 though this method is helpful, it also falls short of perfect reliability and validity. for we can’t be sure how closely preferences track the intensity of experience, especially relative to other evaluative features of conscious experience. insofar as we think that, e.g., evaluative richness contributes to the impact an experience has on wellbeing, it is plausible that richness also partially causes an animal to prefer rich experiences. this problem becomes more obvious when we try to extend this method to interspecies comparisons. in this paper, we aim to find differences in animals’ general capacity for wellbeing. however, we cannot argue for these differences directly in terms of differences in preference strength. it is not clear what it would mean for one animal to behaviorally express stronger preferences generally than another one. plausibly, the behavior of animals is guided by whatever preferences are in combination the strongest at any moment. thus, how preferences affect behavior depends only on the relations between different preferences. there is no stronger behavioral expression for preferences which are not just in relation to others, but absolutely, stronger.20 while observations of animal behavior can be used to construct the preference function of an animal on an interval scale, we need something more demanding (see also: browning, 2023). the claim that an animal has in general stronger preferences than another one can only be derived when it is possible to express the preference functions of both animals on a common scale. based only on behavioral evidence, it seems impossible to construct such a common scale. yet, if we cannot discern absolute differences in preference strength between animals, we cannot detect absolute differences in wellbeing and consequently wellbeing capacity. besides taking preferences as a direct indicator of wellbeing, one may regard preference tests as specific measures of one constituent of wellbeing, i.e., of one feature of conscious experience. yet, the same problem arises. behavior cannot reliably express that one animal has in general much stronger preferences than another, while we are looking precisely for such general differences in capacities for conscious experiences. moreover, preferences seem to be caused by the total evaluation contained within an experience, not by its components in isolation. animals prefer situations which cause good experiences and avoid situations which cause bad experiences, regardless of how different features of experience conspire to create this valence. thus, preference tests are no targeted measure of evaluative intensity or any other individual component of wellbeing. that being said, as suggested earlier, varying one feature of the environment and studying subsequent preference changes while holding other variables constant can provide clues as to how strongly this variation is experienced. 20an example for the assumption that such an absolute difference in preference strength exists would be the claim that the preferences of species s1 are in general three times stronger as the preferences of species s2. this difference cannot be expressed in behavior. however, differences in wellbeing capacity can have this form. thus, preferences cannot both be measurable and map onto differences in the capacity for wellbeing. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 18 more generally, this discussion has shown that our current tools for studying evaluative intensity are, though useful, rather limited. at some point, we will probably add neuroscientific measures to our toolkit to track distinct features of evaluative experience in isolation. in the next section, we will move on to the remaining dimensions concerning synchronic and diachronic unity as well as the self. 4 dimensions of animal wellbeing ii: unity and the self in this section, we are looking at the relation between the remaining dimensions and capacity for wellbeing. this may seem puzzling: we already granted that one plausible view, namely narrow hedonism, holds that wellbeing is entirely constituted by evaluative experience. this begs the question: if narrow hedonism is true, how can the other dimensions intrinsically contribute to an animal’s capacity for wellbeing? the answer has to be that – if narrow hedonism is true – other dimensions constitute wellbeing, if they do, in virtue of modifying the capacity for evaluative experience.21 that is, they either make new types of valenced experience possible or change the properties valenced experiences otherwise (tend to) have. unity and self-consciousness are therefore no distinct third category of consciousness experience but change the character of the other two, i.e., perceptual and evaluative consciousness. 4.1 synchronic unity: mere correlation and split consciousness with that prelude in mind, let’s continue our investigation of the constituents of animal wellbeing. how does synchronic unity, i.e., the integration of conscious contents into one single coherent experience, impact the capacity for wellbeing? prima facie, it seems even unclear whether a higher score on the synchronic unity dimension should be seen as increasing or decreasing capacity for wellbeing. in favor of a positive connection, one may suggest that a high degree of unity indicates a high degree of information integration which is connected to cognitive sophistication. although this matter is far from straightforward (schukraft, 2020b), the typical view is that more intelligent animals which possess abilities such as creativity, imagination, episodic memory, deliberation and agency possess a higher 21one might dispute that these other dimensions actually pick out constituents, rather than a special class of conditions, of wellbeing. this question is not essential, as long as one grants that the way consciousness dimensions contribute to wellbeing differs markedly from other conditions of wellbeing like health and shelter. some conscious experiences are the constituents of wellbeing. differences in consciousness dimensions directly shape which conscious experiences animals can have. the influence of typical wellbeing conditions is much more indirect. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 19 capacity for wellbeing. however, put this way, this suggests merely a correlation between synchronic unity and capacity for wellbeing. this is insufficient because the question at issue is whether synchronic unity partially constitutes wellbeing, i.e., whether it makes a non-instrumental contribution to wellbeing, not whether it correlates with features that do so. as soon as we distinguish between constitution of and correlation with wellbeing, it starts to become unclear why synchronic unity should be involved in animal wellbeing. why should it be good (or bad) for an animal to have more unified experiences? we already clarified that this involvement would probably have to occur mediated through the influence of unity on evaluative experience. but here the same question arises: why should a unified experience feel better or worse than a disunified one? there is no specific reason to think that unity impacts wellbeing. however, since our grasp of what disunified experience even feels like is limited, it may be best to conclude that we don’t know whether synchronic unity is relevant to an animal’s capacity for wellbeing. nevertheless, there is an exception. typical examples of beings which score low on the synchronic unity dimension are split-brain patients and octopodes. sometimes, it is hypothesized that more than one stream of consciousness resides in their mind (carls-diamante, 2017; godfrey-smith, 2020). in other words: they may have two – or arguably in the case of octopodes multiple – conscious perspectives on the world. this raises the intriguing possibility that we would need to ascribe one wellbeing score to each of the conscious selves and thus – when talking about the wellbeing of the whole organism – add them together.22 if we have no systematic reason to believe that these octopus or split-brain consciousnesses individually have a lower capacity for wellbeing than the unified consciousness of other animals, then we should expect the multiple octopus consciousnesses to jointly have a higher wellbeing capacity than animals whose experience is unified into a single conscious stream.23 if that is true, then having very low synchronic unity actually contributes to the capacity for wellbeing. it seems that this connection is non-linear. moderately high synchronic unity does not contribute to wellbeing capacity relative to very high unity. the connection only obtains for types of consciousness which are so disunified that we should treat them as plurality of conscious streams. in conclusion, it is possible that synchronic unity is, to some extent, a ground of wellbeing. 22one may insist on other ways to compute the total wellbeing score of, e.g., a split-brain patient than addition. i have no objection to this but establish the following constraint: if the two hemispheres of a split-brain patient indeed constitute separate streams of consciousness with individual wellbeing levels, then the total wellbeing of that patient should be determined in whatever way is usually used to determine the total wellbeing of a group of animals when various typical individual animals are involved. 23see gottlieb (2022) for a view according to which the octopus brain does not contain multiple distinct wellbeing subjects. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 20 4.2 asynchronic unity: speed of consciousness and mental time-travel next up is asynchronic unity. i think that especially two features regarding the temporal unity of conscious experience are relevant to wellbeing capacity: the first is the rate of conscious experience, the second is the capacity for mental time-travel. in respect to the first feature, i will rely on the detailed treatments of schukraft (2020c, 2020d). as described by schukraft (2020d), “animals with faster rates of subjective experience undergo more subjective moments per objective unit of time than animals with slower rates of subjective experience.” this means that, when a painful stimulus is present for a particular interval of objective time, animals with a higher rate of experience are subjectively longer in pain. there are many reasons to think that differences in the rate of subjective experience exist (schukraft, 2020d). first, there are pervasive differences in neural functioning, especially in the speed with which information is transmitted and processed, between the brains of different animals. second, how fast animals have to experience the world to successfully cope plausibly depends on their ecological niche. obviously, the latter varies extremely between animal species. third, humans frequently report – especially in extreme conditions such as during accidents or under the influence of drugs – that they experience time as passing slower or faster than usual. if there can even be intra-human differences regarding the rate of conscious experience, differences between animals having extremely different bodies and brains and inhabiting different habitats seem almost but guaranteed. fourth, as the second point would suggest, animals differ vastly in how quick they can react to stimuli. fifth, studies of the temporal resolution of perception are at present arguably the most promising way to ascertain an animal’s rate of conscious experience (schukraft, 2020c). temporal resolution is examined by measuring an animal’s critical flicker-fusion frequency (cff). this is the threshold beneath which an animal experiences a flickering light not as series of flashes, but as a continuous stream of light (inger et al., 2014). if animals have a low cff, this means that they are bad at discriminating stimuli which follow each other closely in time, i.e., they have low temporal resolution.24 cff has already been tested for a large range of animals (for an overview, see schukraft, 2020c). surprisingly, humans exhibit merely an average cff of 60 hz. they are, for instance, surpassed by chickens (cff of 87 hz) and tsetse flies (145 hz). in general, even animals which are relatively close to each other evolutionarily can exhibit vastly different cffs. that being said, there is a tendency that birds and insects have high cff. while the temporal resolution of perception does not directly translate into the temporal acuity of conscious 24as it stands, cff is a measure of visual stimulus discrimination. thus, it disadvantages animals which tend to rely on other sensory modalities. for this reason, one needs to transfer the basic insight motivating cff – that the threshold up to which stimulus frequencies can be distinguished points to the temporal resolution of perception – to other sensory modalities. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 21 perception, cff is a (defeasible) indication of the speed of conscious perception nonetheless.25 why does a faster rate of subjective experience increase a creature’s capacity for wellbeing? in reusing an argument pattern familiar from the discussion of richness, we can recognize that a faster rate of consciousness entails that an animal undergoes (ceteris paribus) more conscious experiences than an animal with a slower rate of conscious experience.26 if conscious experiences constitute wellbeing, then a higher number of conscious experiences can constitute more wellbeing. for example, experiencing pleasure for five minutes of subjective time is not as good as experiencing pleasure for 20 minutes of subjective time. thus, if one experiences more subjective moments per objective unit of time, then one can attain higher or lower wellbeing in the (objective) time period in question, i.e., one’s capacity for wellbeing is higher. arguably, the capacity to (consciously) perform mental time-travel, i.e., to relive past events and to imagine future events, impacts an animal’s capacity for wellbeing. there are several mechanisms through which this connection may obtain. first, mental time-travel might increase the number of experiences with strong valence that occur. for instance, this applies when an event that causes extreme fear is not just experienced once, but relived in memory and possibly anticipated in episodic foresight. second, being able to situate one’s experience in time may increase or decrease its impact on wellbeing. for example, knowledge or imagery of the fact that a negative experience will last for a long time or memory of a former traumatic experience of the same kind may make the experience worse. on the other hand, one can imagine that knowledge of the fact that a negative experience will be short-lived can decrease its unpleasantness. in addition, mental time-travel facilitates the ability to detach from the present moment and focus on a potentially better future (schukraft, 2020b). this, again, may decrease how bad negative experiences and how good positive experiences feel. finally, the sense of time accompanied by the capacity for mental time-travel may enable an animal to form interests regarding its own long-term future. the animal may entertain goals like eating a piece of cached food seven days from now on (clayton & dickinson, 1998). this entails that animals able to mentally timetravel might have more interests which can be satisfied or frustrated. given a view which ties wellbeing to the satisfaction of preferences, this directly increases their capacity for wellbeing. even on a hedonist view, the advent of future-directed preferences might increase animals’ wellbeing capacity if animals consciously experience the satisfaction and frustration of their preferences. thus, there are several 25since i do not have space here to defend the significance of cff for the speed of conscious perception extensively, i refer the reader to schukraft (2020c) instead. 26my way of formulating the argument might seem to presuppose a form of temporal atomism (hurley, 1998), i.e., the assumption that experiences decompose into snapshots with minimal temporal duration. however, if one is uncomfortable with this view, one may just as well talk about the subjective duration, not about the number, of conscious experiences. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 22 arguments in favor of a relationship between mental time-travel and capacity for wellbeing. most of them point in the direction that mental time-travel increases an animal’s capacity for wellbeing. in conclusion, both the rate of conscious experience and the capacity for mental time-travel are likely grounds of animal wellbeing. 4.3 self-consciousness and suffering the final dimension concerns self-consciousness. the connection between selfconsciousness and wellbeing capacity is rather unclear, but potentially very significant. the basic intuition is that harms feel worse when they are explicitly experienced as pertaining to oneself, rather than in a sense which does not involve a conscious self-representation. in reverse, the same might be true of positively valenced experiences. it is open how exactly being attributed to a self would make experiences better or worse. one idea could be that self-attribution of experiences causes harms to attain an existential dimension, i.e., to become threats for the continued existence of the self. another is that being attributed to a self makes conscious experience more focused and concrete which might cause them to ground more wellbeing. according to a possible view, wellbeing and self-consciousness are intimately related. namely, one might hold that no conscious experience can constitute wellbeing, unless it involves self-consciousness. according to this line of thought, (wellbeing-relevant) suffering and pleasure require that they are experienced as the suffering and pleasure of someone, i.e., oneself (dennett, 1995; metzinger, 2021). however, it is doubtful that this move is very plausible. when we start to consider thought experiments, it seems clear that an animal which lacks self-consciousness is nonetheless worse of when it is in excruciating pain and agony than otherwise. the only way to undermine the coherence of this kind of thought experiment is to embrace the view that all consciousness necessarily involves a form of (prereflective) self-consciousness (damasio, 1999; duncan, 2019; gallagher, 2010; zahavi, 2014). however, if this view is true, then the demand that conscious experiences must involve self-consciousness to constitute wellbeing is empty. furthermore, if consciousness necessarily presupposes self-consciousness, then the presence of selfconsciousness, as opposed to its absence, cannot be relevant to wellbeing. however, even then, it is still conceivable that the degree of self-consciousness makes a difference to the capacity for wellbeing. for instance, someone may argue that the pain experiences of someone who has a sophisticated self-model which, inter alia, allows for the application of a general theory of mind to oneself, are experienced as worse. however, in the absence of a compelling argument to this effect, i do not endorse this claim. thus, it is unclear whether self-consciousness is a ground of wellbeing. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 23 in the next section, i will shed further light on the strengths and weaknesses of the framework i propose by contrasting it with the account for interspecies comparisons of welfare developed by browning (2023). we will see that both approaches can serve complementary roles in the investigation of animal wellbeing. 5 wellbeing from an actualist welfare and from a possibilist consciousness perspective the preceding considerations are meant to display the potential of a general research program for the study of animal wellbeing. in a nutshell, it comprises two steps: first, researchers should investigate: which consciousness dimensions and which aspects of them are plausible constituents of wellbeing and how much wellbeing do they ground? second, researchers should aim to identify which of these dimensions of consciousness and which of these aspects of them are present in a given animal species and to what extent. in principle, following this procedure tells you the wellbeing capacity of every animal species. obviously, both steps are connected to a multitude of challenges in practice. for the distinctive features of this approach to emerge more clearly, it is instructive to compare it to extant approaches to research on animal wellbeing. 5.1 browning’s framework for interspecies welfare comparisons while there is a thriving research program on the measurement of wellbeing within individual animals or perhaps individuals within the same species (for an introduction, see dawkins, 2021), the challenge that different species might differ in their wellbeing capacity is usually ignored. an admirable exception is browning (2023). i will describe her framework for interspecies welfare comparisons in what follows, as it is the best account of how the tools of animal welfare science can be fruitfully directed to the investigation of differences in wellbeing capacity.27 according to browning, the main problem of interspecies comparisons of animal wellbeing (which take into account differences in wellbeing capacity) is the underdetermination of hypotheses about welfare by the empirical evidence. suppose we want to compare the wellbeing level of two species and have agreed upon a plausible indicator of wellbeing applicable to both species. for purposes of illustration, i will take heart rate and the amount of vocalization as examples of 27i sometimes use the notion ‘welfare’ when discussing browning’s framework, as this expression underlines the connection between her ideas and the methods of so-called ‘animal welfare science’. however, since browning characterizes welfare in terms of subjective experience, the target of her approach is the notion of ‘wellbeing’ as i use it. thus, the two expressions can be treated as synonyms here. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 24 indicators of fear experience. suppose we find that – across many situations – species a manifests faster heart rate or more intense vocalization than species b. there are two possible conclusions: (i) species a has stronger experiences of fear than species b. (ii) the relation between the fear experience and the indicator differs between species a and b: the same fear experience leads to a stronger indicator response in a than in b. the empirical results by itself do not tell us which conclusion to draw, i.e., they do not tell us whether the difference in responses between a and b indicates a difference in wellbeing or not. in the face of this conundrum, browning proposes two methods to proceed: first, in respect to particular animals and particular indicators, one can justify a similarity assumption with respect to the relation between wellbeing and wellbeing indicators. in particular cases, it might be plausible that the relation between wellbeing level and indicator is constant between species. for instance, this might be the case if the indicator response is controlled via deep physiological pathways (arguably in the case of heart rate) and the animal species are closely related evolutionarily. if the relation between wellbeing and indicator response is assumed to be constant, variation in indicator responses can be unambiguously traced back to differences in wellbeing level. second, a special case of the first method, evidence on wellbeing differences that is obtained in virtue of multiple independent indicators which all point in the same direction can defuse the underdetermination problem. if many indicator responses point to the same wellbeing difference and the processes producing the responses are independent, then it seems to become increasingly improbable that all the different relations between wellbeing and indicator responses differ between two species in the same direction. thus, at some point, we might start to trust convergent evidence of interspecies differences in wellbeing level. thus, browning’s framework shows how, in principle, we could use similarity assumptions and convergent evidence to assess wellbeing levels of various animals. if we can ascertain the wellbeing level of animals in diverse situations, we also learn their maximum and minimum wellbeing levels, i.e., their wellbeing capacity. how does browning’s framework differ from the wellbeing dimensions approach i have described and where do respective strengths and weaknesses lie? 5.2 comparison between the wellbeing dimensions approach and browning’s framework there are two main differences between the approach i proposed and browning’s framework. browning’s approach is more direct in two respects. first, browning proposes direct measures of wellbeing while i propose to examine wellbeing indirectly via the dimensions of conscious experience. since browning understands wellbeing as the evaluative character of conscious experience, this differdung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 25 ence might in the end be merely or mostly conceptual, but it is relevant nonetheless. for browning’s approach takes the tools of animal welfare science to be the main instruments for investigating animal wellbeing while i resort to the methods of animal consciousness science. research in animal consciousness science ideally aims to isolate different types of conscious experience and reveal their specific features, e.g., different types of emotional experience. by contrast, research in animal welfare science (insofar as it investigates conscious experience) typically aims to find measures of the integrated evaluative state of the animal, i.e., measures of how (good or bad) the animal experiences its state in total, given all the different experiences of the animal. second, browning’s approach is actualist in the sense that it aims to directly assess the wellbeing level of various species (where these are comparable across species) and use this as a basis to derive their wellbeing capacity. by contrast, my approach is possibilist in the sense that it takes wellbeing capacity as the more fundamental target of research which can then – in conjunction with knowledge about actual living conditions and the tools of animal welfare science (preference tests etc.) – be used to infer wellbeing levels which admit of interspecies comparisons. what are the advantages of browning’s framework? i will focus on what i see as its most important virtue. to make this virtue concrete, we need an example for a plausible indicator of the total wellbeing (at a specific point in time) of an animal. in other work, browning favors judgement biases as such an indicator (browning, 2022). there is ample evidence suggesting that the overall affective state of an animal influences whether it tends to interpret ambiguous stimuli as positive (“optimism bias”) or negative (“pessimism bias”) (lagisz et al., 2020). the strength of the disposition to interpret stimuli optimistically or pessimistically can then be used as an indicator of how good or bad the animal feels overall. assuming that judgement bias is indeed a valid indicator of wellbeing and that we can detect specific features of conscious experience, judgement bias can be used to test whether various features of consciousness contribute to wellbeing (and how much). take our earlier hypothesis that, ceteris paribus, the magnitude of an animal’s (positive or negative) wellbeing is increased the more different types of evaluative experiences it undergoes, e.g., many different emotions. assuming that the total wellbeing of the animal can be measured via judgement biases, we can (in principle) compare the total wellbeing of various animals given changes in the richness of types of evaluative experience to see which impact those changes have. this tells us the impact of the richness of types of evaluative experience on wellbeing. one might say that the judgement of the importance of various features of consciousness to wellbeing does not need to be made by human researchers, because it can instead be deferred to the animals themselves.28 28this method generalizes to some, but not all, consciousness dimensions. for instance, it is unclear how this method could reveal the impact of the rate of consciousness to wellbeing (see below). dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 26 we have just said that the framework proposed by browning provides a way forward for assessing the total wellbeing of an animal and can also illuminate which features of consciousness are relevant to wellbeing and by how much, i.e., complement the wellbeing dimensions approach. this raises the question: given the potential of browning’s framework, why should we pursue the latter approach as well, i.e., take consciousness dimensions as the starting point for investigating wellbeing? the dimensions approach is needed because there are clear limitations to browning’s framework. first, as she readily admits, as soon as we compare species which differ heavily in terms of their physiology, brain anatomy and evolutionary history – think of comparing mammals and invertebrates although cases of less radical dissimilarity are probably also sufficient – we don’t have a good justification for the requisite similarity assumption. in these cases, the relation between wellbeing and indicator response probably differs widely between species.29 second, it seems very likely that the traction convergent evidence provides on animal wellbeing capacity is rather limited. take typical indicators of wellbeing as used in animal welfare science, like judgement biases or heart rate. how can these indicators tell us that there are interspecies differences in wellbeing capacity? presumably, the idea that these indicators actually reveal some interspecies differences in wellbeing capacity entails that the strongest judgement bias, fastest heart rate etc. found in animal species a are stronger than the maximum judgement bias, heart rate and so on found in animal species b. but finding this sort of convergence would be quite strange. it seems relatively clear that the maximum indicator responses animals can have are mostly determined not by their maximum wellbeing level, but by other factors. the maximum heart rate is determined by basic physiological constraints. the maximum judgement bias is most likely determined by basic psychological facts about how judgements are made in an animal. it seems to me that the same goes for other potential wellbeing indicators discussed in animal welfare science. given that the maximum indicator responses seem not to be determined by facts about wellbeing, they are not indicative of wellbeing capacity. thus, we can make the empirical prediction that we will not find a strong convergence of many wellbeing indicators such that no species has a higher maximum indicator response on all of them.30 we have seen that typical indicators of animal welfare (like judgement bias) are not well-suited to pick up on differences in wellbeing capacity. this is were looking at consciousness dimensions provides an advantage: given broad hedonism, differences in wellbeing capacity ultimately need to be based on differences 29similarly, it becomes more challenging to find indicators of wellbeing which can be applied to both species at all. 30if we indeed discover some convergence, then it will be up for debate whether this convergence points to a hidden variable such that the indicators in question are not truly causally independent or shows that the level of the maximum indicator responses is actually a valid indicator of wellbeing capacity. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 27 in capacities for conscious experience. thus, while it is hard to see which kinds of general welfare measures could provide the convergent evidence of wellbeing capacity differences browning’s framework requires, the wellbeing dimensions approach i favor can shed light on differences in wellbeing capacity by systematically investigating different consciousness dimensions. moreover, the wellbeing dimensions approach has two further benefits: first, welfare measures like cognitive biases aim to measure the total wellbeing state of an animal, as we mentioned earlier. this is helpful both because this information is often desired and because we can use it to assess the contribution to wellbeing of individual consciousness dimensions. by contrast, the dimensions approach distinguishes potential constituents of wellbeing which brings its own benefits. for by distinguishing which features of consciousness contribute to wellbeing, it can guide the search for and refinement of relevant measures of total wellbeing. if those measures aspire to measure total wellbeing, they should be supplemented to detect the constituents of wellbeing discovered by the dimensions approach, if they do not already do so. second, the dimensions approach captures constituents of wellbeing which seem like they are inaccessible to measures of animal welfare, as the latter are usually understood. we have seen that the rate of conscious experience of an animal or the question whether its consciousness is split into distinct conscious streams might have a huge significance for its wellbeing capacity. typical measures of animal welfare, even suitably refined, cannot detect such basic structural features of consciousness. to take such features into account, one needs to first investigate an animal’s conscious experience in its own right and, second, reflect – using the means of philosophical value theory – on which of its conscious experiences matter. to summarize, an investigation of animal wellbeing along the lines proposed by browning (2023) promises to uncover valuable information about animal wellbeing and incorporates means to empirically test how much different features of consciousness contribute to wellbeing. at the same time, the wellbeing dimensions approach is better suited to detect differences in wellbeing capacity. also, its insights can lead to the improvement of general welfare measures as favored by browning and it can tackle aspects of consciousness which are inaccessible to the framework of browning. hence, it turns out that both approaches are complementary and should be pursued simultaneously because they can mutually enrich each other.31 31since they look for a wide range of indicators of various kinds of conscious experience, i see the systematic and very comprehensive review conducted by miller et al. (2022) as a successful first empirical implementation of the wellbeing dimensions approach. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 28 6 conclusion and preliminary ethical upshot 6.1 summary i have argued that the study of animal wellbeing will ultimately need to ground its assessments of wellbeing in a framework describing the kinds of conscious experiences various animal species can have. for the capacity for wellbeing animals have is closely related to – or entirely determined by – their capacities for consciousness. in this paper, we have arrived at a list of features of conscious experience that may constitute wellbeing and consequently are or may impact an animal’s wellbeing capacity. definitely or very likely grounds of wellbeing are: 1. evaluative intensity. 2. evaluative token-richness. 3. rate of conscious experience. 4. mental time-travel. possibly grounds of wellbeing are: 1. (non-evaluative) perceptual experience. 2. evaluative type-richness. 3. kinds of accessible evaluative experiences. 4. synchronic unity. 5. self-consciousness. in the search for adequate dimensions of animal wellbeing, these nine are the most promising candidates. it deserves mentioning that it is an open question whether the capacity for mental time-travel increases or decreases an animal’s capacity for wellbeing. this list already indicates the high uncertainty regarding the constituents of wellbeing. this uncertainty rests largely on the immaturity of the science of animal consciousness. presently, the investigation of animal consciousness on an adequately fine-grained level is largely in its infancy. when we accumulate knowledge of what animals consciously feel like in various situations and what these feelings consist in, we can expect that our intuitions regarding the conscious states in which they feel good or bad and to what extent will successively sharpen. in addition, as explained in §5.2, animal welfare science allows us to empirically learn about the contribution different features of consciousness make to wellbeing. nonetheless, the question which conscious experiences are good for an animal and how much involves an inextricable element of ethical reflection as well. 6.2 preliminary ethical upshot in spite of this uncertainty, we can see the contours of a picture where an animal’s capacity for consciousness assumes the role of the dominant arbiter of its moral weight, i.e., how much it matters morally for its own sake. in virtue of its relation to wellbeing, consciousness can serve the requisite normative role, and thanks to the advances of consciousness science, it is also susceptible to empirical investigation. while i take detailed normative assessments of animal wellbeing and moral weight to depend on a research program which must still progress into maturity, i will note at least four early normatively relevant implications of the wellbeing dimensions approach: first, there is a non-negligible chance that some animal species possess a similarly high or higher wellbeing capacity than humans. this dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 29 follows from the preceding sections. there are indications that some of the features we listed as potentially increasing wellbeing capacity, namely rate of conscious experience and synchronic disunity, are more pronounced in some other species than in humans. in regard to mental time-travel, and arguably other cognitive sophistications only available to humans (schukraft, 2020b), there is something to be said in favor of the view that they don’t increase, but decrease the wellbeing capacity of their possessor. in light of the connection between wellbeing and moral weight, depending on one’s collateral ethical commitments, it may have strongly revisionary ethical consequences if humans turn out to have less wellbeing capacity than some of our fellow species. second, some of the ways in which human decisions (explicitly or implicitly) trade-off between the wellbeing of different animals pose potentially disastrous suffering risks. as a first example, forms of animal advocacy, e.g. leafletting, when they emphasize the suffering of large mammals, may cause substitution effects, where humans substitute for their lowered consumption of some meat, e.g. beef, with higher consumption of other meat, e.g. chicken and fish. if the capacity for wellbeing of chickens and fish is not decreased much more than one order of magnitude relative to cows, this substitution probably increases overall animal suffering, i.e. negative wellbeing, since one needs to produce many more chickens and fish to satisfy the same demand for calories than cows. such a suffering risk through substitution effects is also posed by a tax on environmental harms of food products, since beef production causes more co2 emissions than the production of fish and chicken meat (ritchie, 2020) such that the relative price of the latter kinds of meat would decrease.32 a further example is insect farming which is sometimes praised as an ethical alternative to common forms of meat (for an overview, see lambert et al., 2021). however, if the capacity for wellbeing of farmed insects is not numerous orders of magnitude below the wellbeing capacity of other farmed animals, the sheer number of insects which would be necessary to satisfy demand would guarantee that the suffering caused greatly surpasses the suffering stemming from factory farming of larger animals (sebo & schukraft, 2021). knowledge about capacity for wellbeing is necessary to avoid such suffering risks. this brings us to the third implication of the approach proposed here which is that the value of information about animal wellbeing and derivatively the kinds of conscious experiences various animal species are capable of is extremely high. due to our uncertainty, credible estimates of animal wellbeing levels span many orders of magnitude (muehlhauser, 2017). due to the connection between moral weight and animal wellbeing, diminishing our uncertainty would have grave consequences for which interventions and trade-offs we deem worthwhile. neverthe32as remarked by an anonymous reviewer, harmful substitution effects may also occur in animal experimentation where ethics guidelines advise researchers to, ceteris paribus, reduce the use of mammals in research in favor of seemingly less sophisticated animals, such as invertebrates (lee et al., 2020; mandal & parija, 2013). dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 30 less, this kind of research is of course itself subject to ethical constraints (mazor et al., 2022). fourth, we have seen that many actions that are meant to improve the lives of animals have effects on other animal species or humans. sometimes, e.g. for some meat consumption substitution effects, these effects are unintended and subtle. given that these trade-offs between the wellbeing of different species are ubiquitous, the scope of precautionary principles which urge us to prioritize the avoidance of risks of inflicting or allowing animal suffering is extremely limited. in most real-world situations, we have to compare the value of animal wellbeing with other values and ethical constraints. while the manner in which these trade-offs are conducted depends on the ethical framework one subscribes to, almost all theories place at least some weight on animal wellbeing. this is why we urgently have to learn more about it. acknowledgments i thank jonathan birch, eline kuipers, two anonymous reviewers and the participants of the conference on interspecies comparisons of welfare at the london school of economics for helpful comments and advice. references appleby, m. c., & sandøe, p. (2002). philosophical debate on the nature of well-being: implications for animal welfare. animal welfare, 11(3), 283–294. https://doi.org/10.1017/s0962728600024866 beauchamp, t. (2019). the principle of beneficence in applied ethics. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (spring 2019). metaphysics research lab, stanford university. https://plato.stanford.edu/archives/spr201 9/entries/principle-beneficence/ birch, j. (2017). animal sentience and the precautionary principle. animal sentience, 2(16). https://doi.org/10.51291/23777478.1200 birch, j. (2022a). materialism and the moral status of animals. the philosophical quarterly. https://doi.org/10.1093/pq/pqa b072 birch, j. (2022b). should animal welfare be defined in terms of consciousness? philosophy of science, 1–11. https://doi.org/10.1017/psa.2022.59 birch, j., schnell, a. k., & clayton, n. s. (2020). dimensions of animal consciousness. trends in cognitive sciences, 24(10), 789–801. https://doi.org/10.1016/j.tics.2020.07.007 bramble, b. (2016). a new defense of hedonism about well-being. ergo, an open access journal of philosophy, 3. https://doi.org/https://doi.org/10.3998/ergo.12405314.0003.004 browning, h. (2020). if i could talk to the animals: measuring subjective animal welfare. browning, h. (2022). assessing measures of animal welfare. biology & philosophy, 37(4), 36. https://doi.org/10.1007/s10539022-09862-1 browning, h. (2023). welfare comparisons within and across species. philosophical studies. https://doi.org/10.1007/s11098022-01907-1 bruckner, d. w. (2010). subjective well-being and desire satisfaction. philosophical papers, 39(1), 1–28. https://doi.org/10.1 080/05568641003669409 carls-diamante, s. (2017). the octopus and the unity of consciousness. biology & philosophy, 32(6), 1269–1287. https://doi.org/10.1007/s10539-017-9604-0 chalmers, d. j. (2022). reality+: virtual worlds and the problems of philosophy. new york: w. w. norton. clark, r. e., & squire, l. r. (1998). classical conditioning and brain systems: the role of awareness. science, 280(5360), 77–81. https://doi.org/10.1126/science.280.5360.77 dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1017/s0962728600024866 https://plato.stanford.edu/archives/spr2019/entries/principle-beneficence/ https://plato.stanford.edu/archives/spr2019/entries/principle-beneficence/ https://doi.org/10.51291/2377-7478.1200 https://doi.org/10.51291/2377-7478.1200 https://doi.org/10.1093/pq/pqab072 https://doi.org/10.1093/pq/pqab072 https://doi.org/10.1017/psa.2022.59 https://doi.org/10.1016/j.tics.2020.07.007 https://doi.org/10.3998/ergo.12405314.0003.004 https://doi.org/10.1007/s10539-022-09862-1 https://doi.org/10.1007/s10539-022-09862-1 https://doi.org/10.1007/s11098-022-01907-1 https://doi.org/10.1007/s11098-022-01907-1 https://doi.org/10.1080/05568641003669409 https://doi.org/10.1080/05568641003669409 https://doi.org/10.1007/s10539-017-9604-0 https://doi.org/10.1126/science.280.5360.77 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 31 clayton, n. s., & dickinson, a. (1998). episodic-like memory during cache recovery by scrub jays. nature, 395(6699), 272–274. https://doi.org/10.1038/26216 crisp, r. (2006). hedonism reconsidered. philosophy and phenomenological research, 73(3), 619–645. https://doi.org/10.111 1/j.1933-1592.2006.tb00551.x crisp, r. (2021). well-being. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (winter 2021). metaphysics research lab, stanford university. https://plato.stanford.edu/archives/win2021/entries/well-being/ crook, r. j. (2021). behavioral and neurophysiological evidence suggests affective pain experience in octopus. iscience, 24(3). https://doi.org/10.1016/j.isci.2021.102229 crump, a., browning, h., schnell, a., burn, c., & birch, j. (2022). sentience in decapod crustaceans: a general framework and review of the evidence. animal sentience, 7(32). https://doi.org/10.51291/2377-7478.1691 cutter, b. (2017). the metaphysical implications of the moral significance of consciousness. philosophical perspectives, 31(1), 103–130. https://doi.org/10.1111/phpe.12092 damasio, a. (1999). the feeling of what happens: body and emotion in the making of consciousness. harcourt college publishers. dawkins, m. s. (2017). animal welfare with and without consciousness. journal of zoology, 301(1), 1–10. https://doi.org/10 .1111/jzo.12434 dawkins, m. s. (2021). the science of animal welfare: understanding what animals want (1st ed.). oxford university press. https://doi.org/10.1093/oso/9780198848981.001.0001 de vere, a. j., & kuczaj, s. a. (2016). where are we in the study of animal emotions? wiley interdisciplinary reviews: cognitive science, 7(5), 354–362. https://doi.org/10.1002/wcs.1399 dennett, d. c. (1995). animal consciousness: what matters and why? social research: an international quarterly, 62, 691–710. droege, p., weiss, d. j., schwob, n., & braithwaite, v. (2021). trace conditioning as a test for animal consciousness: a new approach. animal cognition, 24(6), 1299–1304. https://doi.org/10.1007/s10071-021-01522-3 duncan, i. j. h., & fraser, d. (1997). understanding animal welfare. in m. c. appleby & b. o. hughes (eds.), animal welfare (pp. 19–32). cab international. duncan, m. (2019). the self shows up in experience. review of philosophy and psychology, 10(2), 299–318. https://doi.org/10.1007/s13164-017-0355-2 dung, l. (forthcoming). preserving the normative significance of sentience. journal of consciousness studies. dung, l. (2022a). does illusionism imply skepticism of animal consciousness? synthese, 200(3), 238. https://doi.org/10.100 7/s11229-022-03710-1 dung, l. (2022b). assessing tests of animal consciousness. consciousness and cognition, 105, 103410. https://doi.org/10.101 6/j.concog.2022.103410 dung, l. (2022c). why the epistemic objection against using sentience as criterion of moral status is flawed. science and engineering ethics, 28(6), 51. https://doi.org/10.1007/s11948-022-00408-y dung, l., & newen, a. (2023). profiles of animal consciousness: a species-sensitive, two-tier account to quality and distribution. cognition, 235, 105409. https://doi.org/10.1016/j.cognition.2023.105409 fischer, b. (2022). theories of welfare and welfare range estimates [ea forum]. https://forum.effectivealtruism.org/posts/wf ewn2x4k8w8nteas/theories-of-welfare-and-welfare-range-estimates fletcher, g. (2013). a fresh start for the objective-list theory of well-being. utilitas, 25(2), 206–220. https://doi.org/10.1017/ s0953820812000453 fletcher, g. (ed.). (2015). the routledge handbook of philosophy of well-being. routledge. https://doi.org/10.4324/97813156 82266 gaffney, l. p., lavery, j. m., schiestl, m., trevarthen, a., schukraft, j., miller, r., schnell, a. k., & fischer, b. (2023). a theoretical approach to improving interspecies welfare comparisons. frontiers in animal science, 3. https://www.fron tiersin.org/articles/10.3389/fanim.2022.1062458 gallagher, s. (2010). defining consciousness: the importance of non-reflective self-awareness. pragmatics & cognition, 18(3), 561–569. https://doi.org/10.1075/pc.18.3.04gal galpayage dona, h. s., solvi, c., kowalewska, a., mäkelä, k., maboudi, h., & chittka, l. (2022). do bumble bees play? animal behaviour, 194, 239–251. https://doi.org/10.1016/j.anbehav.2022.08.013 gibbons, m., crump, a., barrett, m., sarlak, s., birch, j., & chittka, l. (2022). can insects feel pain? a review of the neural and behavioural evidence. in advances in insect physiology. academic press. https://doi.org/10.1016/bs.aiip.2022.10. 001 godfrey-smith, p. (2020). metazoa: animal minds and the birth of consciousness. william collins. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1038/26216 https://doi.org/10.1111/j.1933-1592.2006.tb00551.x https://doi.org/10.1111/j.1933-1592.2006.tb00551.x https://plato.stanford.edu/archives/win2021/entries/well-being/ https://doi.org/10.1016/j.isci.2021.102229 https://doi.org/10.51291/2377-7478.1691 https://doi.org/10.1111/phpe.12092 https://doi.org/10.1111/jzo.12434 https://doi.org/10.1111/jzo.12434 https://doi.org/10.1093/oso/9780198848981.001.0001 https://doi.org/10.1002/wcs.1399 https://doi.org/10.1007/s10071-021-01522-3 https://doi.org/10.1007/s13164-017-0355-2 https://doi.org/10.1007/s11229-022-03710-1 https://doi.org/10.1007/s11229-022-03710-1 https://doi.org/10.1016/j.concog.2022.103410 https://doi.org/10.1016/j.concog.2022.103410 https://doi.org/10.1007/s11948-022-00408-y https://doi.org/10.1016/j.cognition.2023.105409 https://forum.effectivealtruism.org/posts/wfewn2x4k8w8nteas/theories-of-welfare-and-welfare-range-estimates https://forum.effectivealtruism.org/posts/wfewn2x4k8w8nteas/theories-of-welfare-and-welfare-range-estimates https://doi.org/10.1017/s0953820812000453 https://doi.org/10.1017/s0953820812000453 https://doi.org/10.4324/9781315682266 https://doi.org/10.4324/9781315682266 https://www.frontiersin.org/articles/10.3389/fanim.2022.1062458 https://www.frontiersin.org/articles/10.3389/fanim.2022.1062458 https://doi.org/10.1075/pc.18.3.04gal https://doi.org/10.1016/j.anbehav.2022.08.013 https://doi.org/10.1016/bs.aiip.2022.10.001 https://doi.org/10.1016/bs.aiip.2022.10.001 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org leonard dung 32 gottlieb, j. (2022). octopuses (probably) don't have nine minds. ea forum. https://forum.effectivealtruism.org/posts/wb wxlztf5fpzmk8nh/octopuses-probably-don-t-have-nine-minds grahek, n. (2001). feeling pain and being in pain (2nd ed.). mit press. green, t. c., & mellor, d. j. (2011). extending ideas about animal welfare assessment to include 'quality of life' and related concepts. new zealand veterinary journal, 59(6), 263–271. https://doi.org/10.1080/00480169.2011.610283 griffin, j. (1988). well-being: its meaning, measurement and moral importance (1st ed.). oxford university pressoxford. https://doi.org/10.1093/0198248431.001.0001 heathwood, c. (2016). desire-fulfillment theory. in g. fletcher (ed.), the routledge handbook of the philosophy of well-being (pp. 135–147). routledge. heathwood, c. (2019). which desires are relevant to well-being? noûs, 53(3), 664–688. https://doi.org/10.1111/nous.12232 hoffman, k. n. (2020). subjective experience in explanations of enimal ptsd behavior. philosophical topics, 48(1), 155–175. https://doi.org/10.5840/philtopics20204818 hurley, s., l. (1998). consciousness in action. harvard university press. inger, r., bennie, j., davies, t. w., & gaston, k. j. (2014). potential biological and ecological effects of flickering artificial light. plos one, 9(5), e98631. https://doi.org/10.1371/journal.pone.0098631 kagan, s. (2019). how to count animals, more or less (1st ed.). oxford university press. https://doi.org/10.1093/oso/978019 8829676.001.0001 kammerer, f. (2019). the normative challenge for illusionist views of consciousness. ergo, an open access journal of philosophy, 6. https://doi.org/https://doi.org/10.3998/ergo.12405314.0006.032 kammerer, f. (2022). ethics without sentience. facing up to the probable insignificance of phenomenal consciousness. journal of consciousness studies. kremer, l., klein holkenborg, s. e. j., reimert, i., bolhuis, j. e., & webb, l. e. (2020). the nuts and bolts of animal emotion. neuroscience & biobehavioral reviews, 113, 273–286. https://doi.org/10.1016/j.neubiorev.2020.01.028 kriegel, u. (2019). the value of consciousness. analysis, 79(3), 503–520. https://doi.org/10.1093/analys/anz045 lagisz, m., zidar, j., nakagawa, s., neville, v., sorato, e., paul, e. s., bateson, m., mendl, m., & løvlie, h. (2020). optimism, pessimism and judgement bias in animals: a systematic review and meta-analysis. neuroscience & biobehavioral reviews, 118, 3–17. https://doi.org/10.1016/j.neubiorev.2020.07.012 lambert, h., elwin, a., & d’cruze, n. (2021). wouldn’t hurt a fly? a review of insect cognition and sentience in relation to their use as food and feed. applied animal behaviour science, 243, 105432. https://doi.org/10.1016/j.applanim.2021.10 5432 lee, k. h., lee, d. w., & kang, b. c. (2020). the “r” principles in laboratory animal experiments. laboratory animal research, 36(1), 45. https://doi.org/10.1186/s42826-020-00078-6 lin, e. (2014). pluralism about well-being. philosophical perspectives, 28(1), 127–154. https://doi.org/10.1111/phpe.12038 mandal, j., & parija, s. c. (2013). ethics of involving animals in research. tropical parasitology, 3(1), 4–6. https://doi.org/10 .4103/2229-5070.113884 mason, g. j., & lavery, j. m. (2022). what is it like to be a bass? red herrings, fish pain and the study of animal sentience. frontiers in veterinary science, 9. https://www.frontiersin.org/articles/10.3389/fvets.2022.788289 may, a. (2005). cluster headache: pathogenesis, diagnosis, and management. the lancet, 366(9488), 843–855. https://doi.org/10.1016/s0140-6736(05)67217-0 mazor, m., brown, s., ciaunica, a., demertzi, a., fahrenfort, j., faivre, n., francken, j. c., lamy, d., lenggenhager, b., moutoussis, m., nizzi, m.-c., salomon, r., soto, d., stein, t., & lubianiker, n. (2022). the scientific study of consciousness cannot and should not be morally neutral. perspectives on psychological science, 174569162211102. https://doi.org/10.1177/17456916221110222 mee, j., & boyle, l. (2020). assessing whether dairy cow welfare is “better” in pasture-based than in confinement-based management systems. new zealand veterinary journal, 68(3), 168–177. https://doi.org/10.1080/00480169.2020.1721034 mendl, m., burman, o. h. p., & paul, e. s. (2010). an integrative and functional framework for the study of animal emotion and mood. proceedings of the royal society b: biological sciences, 277(1696), 2895–2904. https://doi.org/10.1098/rspb.2 010.0303 metzinger, t. (2021). artificial suffering: an argument for a global moratorium on synthetic phenomenology. journal of artificial intelligence and consciousness, 8(1), 43–66. https://doi.org/10.1142/s270507852150003x miller, r., schiestl, m., trevarthen, a., gaffney, l., lavery, j. m., fischer, b., & schnell, a. (2022). from pigs to silkworms: cognition and welfare across 10 farmed taxa [preprint]. animal behavior and cognition. https://doi.org/10.1101/2022 .11.11.516141 millsopp, s., & laming, p. (2008). trade-offs between feeding and shock avoidance in goldfish (carassius auratus). applied animal behaviour science, 113(1-3), 247–254. https://doi.org/10.1016/j.applanim.2007.11.004 dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://forum.effectivealtruism.org/posts/wbwxlztf5fpzmk8nh/octopuses-probably-don-t-have-nine-minds https://forum.effectivealtruism.org/posts/wbwxlztf5fpzmk8nh/octopuses-probably-don-t-have-nine-minds https://doi.org/10.1080/00480169.2011.610283 https://doi.org/10.1093/0198248431.001.0001 https://doi.org/10.1111/nous.12232 https://doi.org/10.5840/philtopics20204818 https://doi.org/10.1371/journal.pone.0098631 https://doi.org/10.1093/oso/9780198829676.001.0001 https://doi.org/10.1093/oso/9780198829676.001.0001 https://doi.org/10.3998/ergo.12405314.0006.032 https://doi.org/10.1016/j.neubiorev.2020.01.028 https://doi.org/10.1093/analys/anz045 https://doi.org/10.1016/j.neubiorev.2020.07.012 https://doi.org/10.1016/j.applanim.2021.105432 https://doi.org/10.1016/j.applanim.2021.105432 https://doi.org/10.1186/s42826-020-00078-6 https://doi.org/10.1111/phpe.12038 https://doi.org/10.4103/2229-5070.113884 https://doi.org/10.4103/2229-5070.113884 https://www.frontiersin.org/articles/10.3389/fvets.2022.788289 https://doi.org/10.1016/s0140-6736(05)67217-0 https://doi.org/10.1177/17456916221110222 https://doi.org/10.1080/00480169.2020.1721034 https://doi.org/10.1098/rspb.2010.0303 https://doi.org/10.1098/rspb.2010.0303 https://doi.org/10.1142/s270507852150003x https://doi.org/10.1101/2022.11.11.516141 https://doi.org/10.1101/2022.11.11.516141 https://doi.org/10.1016/j.applanim.2007.11.004 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org dimensions of animal wellbeing 33 moore, a. (2019). hedonism. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (winter 2019). metaphysics research lab, stanford university. https://plato.stanford.edu/archives/win2019/entries/hedonism/ muehlhauser, l. (2017). report on consciousness and moral patienthood. open philanthropy. https://www.openphilanthro py.org/2017-report-consciousness-and-moral-patienthood nagel, t. (1974). what is it like to be a bat? philosophical review, 83(4), 435–450. https://doi.org/10.2307/2183914 nieder, a., wagener, l., & rinnert, p. (2020). a neural correlate of sensory consciousness in a corvid bird. science, 369(6511), 1626–1629. https://doi.org/10.1126/science.abb1447 pandit, p. (2021). toward a more credible principle of beneficence. journal of indian council of philosophical research, 38(3), 407–422. https://doi.org/10.1007/s40961-021-00258-2 panksepp, j. (2010). affective consciousness in animals: perspectives on dimensional and primary process emotion approaches. proceedings of the royal society b: biological sciences, 277(1696), 2905–2907. https://doi.org/10.1098/rspb.2 010.1017 park, r. m., foster, m., & daigle, c. l. (2020). a scoping review: the impact of housing systems and environmental features on beef cattle welfare. animals, 10(4), 565. https://doi.org/10.3390/ani10040565 rice, c. m. (2013). defending the objective list theory of well-being. ratio, 26(2), 196–211. https://doi.org/10.1111/rati.12007 ritchie, h. (2020). the carbon footprint of foods: are differences explained by the impacts of methane? https://ourworldinda ta.org/carbon-footprint-food-methane robbins, j., franks, b., & keyserlingk, m. a. g. von. (2018). “more than a feeling”: an empirical investigation of hedonistic accounts of animal welfare. plos one, 13(3), e0193864. https://doi.org/10.1371/journal.pone.0193864 schukraft, j. (2020a). comparisons of capacity for welfare and moral status across species. rethink priorities. https://rethin kpriorities.org/publications/comparisons-of-capacity-for-welfare-and-moral-status-across-species schukraft, j. (2020b). differences in the intensity of valenced experience across species. rethink priorities. https://rethinkpri orities.org/publications/differences-in-the-intensity-of-valenced-experience-across-species schukraft, j. (2020c). does critical flicker-fusion frequency track the subjective experience of time? rethink priorities. https://rethinkpriorities.org/publications/does-critical-flicker-fusion-frequency-track-the-subjective-experience-oftime schukraft, j. (2020d). the subjective experience of time: welfare implications. rethink priorities. https://rethinkpriorities.or g/publications/the-subjective-experience-of-time-welfare-implications sebo, j., & schukraft, j. (2021). don’t farm bugs. aeon. https://aeon.co/essays/on-the-torment-of-insect-minds-and-ourmoral-duty-not-to-farm-them van gulick, r. (2022). consciousness. in e. n. zalta & u. nodelman (eds.), the stanford encyclopedia of philosophy (winter 2022). metaphysics research lab, stanford university. https://plato.stanford.edu/archives/win2022/entries/consciou sness/ veit, w., & browning, h. (2020). perspectival pluralism for animal welfare. european journal for philosophy of science, 11(1), 9. https://doi.org/10.1007/s13194-020-00322-9 višak, t. (2022). capacity for welfare across species. oxford university press. webb, c. e., woodford, p., & huchard, e. (2020). the study that made rats jump for joy, and then killed them. bioessays, 42(6), 2000030. https://doi.org/10.1002/bies.202000030 wilcox, m. g. (2021). the intrinsic value of liberty for non-human animals. journal of value inquiry, 55(4), 685–703. https://doi.org/10.1007/s10790-020-09762-1 zahavi, d. (2014). self and other: exploring subjectivity, empathy, and shame. oxford university press. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. dung, l. (2023). dimensions of animal wellbeing. philosophy and the mind sciences, 4, 9. https://doi.org/10.33735/phimisci.2023.9878 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://plato.stanford.edu/archives/win2019/entries/hedonism/ https://www.openphilanthropy.org/2017-report-consciousness-and-moral-patienthood https://www.openphilanthropy.org/2017-report-consciousness-and-moral-patienthood https://doi.org/10.2307/2183914 https://doi.org/10.1126/science.abb1447 https://doi.org/10.1007/s40961-021-00258-2 https://doi.org/10.1098/rspb.2010.1017 https://doi.org/10.1098/rspb.2010.1017 https://doi.org/10.3390/ani10040565 https://doi.org/10.1111/rati.12007 https://ourworldindata.org/carbon-footprint-food-methane https://ourworldindata.org/carbon-footprint-food-methane https://doi.org/10.1371/journal.pone.0193864 https://rethinkpriorities.org/publications/comparisons-of-capacity-for-welfare-and-moral-status-across-species https://rethinkpriorities.org/publications/comparisons-of-capacity-for-welfare-and-moral-status-across-species https://rethinkpriorities.org/publications/differences-in-the-intensity-of-valenced-experience-across-species https://rethinkpriorities.org/publications/differences-in-the-intensity-of-valenced-experience-across-species https://rethinkpriorities.org/publications/does-critical-flicker-fusion-frequency-track-the-subjective-experience-of-time https://rethinkpriorities.org/publications/does-critical-flicker-fusion-frequency-track-the-subjective-experience-of-time https://rethinkpriorities.org/publications/the-subjective-experience-of-time-welfare-implications https://rethinkpriorities.org/publications/the-subjective-experience-of-time-welfare-implications https://aeon.co/essays/on-the-torment-of-insect-minds-and-our-moral-duty-not-to-farm-them https://aeon.co/essays/on-the-torment-of-insect-minds-and-our-moral-duty-not-to-farm-them https://plato.stanford.edu/archives/win2022/entries/consciousness/ https://plato.stanford.edu/archives/win2022/entries/consciousness/ https://doi.org/10.1007/s13194-020-00322-9 https://doi.org/10.1002/bies.202000030 https://doi.org/10.1007/s10790-020-09762-1 https://doi.org/10.33735/phimisci.2023.9878 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org animal wellbeing wellbeing and consciousness theories of wellbeing and broad hedonism wellbeing capacity and wellbeing variance wellbeing, trade-offs, and moral weight the dimensions framework of consciousness the five dimensions empirical tests dimensions of animal wellbeing i: perceptual and evaluative experience how can there be differences in wellbeing capacity? perceptual consciousness evaluative consciousness empirically accessing evaluative consciousness dimensions of animal wellbeing ii: unity and the self synchronic unity: mere correlation and split consciousness asynchronic unity: speed of consciousness and mental time-travel self-consciousness and suffering wellbeing from an actualist welfare and from a possibilist consciousness perspective browning's framework for interspecies welfare comparisons comparison between the wellbeing dimensions approach and browning's framework conclusion and preliminary ethical upshot summary preliminary ethical upshot look who's talking! varieties of ego-dissolution without paradox look who’s talking! varieties of ego-dissolution without paradox sascha benjamin finka (sfink@ovgu.de) abstract how to model non-egoic experiences – mental events with phenomenal aspects that lack a felt self – has become an interesting research question. the main source of evidence for the existence of such non-egoic experiences are self-ascriptions of non-egoic experiences. in these, a person says about herself that she underwent an episode where she was conscious but lacked a feeling of self. some interpret these as accurate reports, but this is questionable. thomas metzinger (2004, p. 566, 2018), rocco gennaro (2008), and charles foster (2016, p. 6) have hinted at the self-defeating nature of such statements if we take them to be genuine reports: apparently, the reporter (a) explicitly denies her existence during the selfless experience, but (b) implicitly affirms her existence as a witness to that selfless experience in order to give a first-person report about it. so the content of such a report conflicts with the pragmatics of reporting. if all self-ascriptions of non-egoic experiences are self-defeating in this way, then they cannot count as evidence for the existence of non-egoic experiences. here, i map out why such strong conclusions do not directly follow: what look like self-ascriptions of non-egoic experiences may occur for a number of reasons. only some explanations for such utterances rely on a change in consciousness. of those that do rely on a change in consciousness, only one (total ego-dissolution) is incoherent. but its alternatives do not lead to contradictions. i argue that the most likely change in phenomenality that leads to self-ascriptions of non-egoic experiences is not one where a felt self disappears, but where it expands. keywords consciousness ∙ ego-dissolution · ego-expansion · multiple selves ∙ non-egoic experiences · no-self · self-refutation this article is part of a special issue on “radical disruptions of self-consciousness”, edited by thomas metzinger and raphaël millière. we are led to believe that some experiences are non-egoic: they feel like something, but lack a felt self; they are clearly phenomenal, but the ego is not part of this experience; no one is in the experience, but it still has qualitative features. if there was a feeling of self in experience before one went into a non-egoic episode, we call aotto-von-guericke-university magdeburg fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-7535-1882 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 2 this process phenomenal ego-dissolution.1 non-egoic episodes are characterised by (i) the absence of a feeling of self, and (ii) a noticeable contrast to a feeling of self experienced previously. apparently, people reporting selfless experiences did not always lack a feeling of self; instead, it is unusual, something that contrasts with one’s common ways of experiencing, something remarkable. why should we think that there are such experiences? some people claim that they have undergone (or are undergoing) an episode during which they themselves experienced in a non-egoic way. these claims come in the form of a reflexive first-person ascription, where a person ascribes to herself some mental event – in this case, an episode of phenomenal ego-dissolution, either in the past or in the present.2 such self-ascriptions of (episodes of) non-egoic experience are what i focus on here. but because this phrase is cumbersome, we can use the abbreviation sane for them. strictly speaking, sane could be merely mental and need not be verbally expressed. but if they never were expressed, they would not encourage others to think that such non-egoic experiences exist. it is the verbal expression of such self-ascriptions of non-egoic experiences that counts as intersubjective evidence for there being non-egoic experiences. this is what i focus on here: utterances which, in their most likely reading, involve or are reasonably taken to express sane. such verbal expressions are what i mean when i use the abbreviation sane. and they are primarily responsible for the sane paradox i focus on (section 3):3 that sane are self-refuting, that their structure entails their own falsity. this paper presents examples of sane and their phenomenological profile in section 1, characterises their importance for a research programme on non-egoic experiences in section 2, and analyses the argument for them being self-refuting in section 3, where the underlying assumptions for this claim are made explicit. now, i do not think that sane need to be self-refuting. in fact, we have diagnoses available on several levels – pragmatic (section 4.1), cognitive (section 4.2), and phenomenological (section 5) – to explain why people utter sane. the sane paradox only comes about by embracing (i) a connection between the phenomenal feeling of self and the underlying mechanisms facilitating self-reference and (ii) that there is only one unique phenomenal feeling of self. each of these two claims can be rejected separately, leading to competing diagnoses about what happens in those phenomenal episodes some sane refer to. should we keep all of these diagnoses in our tool box? in section 5, i argue against such a pluralism: we should expect at most one of these diagnoses to be accurate. in order to defend the feasibility of researching phenomenal ego-dissolutions 1conceptually, such episodes of phenomenal ego-dissolution can be transient (if one regains a feeling of self afterwards) or (otherwise) permanent. some buddhist monks may claim that they achieved permanent ego-dissolution. but most cases appear to be of the transient kind. therefore, i will restrict myself to these. all theoretical points i make, however, will apply to the permanent kind as well. 2that is, she ascribes to herself the property of experiencing non-egoically at a time 𝑡 and a location 𝑙 of her physical body. 3there are, however, cognitive analogues. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 3 and non-egoic experiences, we must give an account of how sane can offer evidence for some form of phenomenal ego-dissolution. in order to do so, we need to understand better the argument why sane are supposed to be self-defeating. then we can catalogue the ways in which we can reject this argument. these are the goals of this paper. 1 sane introduced: self-ascriptions of episodes of non-egoic experiences and phenomenal egodissolution some people say that they have had episodes in their mental life where they felt like no one at all, where their selves went missing, where there was no feeling of self. these episodes, however, were not of unconsciousness, akin to anaesthesia or deep sleep. instead, they had phenomenal experiences with felt qualities, but these felt as if they were happening to no person, no self. i call such experiences nonegoic experiences, and a process of leading into them a phenomenal ego-dissolution, because, intuitively, there was a prior feeling of self that then vanished. people express their beliefs about having had such experiences in verbalised self-ascriptions of (episodes of) non-egoic experience, which i abbreviate as sane. sane can be found in a wide range of circumstances. in psychiatry, egodissolution is seen as a symptom of certain psychopathologies (sass, pienkos, nelson, & medford, 2013; simeon & abugel, 2006). elyn saks (2007, pp. 12–13) has provided us with one of the most vivid examples in the center cannot hold: and then something odd happens. my awareness (of myself, of him, of the room, of the physical reality around and beyond us) instantly grows fuzzy. or wobbly. i think i am dissolving. i feel – my mind feels – like a sand castle with all the sand sliding away in the receding surf. what’s happening to me? this is scary, please let it be over! i think maybe if i stand very still and quiet, it will stop. this experience is much harder, and weirder, to describe than extreme fear or terror. most people know what it is like to be seriously afraid. if they haven’t felt it themselves, they’ve at least seen a movie, or read a book, or talked to a frightened friend – they can at least imagine it. but explaining what i’ve come to call “disorganization” is a different challenge altogether. consciousness gradually loses its coherence. one’s center gives way. the center cannot hold. the “me” becomes a haze, and the solid center from which one experiences reality breaks up like a bad radio signal. there is no longer a sturdy vantage point from which to look out, take things in, assess what’s happening. no core holds things together, providing the lens through which to see the world, to make judgments and comprehend risk. ranfink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 4 dom moments of time follow one another. sights, sounds, thoughts, and feelings don’t go together. no organizing principle takes successive moments in time and puts them together in a coherent way from which sense can be made. and it’s all taking place in slow motion. of course, my dad didn’t notice what had happened, since it was all happening inside me. saks calls this experience disorganised, marked primarily by a loss of mereological, spatial, or temporal unity (“reality breaks up”, “no core holds things together”, “random moments of time follow another”), a loss of determinacy (“my awareness […] grows fuzzy […] wobbly”), or a loss of coherence. this is tied in with a reduced form of specificity of the self-boundary: “the ‘me’ becomes a haze.” she appears to see the dissolution of her feeling of self as the cause of the other forms of disorganisation: the fact that a core, a centre, a “sturdy vantage point” is missing leads to the lack of unity in these experiences. without a feeling of self, the senses do not cohere among each other or with cognition (e.g. judgments and comprehension of risk). even time seems to lack a steady, appreciable pace or sequence. the feeling of self, it appears, is associated by saks with diachronic and synchronic phenomenal unity. and, notably: this radical change in experience is not behaviourally noticeable for the people around her. this experience appears to be distant from our everyday way of experiencing, hardly describable, alien. the difficulty is apparent in a specific reluctance to use indexicals as ways to orientate and locate oneself in time or space: “i feel” – herself being involved in feeling – is replaced by “my mind feels”, where she is merely the owner of the mind, herself apparently dissociated from what she owns, like a traveller who is far away from the home she owns. then again, “my” becomes replaced by “the ‘me’ ”, a thing one has, not what one is. elsewhere, all indexicals are avoided, when “one’s center” – still intended as referring to an indeterminate someone who has a centre – is replaced by “the center”, which lacks any indication of being the centre of a person. this creative sensitivity to the subtleties of language is what makes the passage so captivating. but we should be wary of the possibility that literary quality may have trumped veridicality here. what supports saks’s description, however, is that other reports converge on similar themes, for example reports of psychedelic experiences (see especially letheby & gerrans, 2017; millière, 2017; see also deane, 2020, this issue, as well as letheby, 2020, this issue): i realized then that i wasnt [sic] myself, i wasn’t anything anymore. i had been broken down into nothingness, into oblivion. i felt as though the darkness was more of a something than i. i’m fucking dead i thought. there is no other explanation. i felt nothingness, like my brain had just been paused and i was still taking in my surroundings unwillingly. i felt comatose.4 4erowid report #87426. https://erowid.org/exp/87426. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 5 during this 5-meo-dmt-induced experience, rew (17 years old at the time), felt like he had dissolved into “nothingness”, that even immaterial darkness was more existent than he was. no feeling of self seems to be at the core of this experience. additionally, even though he was perceptive (“i was still taking in my surroundings”), he apparently did not feel like an agent in this process (he did so “unwillingly”). while such experiences appear to be frightening when they happen spontaneously, some people try to attain them deliberately: non-egoic experiences are the goal of certain spiritual practice, a sign of having mastered certain meditative or contemplative techniques (berkovich-ohana, dor-ziderman, glicksohn, & goldstein, 2013; dor-ziderman, berkovich-ohana, glicksohn, & goldstein, 2013; josipovic, 2010), indicating a high level of enlightenment, awakening, or connection with the divine. what is a symptom of pathology in one context is a pinnacle of spirituality in another. reoccurring features of sane are: (cease) a feeling of self is missing in experience. (access) the absence of a feeling of self is cognitively accessible.5 without cease, the experience would not be non-egoic. without access, we would not have any reports or memories of these experiences. both are common themes in sane. if taken at face value, sane suggest that we can undergo nonegoic episodes in meditation, schizophrenia, or psychedelic highs. if non-egoic episodes occur in such a wide range of circumstances, are they anything special at all? intuitively, we might think so, but some philosophers argue that experiences are mainly egoless, at least to some degree. for example, in la transcendance et l’ego, jean-paul sartre reflects on his problems with husserl and public transport: no one would deny for a moment that the i appears in a reflected consciousness. […] [this reflective consciousness, however,] modifies the spontaneous consciousness. since, in consequence, all the non-reflective memories of unreflected consciousness show me a consciousness without a me and since, on the other hand, theoretical considerations concerning consciousness which are based on intuition of essence have constrained us to recognise that the i cannot be a part of the internal structure of erlebnisse, we must therefore conclude: there is no i on the unreflected level.6 when i run after a streetcar, when i 5especially: accessible for conceptualisation and conceptual processing, e.g. belief and memory formation, reasoning, deliberation, evaluation, etc. only conceptualisation will lead to verbal reports about non-egoic experiences. 6obviously, it does not follow that if 𝑎 modifies 𝑏, then whatever holds for 𝑎 cannot hold for 𝑏, because this would exclude mutual modification. let us read this passage simply as indicating a distinction between “reflective” and “non-reflective” consciousness. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 6 look at the time, when i am absorbed in contemplating a portrait, there is no i. there is consciousness of the streetcar-having-to-be-overtaken, etc., and non-positional consciousness of consciousness. (sartre, 1960, pp. 48–49)7 sartre argues that there is a basic, unreflected level of living consciousness, the erlebnisse. these are experiences of simple perceiving without apprehending (like being absorbed in looking at a portrait) and of actions without explicit intentions (like overtaking a streetcar). on this level, there is no self in the experience. a felt self is introduced only by reflection. here, one has to mark oneself as the thinker of thoughts, the initiator of actions, the person inhabiting a body, and one has to dissociate oneself from other thinkers, agents, and bodies. in a weaker reading, sartre argues that even if it is undeniable that a self exists on a reflected level (where we can remember and report on erlebnisse as being my experiences, think about our lived experiences, and have cognitive access to these experiences), this does not strictly entail that a self is also part of our unreflected experiences. then, non-egoic experiences are a possible kind of phenomenal consciousness outside of self-involving conscious cognition. owen flanagan (1992, p. 178)8 and david hume (1739, p. i.4.vi)9 seem to go even further and claim that experiences standardly lack a felt self, that there is nothing it feels like to be someone, no self-quale or feeling of self. but if experiences are always non-egoic, we cannot account for a very prominent feature of sane: there is a clear change from what is felt before and after to what the experience is like during the non-egoic episode. this is a third recurring feature of sane: (difference) there is a felt difference in the feeling of self between the time before/after and during non-egoic episodes. without difference, a non-egoic experience would not be remarkable, but they clearly seem to be. the moments before and after contrast sharply with those during the episodes of ego-dissolution. we try to capture this difference by questionnaires like the 5d-asc (studerus, gamma, & vollenweider, 2010) or the edi 7in the original: “personne ne songe à nier que le je apparaisse dans une conscience réfléchie. […] la réflexion modifie la conscience spontanée. puisque donc tous les souvenirs non-réflexifs de conscience irréfléchie me montrent une conscience sans moi […]: il n’y a pas de je sur le plan irréfléchi. quand je cours après un tramway, quand je regarde l’heure, quand je m’absorbe dans la contemplation d’un portrait, il n’y a pas de je. il y a conscience du tramway-devant-être-rejoint, etc., et conscience non-positionnelle de la conscience.” (sartre, 1936/1992) 8“the illusion is that there are two things: on one side, a self, an ego, an ‘i’, that organizes experience, originates action, and accounts for our unchanging identity as persons and, on the other side, the stream of experience. if this view is misleading, what is the better view? the better view is that what there is, and all there is, is the stream of experience.” 9“[…] identity is nothing really belonging to these different perceptions, and uniting them together; but is merely a quality, which we attribute to them, because of the union of their ideas in the imagination, when we reflect upon them.” fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 7 (nour, evans, nutt, & carhart-harris, 2016), which specifically probe aspects of ego-distortion. it is this felt difference from everyday experiencing that is unaccounted for by philosophers who deny a feeling of self in general. sartre seems to be providing an account of something other than radical non-egoic experiences because following a street car is hardly as dramatic as the experiences described by saks and rew seem to be. a theory that denies the existence of a feeling of self in general or suggests that non-egoic experiences are extremely widespread does not explain the kind of data gathered in the validated questionnaires. something special appears to be going on during these episodes. 2 sane and their role in researching non-egoic experiences we may be tempted to focus directly on the question: what can we learn from sane about the specific features of non-egoic experience? – and simply proceed with a research programme centred on non-egoic experiences themselves. but maybe we should start by asking about the epistemology of egodissolutions: how can we know what happens in such episodes? how can we know about the specifics, the dynamics, the onset and offset of these episodes? how can we determine which phenomenal features and structures are affected? what evidence do we have that there is a difference in experience, not merely in reporting? what grounds do we have to think that phenomenal ego-dissolution exists and is not some kind of artefact? what could count as primary evidence for non-egoic experiences? probably, introspecting non-egoic episodes ourselves. however, few of us seem to have had them – and, contrary to hume, sartre, and flanagan, even fewer of us describe our everyday experiences as non-egoic. such experiences appear to be rare in the general population but also in the life of any individual. they are also hard to induce reliably in an experimental setting. therefore, we should not expect a great amount of data from introspection. additionally, in states where individuals might experience ego-dissolution, a broad range of cognitive capacities also appear to be altered, so we should not expect data from reliable, rigorous, or systematic first-person methods. these worries strengthen general doubts concerning introspection or first-person data (see e.g. schwitzgebel, 2012, 2008), doubts we should take seriously. introspection – an individual directly registering her own mental states – may lead the introspecting individual herself to believe in non-egoic experiences; but introspection does not provide sufficient reason for the scientific community as a whole to accept the existence of such experiences. the scientific community as a whole will prefer intersubjectively accessible evidence. but phenomenal experiences themselves cannot be intersubjectively observed. so maybe we must consider indirect or secondary evidence – not direct evidence for the presence of such experiences, but evidence for others having evfink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 8 idence of the presence of such experiences. ideally, such evidence comes in the form of non-verbal behaviour. but behaviour appears to be silent concerning the presence or absence of a self, and we currently lack any decisive neural or physiological markers for such states. alvin goldman (1997) argues that in order to establish such markers as reliable indicators of consciousness, we have to calibrate and validate them by using first-person reports. only then could we build a foundation that allows us to use such non-verbal markers as evidence. it seems that there is no way around first-person reports, at least in an initial stage, whether free-form or as questionnaire responses. our core evidence for the existence and nature of phenomenal ego-dissolution is therefore verbalised self-ascriptions of episodes of non-egoic experiences, which i abbreviate as sane. usually, these will be presented as recollections of having been in a non-egoic state, but sometimes (e.g. in psychopathological cases) they might be presented as reports of an ongoing non-egoic episode.10 some contain possessive phrases (e.g. “my self was absent”), while others use the indexical “i” to claim nonexistence (e.g. “i did not exist”). if sane are the foundation for establishing that non-egoic experiences exist and what they are like, then that’s bad news: such reports seem to be selfdefeating, some claim. if so, we have no evidence for the existence of phenomenal ego-dissolution and, a fortiori, any sane actually speaks against phenomenal ego-dissolution: that they exist indicates that they are false, goes the argument. hence, no research programme on non-egoic experiences can get off the ground: the very evidence taken to prove their existence and their nature, sane, indirectly proves their absence. why should sane be self-refuting in this way? here is a rough sketch. grammatically, sane can come in the past tense (“i did not feel like anybody”) or in the present tense (“i do not feel like anybody”). in character and form, past tense sane resemble reports about remembered feelings, dream reports made upon awakening, or retrospective reports about psychedelic trips; sane given in the present tense (e.g. during meditation) resemble introspective reports about ongoing experiences (see section 1). so sane resemble reports. but are they reports? our natural stance is to interpret sane as reports. thus, we treat them as assertions, and therefore: as aiming at truth. we presume that the experiencer witnessed her non-egoic experience and afterwards describes what she witnessed more or less veridically (or, at least, describes it with the full intent to be veridical).11 thus, if we treat them as reports, we treat them as giving us insight into non-egoic states of consciousness. 10for example: at the workshop on radical disruptions of self-consciousness at frankfurt’s fias in october 2018, aviva berkovich-ohana showed a video clip of a meditator talking during meditation about “falling into space”, which may be understood as a report about entering a non-egoic experience: “a sense that there is no need for center […] as if am falling out of the center. […] no need to be located anywhere.” 11otherwise, we would interpret them as fabrications, fictions, or falsehoods. if they are nonveridical assertions, little is gained for a research programme on non-egoic experiences. thus, i focus on whether they can be veridical – or something close enough. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 9 this interpretation is problematic: how can someone witness something that, essentially, does not allow for the presence of anybody to witness it? if the reporter was present, then it wasn’t a non-egoic episode; if it was a non-egoic episode, there cannot be anyone to report on it. thus, if we interpret sane as reports, they seem self-defeating, goes the claim. either we reject this argument or we reinterpret sane as something other than reports. i do not buy into this criticism in general: some sane some may be meaningful, honest, maybe even truthful. but there is something right: the most prominent understanding of ego-dissolution simply cannot be squared with sane being reports, i argue. but this simply tells us what ego-dissolutions cannot be, not that they cannot exist in general. so, what are the hidden assumptions that lead to sane being self-defeating? 3 sane rejected: the self-defeating nature of firstperson reports of non-egoic episodes sane face a very reasonable doubt: they appear to be paradoxical. why? because these reports express a claim that the speaker herself has had a non-egoic episode, but come with a corresponding de se-belief of the speaker that she herself is or has been in such a state. so by reporting, one apparently contradicts the report: if you don’t exist, then who’s talking about being no one? who’s remembering this? who’s reporting this? who was the witness? the notion that sane are self-undermining has been expressed at least three times. first, by thomas metzinger (2004, p. 566) in his being no one (see also his 2018): autophenomenological reports given by human beings about selfless states […] will usually not impress philosophers much, because they contain an inherent logical fallacy: how can you coherently report about a selfless state of consciousness from your own, autobiographical memory? […] such reports generate a performative self-contradiction, because you deny something that is presupposed by what you are currently doing. if metzinger is right and such claims or self-attributive beliefs that one has had a non-egoic experience are self-contradictory, then sane could never be reports and such self-attributive beliefs could never be knowledge, even if there were non-egoic states. non-egoic states, if they exist, are (so he claims) inaccessible for autobiographical memory and self-attributive de se beliefs. if we have them in autobiographical memory, they are skewed by the processes of memory formation and retrieval (see also metzinger, 2018, p. 13).12 therefore, we must distrust such re12note that on page 13 he writes about “full-absorption episodes”, which may differ from non-egoic experiences. what they share is their principled ineffability: “a full-absorption episode cannot fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 10 ports and self-attributive beliefs. however, if we distrust sane, then we have no intersubjective reason to believe that such non-egoic experiences exist. at the very least, we must be agnostic vis-à-vis phenomenal ego-dissolution, despite the existence of sane.13 rocco gennaro (2008) makes a similar point: so then a real problematic case would be one where there is a claim to have a [pure conscious event], and thus a truly introvertive [i.e. looking inside the mind, sbf] mystical experience, but where there is no conscious i or self present. but it is very unclear that there are such cases. […] [w]e do indeed find reference to an “awareness of self” and the conscious employment of the concept “i”. […] more theoretically, it seems to me that anyone having a truly introvertive experience must be consciously employing the i-concept. for one thing, the practitioner is clearly taking the mental state to be her own as opposed to someone else’s. for another thing, it is difficult to understand how practitioners can later remember and describe these events without having employed conscious i-thoughts during the alleged [pure conscious event], that is, without having experienced the event as one’s own. gennaro stresses the necessity of a de se element in such beliefs: subjects must attribute these experiences to themselves. they did not happen to nobody or to someone else. thus, mechanisms for cognitive self-reference must have been active in such episodes, making them egoic. there must be some “awareness of self”. therefore, these claims do not prove the existence of non-egoic episodes. the problem has been raised, more casually, by charles foster (2016, p. 6) in his being a beast: j.a. baker [author of the book the peregrine] pursued his peregrines to the point of assimilation with them. his express purpose was to annihilate himself [and become a peregrine]. […] as a method, dissolution creates great literary difficulties. if j. a. baker really disappears, who is left to tell the story? and if he doesn’t, why should we take the story seriously? be reported, because the self-referential mechanisms of forming an autobiographical memory are suspended. therefore, only the process of entering into it or of emerging out of it can be faithfully represented in the autobiographical self-model; the episode itself is not a part of the subject’s inner life narrative.” 13that sane are self-defeating is good news for those who hold that consciousness is essentially selfinvolving, like e.g. zahavi & kriegel (2015). such a position appears to be a pretty safe bet if there can never be a datum speaking against it thanks to the self-undermining nature of sane. but it can hardly be so easy. first, it seems to make the self a necessary feature of phenomenality for conceptual reasons. but usually, the idea that consciousness is always self-involving is posited as a phenomenological claim, something we have to discover by first-person methods rather than linguistic analysis. second, sane pop up widely: they are symptoms of certain psychotic episodes, appear to capture aspects of psychedelic trips, and are used to describe the essence of deep meditative states. it seems odd to explain all of this away by attesting to widespread conceptual confusion. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 11 indeed, why should we take such reports seriously? an answer ought to be given before we invest time and effort into offering an explanation for phenomenal ego-dissolution itself. i defend the idea that such reports, while they appear selfcontradictory, are not necessarily problematic at their core. so what, exactly, is the problem? let me start with what is not the problem before presenting the problem’s different flavours. contrary to gennaro, the problem is not raised by the first-person indexical “i” or the first-person possessive “my”. consider sartre’s phrase: (sartre) when i run after a streetcar, when i look at the time, when i am absorbed in contemplating a portrait, there is no i. who was running? sartre was! who wasn’t there? sartre. this aftertaste of paradox remains if we replace “i” with its referent: (sartre∗) when sartre runs after a streetcar, when sartre looks at the time, when sartre is absorbed in contemplating a portrait, there is no sartre. what raises the problem is, apparently, that asserting that one does not exist, if true, precludes the possibility that one can assert it. by saying that one does not exist, one marks oneself out as a liar. or a lunatic, because denying one’s own existence precludes one’s ability to be rational. or a driveler, because by saying that one does not exist, one talks nonsense. apparently, so the argument goes, one cannot truthfully assert a sane because its truth undermines any possibility of asserting it. at the heart of the sane paradox is therefore an infelicity of performance, as diagnosed by karl-otto apel (1976, p. 73).14 you say that you don’t exist, but: look who’s talking! the problem is not a logical one because what is expressed in the utterance can be perfectly true: it is possible that a certain person 𝑎 does not exist. but while everybody else can express this, 𝑎 can not. so the paradox is not in what is said, not in the syntax or even the semantics of what is expressed; it arises because of an infelicity in performing the speech act of assertion. we may follow apel and call it a performative fallacy. the contradiction does not arise simply because of the proposition being expressed, but by expressing it. why is it a fallacy at all, not merely an infelicity? because from what is said (“i do not exist.”) and the fact that this is said, we can derive a contradiction: what is said entails that the referent of the name or first-person pronoun does not exist; but the fact that it is said requires the existence of a speaker; and the speaker is the referent of the first-person pronoun or name. it is perfectly possible that apel does not exist; it is just impossible that apel can assert his own nonexistence truthfully. the contradiction arises not as a result of the proposition 𝑝 itself, but only if the person this proposition 𝑝 is about expresses, assents to, asserts, believes 14apel bases his analysis on wittgenstein’s on certainty, hintikka’s cogito, ergo sum, and stegmüller’s metaphysik, skepsis, wissenschaft. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 12 𝑝. any stance of the speaker towards 𝑝 that requires honesty or veridicality is itself precluded by 𝑝. call this the sane paradox. the sane paradox involves some important presuppositions: (unique) for each individual,15 there is one and only one self that is the referent of self-reference and fulfils the functions of self-reference, or only one mechanism facilitating all kinds of self-reference. (phenself) a self (or any mechanism facilitating self-reference) necessarily comes with a phenomenal feeling of self. uniqe is an ontological commitment to selves as concrete entities and their distribution. phenself is a thesis that connects a feeling of self to something that is felt, namely an underlying self. not all feelings-of require that there be something real they are feelings-of: i can have a feeling of being followed without being followed. but according to phenself, the feeling of self is a necessary correlate of a self, so if the feeling of self is missing, the self is gone with it. together with cease, we can see how the problem arises: if there is one and only one self (uniqe) and if this self necessarily shows itself in experience (phenself), then – if a feeling of self is missing (cease) – any self-reference is impossible. but self-reference seems to be necessary in order for this state to be cognitively accessible as one’s own. this immediately suggests several ways of dealing with sane: either we can accept these assumptions or we can reject them. if we accept them, we give a cognitive or pragmatic diagnosis. a pragmatic diagnosis works on the level of communicative practices: what is said in a sane is not what is meant (but what is meant is meant truthfully). a cognitive diagnosis rejects implicatures as tools for explanation and instead targets the mental state of the speaker: what is said in a sane is what is meant, but what is meant cannot be rationally believed. so it is either not believed at all or it is irrationally believed. i will address these diagnoses in a moment. but we may also reject any or all of the assumptions of the sane paradox. this opens the door to phenomenological diagnoses, where the focus is on phenomenal features being present or absent and why they are present or absent. such phenomenological diagnoses therefore often cross the border between mere phenomenological description and theory about the mechanisms underlying phenomenal consciousness. if we reject uniqe, we suggest that many selves can be present in one body, and even if one of these selves does not show itself in phenomenality, other selves may still be present and do their work. call this the strategy of multiplicity. if we reject phenself, we suggest that a feeling of self can be lost without the mechanism of cognitive self-reference being impeded. call this the strategy of a merely cognitive self. if we reject cease, then people retain their feeling of self, but apparently fail to notice that it is still there. i will introduce a 15if we want to qualify this: for each individual at a moment in time 𝑡. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 13 version of this as the strategy of ego-expansion. we may also reject difference and say that there is no felt difference between non-egoic experiences and our normal way of experiencing. i will consider this strategy as well as the others in 5. there, i will also discuss problems with forming autobiographical memories of non-egoic experiences. but before we go so far into diagnosing radical disruptions of self-consciousness, let me start with the pragmatic and cognitive diagnoses. 4 sane diagnosed: what is said vs. what is meant, believed, rational, true 4.1 a pragmatic diagnosis one conclusion one might draw from the performative fallacy is that no sane has any meaning because sane cannot have truth conditions. so either they are nonsense or they share their meaning with all other contradictions. this is the wrong conclusion to draw. yes, if we accept all five assumptions, what is said in a sane leads to a contradiction, but this does not entail that they are meaningless. remember that the contradiction follows not from what is said, but from saying it. therefore, the proposition expressed by what is said can be the case. the utterance is perfectly intelligible – it simply is necessarily false when i say it. the question is then: why should someone say something that is immediately falsified by saying it? maybe what is said is not what is meant. if people assert that they don’t exist, they violate conversational principles (grice, 1991), namely the maxim of quality: “try to make your contribution one that is true.” if a sane cannot be true, it triggers conversational implicatures: the sane must mean something other than what the speaker says. just as obviously false statements like “i wasn’t myself that night”, “there is no time like the present!”, or “we’re busy doing nothing” can be uttered in a meaningful way, so can sane. what might a speaker mean with a sane? they may intend to convey things like “i don’t feel well” or “nobody is noticing me”, or express desires like “i wish i weren’t here”, or boast that “i got totally lost in what i was doing, paying no attention to myself”, or “look, i have acquired enlightenment, admire me!”, and so on. none of this requires any form of actual ego-dissolution. instead of being a truthful report, what appears to be a report about a non-egoic state performs a different conversational function. it therefore does not require anybody to make the ontological commitment that non-egoic experiences exist. in fact, the hearer can decode the message only because such reports are openly self-contradictory, similar to a case where a speaker excuses her previous erratic behaviour by saying “i wasn’t myself”. some variants of sane might then be a complex way to communicate one’s unease with oneself. this is a good fit for cases of apparent ego-dissolution in depression. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 14 however, this pragmatic diagnosis has two drawbacks. first, it does not explain the details with which such reports are usually embellished. reconsider saks talking about herself dissolving like “a sand castle with all the sand sliding away in the receding surf”. these metaphors fulfil no further pragmatic function. rather, they fit our understanding of sane as detailed reports. second, this diagnosis hardly explains cases of written and kept-hidden reports of non-egoic states, e.g. in diaries. given that these writings exist and may be revealed after an author’s death, we should be open to the possibility that some of them are intended as earnest reports, not as pragmatic play. if we can reasonably rule out the intention of triggering drastic pragmatic implicatures and deviations of what is meant from what is said, we need another diagnosis for what underlies the expression of a sane for cases that do not lend themselves to pragmatic explanations. this leads us to cognitive diagnoses. 4.2 cognitive diagnoses if all the signs suggest that a speaker actually means what she says in a sane, despite the apparently self-defeating nature of sane, then we have to ask what brought the speaker to make such an assertion. the possibilities include: (a) the speaker is confused and does not actually believe what she says; (b) if the speaker shows signs of believing what she says, she must be irrational; or (c) if the speaker shows signs of believing what she says and signs of being rational, she intends to deceive us. 4.2.1 unbelievable? usually, we might think that a sane cannot be believed because believing means taking something to be at least possibly true. but it cannot be true that one believes any 𝑝 when one does not exist. therefore, sane cannot be believed. but, as roy sorensen (2004, pp. 70–71) argues, we are sometimes led to believe necessarily false statements like contradictions, just as badly designed calculators used to. the old calculators simply rounded off after the last number they were able to display: if you divide 1 by 3 and multiply by 3, you get 0.99999, but not 1. one might defend this by ascribing to the calculator the clever “belief” that 1=0.99999… but this ascription fails if we magnify the miscalculation. a calculator should believe that 1 ÷ 3 × 9, 999, 999 = 1 × 9, 999, 999 ÷ 3 thanks simply to the commutative properties of multiplication.16 but if we do the calculation on these old machines, we get 3,333,333 for the right-hand side of the equation, and 3,333,332.97 for the left. obviously, 3,333,332.97 ≠ 3,333,333. we 16most old calculators do not have brackets, but perform calculations in the sequence we type. the reading with brackets is (1 ÷ 3) × 9, 999, 999 = 1 × (9, 999, 999 ÷ 3). fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 15 also know that we may have contradictory preferences. obviously, preferences should be transitive: if we like 𝑎 more than 𝑏, and 𝑏 more than 𝑐, we should like 𝑎 more than 𝑐. but, as amos tversky (1969) showed, we apparently do not work like this. instead, we sometimes prefer 𝑐 to 𝑎. under some circumstances, our inbuilt tendencies to accept some contradictions become apparent. but unless this is made obvious to us, we may still assent to self-contradictory statements with conviction. believing contradictions is not unusual, either for us or for machines. so just because sane are contradictory, a person need not be confused about what she said in a sane. she can believe what is said. we simply have to ascribe to her a false belief. falsity due to self-contradiction does not add a lot to this because we already know that delusions can be self-contradictory (bortolotti, 2009). so nothing speaks against explaining away non-egoic states by declaring “reporters” of sane to be delusional. and because the falsity of these beliefs is apparently so obvious, we might as well question their rationality. but do we have to? 4.2.2 irrational? if a person offers a sane, she apparently disqualifies herself from being rational – only the insane come up with sane. why? in order to rationally believe a proposition, one needs to have adequate reasons to believe it. while one can have a practical reason to believe something self-contradictory (e.g. someone may pay me a large sum if i do), one cannot have a theoretical reason to believe a contradiction, because logical consistency trumps other reasons. therefore one cannot rationally believe something self-contradictory. this suggests that our reporters of sane must be irrational or delusional. they have gained false beliefs about their own past or present and do not let go of them despite strong and obvious inconsistencies. but someone need not be irrational just because they believe in contradictions. roy sorensen (2004, pp. 146–147) even argues that reason demands belief in infinitely many contradictions. consider the very reasonable meta-belief that one of my beliefs is false. this precludes my belief system from being coherent and, taken as a whole, fully true. it is therefore rational to believe that one believes a contradiction, if one believes that one believes a contradiction or something false. this merely shows that one can rationally believe in contradictions, but it does not identify which of one’s beliefs is inconsistent. yet “detected contradictions are instantly abandoned” (sorensen, 2004, p. 155). however, this is not what happens in the case of a sane: meditators likely remain steadfast that they have had non-egoic experiences even if they are aware of the sane paradox; saks still reports on her episodes of ego-dissolution, and tries different literary devices to sidestep the paradox. neither she nor the meditators abandon their beliefs in their past non-egoic experiences. we may take this as strong evidence of irrationality. but it raises more issues to explain. we need to give an explanation for the stability of such beliefs despite their obvious self-contradictory nature. what is also needed is an explanation fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 16 for their ubiquity in a broad range of different circumstances, from meditation to psychedelics to psychopathologies. irrationality will not likely be a one-size-fitsall solution here. should we really think that meditators try to achieve irrationality as a goal? this is easily achieved without countless hours of sitting still. i am suspicious of the idea that all sane are best interpreted as irrational delusion. delusions are often highly idiosyncratic: i might believe that i am napoleon, while you may believe that you are jesus. even when there is some similarity (you and me both believe we are famous dead people), our delusions differ massively in their details. in contrast, sane appear to be very homogenous across individuals. what is therefore also needed is an explanation of the specificity of this form of irrationality. we need some explanation: why do specifically de se beliefs about non-egoic episodes arise, and not other self-contradictory beliefs like “i am saying nothing”? why are they so symptomatic of specific pathologies, but not of all pathologies? while some sane might be due to delusion or irrationality, it is unlikely that all are. but if sane cannot be true and the speaker is neither delusional nor irrational, then are they merely acting as if they believed in non-egoic episodes? 4.2.3 untrue one diagnosis of sane is that they are not made in honesty. if we believe this, we may feel pressured to call a sane “reporter” a liar or bullshitter: what look like sane are false assertions. moreover, they are not contingently false, but necessarily so, whenever they are stated. why should people assert something necessarily false if they do not want to trigger the corresponding implicature? one explanation is that such speakers are trying to deceive. given that having non-egoic experiences is sometimes seen as a mark of high levels of spiritual enlightenment, which come with high social status in some religions, there are high incentives to lie about having non-egoic experiences in order to gain such status. however, the lie is extremely blatant, for it is logically impossible to make such an assertion in earnest. this makes it a bad candidate for a lie. a good lie expresses something that can at least possibly be the case. otherwise, nobody buys into the deception. it simply is too obvious a lie if one claims that 2 + 2 = 5. yet some impossibilities are harder to track. depending on context, even an impossibility might be bought as being the truth by a specific audience. for example, one might try to impress a stranger by claiming to be that famous mathematician who discovered the largest prime number; or one might claim that there is an all-powerful, all-knowing, and all-loving being that allows for suffering in the world but cannot create a stone too heavy for it to lift. some sane might be similar deceptions. however, seeing such “reports” as deceptions for the reporter’s own benefit fails to explain why such reports regularly appear in certain psychopathologies. social status is largely reduced here. thus, there is some incentive to hide one’s mental illness, not advertise it by spewing sane. before i get into this, let me emphasise: it would be foolish to think that all fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 17 sane are to be judged equally. some may be lies to achieve higher social status, but not all; some may be delusions, but not all; some may stem from irrationality, but not all; some may be nonsense, but not all; some may be made in order to express by implicature that one does not feel at ease with oneself, but not all. we have to look at the specific context and speaker to decide. if we accept the sane paradox raised by metzinger, gennaro, and foster, our research programme on non-egoic experiences reduces to a study of pragmatics, delusions, lies, and doxastic failures. but it is not a study of phenomenal experiences. in order to give such a phenomenological diagnosis, we will need to reject some of the presuppositions underlying the paradox. if we can, then these utterances may express something that can be the case. so, can sane actually be accurate descriptions of certain phenomenal experiences? 5 sane revived: four ways to lose one’s self 5.1 rejecting the sane paradox the pragmatic and the cognitive diagnoses are all available to explain why a sane may be given by a person. each of these diagnoses allows us to accept the presuppositions of the sane paradox and remain agnostic as to whether phenomenal ego-dissolutions exist at all. even though we should accept these diagnoses as parts of a toolbox, they do not lend themselves to building a research programme focused on the phenomenality of ego-dissolution. for this, we have to reject the paradox. let me re-address the presuppositions behind the sane paradox. non-egoic experiences are supposedly marked by the following features: (cease) a feeling of self is missing in some experiences. (access) the absence of a feeling of self is cognitively accessible. (difference) there is a felt difference in the feeling of self between the time before/after and during non-egoic episodes. two additional assumptions give rise to the sane paradox. (unique) for each individual, there is one and only one self that is the referent of self-reference and fulfils the functions of self-reference, or only one mechanism facilitating all kinds of self-reference. (phenself) a self (or any mechanism facilitating self-reference) necessarily comes with a phenomenal feeling of self. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 18 only together do they preclude the possibility of reporting on non-egoic experiences. but we might reject uniqe, cease, or phenself and thereby preempt the sane paradox. if not the feeling of self, but merely a feeling of self dissolves, then another mechanism facilitating self-reference might bring about relevant de se beliefs. if only one feeling of self exists but it doesn’t dissolve, then de se beliefs can occur. if self-referencing mechanisms need not show themselves in experience, then the relevant de se beliefs can occur. but before i go into these non-paradoxical accounts, is there maybe a way to account for all the features of sane while accepting the assumptions leading to the paradox? someone may defend the idea that one can provide sane from episodes of full ego-dissolution by relying on specific forms of memory. i do not think that this works well. let me explain why. 5.2 can we defend total ego-dissolution against the sane paradox? total ego-dissolution would be the case if uniqe, cease, and phenself were all accepted. so there is only one singular self or mechanism facilitating self-reference and self-consciousness which necessarily manifests itself in a specific feeling of self and which is inactive or nonexistent during such episodes. if this were the case, then we should not get any sane when a feeling of self ceases. in this framework, if we ascribe a non-egoic experience to ourselves, then there must have been a self or a mechanism facilitating self-reference in order to bring about this de se ascription. some may accept these theses, but still find the sane paradox unconvincing, specifically when it comes to memory. it might be easy to see why one cannot report on oneself as currently undergoing an episode of total ego-dissolution. one would need to speak of oneself in the third-person singular or use a definite description that happened to refer to oneself (e.g. “the speaker of this sentence”). for the type of de se reference necessary for a self -ascription, however, we would need some way of modelling ourselves in order to refer to ourselves as ourselves – and not merely to ourselves by accident. but one might think this: even though one could not report on such experiences while one is undergoing them, one might remember them and report on them afterwards as having happened to oneself. metzinger (2004) apparently rejects this possibility, because “how can you coherently report about a selfless state of consciousness from your own, autobiographical memory?” certainly, autobiographical memory relies on self-reference across time, which relies on a model of the progression of a self. but why shouldn’t some other form of memory suffice? who else could this experience have happened to, when all the experiences that i can remember are necessarily my own? could i not validly infer later on that this experience was my own? if so, we need not reject any of the premises. instead the sane paradox is restricted: only present-tense sane fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 19 are impossible, but we may still report veridically about non-egoic experiences, if only retrospectively. this might seem like a reasonable reply, but only because it relies on a misunderstanding: semantic memory is not the same as episodic and autobiographic memory. that is, some memories of facts will not be de se memories (conway, 2005; conway & pleydell-pearce, 2000; klein & loftus, 2014; klein & nichols, 2012). i remember a lot of things that are facts, but not about me, e.g. when the second world war ended, that paris is the capital of france, that joaquin phoenix had a brother called river. even if a fact concerns ourselves, a memory of that fact is not automatically a de se memory (conway, 2005; klein & loftus, 2014). for example, you might remember that someone knocked over the lamp during the party last night, but you might fail to remember that it was you who knocked it over.17 as david lewis (1979) argues: no amount of knowledge of facts, even up to divine omniscience, can replace knowledge de se. and it is only knowledge de se that allows us to locate ourselves in the world. god would fail to know where and who she was if she had only factual omniscience. if one wants to defend the possibility of veridical reports from total ego-dissolution, one has to show how one can form de se memories of one’s own episode of total ego-dissolution. how could that be? two stages are the most promising candidates during which memories may be imbued with a de se aspect: the formation and the retrieval (or recall) of the memory.18 in cases of total ego-dissolution, the de se aspect of memory cannot arise during formation, because at that point a self (or a mechanism facilitating selfreference) and therefore the basis for self-reference is missing. this is metzinger’s, gennaro’s, and foster’s point. therefore, the de se aspect must be inserted during reconstruction, injected during retrieval, added on during recall. unfortunately, there is no specific noticeable trace of what exactly is “added on”: a false memory of me knocking over a lamp feels just as real as a veridical memory of me knocking over a lamp. sometimes, it does happen that we “add ourselves” to memories of affairs that did not involve us (hyman jr, husband, & billings, 1995; loftus & pickrell, 1995). someone might reply that these are therefore not memories at all, but pseudo-memories that are simply phenomenologically indistinguishable from real ones; what makes something a memory is not how something feels, but whether what is remembered happened as it was remembered. yes, there can be false memories with a de se component added later; but there can also be true memories with a de se component added later. this is obvi17quite a few thrillers involving amnesia rely on this distinction: one remembers that something happened, but one forgets that it happened to oneself. for example, shattered (1991), directed by wolfgang petersen with tom berenger, bob hoskins, and greta scacchi, distributed by metrogoldwyn mayer, uses this plot. 18memory storage does not seem like the right kind of stage for a de se aspect to creep in, mainly because adding the de se aspect seems to involve specific mechanisms, namely those facilitating self-reference. storage, usually understood as a passive, even if memory-altering, process, is usually too unspecific to introduce this specific aspect. still, most of what i say about formation and retrieval would apply to de se introduction during storage as well. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 20 ously right. as an example: i cannot remember things from my early childhood. most of what i know from that time, i know from pictures and hearsay. say i see myself on one of these photos in the arm of my uncle karl smoking a pipe. i know that it happened to me, but i know this not by remembering the experience itself. however, after a while, i might forget how i formed this knowledge that my uncle karl held me while smoking a pipe. i might form a de se memory about this fact that i take to be one formed in my early childhood. earlier, it was a fact about myself that i knew, stored in semantic memory; now i remember it as a de se fact with a phenomenological de se component that was added on later – and thus it has been turned into an episodic memory. the problem is not that there cannot be true or accurate memories of facts with a de se component added on later. the problem arises from the “adding on” itself. for if this “adding ourselves in” is part of the reconstruction, then such memories fail to show that there actually was an episode of ego-dissolution – for it is indistinguishable from the outside or the inside whether the de se aspect was added on later or accurately remembered as being present during the experience. if the de se aspect was previously present, then the experience was egoic. if one agrees that it was a later mis-reconstruction, addition, or embellishment, then one should be doubtful that the rest of the experience was remembered accurately. we cannot be sure that we are learning something about the deep structure of nonegoic experiences if the reports are edited. but one cannot distinguish between these cases. but what if the de se addition is justified by another memory system? what if one does not remember a proposition about someone and “add on” the de se component; what if instead one remembers imagistically?19 that is, one remembers the sensations felt during that episode as sensations, as one conjures up the scene of what it was like to sit in one’s mother’s kitchen as a toddler. in this remembered stream of sensations, a felt self could be missing. but because sensations always have to be mine (i don’t feel anybody else’s), i can reasonably infer that these sensory impressions with a lack of felt self must have been mine. what is added is a possessive de se component, not a phenomenal one: the experiences were mine, but i was not in these experiences. but again: what ensures that the de se component was not there in the remembered experience? 19what i am talking about, in effect, could be interpreted as iconic memory. but i am concerned about the distinction between visible and informational persistence in iconic memory (coltheart, 1980). visible persistence is supposedly very short-lived (less than 300 ms) and therefore is not suited to playing the role attributed to “imagistic memory” here. informational persistence is longer-lived, but preserves only the information in the visual stimulus. it is about something visual, but is not visual itself. it is therefore not like remembering in the form of a mental image, which is specifically what would be needed for this type of defence; otherwise, the objections against semantic memory of such episodes can be applied. it seems that there is no obvious candidate in psychology that is able to play the role of “imagistic memory”. but one would need such a form of non-semantic memory to get this defence off the ground. therefore, i discuss the issue despite a lack of match to empirical psychology. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 21 what if someone holds that a self is a necessary requirement for experiences in general? any form of consciousness, we are often told, involves a first-person perspective. and what is a first person other than a self? maybe it is the perspectival nature of experience which must point at someone, but not at anyone: at a self as the owner of the experience, as that whose experience it is, as the centre of the perspective. the feeling of self is not something extra, but is somehow ingrained in the structural composition of phenomenal sensations because of their perspectival nature. maybe this justifies the inference that those were my experiences if i become aware of the perspectival nature of my imagistically remembered sensations. i agree that our sensations, especially our auditory and visual experiences, are perspectival. but i do not think that this forces us to accept a self outside experience, a mechanism facilitating self-reference, or even a feeling of self as part of our experiences. our experiences are perspectival and thereby indicate a centre in experience. but the perspectival nature of experiences does not necessarily indicate oneself as the centre; it just indicates someone. consider, as an analogy, being john malkovich.20 in this movie, people can enter the mind of john malkovich through a door in the 7 12 th floor of a manhattan building. when a person enters through that door, we see the succeeding scene unfold from john malkovich’s firstperson perspective. obviously, the visual imagery suggests that these experiences are someone’s, but they need not be mine. neither set-up nor framing specify whether these visual experiences are those of the person entering the mind of john malkovich or of john malkovich himself. and when seeing the scene as an observer, i need not feel like i am john malkovich. the perspectival nature of the visual sensations is preserved, but they do not necessarily therefore feel like my own perspective, when i see the scene. the same goes for the people undergoing the experience in the movie: they do not confuse themselves with john malkovich, but simply partake in his stream of consciousness – they experience this stream as john malkovich’s. therefore, conceptually, the perspectival nature of sensory imagery is not sufficient to indicate that these sensations are one’s own, that the vanishing point of this perspective coincides with my self. the difference between experiencing a perspective and experiencing a perspective as being my own is then not one of sensory impressions at all. selfhood is not a sensory quale and not part of the imagistic array i remember or experience. this does not mean that there is no feeling of self at all, but simply that there is no sensation in any sensory modality that is that feeling of self. if the perspectivity of our sensory experiences does not suffice for a feeling of self inside experiences or necessarily indicate a self outside our experiences, then attributing a remembered sensory experience to myself is something problematic for allegedly non-egoic episodes. in order to remember this stream of consciousness as being my own, a de se component is again “added on” to the perspective in recall. and again, we – whether as memorisers or as external observers – cannot 20universal 1999, directed by spike jonze, written by charlie kaufman. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 22 determine whether the de se component was present in the first place or added on later. so if there are non-egoic experiences, then we cannot have first-person reports of them, for this requires reference to oneself as oneself. if there are such experiences, then we cannot have autobiographical memories of them for the same reason. if they exist, then no purported autobiographical memory would prove their existence, for these would be indistinguishable from mis-reconstructions or mis-categorisations of egoic episodes as non-egoic ones. while this does not prove the nonexistence of episodes of total ego-dissolution, there is no way to demonstrate their existence by providing “reports”. this puts them in the same category as unicorns, whose existence one also cannot prove by pointing to “reports” and drawings of them. we should therefore consider alternatives to total ego-dissolution where we do not need to posit something being “added on” to a memory. then, it might be less controversial how we can see sane as indicators, in some sense, of non-egoic episodes. 5.3 multiple feelings of self: plurality or modularity let us accept cease and phenself but reject uniqe: a feeling of self can temporarily cease to exist but there could be many differentiable phenomenal “selves” in our experiences. these might include a feeling of ourselves as a body, another as a thinker of thoughts, another as a person in social relations, and so on – each likely associated with a dissociable cognitive mechanism that facilitates this form of self-consciousness. obviously, these distinct feelings of selves in each of us may influence one another, but they could still be dissociable. if they are dissociable, one feeling of self can cease to exist while others persist. then, one can reasonably say that one of these – a, but not the, phenomenal self – dissolves while a system retains other mechanisms sufficient for de se beliefs or cognitive self-reference. even though he does not endorse it, this position may be developed from work by raphaël millière (2017), who, in order to give an account of what happens in egodissolution, helpfully distinguishes between different forms of self-consciousness, e.g. the disruption of self-referential thoughts, of narratives about oneself, of body ownership, of bodily self-awareness, and of bodily self-location (and extension).21 we can construe the thesis of “multiple selves” (i.e. multiple feelings of self at one time in one system) in two ways: as modularity or plurality. in the case of modularity, there is no single homogeneous mechanism bringing about a feeling of self; different modules specialise in different tasks, among them those mentioned by millière (2017). each module can bring about a kind of self-reference and selfrelatedness independently from others, and each of these forms may, if we accept phenself, manifest itself in phenomenal consciousness in a distinct way. but just as the knife, saw, and corkscrew still contribute to one swiss army knife, each 21see also gennaro (2020, this issue), as well as millière (2020, this issue). fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 23 self-module contributes to a larger, integrated, and unique feeling of self, which allows for mediation between these modules and a coherence between these feelings as indicating one and the same self. a disturbance in the modules affects the character of a feeling of self, but there is still one feeling of self. because each self-module can fail independently and because such disturbances are noticeable phenomenologically, there can be reports of such partially non-egoic experiences – but only if some of the mechanisms necessary for self-reference are intact, for they enable the formation of first-person reports (e.g. the modules for cognitive or linguistic self-reference) or autobiographical memories. in the case of plurality, these modules are not part of a larger, integrated feeling of self. each contributes to forms of self-reference, but each produces its own distinct feeling of self – without any overarching unique feeling of self that they are part of. some of these can be disturbed while others remain unaffected and therefore enable first-person reports and autobiographical memories of (partially) non-egoic experiences. simply, there is no unique feeling of self, but multiple feelings of self, each associated with its own distinct mechanism facilitating some form of self-reference. in both cases, the expression “i” in first-person reports will be ambiguous: it can refer to the bodily self, the social self, the cognitive self, or the reflexive self. if so, then the performative self-contradiction could be prevented by disambiguating. consider sartre’s statement again: (sartre) when i run after a streetcar, when i look at the time, when i am absorbed in contemplating a portrait, there is no i. in this example, the first two instances of “i” obviously refer to a physical body that runs and looks, the third “i” refers to a cognitive self, while the fourth “i” refers to the reflexive self – some being present in experience, some not. if we disambiguate in this way, the paradox ceases to arise: (sartre∗∗) when body-i run after a streetcar, when body-i look at the time, when cognitive-i am absorbed in contemplating a portrait, there is no reflexive-i. this partial breakdown in the chorus of the selves could be what is expressed in sane. non-egoic experiences then need not be total ego-dissolutions; they are only partial disruptions of feelings of self and their associated self-reference-enabling modules. but i see some drawbacks. first, this explanation does not seem to fit saks’s report, where she claims that the centre (singular) does not hold. the definite article “the” here suggests a uniqueness: one and only one centre. apparently, there is one coherent feeling of self for her, and it has gotten lost. nor did only some of the feelings of self (as in the case of plurality) get lost; nor did the feeling of self change its character (as in the case of modularity). if the uniqueness expressed fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 24 by saks is not just loose talk, we should search for alternative explanations that do not rely on a multiplicity of selves or feelings of self. the second drawback i see is: the account needs to explain how different feelings of self interact in order to give rise to the appearance of one integrated self, a feeling of self as inhabiting this body rather than that, and as the thinker of thoughts, and also as the experiencer of sensations, plus as the person addressed by others, and so on. if such an explanation is not provided, this approach remains unconvincing because this high degree of integration is the most prominent features of selves: their individuality – their lack of partition. a feeling of self should share this undividedness in order to be a feeling of one self as an individual. if such an explanation for overarching integration of feelings of self into one is provided, the multiplicity thesis loses some of its drive. because now, there is apparently one overarching process that forms a coherent feeling of self out of several. the multiplicity thesis also becomes ambiguous: it is unclear what is disturbed in non-egoic states – is it the integrating process or the parts being integrated? a third drawback i see is that because this explanation relies on multitudes, either of modules contributing to a larger feeling of self or of unintegrated feelings of self, it is not very parsimonious. with occam’s razor, we should prefer an explanation that covers the same territory with fewer entities postulated. we can also consider an evolutionary take: the amount of sugar one’s body burns is a constraint on one’s evolutionary success. finding the right proportion, the sweet spot, between energy conservation on the one hand and access to resources (like food, status, or mates) on the other hand is an evolutionary imperative. so we have to show what an organism gains something by spending energy. if each of the multiple mechanisms for self-reference bringing about a feeling of self demands energy to be active, then the multiple-selves account looks like a very costly model in comparison to one that claims only one unique mechanism facilitating all feelings of self. an explanation without such pluralism might then be preferable, if not for theoretical reasons then for reasons of energy conservation.22 both of these reasons for rejecting the multiplicity account are comparative: of two or more accounts, they tell us which to prefer. so we should look at further alternatives. if multiplicity is the only game in town, it wins by default. but wouldn’t it be cool if there could be more to ego-dissolution than just the disturbance of some-but-not-all feelings of self? 22there is also an evolutionary reason speaking against a unique mechanism facilitating selfreference: robustness. one mechanism, if disturbed, would lead to a global break down of selfreference. given that locating oneself, discriminating one’s body from others, and other forms of self-reference have an evolutionary advantage, any disturbance of a unique mechanism facilitating self-reference would come with high costs to the organism. in comparison, if there are many independent mechanisms, a break down in one could leave a sufficient level of functionality for continued survival. the question is how to weigh robustness against energy conservation. for this, we would need to compare theories. what matters here is that there are comparative constraints on theory selection (e.g. occam’s razor), and that they suggest looking for alternatives to compare to. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 25 5.4 ego-expansion let us accept uniqe and phenself but reject cease: there is only one feeling of self but it does not dissolve. it is the dissolving of a unique mechanism for selfreference or a unique self-model, indicated by the loss of a feeling of self, that leads to the sane paradox, so rejecting cease disarms the paradox. so what happens in such states if it is not a feeling of self dissolving? here is one option:23 rather than contracting into nothingness, the feeling of self expands until everything in consciousness becomes part of it or attached to it. instead of being nothing, in these states the feeling of self is associated and involved with everything. in this way, our engine for self-reference is still active – indeed, it is hyperactive. if the feeling of self is still there and an indicator of the activity of a mechanism facilitating self-reference, then we can get first-person reports and self-ascriptions of experiences of ego-expansion. but why would they be about ego-dissolution if the feeling of self is not dissolving? the problem might arise from an ambiguity in how we conceptualise such experiences: in both the case of ego-dissolution and ego-expansion, the self/other distinction loses its unique meaning. why? an empirical concept is uniquely meaningful only if there are some things that fall under its extension and some that do not. if everything falls under the extension of “feeling of self”, it is not the feeling of self that dissolves, but the boundary between a feeling of self and the feeling of anything else in consciousness. it is the concept of “self” that becomes meaningless, because it lacks any boundary and allows for no distinctions in one’s mind during full ego-expansion. the categories self or belongs to me become meaningless if everything is experienced as belonging to me. in a common but naïve stance, where one projects what is in consciousness onto the world, one is everything and one with the universe. one’s concept of “self” would be extensionally indistinguishable from the concept of things being identical to themselves. the same holds in the case of ego-dissolution: the category of other or not belonging to oneself becomes meaningless because it encompasses everything in one’s mind. so the self/other distinction loses its meaning under both circumstances. “self” needs “other” to make sense because the two are mutually exclusive opposites. in cases where everything falls under self or everything falls under other, it is just as meaningful to claim “the self ceased to exist” as it is to claim “the other ceased to exist”. in both total ego-dissolution as well as full ego-expansion, the term self becomes empty just as the term other becomes empty. thus, first-person reports of “egodissolution” might be adequate descriptions of a conceptual dichotomy dissolving. so in cases where there is no empirical distinction between the concepts “self” and “other” available to oneself, it might make just as much sense to say that everything is other (report of ego-dissolution) or to say that everything is me (report of ego-expansion). but which is the experience underlying this report? egodissolution or ego-expansion? total ego-dissolution is disqualified because it does 23i am grateful to franz x. vollenweider for suggesting this diagnosis. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 26 not easily square with sane. (or, at least, it makes assumptions one need not share.) therefore, if sane are indicators of a feeling of self-alteration, and if there is only one feeling of self, we are left with ego-expansion as a suitable alternative. in the extreme case of ego-expansion, everything going on in consciousness would become part of the self-model, such that nothing could be meaningfully referred to as “other”. only unconscious processes would be exempt from such a hypertrophying self. we would feel as involved in pushing the clouds across the sky as we would in moving our legs. reports collected by berkovich-ohana et al. (2013, pp. 5–6) reflect this tendency for ego-expansion when meditators speak of the bodily space being larger, bodily sensations being wider, a “sense of expansion”, a “center of space [becoming] endless”, that there were “little bodily boundaries compared to the usual feeling”, and so on. data by preller et al. (2018) are suggestive of an explanation: they found that “lsd induces hyper-connectivity predominately in sensory and somatomotor areas” (preller et al., 2018, p. 3). together with an increased hyperconnectivity in the sensory thalamus, a classic neural hotspot for consciousness and its unified nature, this can be interpreted as: more and more active neural correlates of sensory conscious states are incorporated into and associated with one’s own bodily self.24 this oceanic explanation of non-egoic states clearly marks such episodes as egoic: a feeling of self would be present! and it explains the misconceptualisation that happens in sane as benign. it might also explain why certain people who claim to have achieved self-annihilation do not simply stop speaking but instead use a plural form. consider foster (2016) talking about baker, the author who wanted to capture what it is like to dissolve and become one with a flock of peregrines: “if baker is to be believed, it worked […] the pronouns changed from ‘i’ to ‘we’.”25 this change to the first-person plural may indicate that a felt self has expanded and now encompasses more than what the speaker usually associates with himself. ego-expansion might thus be a suitable alternative to the multiplicity account and to total ego-dissolution. an obvious problem for the ego-expansion account is that it does not straightforwardly explain differences in describing these experiences. if we believe that the feeling of self is unitary, then only two options remain for attaining experiences without a self/other distinction, namely total ego-dissolution and ego-expansion. total ego-dissolution, if we buy into the sane paradox, is undetectable or ineffable. but still, we have to explain why subjects differ in their 24note that “ego-inflation” in the edi is associated with “unusually elevated self-assuredness and confidence” (nour et al., 2016, p. 3), reflected in claims like “i felt more important or special than others” or “i felt especially sure-of-myself”. “ego-inflation” therefore captures something different than “ego-expansion”. “ego-expansion” is a claim about an increase in experiences associated with oneself, but this need not come with self-assuredness or confidence. thus, phenomenal egoexpansion (in the sense i in which use the term) would not be captured by “ego-inflation” in the edi questionnaire. 25in the peregrine, baker does shift from using “i” as a way of referring to the narrator to using “we”. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 27 replies to questionnaires or how they describe their experiences: some answers suggest an “anxious ego-dissolution”, some an “ego-inflation” or an “oceanic boundlessness”. in this account, we have to explain all in terms of ego-expansion. how might this work? i am sympathetic to ego-expansion, so i propose the following empirical hypothesis: subjects conceptualise their experience of ego-expansion as “no self” or “all self” depending on where they are focusing when the expansion happens. consider this: our attention can be on our thoughts, our body, our breathing, our digestion, or other aspects of our body and mind. but it can also attach itself to external events, like the sky, water flowing, sounds of neighbours arguing, and so on. it can be on self-related thoughts (i still have to feed the cat) or thoughts where the content is unrelated to us (there is always a prime number between 𝑛 and 2𝑛). we know that under the influence of lsd, our attention is altered. subjects who have taken lsd often focus with extreme intensity on small details, often at the cost of ignoring everything else. say that while the feeling of self is slowly expanding, one focuses not on one’s own hand or breathing but on a feature of the environment (e.g. the veins of a leaf) or a mathematical truth. when attention is shifted again, as the person tries to get back to herself, she realises that she does not find herself where she expected herself to be; she has lost the familiar feeling of her usual, expected self-boundaries because they slowly expanded while she was distracted. the mismatch of predicted self-boundaries in experience and the lack of felt self-boundaries is then interpreted as ego-dissolution – because if it is not where it always is, it must be gone! compare this to a case where people who have taken lsd focus on their own body or on their body in relation to the environment. the process of a feeling of self expanding is felt as a process: we can attend to and appreciate the dynamics and characteristics of its unfolding. as the endpoint of maximal ego-expansion is reached, we (not i) know how we got there. the mismatch of prior felt self-boundaries and the lack of felt selfboundaries is interpreted as an expansion of self because we witnessed our feeling of self expanding. what i suggest is that whether we categorise such an episode as ego-dissolution or oceanic self-boundlessness depends solely on how aware we were of the process of expansion. this suggests that people who are more introspective or sensitive, focusing on their body and mind, or more trained in keeping their attention partially on themselves might experience more episodes of oceanic boundlessness than people who lack such training or are more externally focused or do more rumination (i.e. thinking non-self-related thoughts and getting lost in them).26 coming back to the sane paradox: we can expect first-person reports of such 26i suspect that experienced meditators should feel more ego-expansion rather than selfannihilation due to their mental training in attending to their attentional focus. unfortunately, buddhist literature appears to talk more about ego-dissolution. this might be a cultural artefact attributable to the status of specific writings and ways of describing one’s experience in this spiritual practice. or it may also speak against my account. here, unfortunately, i lack the space to clear up this apparent mismatch. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 28 episodes of ego-expansion. and we can understand why some of them might rather be phrased in terms of ego-dissolution. so this interpretation can explain the specific structure and existence of sane if ego-expansion underlies them. ego-expansion is comparatively more parsimonious than the multiplicity account. however, there is an even more parsimonious solution available. 5.5 no-ego revelation for total ego-dissolution, we accepted both that there is only one feeling of self (uniqe) and that it dissolves (cease), which got us the sane paradox on the condition that mechanisms for self-reference must show themselves in phenomenal consciousness (phenself). for the multiple-selves hypothesis, we rejected uniqe and accepted cease, which led to a non-parsimonious but acceptable diagnosis. for ego-expansion, we accepted uniqe and rejected cease, leading to a more parsimonious explanation in comparison to the multiplicity account. logically, we could reject both uniqe and cease. however, there are two ways in which we can reject cease: first, there are many feelings of self and everything stays the same; or, second, there is no feeling of self to begin with, so no feeling of self can dissolve. this second interpretation denies, to some degree, phenself: a mechanism facilitating self-reference does not need to show itself in phenomenal consciousness at all. this, the no-ego account, is the most parsimonious attempt at a diagnosis. so, what if there never was a phenomenal self? here, we deny the basic presupposition of phenomenal ego-dissolution, because something has to exist before it can cease to exist. but i argued in the beginning, against some interpretations of hume and sartre, that this account fails to capture the difference between being inside and outside a non-egoic episode. so, how can we reply to this challenge? at its core, the no-ego account explains this apparently felt difference as a difference in beliefs about one’s experience. certain beliefs, e.g. about the proper grammatical construction of sentences, may coerce us into believing that there is a self in consciousness. because how else could we make sense of statements like this: (dream) i dreamt i was napoleon. the second instance of “i” in dream appears to pick something out in the dream, i.e. it is about what i dreamt of, not who dreamt it. this suggests that there is a feeling of self in experiences like dreaming. but this leads to paradoxes similar to the sane paradox for cases like “i dreamt i was dead and gone.” or “i sometimes imagine myself as never having been born.” how do we explain these away? i suggest: grammar bewitches us. as an analogy, think of mark twain complaining about “the awful german language” (1880). in german, articles indicate a gender: masculine (der), feminine (die), or neuter (das). according to twain, these grammatical genders do not make fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 29 any sense: “in german, a young lady [das mädchen] has no sex, while a turnip [die rübe] has. think what overwrought reverence that shows for the turnip, and what callous disrespect for the girl.” we can imagine someone so confused by this grammatical artefact that they really think that any young lady is genderless and acquires a gender only after puberty, when she turns into a woman, or that turnips are actually all female. this odd fellow has been misled by grammar. just like our odd fellow who superimposed grammatical gender onto external reality, it might be that we naturally force our conceptions about proper grammar onto experiences themselves. it is not that we find a self in experience; rather, a self is introduced in the way our language forces us to report on experiences. but, as i said before, such a theory fails to explain a reported change: things appear to be different before and after in comparison to during the non-egoic episode. if there never was a self in experience, what is it that happens in such episodes of falsely labelled “ego-dissolution”? i suggest an ironic solution: the fundamental irony of spiritual enlightenment is that it is a basic and general fact that phenomenal experiences are always selfless but we hardly ever notice this. it takes special circumstances for one to cognitively register that one never felt like anybody. nothing changes in phenomenality if we attain these states. we just recognise that something has always been lacking: a feeling of self. this realisation that the idea of being a self is only a side-effect of grammar and not something we actually experience would be deeply ironic: becoming enlightened by attaining a non-egoic state of experience is nothing special or extraordinary – it simply is our basic mode of being. but one does change, if only doxastically: before and after the episodes, one has false beliefs about the structure of one’s experiences; during these episodes, one gains true beliefs about the structure of one’s experiences. the difference is a difference in our belief system. and because all we claim is that there is no selfreference in phenomenal consciousness, we can still hold that there are unconscious processes that facilitate self-reference. we can therefore have purely cognitive de se beliefs without any phenomenal component indicating a lack of self-reference in phenomenal consciousness. despite claiming that there is no feeling of self, the no-ego account still belongs to the phenomenological diagnoses because it makes a claim about phenomenality: there is no feeling of self and there never was. in these episodes, we realise that we confused our way of speaking with our ways of experiencing. this is the revelation one attains in non-egoic experiences. revelations can be short-lived or persistent, frightening or enlightening. but they need not come with a phenomenal difference. people may react differently to them: some may feel as if they have lost something because a welcomed illusion has been revealed as just that – an illusion. language still has them under a spell, and they may therefore think that something must have been there before.27 27if our odd fellow fails to find a turnip’s ovaries, he may also think that this turnip has lost its gender – even though it never had one to begin with. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 30 others might truly feel the glow of epistemic progress and accept that language is confusing. they either stop speaking or start using a form of language that allows them to avoid a first-person pronoun, like speaking japanese, or in english using only plural forms, third-person pronouns, or definite descriptions (e.g. “the speaker”) to avoid being bewitched again by language. this doxastic difference between before/after and during the episode can have widespread cognitive effects, like most doxastic differences. obviously, we can have reports here because this account of non-egoic experiencing does not exclude the possibility of any form of cognitive self-reference. it just claims that there never was a phenomenal self, not that there never was any form of self-reference. all it rejects is that forms of self-reference ever show themselves in experience. non-egoic states aren’t anything special. the no-ego account is obviously highly counterintuitive. for many, it is an obvious fact not just that they are someone, but that it feels like something to be someone. that this belief is so widespread may merely underline how deep the confusion runs. we need a convincing error theory for why this delusion that there is a feeling of self is so common. pointing to grammar may get us only so far. but a proponent of the no-ego account may offer the following explanation, which relies on two theses. first: in any language, we track cognitive differences in speaker and audience. according to a pragmatics-first approach (see e.g. bar-on, 2013; moore, 2016, 2017, 2018; scott-phillips, 2015), language was developed for communication and coordination, not necessarily for cognition. in communication, we have an embodied speaker and audience. both are co-represented during an act of communication. however, communication happens in the real world, not just in our minds – and in the real world, speaker and audience differ. it is good communicative practice to keep track of the intentions and beliefs of the different individuals involved in a communicative act. only if we are aware that someone differs from us in their beliefs may we be able to deceive or teach. in the framework of gricean communication, tracking audience and speaker separately is in fact basic to any communicative act. and because language is one of the most important tools we have for categorising and capturing the structure of our experiences, we transpose into descriptions of experience this difference between speaker (i.e., us) and audience (the others). second, we are by nature naïve realists: in our natural stance towards our experiences, we take ourselves to be in contact with the real world, not with mental representations or phenomenal experiences standing in for the real world, unless we reflect on or are made aware of specific illusions. for this reason, we often pick out experiences by what we take them to stand in for in the world. for example, we may say that we had an experience as of a lamb on the horizon – although we only experience a hazy fleck of white. could be a lamb, could be a hungarian sheepdog (a komondor), or even a white car. so we picked out this experience based not on how we experienced it, but on what we think it relates to in the world outside our fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 31 minds. in such cases, there is a “crossing over” from what we believe about the world onto how we categorise our experiences themselves. in this process, we may come to believe that there are certain structures in our experiences even though these structures pertain only to the world outside. here is an illustrative example adapted from palmer (1999, p. 209): we know that each of us has only one left and one right hand. and we think that our visual experiences present us with only one left and one right hand when we look at either one. but because of the set-up of our eyes, anything outside of where we focus produces non-aligned images on our retinas. this non-alignment shows itself as “seeing double” outside of where we fixate (including an area called “panum’s fusional area”), which is sometimes interpreted as depth. the effect is most striking if we stretch an arm out and focus on our thumb or on the background. if we fixate on the thumb, we see double in the background; if we fixate on the background, we visually experience two thumbs. but this is true for all visual experiences, thanks to the architecture of our visual system. so we do not actually experience one left hand if we fixate on the right; we experience two. yet in forming our beliefs about experience, we superimpose how we think the world is – where we actually only have one left and one right hand – onto what we think our experiences are like. this forcing-onto-experience might also happen with “fundamental” structures: structures that stem from a basic feature of our dominant form of communication, namely tracking which intentions and beliefs belong to us as speakers and which ones belong to our audience. it might be that there is no felt self in phenomenal consciousness, but merely cognitive tracking of which mental states are under our control and which ones are outside of it. obviously, today we exploit language for cognition, not merely for communication. still, tracking which mental states belong to the speaker and which to others remains important. and this tracking of our own mental states as our own could be superimposed onto consciousness even though it is not in consciousness. in this view, there is no feeling of self. there is, however, a self superimposed on an experience, necessary for registering this experience as one being had by the speaker, and not necessarily shared by the audience. in most cases, the speaker will fail to notice this superimposition as a purely cognitive add-on and will instead take it to be an accurate apprehension of the experience itself. i am uncertain whether i buy into this sketch of an error theory, but it certainly is a plausible one which a proponent of a non-egoic account might offer. this error theory explains why the ego-delusion – the delusion that there is a felt self – is so widespread: it exploits a mechanism at the core of every gricean communicator. 5.6 summary how should we proceed if we captured a sane in the wild? if we have reason to believe that (concerning this specific sane) what is said reflects what is meant, we reject the pragmatic diagnoses. then, if we have reason to believe that the fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 32 utterer of this sane is neither obviously deluded, nor irrational, nor deceiving us, we exclude the cognitive diagnoses. that leads us to the phenomenological diagnoses, where four options are available. in each, the speaker uses this sane to express something concerning her experiences. of these, only one (total egodissolution) leads to the sane paradox. this leaves three live options. must we favour one of these phenomenological diagnoses over the others for all cases? or should we expect that each may apply to some but not all cases of “non-egoic” ego-alteration? should we therefore seek one general phenomenological diagnosis, or be pluralists who proceed on a case-by-case basis? this is what i focus on in the last section. 6 many phenomenological diagnoses or only one? my goals for this article were twofold. first, i have tried to show that there is not only one diagnosis available to us to explain the occurrence of verbalised selfascriptions of non-egoic experiences (sane). pragmatic and cognitive diagnoses are available to us. in these, we explain why someone says something apparently self-defeating without making any commitment to any unusual experiences. second, i hope to have demonstrated that the sane paradox relies on controversial claims: the first, called uniqe, that there is one and only one mechanism for self-reference; the second, called phenself, that each mechanism facilitating self-reference necessarily manifests itself in phenomenal consciousness; the third, called cease, that a feeling of self dissolves in these episodes. but we can reject either one and give explanations of how this is compatible with the occurrence of honest and somewhat veridical sane: in phenomenality, a self may be modular, there may be a plurality of selves in each person, the self may expand to the point where the self/world distinction loses all meaning, or there never was a phenomenal self – a fact we come to cognitively realise in such episodes. total ego-dissolution, however, may be a state achieved only in death, not while we continue to cognise. now, are these phenomenal hypotheses in opposition? do some entail others? are some incompatible? one might think not, and thereby accept that egodissolution is a heterogeneous phenomenon. this would raise many methodological questions about how to differentiate these phenomena empirically. but i believe that such pluralism of diagnoses is misguided: we should expect a homogeneous phenomenological diagnosis for all cases. why only one account? if there is a feeling of self, it is a highly abstract feature of consciousness. we should not expect massive variance among individuals, just as we do not expect massive inter-individual differences in the way we experience space, time, or unity (cf. hohwy, 2011; fink, 2018). thus, we should not expect more than one account to be true, but we should expect more than one to be false. does one entail another? no. all phenomenological diagnoses differ tremendously. let me illustrate. is there any feeling self at all phenomenally present fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 33 or not? here, all theories agree except for the no-ego account. if there is a feeling of self, is it unique and monolithic? total ego-dissolution and ego-expansion subscribe to this view, which distinguishes them from modularity and plurality accounts. the no-ego account denies the antecedent here. lastly, if there is a feeling of self, does it dissolve during these episodes? here, ego-expansion says no, while total ego-dissolution and the multiplicity account both say yes. so none of these theories can be reduced to any of the others. therefore, at most one can be true. and if we cannot expect inter-individual variation on such a fundamental level, these accounts are actually in competition. if the phenomenological diagnoses are mutually irreducible, which one should we prefer? due to the sane paradox and the problems with de se memory formation, i reject total ego-dissolution as a live option. what of its alternatives? so far, the data do not tell, and each hypothesis fails to account for some aspects of certain reports.28 authors like millière (2017) may favour a multiplicity account. but this account is not very parsimonious because it postulates a plenitude of feelings of self. additionally, it still needs to explain how they interact in order to give rise to a feeling of one unique and undivided self, a self in the sense of an individual. these are reasons why i prefer its alternatives. what of the other two? i have a place in my heart for the no-ego account, mainly due to its ironic nature: thousands of hours of sitting and attending and all you get is just an insight about what was there from the start. but, unfortunately, irony is not evidence. we should, however, worry about attributing to ourselves a massive number of false beliefs. yet, i find the error theory of why we believe that there is a felt self convincing. it relies, however, on the pragmatics-first account of language and on us being naïve realists. both are controversial theses. if we reject either, we should favour ego-expansion as the underlying process for “ego-dissolution”. here, episodes of an inflated sense of self are misapprehended, leading to a report either about ego-expansion or about ego-dissolution. this is the account i deem most likely. however we pick our favourites from amongst the phenomenological diagnoses, we have already taken a step forward. we now have several hypotheses that compete; each makes different predictions and can be supported by different kinds of evidence; and we may have a ranking of which has the highest prior probability. as for which of these will explain the incoming data best – time will tell. 28this may be explained to some degree by cultural and social biases in reporting. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 34 acknowledgments this article benefitted tremendously from the input of numerous individuals. i am grateful to all of them. among these are the participants of the workshop on radical disruptions of self-consciousness at frankfurt in october 2018. especially, i want to thank thomas metzinger and raphaël millière for their great and detailed input and criticism, which improved the paper tremendously, and also for their work in making this special issue happen, and their trust and courage to publish with philosophy and the mind sciences, making this our inaugural issue. i also wish to thank two anonymous reviewers and three lessthan-anonymous commentators (chiara caporuscio, adrian kind, fabian fuchs) for their supportive and helpful comments. also, i am grateful to my certainly-not-anonymous coeditors-in-chief, wanja wiese and jennifer windt, for help, support and guts to simply go and run with the idea of a free open-access-journal. lastly, i thank emily troscianko for her great help and input as lector. all remaining mistakes are mine. part of this research was funded by the dfg-rtg-2386 extrospection. references apel, k.-o. (1976). das problem der philosophischen letztbegründung im lichte der transzendentalen sprachpragmatik: versuch einer metakritik des “kritischen rationalismus”. in b. kanitscheider (ed.), sprache und erkenntnis (festschrift für gerhard frey zum 60. geburtstag) (pp. 55–82). innsbruck: inst. f. sprachwissenschaft d. univ. innsbruck. bar-on, d. (2013). origins of meaning: must we “go gricean”? mind & language, 28(3), 342–375. https://doi.org/10.1111/mila.12021 berkovich-ohana, a., dor-ziderman, y., glicksohn, j., & goldstein, a. (2013). alterations in the sense of time, space, and body in the mindfulness-trained brain: a neurophenomenologically-guided meg study. frontiers in psychology, 4, 912. https://doi.org/10.3389/fpsyg.2013.00912 bortolotti, l. (2009). delusion and other irrational beliefs. oxford: oxford university press. coltheart, m. (1980). iconic memory and visible persistence. perception & psychophysics, 27(3), 183–228. https://doi.org/10.3758/bf03204258 conway, m. a. (2005). memory and the self. journal of memory and language, 53(4), 594–628. https://doi.org/10.1016/j.jml.2005.08.005 conway, m. a., & pleydell-pearce, c. w. (2000). the construction of autobiographical memories in the self-memory system. psychological review, 107(2), 261–288. https://doi.org/10.1037//0033-295x.107.2.261 deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 dor-ziderman, y., berkovich-ohana, a., glicksohn, j., & goldstein, a. (2013). mindfulness-induced selflessness: a meg neurophenomenological study. frontiers in human neuroscience, 7(582), 1–17. https://doi.org/10.3389/fnhum.2013.00582 fink, s. b. (2018). introspective disputes deflated: the case for phenomenal variation. philosophical studies, 175(12), 3165–3194. https://doi.org/10.1007/s11098-017-1000-8 flanagan, o. (1992). consciousness reconsidered. cambridge, ma: mit press. foster, c. (2016). being a beast. london: profile books. gennaro, r. j. (2008). are there pure conscious events? in c. chakrabarti & g. haist (eds.), revisiting mysticism (pp. 100–120). newcastle: cambridge scholars press. gennaro r. j. (2020). cotard syndrome, self-awareness, and i-concepts. philosophy and the mind sciences, 1(i), 4. https://doi.org/10.33735/phimisci.2020.i.41 goldman, a. i. (1997). science, publicity, and consciousness. philosophy of science, 64(4), 525–545. https://doi.org/10.1086/392570 grice, h. p. (1991). logic and conversation. in m. ezcurdia & r. j. stainton (eds.), studies in the way of words (p. 47). cambridge: harvard university press. hohwy, j. (2011). phenomenal variability and introspective reliability. mind & language, 26(3), 261–286. https://doi.org/10.1111/j.1468-0017.2011.01418.x hume, d. (1739). a treatise of human nature. london: john noon. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1111/mila.12021 https://doi.org/10.3389/fpsyg.2013.00912 https://doi.org/10.3758/bf03204258 https://doi.org/10.1016/j.jml.2005.08.005 https://doi.org/10.1037//0033-295x.107.2.261 https://doi.org/10.33735/phimisci.2020.i.39 https://doi.org/10.3389/fnhum.2013.00582 https://doi.org/10.1007/s11098-017-1000-8 https://doi.org/10.33735/phimisci.2020.i.41 https://doi.org/10.1086/392570 https://doi.org/10.1111/j.1468-0017.2011.01418.x https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org look who’s talking! varieties of ego-dissolution without paradox 35 hyman jr, i. e., husband, t. h., & billings, f. j. (1995). false memories of childhood experiences. applied cognitive psychology, 9(3), 181–197. https://doi.org/10.1002/acp.2350090302 josipovic, z. (2010). duality and nonduality in meditation research. consciousness and cognition, 19(4), 1119–1121. https://doi.org/https://doi.org/10.1016/j.concog.2010.03.016 klein, s. b., & loftus, e. f. (2014). the mental representation of trait and autobiographical knowledge about the self. in t. k. srull & r. s. wyer jr (eds.), the mental representation of trait and autobiographical knowledge about the self : vol. v (2nd ed., pp. 1–50). new york: psychology press. klein, s., & nichols, s. (2012). memory and the sense of personal identity. mind, 121(483), 677–702. https://doi.org/10.1093/mind/fzs080 letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 letheby, c., & gerrans, p. (2017). self unbound: ego dissolution in psychedelic experience. neuroscience of consciousness, 2017, 1–11. https://doi.org/10.1093/nc/nix016 lewis, d. k. (1979). attitudes de dicto and de se. the philosophical review, 88(4), 513–543. https://doi.org/10.2307/2184843 loftus, e. f., & pickrell, j. e. (1995). the formation of false memories. psychiatric annals, 25(12), 720–725. https://doi.org/10.3928/0048-5713-19951201-07 metzinger, t. (2004). being no one. cambridge, ma: mit press. metzinger, t. (2018). minimal phenomenal experience. mindrxiv. https://doi.org/10.31231/osf.io/5wyg7 millière, r. (2017). looking for the self: phenomenology, neurophysiology and philosophical significance of drug-induced ego dissolution. frontiers in human neuroscience, 11(245), 1–22. https://doi.org/10.3389/fnhum.2017.00245 millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 moore, r. (2016). gricean communication and cognitive development. the philosophical quarterly, 67(267), 303–326. https://doi.org/10.1093/pq/pqw049 moore, r. (2017). pragmatics-first approaches to the evolution of language. psychological inquiry, 28(2/3), 206–210. https://doi.org/10.1080/1047840x.2017.1338097 moore, r. (2018). gricean communication, joint action, and the evolution of cooperation. topoi, 37(2), 329–341. https://doi.org/10.1007/s11245-016-9372-5 nour, m. m., evans, l., nutt, d., & carhart-harris, r. l. (2016). ego-dissolution and psychedelics: validation of the ego-dissolution inventory (edi). frontiers in human neuroscience, 10(269), 1–13. https://doi.org/10.3389/fnhum.2016.00269 palmer, s. e. (1999). vision science: photons to phenomenology. cambridge, ma: mit press. preller, k. h., burt, j. b., ji, j. l., schleifer, c. h., adkinson, b. d., stämpfli, p., et al. (2018). changes in global and thalamic brain connectivity in lsd-induced altered states of consciousness are attributable to the 5-ht2a receptor. elife, 7(e35082), 1–31. https://doi.org/10.7554/elife.35082 saks, e. r. (2007). the center cannot hold: my journey through madness. new york: hyperion. sartre, j.-p. (1960). the transcendence of the ego. new york: hill; wang. sartre, j.-p. (1992). la transcendance et l’ego. paris: vrin. (original work published 1936) sass, l., pienkos, e., nelson, b., & medford, n. (2013). anomalous self-experience in depersonalization and schizophrenia: a comparative investigation. consciousness and cognition, 22(2), 430–441. https://doi.org/10.1016/j.concog.2013.01.009 schwitzgebel, e. (2008). the unreliability of naive introspection. philosophical review, 117(2), 245–273. https://doi.org/10.1215/00318108-2007-037 schwitzgebel, e. (2012). introspection, what? in d. smithies & d. stoljar (eds.), introspection and consciousness (pp. 29–47). https://doi.org/10.1093/acprof:oso/9780199744794.003.0001 scott-phillips, t. c. (2015). nonhuman primate communication, pragmatics, and the origins of language. current anthropology, 56(1), 56–80. https://doi.org/10.1086/679674 simeon, d., & abugel, j. (2006). feeling unreal: depersonalization disorder and the loss of the self. new york: oxford university press. sorensen, r. a. (2004). vagueness and contradiction. new york: oxford university press. studerus, e., gamma, a., & vollenweider, f. x. (2010). psychometric evaluation of the altered states of consciousness rating scale (oav). plos one, 5(8), 1–19. https://doi.org/10.1371/journal.pone.0012412 tversky, a. (1969). intransitivity of preferences. psychological review, 76, 31–48. https://doi.org/10.1037/h0026750 zahavi, d., & kriegel, u. (2015). for-me-ness: what it is and what it is not. in d. dahlstrom, a. elpidorou, & w. hopp (eds.), philosophy of mind and phenomenology (pp. 36–53). new york: routledge. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1002/acp.2350090302 https://doi.org/https://doi.org/10.1016/j.concog.2010.03.016 https://doi.org/10.1093/mind/fzs080 https://doi.org/10.33735/phimisci.2020.i.47 https://doi.org/10.1093/nc/nix016 https://doi.org/10.2307/2184843 https://doi.org/10.3928/0048-5713-19951201-07 https://doi.org/10.31231/osf.io/5wyg7 https://doi.org/10.3389/fnhum.2017.00245 https://doi.org/10.33735/phimisci.2020.i.48 https://doi.org/10.1093/pq/pqw049 https://doi.org/10.1080/1047840x.2017.1338097 https://doi.org/10.1007/s11245-016-9372-5 https://doi.org/10.3389/fnhum.2016.00269 https://doi.org/10.7554/elife.35082 https://doi.org/10.1016/j.concog.2013.01.009 https://doi.org/10.1215/00318108-2007-037 https://doi.org/10.1093/acprof:oso/9780199744794.003.0001 https://doi.org/10.1086/679674 https://doi.org/10.1371/journal.pone.0012412 https://doi.org/10.1037/h0026750 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sascha benjamin fink 36 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.40 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org sane introduced: self-ascriptions of episodes of non-egoic experiences and phenomenal ego-dissolution sane and their role in researching non-egoic experiences sane rejected: the self-defeating nature of first-person reports of non-egoic episodes sane diagnosed: what is said vs. what is meant, believed, rational, true a pragmatic diagnosis cognitive diagnoses unbelievable? irrational? untrue sane revived: four ways to lose one's self rejecting the paradox can we defend total ego-dissolution against the paradox? multiple feelings of self: plurality or modularity ego-expansion no-ego revelation summary many phenomenological diagnoses or only one? laying down a forking path: tensions between enaction and the free energy principle. laying down a forking path: tensions between enaction and the free energy principle ezequiel a. di paoloa,b,c (ezequiel.dipaolo@ehu.es) evan thompsond (evan.thompson@ubc.ca) randall d. beere (rdbeer@indiana.edu) abstract several authors have made claims about the compatibility between the free energy principle (fep) and theories of autopoiesis and enaction. many see these theories as natural partners or as making similar statements about the nature of biological and cognitive systems. we critically examine these claims and identify a series of misreadings and misinterpretations of key enactive concepts. in particular, we notice a tendency to disregard the operational definition of autopoiesis and the distinction between a system’s structure and its organization. other misreadings concern the conflation of processes of self-distinction in operationally closed systems and markov blankets. deeper theoretical tensions underlie some of these misinterpretations. fep assumes systems that reach a non-equilibrium steady state and are enveloped by a markov blanket. we argue that these assumptions contradict the historicity of sense-making that is explicit in the enactive approach. enactive concepts such as adaptivity and agency are defined in terms of the modulation of parameters and constraints of the agent-environment coupling, which entail the possibility of changes in variable and parameter sets, constraints, and in the dynamical laws affecting the system. this allows enaction to address the path-dependent diversity of human bodies and minds. we argue that these ideas are incompatible with the time invariance of non-equilibrium steady states assumed by the fep. in addition, the enactive perspective foregrounds the enabling and constitutive roles played by the world in sense-making, agency, development. we argue that this view of transactional and constitutive relations between organisms and environments is a challenge to the fep. once we move beyond superficial similarities, identify misreadings, and examine the theoretical commitments of the two approaches, we reach the conclusion that far from being easily integrated, the fep, as it stands formulated today, is in tension with the theories of autopoiesis and enaction. keywords agency ∙ autopoiesis ∙ enaction ∙ free energy principle ∙ sense-making aikerbasque, basque foundation for science, bizkaia, spain bias-research center for life, mind and society, university of the basque country, donostia, spain ccentre for computational neuroscience and robotics, university of sussex, brighton, uk ddepartment of philosophy, university of british columbia, canada ecognitive science program, indiana university, bloomington, usa di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-3296-5021 https://orcid.org/0000-0003-0084-8477 https://orcid.org/0000-0001-7279-3028 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 2 1 introduction the free energy principle (fep, friston et al., 2006; friston, 2012) has spurred a large body of literature in the sciences of mind. a growing subset of this literature concerns a series of claims about the compatibility between fep and theories in embodied cognitive science, in particular, the enactive approach (e.g., di paolo et al., 2017; thompson, 2007; varela et al., 1991). several articles (to date over two dozen) have appeared that make some connection between fep, autopoiesis, and enaction. we are interested in examining these claims. they are varied and presented with justifications that range from pointing out general similarities to attempting to demonstrate deeper, principled connections between the two approaches. these claims have been interpreted in different ways. some welcome the prospect of compatibility as positive to both the enactive and fep positions, leading potentially to a more powerful and unified theory of embodied cognition. they speak of elaborations, interpretations, extensions, syntheses, etc. (e.g., bruineberg et al., 2018). others think fep may serve to reconcile enaction with positions enactivists criticize, such as modern versions of representationalism (clark, 2015; constant et al., 2021; wiese & friston, 2021). others adopt a different tone that suggests that fep has overcome the limitations of the enactive approach, or subsumes or absorbs autopoietic and enactive theories (e.g., allen & friston, 2018; “fep subsumes autopoiesis,” korbak, 2021, p. 2747; “fep provides an implementation of enactivism, and in a sense supersedes or absorbs classical (i.e., autopoietic) formulations,” ramstead et al., 2021, p. 59). finally, in contrast to the latter, some propose that enactive ideas can be used to fill in gaps and resolve problems in the fep framework (kirchhoff & froese, 2017; van es & kirchhoff, 2021). given this variety, it is difficult to examine this literature as a coherent whole. moreover, claims continue to evolve, so that sometimes important details absent in one piece of work become relevant in another. nevertheless, we think that there are a few core shared claims that do not seem to change. in particular, there is the claim that the extension of fep from its original domain of application in neuroscience to living systems in general serves as the basis of a broad theory of biology and cognition, something that resonates with enactive discourse about the continuity between life and mind (thompson, 2007). a more specific common claim concerns the need to attend to the active role of the cognitive agent in engaging the world by following vital norms. this claim underlines the close connection between internal processes, processes in the environment, the agent’s activity, and the agent’s viability. at first sight, enactivists would agree with these claims. but are these parallels merely superficial? the literature we examine here claims that the connection between the two perspectives runs deep. we disagree. we will argue that the apparent compatibilities are based on quick readings, or even misreadings, of enactive ideas and what they entail, while central aspects of autopoiesis and enaction are left unacknowledged. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 3 clearing up misinterpretations may move the discussion forward, but will not by itself resolve the deeper tensions between fep and enaction as theories of cognition. many commentators focus on what they see as the main source of tension, which has to do with the representational and cognitive-loaded bias that instantiations of fep and related ideas tend to adopt, particularly in the form of predictive processing models (hohwy, 2013). they argue that fep admits wider interpretations of terms such as “model” and “inference,” and that these interpretations are compatible with enaction’s rejection of internalism and computationalism. while these arguments have merit, we think they miss much more basic tensions between the two theories. these tensions have to do with how the enactive approach conceives of agents as precarious, self-constituted entities in ongoing historical development and capable of incorporating different sources of normativity throughout their development, a world-involving process that is co-defined with their environment across multiple spatiotemporal scales and together with other agents. the enactive view, as we shall explain, is at odds with universalizing the ideas of nonequilibrium steady states and markov blankets that serve as key assumptions in fep. underlying these worries are divergent fundamental conceptions of materiality and temporality. these incompatibilities lead to a series of forking claims about embodied agents in general, claims that take on a sharper contrast in the case of human beings due their diversity and historicity. nonrepresentational interpretations of fep do not rejoin these divergences. before we proceed, a few caveats. first, we do not intend to critically evaluate fep altogether (for this, see e.g., aguilera et al., 2021; baltieri et al., 2020; biehl et al., 2021; colombo & wright, 2021; litwin & miłkowski, 2020; raja et al., 2021; van es, 2021). we are exclusively concerned with examining claims about the compatibility between fep and enaction. second, in specifying what we see as points of theoretical tension between the approaches, we work on the basis of the fep as currently formulated. we make no claims that clarifications, extensions, or elaborations of the fep could not address some of the divergences between the approaches. our argument will proceed roughly as follows: (1) there are certain principles and phenomena that fep needs to be able to account for if it claims to make contact with enaction; (2) to date, fep has not successfully accounted for these enactive principles and phenomena, despite some claims to the contrary in the literature; (3) there are reasons to think that it might be difficult for anything like the fep to provide such an account, although time will tell and we are certainly not claiming any impossibility proof. finally, we do not reject specific applications of fep ideas or techniques to make sense of empirical data (walsh et al., 2020). insofar as they offer useful methods for analysis, fep and predictive processing models can provide helpful tools in specific cases. our discussion concerns fep as a general approach to explaining life and mind. it is at this level that we find significant and hard to reconcile differences with the enactive approach. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 4 2 an enactive fep? enactive readings of fep and claims of compatibility with ideas such as autopoiesis are varied. some attempt principled unifying arguments (ramstead et al., 2021), others provide readings of fep from enactive and ecological perspectives (bruineberg et al., 2018), and others simply remark on apparent resonances and similarities (clark, 2015). despite this variety, the motivations for these diverse claims seem fairly aligned. we can roughly describe them as a concern to establish embodiedcognition interpretations of fep that do not necessarily entail a traditionally internalist or representationalist perspective (as predictive processing models tend to do, hohwy, 2013), or that at least recast notions of representation in actionoriented terms (clark, 2015). some authors are also motivated by the potential extensions of fep, originally formulated within the confines of neuroscience, to a wider range of biological phenomena—indeed, to the status of a general principle of theoretical biology and psychology. demonstrating compatibility with the enactive approach, with its emphasis on the continuity between life and mind, could be considered as a step towards this goal. because of these aligned motivations, it makes sense to look at this literature as a whole. we group its claims into two broad sets. we are aware that we may miss some nuances, but we do not think these will be central to our purposes. in particular, we are concerned with the following two groups of ideas: (1) proposals for a link between fep and the theory of autopoiesis, and (2) discussions of various compatibilities between fep and enactive ideas. we present only selected examples in this section and offer a critical analysis in the next two sections. we will not occupy ourselves with other discussions that point to broad similarities between fep and “enactive-like” ideas (e.g., situated embodied activity, action-perception co-dependence) that are shared by various other approaches (e.g., pezzulo et al., 2015). it is important to keep in mind that autopoiesis and enaction are obviously related and the division into two groups is merely one of convenience. the unspoken assumption is that the enactive approach has developed historically from readings and elaborations of the ideas first presented in the classical theory of autopoiesis (maturana & varela, 1980, 1987), and that to this day many central concepts in enaction find their roots in this theory. however—and this may be a confounding factor to keep in mind—the enactive approach is in many ways a branching development of the theory of autopoiesis and is critical of important aspects of classical autopoietic theory (e.g., di paolo, 2018). contemporary defenders of classical autopoietic theory, in turn, have criticized the enactive approach precisely for this reason (e.g., villalobos & ward, 2015). our first group concerns the goal championed by karl friston and others of extending the reach of the fep to biology. this work discusses fep’s relation to the theory of autopoiesis and concerns enactive ideas only indirectly. nevertheless, its di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 5 claims form a kind of backbone for work looking at enaction, the reasoning being that enactive theory is an elaboration of the theory of autopoiesis, as we have just said. friston (2013) provides an early claim concerning the relation between fep and autopoiesis (for more up-to-date versions of this claim, see parr et al., 2019; ramstead et al., 2021; wiese & friston, 2021). this paper describes the applicability of fep to biological systems (see also friston, 2012) by explicitly referring to the concept of autopoiesis. briefly, autopoiesis is a property of the organization of living systems (maturana & varela, 1980, 1987). maturana & varela (1980) originally postulated it as the defining property, necessary and sufficient for the living organization, but subsequent researchers have argued that autopoiesis is necessary but not sufficient for the living organization (bitbol & luisi, 2004; bourgine & stewart, 2004; di paolo, 2005; see thompson, 2007, for discussion). “autopoiesis” means that living systems are organized as networks of biochemical processes such that two conditions obtain: (1) (self-production) the operation of the processes in the network regenerates the set of relations between processes in the network; and (2) (self-distinction) the network emerges as a distinct topological unity in the domain of biochemical interactions. an autopoietic system is therefore materially self-producing and selfdistinguishing. due to the circularity (recursive self-production) to which this idea refers, maturana & varela (1980, see also 1987; varela, 1979) described an autopoietic system as manifesting “organizational closure” or “operational closure.” we will come back to these ideas in the next section. friston (2013) considers an ergodic random dynamical system. according to him, ergodicity in this context “means that the time average of any measurable function of the system converges (almost surely) over a sufficient amount of time” (ibid., p.2) and that “one can interpret the average amount of time a state is occupied as the probability of the system being in that state when observed at random” (ibid.). such a system will tend to evolve towards a regime where the probability density of its states (the chance that a particular state will be visited) does not change over time. if the density of states is unchanging, that is, if the system reaches a nonequilibrium steady state (ness), one can derive a series of relations between internal and external variables. the distinction between these sets of variables is defined by assuming the presence of a markov blanket that statistically “insulates” one set from directly affecting the other.1 instead, effects propagate through action and sensory variables that constitute the blanket and mediate between internal and 1biehl et al. (2021) clarify that markov blankets are presented as a combination of two independent conditions. one concerns the vector field underlying the stochastic differential equation of the system forcing any systematic influences between external and internal variables through the sensory and active variables. the other is an actual markov blanket condition on the stationary distribution saying that according to this distribution the internal and external states are conditionally independent given sensory and active states. for our purposes we take markov blankets to imply a combination of the two conditions. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 6 external variables. according to friston, these relations demonstrate the central claim that “any ergodic random dynamical system that possesses a markov blanket will appear to actively maintain its structural and dynamical integrity” (friston, 2013, p. 2, emphasis removed). friston sees the maintenance of structural and dynamical integrity as the key property of biological systems, from which he conjectures that, to exist, organisms must be ergodic2 (or converge to a ness, e.g., costa et al., 2021; parr et al., 2019) and must be in possession of a markov blanket.3 friston refers to autopoiesis as an attribute of biological self-organization and claims that it can be derived from his analysis. he sees the maintenance of a steady state as a kind of structural integrity; and he sees such structural integrity as a form of homeostasis sustained by the system appearing to act against the dispersion or entropy of its states (friston, 2013, p. 5). the same argument is repeated in more recent publications (e.g., parr et al., 2019; wiese & friston, 2021). the parallel is reaffirmed elsewhere by suggesting that markov blankets play the role of the autopoietic system’s boundary-subserving processes of self-distinction. allen & friston (2018) further claim that “one can formulate this in another way; the organism’s internal states constitute probabilistic beliefs about what actions are the most likely to provide evidence for the organism’s existence (survival)” (ibid. p. 2474), and they state that “this notion is at the heart of autopoietic views of life and mind, insofar as it induces a deeply circular causality between internal and external states, to provide a normative principle by which to understand all action and perception” (ibid. p. 2474). 2the phrase “biological systems are ergodic” found in (friston, 2013, p. 5) could be interpreted as involving both the organism and its environment. similarly with the phrase: that “all living systems revisit a bounded set of states repeatedly (i.e., they are locally ergodic)” (ramstead et al., 2018, p. 3) although from the context “living systems” seems equivalent to “organisms.” friston (2012, p. 2105), however, makes reference to systems “in a changing (and possibly non-ergodic) environment” suggesting that the ergodicity claim is made about organisms. 3friston’s argument bears a resemblance to an idea presented by w. r. ashby (1962) if we interpret his use of the term “equilibrium” to mean a steady-state flow. “so the answer to the question: how can we generate intelligence synthetically? is as follows. take a dynamic system whose laws are unchanging and single-valued, and whose size is so large that after it has gone to an equilibrium that involves only a small fraction of its total states, this small fraction is still large enough to allow room for a good deal of change and behavior. let it go on for a long enough time to get to such an equilibrium. then examine the equilibrium in detail. you will find that the states or forms now in being are peculiarly able to survive against the changes induced by the laws. split the equilibrium in two, call one part ‘organism’ and the other part ‘environment’: you will find that this ‘organism’ is peculiarly able to survive against the disturbances from this ‘environment.’ ” (ibid. p. 272). this resemblance may explain friston’s insistence on using the language of “appearances,” e.g., internal states appearing to minimize energy, appearing to engage in active inference, and appearing to model the external world, to mention three instances appearing in the abstract in friston (2013). ashby’s exercise, like friston’s, is ambiguous between deflating the notion that there is a real distinction between organism and environment, when in fact the boundary is drawn by the observer (or the choice of one of possibly many markov blankets), and suggesting that to speak of actual distinctions and of their appearances is ultimately the same thing, a view that enactivists tend to reject (e.g., barandiaran et al., 2009; di paolo et al., 2017, p. 39). di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 7 the central claim of these arguments is that random dynamical systems with one or several markov blankets that converge to a ness conserve themselves in that stationary condition (or appear to do so depending on the perspective) by minimizing free energy, and that these conditions are equivalent to the condition of a system being autopoietic. as we discuss later, friston’s (2013) use of the concept of autopoiesis is rather loose. he does not attempt to substantiate his claim by considering the definition of autopoiesis; instead, he bases the claim on a broad analogy between the conservation of systemic properties in the random dynamical systems he proposes and in autopoietic systems. further elaborations of his argument move into more specific enactive territory and lean more explicitly on ideas such as operational closure, adaptivity, agency, and sense-making (di paolo, 2005, 2009; thompson, 2007). for instance, kirchhoff et al. (2018) discuss the significance of the concept of the markov blanket. after pointing out that metabolic cell production is a process enabled by the existence of cellular boundaries, the authors conclude that “living systems can therefore be construed as a process of boundary conservation, where the boundary of a system is its markov blanket” (ibid. p. 6). the dependencies induced by the markov blanket act as a “kinetic barrier” that keeps the system “far removed from thermodynamical equilibrium.” a markov blanket does not fully isolate the system: “external states may influence internal states even if the former are not constitutive parts of an operationally closed system” (ibid). ramstead et al. (2021) draw similar analogies between markov blankets and the boundaries of (unicellular) organisms. more specifically, they claim in agreement with kirchhoff et al. that “it is fairly straightforward to establish that the markov blanket formalism provides a statistical formulation of operational closure” (ramstead et al., 2021, p. 55). our second group in the literature we are discussing corresponds to arguments in favour of a broad spectrum compatibility between fep and enaction, not just the concept of autopoiesis, though in all cases the claims tend to rely on the work in the first group. for instance, bruineberg et al. (2018) accept that fep is applicable to biological systems, but they reject its association with helmholtzian approaches, in which perception is mediated by unconscious inferences about the causes of sensation. a typical version of such approaches is the postulation of predictive processing models implemented in the brain. bruineberg and colleagues argue that fep is more general and one need not be committed to a helmholtzian perspective (though even when conceiving of strictly biological integrity in terms of the fep, friston already talks about active inference or the semblance of it). the authors articulate an ecological-enactive interpretation of fep, avoiding the helmholtzian sharp separation between organism and environment, and placing the locus of active inference in the whole situated organism and not just its brain. they offer charitable interpretations, such as construing the internal dynamics of active inference as states of action-readiness or expounding friston’s claim that an organism embodies an optimal model of its ecological niche (friston, 2011) in terms of the organism possessing adequate skills that allow it to reach conditions di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 8 of grip.4 these interpretations are indeed more palatable to enactivists and ecological psychologists than the helmholtzian picture. applying a similar ecological-enactive interpretation of fep to questions concerning the self and the feeling of mineness, kiverstein (2020) offers a more detailed comparison with enactive technical concepts. like ramstead et al. (2021), he describes the concept of autonomy in terms of precarious operational closure and suggests that this concept can be cashed out in fep terms. according “to the free energy principle, the autonomy of living systems is a consequence of the inferential processes of free energy minimisation” (kiverstein, 2020, p. 565), and “the free energy principle tells us how living systems might sustain their own operational closure under precarious conditions in their dynamic coupling with the environment” (ibid. p. 566). kiverstein also makes reference to key aspects of the enactive concepts of agency and sense-making: any system that has autonomy will also qualify as an agent that has its own individual point of view upon the world (di paolo et al., 2017). relative to this point of view the environment has affective significance in terms of how it bears on the organism’s self-produced identity. organisms enact values, purposes and norms which are of their own making in the sense that they originate in processes of self-individuation (i.e. free energy minimisation) to which the organism owes its existence. […] perception and action are thus laden with affect […]. (kiverstein, 2020, p. 566) except for the parenthetical reminder that he accepts the equating of selfindividuation and free energy minimization, kiverstein in these passages does not seem to be arguing for a parallel between enaction and fep as much as for what an enactive interpretation of sense-making could add to fep. kiverstein also suggests that a more embodied and situated understanding of active inference should be in terms of readiness for action, and he uses recent developments in enactive theory to expand on what this idea means. as the organisation of the autonomous system becomes less bound to its immediate metabolic needs, so the possible meaningful relations the organism can stand in to the environment becomes less tightly bound to the here and now. the organism becomes sensitive to tendencies and trajectories that constitute the dynamical configurations of the organism–environment system, and their consequences for its precarious existence […] the nervous system can then be thought of as generating and sustaining stable and recurrent patterns of sensorimotor engagement with the environment. these patterns of engagement 4recent interpretations of friston’s claim de-emphasize the idea that the organism implements or is itself a model of its environment, and argue instead that these models correspond to descriptions available to the observer (see, e.g., baltieri et al., 2020; van es, 2021). di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 9 with the environment exhibit just the same properties of operational closure and precariousness as we find in the more basic processes of homeostasis. the argument [here] is that patterns of sensorimotor engagement owe their operational closure to processes of free energy minimisation. (kiverstein, 2020, p. 566) in contrast to kiverstein’s attempt at a careful inter-theoretical comparison, other commentators offer their own particular reading of enactive ideas from an fep standpoint. for example, clark (2015) defends an embodied and action-oriented construal of predictive processing, which he uses to interpret varela et al.’s (1991) notion of “enacting a world.” despite there being several elaborations of this notion in the enactive literature (stewart et al., 2010; thompson, 2007; varela, 1991a, and others), clark idiosyncratically suggests that one way to interpret phrases like “bringing forth a world” is through processes such as active data sampling used to sustain and update a cognitive agent’s predictive models in an action-dependent manner. kirchhoff (2018) follows the thrust of the argument by friston (2013), adding a few glosses: “fep adds more to life than at [autopoietic theory] […] for the fep, living systems minimize surprise on the basis of embodying a probabilistic model of themselves and their environment. this is a step beyond the appeal to mere at in explaining the process of self-maintenance” (kirchhoff, 2018, p. 2526). following enactive arguments about the insufficiency of bare autopoiesis for sensemaking, kirchhoff states that the fep “incorporates adaptivity—viz., the futureoriented aspect of cognition—from the very beginning.” this point justifies the perspective that, more than just being compatible with autopoiesis, fep is also compatible with the enactive approach. adaptivity (di paolo, 2005) is taken as a complexification of autopoiesis. “this shoehorns nicely with the fep. given that the fep can be shown to apply to systems that are, at least arguably, not cognitive, it follows that mentality arises when organisms minimize free energy to a certain degree—viz., in the context of active inference” (ibid. p. 2535). kirchhoff continues, however, by raising the potential worry that fep might not fully account for the enactive notion of sense-making and the existentialphenomenological sense of life-mind continuity: “nothing in the fep is able to account for the constitution of a meaningful perspective” (kirchhoff, 2018, p. 2535). after describing the idea of sense-making as the most general aspect of mindedness and acknowledging that for enactivists this idea is central to an explanation of life-mind continuity, kirchhoff asks whether this concept can fully be captured by the minimization of free energy. this is one of the few instances in the literature we are examining where a potential incompatibility (rather than a mere difference) between fep and enaction is explicitly signaled: “if the fep gives up on sense-making, then how can it explain what is central to life and mind?” (ibid. p. 2536). kirchhoff observes that fep does not negate the idea of sense-making (so the relation between them is not so much compatibility as non-incompatibility), and that this idea has itself been criticized by so-called di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 10 radical enactivists (hutto & myin, 2013) for possibly entailing that basic cognition has meaningful content (content subject to correctness conditions), contrary to the nonrepresentationalist aims of the enactive approach (but see thompson, 2018, for a critical response to radical enactivism). in a strange reversal of roles, kirchhoff argues that if there were such an implication leading from sense-making to content, then the concept of sense-making, and hence the enactive approach, would be in tension with the nonrepresentational, leaner version of fep he defends. here an enactive explanation of meaning moves from something that fep as a theory of the mind lacks to something it should not have because it seemingly invites representations back. elsewhere we see similar worries about fep’s potential lack of cognitive specificity, as in kirchhoff & froese (2017), who suggest that enactive theory can constrain overly broad interpretations of fep. as we mentioned, friston’s (2013) argument applies to any random dynamical system in a ness with a markov blanket, and so the conclusions should be valid for a wide range of systems, including systems traditionally considered non-living or non-cognitive. kirchhoff and froese argue that friston should therefore be committed to a view that sees mind everywhere, as part of any system that minimizes free-energy, including non-biological ones (or, given some kind of life-mind continuity, to a view that finds both life and mind nearly everywhere). given the existence of systems, such as a candle flame, or in bruineberg et al.’s (2018) example, two synchronized pendulum clocks, whose stable features are more parsimoniously explained through dissipative selforganization or dynamical couplings than by supposing that internal variables are performing active inferences, kirchhoff and froese are right to worry about a thinning out of the explanatory power of the fep. they suggest that life and mind should be seen as strongly continuous, and that this continuity places cognitive specificity demands that fep does not meet. to remedy this problem, they suggest that enactive versions of life-mind continuity, defined in terms of autonomy, adaptivity, and sense-making, should be used to constrain an otherwise too liberal fep. in this quick exposition we have seen a sample of how fep, autopoiesis, and enaction have been discussed almost exclusively in terms of important compatibilities. these compatibilities are of different kinds, ranging from broad analogies between free energy minimization, markov blankets, and aspects of autopoiesis to more elaborate readings using enactive concepts such as autonomy, adaptivity, and sense-making. these compatibilities serve to indicate the possibility of a more detailed inter-theoretical relation or to supplement what some authors see as problematic aspects of fep as a candidate universal approach to life and mind. we find very few instances where serious incompatibilities are mentioned, making things look as if, except for some details and clarifications, there is a broad consensus about the match between the approaches. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 11 before we discuss why we strongly disagree with this picture and we examine deeper problems that have not been raised by any of these authors, we will mention a few misreadings that we consider important to clarify for this and future discussions. 3 misreadings of autopoiesis and enaction the enactive approach owes much to the classical theory of autopoiesis but it is not identical with it. enaction inherits several concepts from this theory, and more importantly, a sensitivity to the need to provide operational definitions of its key ideas. as well as serving as a constraint on theory development, such definitions should also facilitate inter-theoretical debates. concepts such as autopoiesis, autonomy, operational closure, sense-making, and agency are defined in operational terms in the enactive literature to avoid vague interpretations. it is surprising that very few attempts to link fep, autopoiesis, and enaction have taken advantage of these operational definitions and that technical terminology is sometimes used loosely or interpreted narrowly. this sloppiness results in a series of misreadings that, as we argue later, can signal deeper incompatibilities between the approaches. here we indicate some of these misreadings. 3.1 autopoiesis let us start by re-stating the classical definition of autopoiesis (after maturana & varela, 1980, pp. 78–79): an autopoietic system is organized (defined as a unity) as a network of processes of production (transformation and destruction) of components which: (i) through their interactions and transformations continuously regenerate and realize the network of processes and relations that produces them; and (ii) constitute the system as a concrete unity in the space in which the processes exist by specifying the topological domain of its realization as a network. we call the first and second requirements “self-production” and “self-distinction,” respectively. it is possible to show that both requirements are dialectically related to each other by using the concept of adaptivity (di paolo, 2018), which is important in recent enactive literature (di paolo & thompson, 2014; thompson, 2007). we return to this concept later. friston (2013, p. 5) refers to autopoiesis as an attribute of biological selforganization and homeostasis in which “active states will appear to maintain the structural and functional integrity of biological states.” others make similar remarks, such as “the internal and blanket states that constitute a subsystem are di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 12 autopoietic, because their (nonequilibrium steady-state or ergodic) probability density is maintained by the flow of the subsystem’s internal and active states.” (palacios et al., 2020, p. 5). autopoiesis is here understood as sufficiently characterized by self-organization and integrity, entailing persistence over time. however, autopoiesis is not defined as self-organization, integrity, or persistence over time. the definition describes the organization of a system that jointly fulfils the requirements of self-production and self-distinction. a system may selforganize and not be autopoietic. it may persist over time and not be autopoietic. and it may spontaneously achieve integrity and still not be autopoietic. the dynamic processes and conditions that form a crystal in a supersaturated liquid solution result in integrity, self-organization, and persistence over time but they do not constitute an autopoietic system (it does not continuously regenerate and realize the network of processes and relations that produces the crystal; it simply undergoes a one-off individuation event until it reaches a steady state). autopoiesis may in many cases entail integrity, self-organization, and persistence over time. when present, these properties are realized in specific ways in autopoietic systems, i.e., by fulfilling the requirements of self-production and self-distinction. but the presence of these properties in a given system, as we see in the crystal counter-example, does not, in any combination, entail autopoiesis. the literature linking fep and autopoiesis often takes these attributes as sufficient and in doing so disregards important aspects of the definition. to clarify these aspects, classical autopoietic theory introduces a distinction between the structure and the organization of a system (maturana & varela, 1980). the structure is the system’s actual realization, the concrete components that constitute a system and the actual and concrete relations between them. the system’s organization is the abstract set of relations that define the system as belonging to a class. autopoiesis is the description of a class of systems, i.e., a description of the organization that defines this class. concrete autopoietic systems may be instantiated in a wide variety of structures, and a given structure may belong to more than one class of organization (fido is a dog, a mammal, a living organism). structures also change over time, even if the organization remains invariant (fido was a puppy, is an adult hound, will be a lazy senior). the distinction between structure and organization, as well as other technical concepts, such as structural coupling and operational closure, are further clarified by beer (2015; beer, 2020) using a toy model that reveals other subtleties about these ideas, such as the relevance of their spatiality and time-extendedness. to come back to the relation between autopoiesis and integrity, selforganization, and persistence over time, it should now be clear that we need to be careful about whether we are talking about structure or organization. for instance, the autopoietic organization of a living organism persists over time as long as it lives, but its structure most often does not. statements about the structure of a system do not obviously translate into statements about its organization, and vice versa. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 13 setting aside this misreading of the idea of autopoiesis we may still ask whether the formal proposal made by friston and colleagues can be expected to lead to systems that specifically meet the requirements of self-production and self-distinction. we address this question in more detail in section 4, but here we discuss why friston and colleagues might think that this is the case. autopoiesis entails the conservation of biological organization and an adaptive relation to the environment, a relation that allows the autopoietic system to remain viable. conservation of organization has sometimes been illustrated in the classical autopoietic literature as a special form of homeostasis, lending some credence to friston’s interpretation of biological systems as resisting a tendency to disorder (e.g., friston, 2011, p. 92). for instance, “an autopoietic machine is an homeostatic (or rather a relations-static) system which has its own organization (defining network of relations) as the fundamental variable which it maintains constant” (maturana & varela, 1980, p. 79). the qualification (“relation-static”) should be read as a warning that we are not talking here of homeostasis in the traditional dynamical sense. indeed, the idea of an autopoietic organization as a regulated variable is problematic and not generally accepted. at the organizational—as opposed to the structural—level there is no gradient to the condition of being or not being autopoietic that could be thus regulated (di paolo, 2005). the sense of homeostasis to which friston and colleagues refer correspond to structural properties not organizational ones. while several references to “structural integrity” and to “self-organization” can be found in friston (2013) and other work, none of them corresponds to the structure/organization distinction elaborated in autopoietic theory. the structural integrity of systems with a markov blanket as they converge to a ness bears some resemblance to the idea of a system that actively conserves its organization. but it is not the same idea. a contrast with the definition of autopoiesis can verify this claim. several discrepancies become apparent. there is no obvious organization-level equivalent in friston’s systems to the network of processes that through transformations realize the conditions of its own production. in fact, the assumptions (of an invariant density of states and the presence of a markov blanket) are such that no regeneration is needed after the ness is reached; the “organization” of the system simply endures. it is unclear in what sense the components of the systems, e.g., its markov blanket, are materially produced by other processes in the system, instead of just being there by assumption (see also raja et al., 2021 for a similar point). friston himself may be thinking of this problem when he recognizes that more than entropy minimization is needed to distinguish a biological system from a petrified stone (friston, 2013, p. 11). in summary, autopoiesis has been loosely interpreted and indeed often misread to different degrees by friston and colleagues. attending to the operational definition of autopoiesis and the distinction between structure and organization, it is possible to show that the characterization of autopoiesis presented by friston and colleagues (structural integrity, persistence, self-organization) does not entail di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 14 autopoietic organization; such properties can be found in non-autopoietic systems. nor do autopoietic systems necessarily realize such properties in their structure. lack of attention to the structure/organization distinction can lead to readings of autopoiesis where both are conflated. 3.2 boundaries the concept of operational closure under precarious conditions (explicated in di paolo & thompson, 2014) underpins the enactive idea of autonomy (derived from varela, 1979), another technical term that extends the concept of autopoiesis to more general domains: [an] autonomous system is defined as a system composed of several processes that actively generate and sustain an identity under precarious circumstances. in this context, to generate an identity is to possess the property of operational closure. this is the property that among the conditions affecting the operation of any constituent process in the system there will always be one or more processes that also belong to the system. and, in addition, every process in the system is a condition for at least one other constituent process, thus forming a network. (di paolo, 2009, p. 15) in other words, the processes that constitute an operationally closed network relate to each other such that they form a set of mutually enabling relations. they do so under precarious conditions, meaning that in the absence of these mutually enabling relations, the same processes would tend to run down. no component is, in other words, strictly self-standing in the absence of the whole network. operational closure does not mean that processes external to the network cannot influence those that belong to it, or that processes in the network cannot influence processes external to it. nor does it mean that these influences cannot have an enabling character, i.e., by being causally necessary for internal processes to continue to exist. in other words, an operationally closed system is open not only to informational external influences but also to all kinds of material and formative relations, including, in the case of living systems, the transport of heat, mass, biomatter, genetic material, microorganisms, and so on. internal processes can literally depend on such enabling exchanges to continue to operate. ramstead et al. (2021) and kirchhoff (2018) make reference to operational closure. these authors see in this concept an enactive analogue of a markov blanket. for ramstead et al. (2021, p. 555), as we have indicated earlier, “it is fairly straightforward to establish that the markov blanket formalism provides a statistical formulation of operational closure.” for kirchhoff (2018), fep accommodates, through the idea of markov blankets, the requirement of operational closure and goes beyond it. confusingly, kirchhoff, following friston (2013), also states that it “can be shown that markov blankets operate in much the same way as a cell boundary” (kirchhoff, 2018, p. 2527; see also allen & friston, 2018, p. 2473; kirchdi paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 15 hoff et al., 2018, p. 6). we also read that a “cell therefore has a markov blanket—its plasmalemma [cell membrane]” (ramstead et al., 2021, p. 551). more complex organisms are described as bounded by an ensemble of nested markov blankets (kirchhoff et al., 2018). in addition, markov blankets are said to act as the epistemological and ontological boundaries of cognitive, not just biological, systems (ramstead et al., 2021, p. 551). similarly, kirchhoff & kiverstein (2021) propose using the formalism of markov blankets to delineate flexible boundaries for the mind. such statements are speculative and perplexing. they again confuse the organization and the structure of living and cognitive systems. we see no straightforward relation between operational closure (an organizational statement about how a network of processes actively produces and distinguishes itself) and markov blankets (a statement about statistical conditional independence between sets of variables). very few commentators seem to have remarked on this discrepancy (e.g., bruineberg & hesp, 2018, p. 38; see also bruineberg et al., 2021). as we have said, an operationally closed system is open to all kinds of interactions and exchanges with the environment, as long as its organization is not destroyed. these exchanges are typically regulated by the system, but not always (think of exposure to ionizing radiation or the effects of gravity), and not always successfully (think of a viral infection). potential breakdowns caused by unregulated exchanges are not necessarily fatal, meaning the condition of operational closure can still be sustained. because operationally closed systems may be realized in such complex unmediated relations between internal and external variables, it seems unreasonable to assume that operational closure is therefore equivalent to the statistical conditions imposed by markov blankets. moreover, because operational closure is an organizational concept, its relation to biological or cognitive structural “boundaries” is not obvious nor are there obvious relations between the boundaries of biological and cognitive systems, as discussed in di paolo (2009). if we know how the organization of an operationally closed system is structurally instantiated, we could in principle point to processes that belong to it and processes that do not. but this epistemic operation is not replaceable by the act of pointing to a physical structure, such as a cell membrane. maintaining the condition of self-distinction can lead to processes that regulate complex, structural and spatial boundaries, as in the case of cell membranes, but also to other processes, such as the prokaryotic crispr-cas immune system that defends unicellular organisms from viruses (rath et al., 2015). importantly, these self-distinction processes, including membranes, are not in any organizational sense at the boundary of the operationally closed system; rather, they are part of it, they are in it. in sum, the status of the relations between different concepts, such as the systemic distinction enabled by operational closure, the structural processes that subserve self-distinction, the presence of actual spatial boundaries, and the conditional statistical independence between internal and external variables (markov blankets) is anything but “fairly straightforward” (ramstead et al., 2021, p. 555). di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 16 4 points of tension between the two approaches attempts to establish some sort of connection between fep and enaction continue to appear as we write these lines. some claims change in tone and detail, but so far we have not seen many cases in which the misreadings we have indicated are not present in some form or other. in this section, we raise points of deeper tension between the two perspectives. as stated in the introduction, our argument is that there are certain principles and phenomena that fep needs to account for if it claims to make contact with or encompass enaction, and that, to date, fep has not successfully accounted for these enactive principles and phenomena. there are reasons to think that it might be difficult for future developments of anything like fep to account for these enactive principles and phenomena, although we are not claiming any impossibility proof. let us clarify some terminology. the comparison between different theoretical approaches can be contentious precisely because it is difficult to agree on a common ground about what aspects should be contrasted. these aspects can include explicit and implicit assumptions, ontologies, types of causality, semantics, categorizations, standards, methodologies, formalizations, models, interpretations, explanatory goals, intended scope, and applications. the criteria for declaring two theories to be incompatible vary and can be contentious if we are not clear about exactly what we are comparing. to facilitate the reading of this section we will use the term tension to describe an apparent contradiction between theories. a tension might turn out to be a challenge, i.e., a contradiction that might be resolved by further work and conceptual coordination not involving a change to fundamental assumptions (e.g., corrections, extensions, clarifications, or making explicit something that was only implicit up to this point). we will qualify tensions as incompatibilities when, within the scope of our analysis, statements describing assumptions, categories, formalizations, and/or interpretations in each theory cannot be held jointly together. claims of incompatibility in this section can be contested by pointing to limitations of the analysis or by offering reformulations, possibly turning them into challenges. however, the point is that such reformulations would require major theoretical accommodation, such as abandoning basic assumptions. because of this, to show an incompatibility is not the same as presenting an impossibility proof, since the latter requires shared terminology and frameworks that may not be currently in place, as in the case here. a stronger difference between theories is that of incommensurability (kuhn, 1962). this refers to the difficulty of comparing the holistic nature of scientific theories (involving in addition to assumptions, formalizations, interpretations, standards, etc., also technological, institutional, economic, and cultural aspects). we do not attempt to claim incommensurability between fep and autopoiesis/enaction, but we do note that it is sometimes possible to appreciate a divergence in the style and orientation of discourses between these approaches. these lead to a difficulty in “neutrally” comparing assumptions, semantics, and formal aspects because these hold different value to different researchers, a difficulty we acknowledge in our case. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 17 our task is not helped by the ongoing revisions and refinements to the fep. to be as fair as possible, we will attend to the more recently published versions and variants of these claims. 4.1 non-equilibrium steady states versus history the work presented in friston (2013) is central to practically all the proposals comparing fep and enaction. it continues to be cited approvingly to this day. it has been argued that the formal argument presented in this work has technical problems (biehl et al., 2021; see also friston et al., 2021), and that when applied to concrete systems “the assumptions of the fep do not hold for a very broad class of systems, namely linear, weakly coupled systems, except for the limited case of fully symmetric agent-environment interaction” (aguilera et al., 2021, p. 18). we will not focus on these problems. our goal is to examine claims of compatibility with the enactive approach. to do this, it turns out that it is sufficient to examine the central assumptions of the fep because we already find tensions between the two theories at this starting point. here we question whether the convergence to non-equilibrium stationary states, or ness, which has been and remains central to the fep argument, is compatible with the theoretical claims of the enactive approach. the assumption we are concerned with is that the random dynamical system in question will converge to an invariant set of states: “because the system is ergodic (and weakly mixing) it will, after a sufficient amount of time, converge to an invariant set of states called a pullback or random global attractor” (friston, 2019, p. 10). in other words, the system reaches a ness with an invariant density of states (costa et al., 2021; friston, 2019; palacios et al., 2020; parr et al., 2019). this enables an important step in the formal derivation of fep mathematical relations. the invariant ness density describes the probability of visiting particular states even if the variables continue to change. as a consequence, the flow of the system can be expressed as a time-independent operation on this invariant probability; see e.g., equation 2.3 in friston (2013), equations 3.2 in parr et al. (2019), equations 1.8 in friston (2019), equations 2 in palacios et al. (2020). the time invariance of these expressions is key: setting “the rate of change of the density to zero” is “useful in formalizing the notion that a system maintains its form over time” (parr et al., 2019, p. 6). and the already quoted: “the internal and blanket states that constitute a subsystem are autopoietic, because their (nonequilibrium steady-state or ergodic) probability density is maintained by the flow of the subsystem’s internal and active state” (palacios et al., 2020, p. 5). the arrival at an unchanging set of relations describing the system in a ness serves as the conceptual link between fep and the properties its proponents associate with autopoiesis (self-organization, integrity, persistence, as we saw in section 3). di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 18 convergence to a ness is an important point of contrast between fep and enactive theory, one which few researchers have remarked upon. like organizational approaches in biology (e.g., longo & montévil, 2014), enaction emphasizes the historical nature of life and mind (even more than the classical theory of autopoiesis does, see di paolo, 2018; thompson, 2007; and varela, 2011)5. key technical enactive ideas, such as the concepts of adaptivity and sense-making, operate through mutually defined transformations of agent and environment that contradict the ness condition. these ideas theorize precisely what happens in transitions between different relatively metastable situations (which might in some cases be approximately stationary in themselves) or in enduring situations where metastable conditions are not reached for significant periods. varela uses the word “enaction” to refer precisely to such processes that happen “between one behavioral moment and the next” (varela, 1991a, p. 106); and di paolo et al. (2017) describe such concrete acts of sense-making as the open transition between different microworlds of activity. in that sense, history is built into the very definition of the term enaction. enactive ideas describe transitions between different regimes that typically require different dynamical descriptions, leading to changes in dynamical flows. in other words, these ideas concern what happens when a dynamical description suitable for one situation becomes unsuitable for the next; i.e., when we are no longer in a situation of time translation symmetry.6 we can say that as a matter of definition, enaction focuses its theoretical apparatus on the history of mutual change between agents and their environments and for this reason enactive ideas are in tension with the time-invariant distribution of states that obtains in the ness condition. we will argue next that this tension is in fact an incompatibility between the two approaches. 5in 1994, maturana and varela wrote individual prefaces to a new edition of de máquinas y seres vivos, the first book on their joint work on autopoiesis. in his preface, varela (2011) acknowledges that an important criticism of this early work was that autopoiesis seemed to imply a form of solipsism. maturana and varela had replaced the idea of mental representations with the idea that the environment triggers perturbations to the ongoing and operationally closed processes of the living system. this proposal included the notion of structural coupling, the idea that the living system and environment act as mutual sources of perturbation, triggering changes of state in each other. varela notes that this notion seemed to many researchers to be a weaker alternative to the input-output approach of the information-processing metaphor, “because it seems to leave the phenomenon of interaction in a grey area of being a ‘mere’ perturbation” (ibid. 614). structural coupling “does not properly take the account of the emerging regularities in the course of a history of interactions […] over these years i have developed an explicit alternative […] turning the historical reciprocity into the clue of a co-definition between an autonomous system and its environment. i propose to call this point of view in both biology and cognitive science, enaction” (ibid.). history and organism-environment co-definition have been explicit concerns of the enactive approach since its beginnings, and they constitute a change in emphasis and orientation with respect to classical autopoietic theory. 6time translation symmetry simply expresses that the laws of change describing a system remain unchanged if we apply a time displacement from t to t + dt. when talking about stationary regimes, distributions, densities, we use the term time-invariant, instead. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 19 let us first clarify what we mean by historicity in this context.7 historical processes and historical explanations abound in biology and the human sciences (e.g, arthur, 1994; byrge et al., 2014; fischer & bidell, 2006; gould, 2002; oyama et al., 2001; thelen, 2005; vygotsky, 2012; waddington, 1957). we associate historical processes with qualitative changes in the space of possibilities for a given system. these changes go beyond changes in states. in dynamical systems terms, historical changes can be described as changes in parameters, whole sets of variables, in the laws of change, constraints, and boundary conditions that result in qualitative dynamical differences (e.g., a reshaping of the attractor landscape due to new (de)couplings to other systems; the transformation of variable sets due to changes in the body, the adaptation to novel external constraints that induce a reduction in degrees of freedom in the system, and so on). we also expect historical processes to be dissipative and often to inhabit extended critical regimes (with scale-invariant distribution of changes, i.e., not only manifested at longer timescales), and to be subject to changing nonholonomic constraints leading to phase spaces under persistent change. phenomena that may be observed in historical processes can be hard to predict and include long-term memory, founder effects, locked-in retention, catastrophic forgetting, path-dependence, and broken symmetries of different kinds leading to diversity and variability (desjardins, 2011, 2015; di paolo, 2001; longo, 2018; longo & montévil, 2014; montévil, 2020). in sum and for our current purposes, historicity fundamentally entails broken time symmetries, i.e., changing laws, couplings, parameters, constraints, boundary conditions and variable sets. systems in a stationary condition, in contrast, cannot accommodate this kind of historical change: they remain in their stationary condition unless something knocks them out of it and eventually forget their history due to their time invariant properties.8 the unchanging density of states 7we follow authors such as longo, montévil, and others and use historicity to describe the historical character of some systems and processes, as elaborated in this paragraph. in other contexts, the term has a different meaning and refers mostly to the actuality or authenticity of historical events, persons, facts, etc. 8in this context, it is worth clarifying that models of active inference based on the fep can be applied to the tracking and representation of nested sequential structures (friston et al., 2018, 2020). these models can encode/generate relations in a hierarchy of nested sequences and in this way capture patterns over several timescales assuming a correlation between scale and model level. but the relation between these models and the question of historicity as defined above is, in our opinion, tenuous. these models describe the statistical relation of discrete nested sequences in order to generate a hierarchy of expectations to drive time-extended behavior, as in the case of reading a text, which may be described as organized hierarchically as sequences of sequences (letters, words, phrases, sentences, paragraphs, etc.). this does not entail that the system being modelled constitutes a historical process or that the model itself instantiates a historical process in the sense of changing spaces of possibilities and broken time-symmetries. we do not deny that deep temporal or sequential generative models may be useful in approximating historical changes, but they do so by tracking relations between nested sequences that are sufficiently close to being stationary. if the sequences change historically, say, as an infant transitions from crawling to walking (adolph & hoch, 2019), a nested sequential model tracking the muscle activation sequences of one kind of behavior does not by itself generate the new sequential relations needed di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 20 arrived at in fep systems is definitionally time invariant. this is explicitly stated in the technical literature. for this reason, the relations between variable sets derived from such a condition cannot accommodate the possibility of the broken time symmetries that define historicity. for example, the appearance of a novel constraint induces novel correlations between variables and these new correlations provoke changes in the flow, thus making the system abandon its stationary condition. in other words, the conservation of time invariant conditions cannot be guaranteed if time translation symmetry is broken. the tension between fep and historicity has already been noted by colombo & wright (2021; see also colombo & palacios, 2021). is this a minor tension or does it signal an incompatibility between the two approaches? we have already claimed that historicity can be said to be part of the very definition of the term enaction. in the rest of this subsection, we will show how important historicity is for enactive theorizing at all scales. historicity in life and mind can be found in many important cases that are central areas not only of enactive research but also of biology and psychology in general, such as the formation of sensorimotor habits (di paolo et al., 2017) and the formation of shared repertoires for coordinated action (di paolo et al., 2018). historical processes are also manifested at shorter, behavioral and neural timescales, as in cases of soft assembly and critical agent-environment integration (e.g., anderson et al., 2012), or the long-range coordination of neural populations at the moment of perceptual awareness (varela et al., 2001). at multiple scales, living agents are constantly undergoing regulation in relation to the set of constraints that redefine their structural dynamics (phase spaces) moment to moment. evidence shows that in living and cognitive systems organizations can endure even if structures change. without attempting to review the many cases in biological, neural, and cognitive systems where processes of historical transformation are at play, we can mention some examples. in biological processes: embryogenesis, life-cycle patterns, epigenetic variability, symbiosis, and metamorphosis.9 at behavioral levels: fluid, critical agent-environment integration across scales in the development of perceptual learning, skill acquisition, expert tool use, and habit formation (anderson et al., 2012). at the neurocognitive level: developmental plasticity, possibly many-many mappings between neural networks and cognitive functions (pessoa, 2014), and more generally what anderson (2014) calls “neural reuse,” the continued acquisition of new uses for neural circuits in evolution and development without those circuits losing their original uses. for the newer kind. at most, it can track them and reconstruct them a posteriori once the new situation is sufficiently stable. 9it is odd to find very few references in the fep literature to these cases and the potential challenges they offer to the premises of the approach. one exception is clark (2017), who realizes that cases such as metamorphic insects can be puzzling from a fep perspective. the solution, he suggests, is to look at whole life cycles as the free energy minimizing strategy of complex organisms. this idea, however, challenges the assumption of convergence to a ness, as clark himself notes in a footnote (16). di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 21 enaction attempts to accommodate many of these phenomena, so it is concerned with historical processes. is historicity itself to be found at the core of its theoretical apparatus? the possibility of historicity is introduced unequivocally and explicitly in enactive theory through the concept of adaptivity (di paolo, 2005, p. 444). adaptivity is the system’s capacity to regulate its states and relation to the environment in ways that result in the avoidance of trajectories that move towards loss of viability. this definition of adaptivity entails that adaptive interventions necessarily modulate the dynamical landscape, for instance, by changing parametric relations to the environment, (un)coupling to other systems, or altering constraints or boundary conditions. these changes must necessarily result in an altered phase space of the adaptive system. by definition, this is the only way in which it is possible to change the tendencies of dynamical trajectories bound to break through the viability boundary in a state-dependent system. to elaborate our critical point, adaptivity, as defined and used in the enactive approach, implies broken time symmetries and so, as we have explained, is incompatible with assuming that a system remains in a ness. time invariant properties of the flow cannot be maintained if parameters, constraints, and even the system’s variable sets are modified in the course of an adaptive event. thus, two ideas that are key for each approach (adaptivity and ness, respectively) contradict each other. adaptivity makes explicit the aspects of time direction and time granularity of agency (di paolo, 2005). an adaptive act is in itself not a conservation, but a modulatory deviation from an existing tendency that would lead to eventual loss of viability if left unchecked. an adaptive act takes time, has a particular timecourse with different phases, and must occur within appropriate time constraints (given by requirements of speed, deadlines, by the relevant embodied and social norms, and so on). in important ways, things are just not the same before and after an adaptive intervention. history is made possible by these properties, especially as organisms become more complex, and adaptive acts relate to one another in networks of mutual influence and triggering, affecting, moment to moment, the relational constraints that shape the dynamics of the agent. the historicity of biological systems (longo & montévil, 2014) is precisely the ongoing broken symmetries in dynamical constraints, parameters, and even variable sets throughout a lifetime of changing mutual dependencies between organism and environment (including other organisms). historicity is not just present in enactive technical concepts but also elaborated in enactive research, particularly to theorize minimal agency, mastery of sensorimotor contingencies in action and perception, and the development of sensorimotor repertoires as the theoretical basis for the concept of sensorimotor agency and linguistic bodies (di paolo et al., 2017, 2018). one of the definitional requirements for agency proposed by enaction is that of interactional asymmetry. in dynamical systems terms, this requirement is meant to capture that not only are agent and environment coupled systems, but also that processes in the agent can trigger modulations (changes in parameters and constraints) of this coupling. again, such modulations of parameters and constraints cannot guarantee that dynamical di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 22 flows will remain in a ness. they typically will not, and these changes are what allow the system to avert the potential loss of viability were it to remain in such a regime. as in the case of widespread extended critical transitions in biological systems (e.g., longo & montévil, 2014), the regulation performed in the acts of an agent can give rise to path-dependent changes on the constraints affecting the dynamical flows of the agent as a whole, essentially altering the configuration of the phase space of organism and environment, and even rendering notions such as attractive sets and ness inadequate. the enactive concept of agency reaffirms for cognitive systems what longo and montévil predicate about biological systems: “biological processes are more ‘history based’ than physical [non-biological] processes. usual physical processes preserve invariants, whereas extended critical transitions [characteristic of biology] are a permanent reconstruction of organization and symmetries, i.e., of invariants” (ibid, p. 175).10 as far as human beings are concerned, evidence strongly suggests that processes of historical change cannot be disregarded. ergodic and stationary conditions in psychology and neuroscience may sometimes make methodological sense in laboratory situations. but such conditions are not always made explicit and these assumptions can lead to systematic problems in interpreting the generality of empirical results or making group-to-individual inferences (fisher et al., 2018). indeed, several researchers interpret the current crisis of replication in psychology precisely as a crisis provoked by assuming ergodicity by default, a practice that be10it is important not to misread this quotation as stating that historical processes can occur only in biological systems. broken symmetries occur in all kinds of physical systems, but usually in the form of single critical transitions. longo and montévil argue that biological systems are characterised by continuous extended critical transitions, and this is what makes them historical: “in our approach to biological processes as ‘extended critical transitions,’ ‘extended’ means that every point of the evolution/development space is near a critical point. more technically, at the mathematical limit, the critical points form a dense subset of the multidimensional space of viability for the biological process. thus, criticality is extended to the space of all pertinent parameters and observables (or phase space), within the limits of viability […] in terms of symmetries, such a situation implies that biological objects (cells, multicellular organisms, species) are in a continual transition between different symmetry groups; that is, they are in transition between different phases, according to the language of condensed matter. these phases swiftly shift between different critical points and between different physical determinations through symmetry changes” (longo & montévil, 2014, p. 173). an example of extended criticality are the broken symmetries induced by the millions of mitosis events a multicellular organism undergoes every day where the resulting distributions of proteins, etc. are not identical and can lead to different integrated cellular dynamics which play a role in the development of different cell types and tissue differentiation. dense sets of critical transitions make it difficult to explain biological systems via minimization or maximization principles: “as a further consequence of our approach, phylogenetic or ontogenetic trajectories cannot be defined by the geodesic principle. indeed, they are not theoretically determined by invariants and their associated symmetries. trajectories are continually changing in a relatively minor but extended way. moreover, we expect the rate of these changes themselves not to be regular with respect to physical time, so that some temporal region can be ‘calm’ while others correspond [to a] sudden burst of changes” (ibid.). the existence of “critical surfaces” in relation to conservation of integration/viability can be shown in relatively simple models of information integration under environmental diversity (aguilera & di paolo, 2019). di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 23 lies the lack of an underlying theory that would grant validity to the assumption for each given case rose et al. (2013). for instance, consider the formation of habits. we could say that there can be no habit without history, or that history is constitutive of the concept of habit. enactive literature is very clear about this point: “to say that the habitual body acts as guarantee for the body at this moment is to say that one’s lived body is a developmental being thick with its own history and sedimented ways of feeling, perceiving, acting, and imagining. these sedimented patterns are not limited to the space enclosed by the body’s membrane; they span and interweave the lived body and its environment, thereby forming a unitary circuit of lived-body-environment” (thompson, 2007, p. 33). importantly, history and contingency lead to diversity, which a theory of human bodies and minds cannot afford to leave without a proper theoretical grounding. the beginnings of such a theory are offered in (di paolo et al., 2018). cultural and interpersonal variability in human beings are not statistical noise, but rather necessary consequences of what it means to undergo human becoming. some events may enter the developmental, cognitive, and emotional history of a person as accidents, but if their effects endure or get amplified, this occurs by triggering a swerve in dynamical paths and reshaping dynamical landscapes. fluctuations can become locked-in and change the course of subsequent history. none of this entails a loss of the system’s integration, but it may entail transformations that redefine skills, sensitivities, meanings, and norms. widespread biological and psychological phenomena of the kind fep proponents intend to cover with their theory, (e.g., veissière et al., 2019) are inherently historical. the enactive approach provides accounts of perception and perceptual learning in terms of mastery of sensorimotor contingencies linking multiple scales from neural to developmental processes. mastery, or acquisition of a know-how, is operationalized using a dynamical theory of equilibration that does not require systems to be in a ness to work (di paolo et al., 2017). in fact, one of the phases of equilibration, accommodation, demands plastic changes in the agent and the environment such that the corresponding phase spaces are modified as a consequence, otherwise equilibration could not occur. phenomena such as developmental spurts in skill level (fischer & bidell, 2006) are large scale manifestations of such changes in dynamical configurations (novel constraints, emergent parameters, changing variable sets). the variability entailed in changing dynamical configurations has been postulated as the origin of motor creativity (orth et al., 2017), the very idea of which is rendered problematic without an account of historical change. all of these processes can be accommodated by the enactive theory of sensorimotor learning, which can also account for the existence of multiple developmental pathways, path-dependence, and the intrinsic variability found in all kinds of human skills (adolph & hoch, 2019; kostrubiec et al., 2012; thelen, 2005; thelen et al., 1996). a central claim of the enactive approach is that sense-making is the activity of an autonomous agent that contrasts with heteronomous information processdi paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 24 ing. to illustrate the sort of neural processes that underpin sense-making, varela (1991a) and thompson (2007) make repeated references to the work of walter freeman, who explicitly acknowledges history-dependence in the most basic neuroscientific scenario of stimulus processing: “the emerging [neural] pattern […] is a state transition that is induced by a stimulus, followed by a construction of a pattern that is shaped by the synaptic modification among cortical neurons from prior learning. […] owing to dependence on history, the patterns created in each cortex are unique to each subject” (freeman, 1999, pp. 149–150). such history dependence of neural processes has been recognized as posing a methodological challenge of non-ergodicity for neuroscience (e.g., medaglia et al., 2011). historicity is manifested over very different timescales, not just the longer timescales associated with development. in discussing the fine temporal structure of cognitive action, varela proposes that coherent patterns of fast neural oscillations emerge at moments of concretion (significance) in action and perception (varela, 1991a). this idea is empirically supported by studies of long-range neural (de)synchronization at moments of perceptual awareness (recognition) and action initiation, evidencing emergent processes that rapidly constrain and free dynamical flows contingently on the actions of the perceiver and her situation (varela et al., 2001). this perspective on brain function puts the emphasis on neural processes radically altering their dynamics in fluid and adaptive ways so as to meet the demands of a concrete sociomaterial situation (e.g., anderson, 2014; fuchs, 2017; pessoa, 2014). such rapid moments of high dimensional dynamic expansion followed by lower dimensional coordination is one of the meanings given to the term enaction itself, as we have said. accordingly, “the hinge that articulates enaction consists of fast non-cognitive dynamics wherein a number of alternative microworlds are activated. these hinges are the source of both common sense and creativity in cognition” (varela, 1991a, p. 109). the latter remark indicates that these fast neural processes are not just history-dependent (“common sense”) but also history-making (“creativity”). the idea of inhabiting and transitioning between microworlds has been further developed by di paolo et al. (2017), where the authors discuss the possibility of extended periods of high-dimensional exploration of sensorimotor possibilities when the transition between activities does not occur or for whatever reason is arrested. examples include periods of hesitation or confusion as to the relevant activity one is engaged in, either by lack of determination (not knowing exactly what comes next out of several options) or overdetermination (current activity continues pre-reflectively while we are reflectively aware that it is no longer the relevant activity). extended exploratory periods that do not settle into metastable modes of activity also include situations in which the required adaptations to some radical sensorimotor disruption have not yet been achieved (changing habits, experiments involving sensorimotor alterations, major injury, trauma, etc.). the authors call these moments of uncommitted dynamical engagement the zero-mode of activity. using a network metaphor to describe the functional and structural relations found di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 25 within a repertoire of sensorimotor schemes, the authors remark “through regional developmental differentiation and integration of schemes, integrated subnetworks can be formed that correspond well to notions such as activities, microworlds, and sensorimotor genres. the structure of the sensorimotor network reflects the history of the agent. in humans the possibilities are open-ended and path-dependent, as we would expect, leading to a way of characterizing otherwise vague concepts such as sensorimotor styles” (di paolo et al., 2017, pp. 177–178). in sum, whether at biochemical, neural, behavioral, or developmental scales, enactive ideas all share in common the expectation that biological and cognitive systems are historical, changing in path-dependent ways, not only at some points in time, but regularly, at different scales, and throughout their lifetime. for this reason, historicity is definitionally built into its key concepts (enaction, adaptivity, agency, sense-making). history is also a thread that drives current directions of the enactive approach (e.g., see recent articulations of the notion of open-ended human becoming in de jaegher, 2021; and di paolo, 2021). could this incompatibility between key ideas in each approach be turned into a reconcilable tension, i.e. a challenge for fep? a possible attempt might take the form of an argument that renders “historical” changes as only apparent (i.e., not really involving broken time symmetries) and explains them as transitions between existing and pre-established “micro-regimes” within what is otherwise a system in a ness (when seen at appropriately long scales). this is similar to clark’s (2017) idea to regard whole life cycles as a free energy minimizing strategy. thus a transition between one microworld and another could be described as a switch between metastable “micro-regimes” within the larger, lifetime ness. this explanation could accommodate some cases of switches between routine activities, but if it were generally the case, it would imply that all of our activities are routine, statistically speaking. the stationary condition must translate into a steady distribution of transits between “micro-regimes.” engaging in novel activities (e.g., learning to drive) or abandoning old ones (because of loss of interest or other factors), however, cannot be explained in this way. the idea of a whole life cycle characterised by a ness does not seem applicable to the open-endedness of the human case. another way a fep account might accommodate historicity could be by not assuming that the system overall must behave in a stationary manner, but that it frequently inhabits different steady-state “micro-regimes” and sometimes is able to transit from one to another without the whole set of steady states being predetermined or fixed over time. this would free the account from the requirement that transitions between “micro-regimes” must also follow a stationary distribution because stationarity is only required locally within a “micro-regime.” but such a refinement still would not answer the question of what exactly happens during such transitions, what triggers them, how frequent they are, how long they are expected to last, what norms regulate the adaptive mechanisms at play in them (since free energy minimization is not guaranteed during a transition), and so on. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 26 moreover, while this idea may bring the fep closer to some aspects of history, it would not be suitable for the extended criticality and dense transitions sets postulated by longo & montévil (2014) as characteristic of biological systems, which in some cases exhibit symmetry breaking of the kind that would not even guarantee steady-state “micro-regimes.” we repeat our point that given the current state of affairs, and the respective importance of the ness condition for fep and of historicity for enaction, this tension counts as a theoretical incompatibility (in the sense that one cannot assert both simultaneously). if new variants of fep were developed replacing the need for the system to converge to and remain in a ness, the situation might change. but the problem of accounting for the historicity of life and mind cannot be solved by admitting that the need for a ness condition is merely an idealization that can sometimes be relaxed or is relative to a timeframe of relevant observation (e.g., wiese & friston, 2021, p. 7). to admit this is, again, to concede that the fep does not have the claimed status of a universal principle, since its domain of applicability would remain limited to relatively simple cases. such a move begs the question of what happens to biological and cognitive systems at the moments the steadystate assumption does not hold, moments that, as we have argued, are copiously found at all scales of biological and psychological phenomena. a theoretical biology and a theoretical psychology should be able to answer when this assumption is sufficiently valid for its application to be useful and what happens when it is not met. if important phenomena (e.g., ontogeny, acquired immunity, the time course of illness or injury and recovery/compensation, skill acquisition, habits, development towards emotional maturity, trauma, personality, language, abstract thought, social interactions, expert use of tools, cultural history, evolution of technology, changes in modes of production, and innumerably other path-dependent processes) fundamentally break the time invariance of ness, it does not follow that we should discard fep ideas as useful research tools, provided we can specify the conditions under which their use makes sense. but, as currently formulated, fep does not answer this question by itself. it would be as if we were expected to deduce the shape of the earth from a theory that assumes its surface must be flat just because under some conditions this is a good approximation but without specifying what these conditions are or what happens when they are not met. the necessary specifications must come from a theoretically-loaded account that tells us when to expect assumptions that lead to a ness to hold and when not to expect them. elaborations to the existing formulation of the fep may provide some answers, but they may require seeking different, perhaps less restrictive starting points. it goes without saying that the same theory cannot adopt an assumption as fundamental and deduce when it does not hold as a corollary. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 27 4.2 mutual enablement and co-constitution of internal and external domains in section 3, we noted the recurrent and puzzling confusion between markov blankets and what autopoietic and enactive theories specify as operationally closure, and in particular processes of organizational self-distinction. these processes of self-distinction do occasionally—but not exclusively—take the form of processes of structural boundary formation, regulation, and maintenance. they can also take other forms (e.g, immune responses against invading pathogens, temperature regulation in warm-blooded animals, musculoskeletal systems of tension and compression maintaining bodily shape). in all instances they involve adaptive regulations of exchanges between organisms and the environment. we can ask whether the confusion between such processes and markov blankets is symptomatic of another tension between the two approaches, a tension in how they conceive of relations between organism and environment and relations between different organisms; whether these relations are merely informational (i.e., contextual), or can also be transactional (e.g., mutually enabling) or even constitutive of biological and cognitive/affective processes. enactivists assert a strong notion of world-involvement, i.e., processes in the environment play more than informational roles in the constitution and actualization of life and mind (di paolo et al., 2017; thompson & stapleton, 2009). to enact a world of significance is to engage in actual acts, which are material events with spreading consequences that are both world-changing and agent-changing. environmental and biological/cognitive processes are mutually enabled and mutually constituted. they interpenetrate at all scales and they coordinate across scales. can these claims be accommodated by the idea of markov blankets?11 here we will re-examine the concept and processes of self-distinction and why these processes should not be quickly equated with markov blankets. we will argue that many important exchanges between organisms and the environment do not follow the statistical constraints of markov blankets and can in fact be unmediated or unconditioned, only to be dealt with adaptively after the fact. after this, we turn to what enactivists describe as world-involvement, i.e., the ways in which relational processes involving both organism and environment play enabling and constitutive roles in the activity of an agent as well as in its ongoing individuation. these relational processes affect the operation of this activity and contribute to establishing the norms of viability that this activity follows. in particular, they help establish which encounters can or should be mediated by states of the organ11for recent critical discussions about markov blankets, see bruineberg et al. (2021) and raja et al. (2021). bruineberg and colleagues argue that a simple statistical idea applied in bayesian networks and describing how variables may be shielded from variations in other variables has been made to do some heavy conceptual lifting in the fep framework, being used, as we discuss in this paper, to play the role of boundaries and sensorimotor interfaces. raja and colleagues, similarly note this problem and the fact that the drawing of a markov blanket is often ad hoc, rather than following a consistent method. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 28 ism. for this reason, world-involving relational dynamics operate at a level prior to markov blankets and cannot be established by them. in other words, these two points indicate that markov blankets insufficiently specify what enactivists highlight about the relation between organism and environment. it is possible, however, that the tensions we discuss in this subsection may take more the form of a challenge than that of a strict incompatibility, as we shall see. historicity and the co-constitution of organism and environment are internally related in the enactive approach. concerns about the conservation of organization are mostly linked to the self-production requirement of autopoiesis (the regeneration of the conditions that continuously give rise to the operationally closed network of processes making up the organism). concerns about barriers, boundaries, and in general about an organism’s relation to its environment are mostly linked to the condition of self-distinction in the definition of autopoiesis. from an enactive perspective, self-distinction and self-production are dialectically related (di paolo, 2018), that is, they are mutually dependent, though distinct, moments of autopoiesis (thompson, 2007; varela, 1991b). you cannot have one set of processes and not the other as long as the organism lives, yet the processes are not the same. all processes subserving self-distinction are themselves products of self-production. in contrast, markov blankets in fep systems are there by assumption.12 in other words, markov blankets are not produced by the system the same way that self-distinction processes and structures are. hence there is nothing in the markov blanket that necessarily links it to processes of organismic constitution. this is a discrepancy also noted by raja et al. (2021), who argue that the choice of where a markov blanket should be is rather ad hoc and follows the convenience of each case (see also bruineberg et al., 2021). this is not a problem in itself; it might even be an advantage in some cases. but it points to a difference between fep and enaction. 12the assumption is considered self-evident: “clearly, one needs to differentiate between the system and its environment […] to do this, we have to introduce a third set of states that separates internal from external states. this is known as a markov blanket.” (ramstead et al., 2018, pp. 3–4). this explains the wide applicability of markov blankets in examples and models of fep, sometimes taking the role of boundaries, other times of action and sensory states, and so on. as raja et al. (2021) argue the choice of where a blanket is located seems to be relatively arbitrary and made to fit the convenience of an fep interpretation. in an attempt to provide a more principled interpretation of markov blankets, friston (2019) associates them with the structured (ordered) flows in far-from-equilibrium dissipative structures (nicolis & prigogine, 1977). but he acknowledges that this poses a problem for the fep formalism, because such structures are in constant material and energetic flux, which contradicts the assumption of a random dynamical attracting set, and confirms that the two central assumptions of fep are independent, unlike self-production and self-distinction for enaction. friston considers this problem an unresolved challenge for fep and its generalization to so-called wandering sets (which would not suffice to address the cases of historicity discussed in this section, as these include the possibility of changing sets of variables, such as emerging novel agent-environment relations). the whole formalism, as it currently stands, rests on “the simplifying assumption that over a suitable time scale, blanket states are well defined—as a subset of attracting states” (friston, 2019, p. 50). di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 29 we may still ask whether markov blankets, even if they are not produced by the organism itself, could still provide a sufficiently informative description of the organism-environment relation. in response to this question, we find that selfdistinction is not sufficiently specified by a statement about statistical dependencies or sparsity of connections between variables, nor does it entail any such statement. let us see why. we confront a semantic difficulty at this point. a markov blanket helps define the sets of internal and external variables in the fep. what counts as internal and external variables in the enactive approach is less straightforward. it is in principle possible to identify what processes belong or do not belong to the operationally closed network of the organism (though in practice there may be some ambiguities; consider for instance the layer of warm air that surrounds the skin or fur of warmblooded animals). but this distinction does not translate into a straightforward topological relation as might be postulated by pointing to boundary processes such as membranes, cell walls, skin, etc. the network of operationally closed processes can extend beyond these boundaries and environmental processes can transverse them. there is a semantic difficulty in comparing apparently simple terms such as “internal” and “external” in the two theories. from an enactive perspective and for the purposes of this discussion, we propose to treat as internal those variables that are topologically contained by an organism’s boundary surfaces in cases where these are well-defined by boundary processes of self-distinction, and as external those variables that lie outside boundary surfaces. the meaning approaches the usage in the fep literature but is not exactly the same, because this distinction between internal and external does not exactly map to that between organism and environment in the enactive view, nor does it presuppose any statistical relation between internal and external variables. keeping this in mind, we repeat that self-distinction is not supported structurally only by the formation of organismic boundaries. “non-boundary” processes of self-distinction include the immune system, crispr-cas anti-virus defense, sweating, shivering and other processes that sustain temperature regulation, processes of maintenance of shape, such as cytoskeletons, musculoskeletal tensegrity, tissue formation and regulation of the extracellular matrix, and so on. in other words, they include internal processes (in the sense just defined). because these “non-boundary” self-distinction processes do not lie at the topological interface between organism and environment, they do not generally entail any particular kind of conditioned independence between internal and external variables; on the contrary, they are regulations of encounters between internal and external processes, where internal effects can follow external factors quite directly. even if they operate internally, these processes of self-distinction adaptively regulate the relation between agent and environment. however, because they operate internally, they seem difficult to describe by markov blankets. let us consider “boundary” processes of self-distinction, such as cell membranes. we find a large number of processes that cross boundaries and barriers di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 30 in both directions without any statistical mediation. for example, osmosis and other forms of diffusion, the ergodicity-breaking formation of protein nanoclusters affecting ion-channels in neurons (weron et al., 2017), horizontal transfer of plasmids in bacteria (soucy et al., 2015), the cumulative effects of diet and adaptation to toxins or their expulsion (landecker, 2011), direct energy transfer in photosynthesis, temperature equilibration and temperature-induced neural plasticity in cold-blooded animals (e.g., beltrán et al., 2021), and “externalized” physiological processes such as the extraction of gaseous oxygen underwater by insects trapping air bubbles in their abdomen (e.g., turner, 2000). not to mention more complex processes such as cell division, sexual reproduction, host-parasite relations, gut microbiota, and symbiosis in general. the list does not include external processes that are hard or impossible to regulate, such as the effects of ionizing radiation, gravity, or unavoidable physical impacts, but that nevertheless affect internal variables directly. external processes affect internal variables without mediation in all of these cases. this range of exchanges between internal and external processes does not pose a problem for enaction. some of these exchanges are avoided or rejected, some are mediated by the activity of the organism, others remain unmediated in the encounter and regulated only in their effects. the condition of operational closure is not necessarily lost even in the latter case, provided the effects of an unmediated exchange can be adaptively assimilated (e.g., a fever response to bacterial infection). these exchanges, however, do not require that internal and external variables relate through a conditional statistical independence and in fact in many of the cases mentioned, this will not be the case. we suggest that processes of self-distinction, as conceived in the enactive approach, do not generally behave like markov blankets do. agency, in the enactive view, involves not just the regulation of boundaries, but also the a posteriori regulation of more direct, boundary-crossing influences of the world, by taking advantage of these influences when useful (e.g.. nourishment) and compensating for their negative effects when not (e.g., fever response). from an enactive perspective, markov blanket conditions are too strong and too undiscriminating to account for the complexity of the organism-environment relation, and this is a point of tension. a related point of tension concerns world involvement, a term enactivists sometimes use to capture the relational aspects of the agent-environment coupling (di paolo, 2014; di paolo et al., 2017). world involvement stresses the fact that the environment is involved in what an agent is and what it does in more than contextual ways, that is, its involvement goes beyond being a source of information. the world is also a material enabler of cognitive acts and may also play constitutive roles in such acts, particularly in contributing to determine their normative aspects (what counts as an act succeeding or failing, what makes an act preferable to another).13 13the distinction between contextual, enabling, and constitutive factors was introduced by de jaegher et al. (2010) to ask the question of whether the dynamic patterns that we observe in di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 31 the world enables and constitutes embodied agency over and above any information it provides to the cognitive processes of the agent. this is not to say such information is irrelevant (provided we describe clearly what we mean by this term, and provided we can establish what counts as the frame of relevance in each case, something that can change historically). it means that other, non-informational, aspects of material processes are also enabling and constitutive factors in cognitive acts and in sustaining the autonomy of the agent. this is what world involvement means. going on a diet, training for a new skill, and so on—the environmental processes in such activities do not merely inform our bodies through sensory-screened inputs about how they should change as a consequence, more often than not they change our bodies directly. drinking a glass of water quenches our thirst in a way that is different from encountering a useful piece of information. a zero-g environment does not inform the body of a cosmonaut that it should grow less dense, it “makes it” do so. in addition to these enabling and constitutive roles in the activity and ongoing self-constitution of the agent, the relational processes of world involvement also contribute to the establishment of the norms of an agent’s activity, as well as its conditions of viability. for instance, by engaging in a novel manual skill, we learn through confronting obstacles to accommodate new norms that emerge from these material encounters, as in the different ways an expert potter prepares different kinds of clay. we learn to establish a new frame of relevance as a historical transformation of our bodies and our relation to the environment. this frame is necessary to determine what counts as meaningful information (e.g., cues we used to be indifferent to, but now tell us whether this is good wood for sawing, or well-prepared clay for shaping a pot, or the age of a wine from its bouquet). relevance, meaning, norms are irreducibly relational properties. because they establish what counts as informational for the agent, these material and relational transformations are prior to any informational account we can provide. both of these points—the enabling and constitutive roles environmental processes play in sense-making and in agent constitution, and the role they play in establishing norms for the activity of an agent—indicate tensions with the idea of markov blankets. these tensions, however, partly rely on how we choose to define internal and external processes. for this reason, it might be possible to work on interpretations of enaction and fep that avoid these apparent contradictions, so these tensions are at least challenges to the fep; we do not claim they are incompatibilities. social interactions can (sometimes) be said to be constitutive of social cognitive performance, i.e., whether there is social intelligence in the dynamics of the interaction itself and not just within the interactors. a contextual factor merely modulates a phenomenon under observation; it changes it, but is not required for the phenomenon to exist or to be the kind of phenomenon that it is. an enabling factor is one without which the phenomenon would not occur. a constitutive factor is one without which the phenomenon would not be the kind of phenomenon that it is (see also de jaegher et al., 2016). di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 32 these are not minor challenges, though. they can lead to divergences in how each approach conceives action and perception. the enactive approach postulates a theory of sensorimotor mastery (know-how) based on equilibrated relations between sensorimotor schemes (di paolo et al., 2017). these schemes are normatively integrated flows of coordinated body and environmental processes in coupling (e.g., the movement of water from a glass into the mouth, the activation of swallowing muscles, its flow within the body, and so on). they combine internal and external processes. because traffic across a markov blanket is conceived as mediated by sensory and effector states there is a tendency of conceiving of the organism-environment relation in informational terms and this is verified by terms such as evidence, surprise, inference, etc. this promotes (not necessarily compels) internalist views of sensorimotor know-how that leave out precisely the transformative effects of the world on our bodies. an informational interpretation assumes that such effects have already occurred, such that the norms that determine what counts as relevant information are already in place and the relevant distinctions already established. but explaining how we can reach this stage is what enactive theories of action and perception are about. for a theory of perception inspired by free energy minimization, such as seth’s (2014), embodied factors such as sensorimotor contingencies play a role in the informational economy of hierarchical predictive models, structuring data, selecting sampling strategies, and generating error-based corrections. such proposals remain all “in-the-head” and reduce the rich materiality of the worldly constituents of perceptual experience to error signal generation (di paolo, 2014). for the enactive approach, in contrast, it is from the material constraints introduced by worldinvolvement in action and perception that the norms of such activities arise (what to do? what counts as doing it right?). these emerging norms result from agentenvironment transactions and crucially frame what counts or does not count as relevant information for the agent and not for the external observer; what is relevant and what is not, what needs to be optimized and what may be safely ignored. in this sense, the mastery of these norms, a transactional process not bound by the brain, is constitutive of perceptual experience (di paolo et al., 2017). whatever one may think of the merits of these theories, the fact remains that they are different in important aspects (this is sometimes acknowledged by those who would still try to reconcile them; see clavel vázquez, 2020). 5 conclusions it is important that our motivation in this article should not be misconstrued as a general statement against intellectual synthesis or against the cross-fertilization of ideas. enactivists generally welcome and celebrate such exchanges. nor is our discussion an exercise in gatekeeping. the enactive approach has undergone important changes over the last decades and will continue to evolve. in part, these changes have resulted from the meeting (sometimes the confrontation) of enactive di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 33 ideas with other theories, disciplines, and practices. a case in point is the ongoing discussion of the relations between enaction and ecological psychology (chemero, 2009; di paolo et al., 2021). nevertheless, the process of inter-theoretical comparison cannot be bypassed or replaced by rhetorical devices, such as stating that some similarities are “fairly straightforward” or by cherry-picking what to keep and what to ignore about a theoretical framework. attempts at discussing the compatibility or complementarity between different theories are initially motivated by what look like interesting similarities and resonances between them. this is fine. but the work of inter-theoretical comparison cannot stay at the level of pointing out these similarities. it requires critical examination to determine whether similarities are only apparent and whether they are not based on misinterpretations. this work demands a deeper and detailed engagement with the assumptions behind the theories being compared, as well as their contexts and goals. despite its abundance, the literature defending compatibilities between the fep and autopoietic and enactive theories has advanced very little beyond bald statements of apparent similarities between these approaches. such statements remain at the beginning of the work that needs to be done. in this article, we have shown that once we take the required steps of checking for misinterpretations and comparing fundamental assumptions, the fep, as it stands formulated today, presents a series of tensions with the theories of autopoiesis and enaction. not only do we find misreadings—such as confusing statements about organization with statements about structures, or conflating notions of self-distinction with material boundaries and the latter with markov blankets— but we also suggest that at the root of these misreadings may lie conflicting views about the constitution of agency and self-individuation in biological and cognitive systems. our conclusion is that, for these reasons, the fep—at least in its present form—and the enactive approach are not as compatible as has been suggested. the inter-theoretical differences we have shown are not minor, and spread respectively through each project as a whole. in some cases, we argued that apparent contradictions can be shown to lead to incompatible statements and claims. in other cases they at least indicate a challenge for proponents of fep interested in connecting to enactive theory to develop extensions, clarifications, or make explicit how the tension can be resolved. in particular, we have shown that enaction conceives of bodies and sensemaking as irreducibly historical (both in the sense of depending on the past and in the sense of changing, or enacting, the present) and co-constituted by the sociomaterial world in ways that go far beyond the processing of information. the centrality of the historical transformation of agents and worlds for the enactive approach is in plain view and a definitional matter in its technical concepts. it is also a direct reading of its classical slogan: laying down a path in walking (thompson, 2007; varela et al., 1991). in contrast, the key assumptions of the fep, such as di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 34 ness, make it difficult for it to reach a similar conception of life and mind, as we have shown. because of this, we consider this to be an incompatibility between the two approaches. other tensions turn around how each theory conceives of the relation between agent and environment. we have suggested that markov blankets do not seem to describe the possible interactions that can be accommodated by the enactive idea of operational closure, particularly those that take place “internally.” however, this tension revolves around tricky semantic issues such as definitions of seemingly obvious terms such as “internal” and “external.” if semantic interpretations were modified, then this tension might turn out to be simply a challenge and not an incompatibility, so for now, we take it as such. there are other differences between the two approaches that we have not explored here. for example, there are aspects of the temporality of sense-making and lived experience apart from historicity that can be said to be in tension with notions such as predictions and active inference. sense-making by its technical definition is future-oriented activity, but this orientation can take a variety of forms, from general states of readiness, trust, or alert and their opposites, to conditions of commitment to action and focus or states of openness at the transition between activities. of course, sometimes sense-making can also take the form of expectations, inferences, and predictions. but the varieties of future orientation do not reduce to the latter cases only. establishing this point, however, would require careful conceptual, phenomenological, and terminological analysis, which we cannot pursue here. another difference we have not explored concerns the social dimension. human bodies and minds, apart from being constitutively historical and diverse, are irreducibly social. any approach that hopes to provide explanations for human cognition must acknowledge this fact and supply the right theoretical articulations to work with it. enactivists have proposed the concept of participatory sense-making (de jaegher & di paolo, 2007) as the keystone for developing an enactive approach to social, linguistic, and ethical agency (di paolo & de jaegher, 2021). this idea, in our view, is at odds with the notions of blanketed individual persistence of cognitive individuals at the heart of the fep. in participatory sense-making, actions and intentions become socially constituted and transformative of individual agency; they involve not only other participants but also the relational dynamics of the interaction (de jaegher et al., 2010). we suggest that it would be difficult to account for these mutually transformative processes if each interactor were subject to the conditional statistical independence that is forced on them by markov blankets and the need to remain in a ness during the participation process. again, a full argument, which space prevents us from providing here, would be necessary to establish this point and decide whether it is only a challenge or an incompatibility. we have not dwelled on the question of what motivates researchers to attempt to make the fep compatible with enaction and autopoiesis. one driving motivation seems to be the need to provide more embodied and less neurocentric interpretations of free energy minimization and active inference. although this is a di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 35 good motivation, we think that elaborating this interpretation is an uphill task, in no small measure because of the neurocentric and computational origins of fep and predictive processing. a lot of intellectual effort in the literature that we have discussed is spent, almost apologetically, in revisionist attempts at reinterpreting computationally loaded ideas, such as model, inference, and prediction, in terms of embodiment, dynamics, and agent-environment couplings. having a theoretical approach at hand, such as enaction, that already starts from an embodied and situated perspective could perhaps have given the impression that the uphill task was achievable by establishing a few bridges between the theories. we have argued that things are not so easy. the styles of thinking are different and the ideas difficult to harmonize. whether there are other ways to make the fep coherently more embodied is not a question we have set ourselves to answer here. acknowledgments we are grateful to miguel aguilera, maxwell ramstead, and the reviewers of this paper for their comments and suggestions. we also thank wanja wiese for his editorial advice and support during the review process. references adolph, k. e., & hoch, j. e. (2019). motor development: embodied, embedded, enculturated, and enabling. annual review of psychology, 70(1), 141–164. https://doi.org/10.1146/annurev-psych-010418-102836 aguilera, m., & di paolo, e. a. (2019). integrated information in the thermodynamic limit. neural networks, 114, 136–146. https://doi.org/10.1016/j.neunet.2019.03.001 aguilera, m., millidge, b., tschantz, a., & buckley, c. l. (2021). how particular is the physics of the free energy principle? physics of life reviews. https://doi.org/10.1016/j.plrev.2021.11.001 allen, m., & friston, k. j. (2018). from cognitivism to autopoiesis: towards a computational framework for the embodied mind. synthese, 195(6), 2459–2482. https://doi.org/10.1007/s11229-016-1288-5 anderson, m. l. (2014). after phrenology: neural reuse and the interactive brain. mit press. anderson, m. l., richardson, m. j., & chemero, a. (2012). eroding the boundaries of cognition: implications of embodiment. topics in cognitive science, 4(4), 717–730. https://doi.org/10.1111/j.1756-8765.2012.01211.x arthur, w. b. (1994). increasing returns and path dependence in the economy. the university of michigan press. ashby, w. r. (1962). principles of the self-organizing system. in h. v. foerster & jr. g. w. zopf (eds.), principles of selforganization: transactions of the university of illinois symposium (pp. 255–278). london: pergamon press. baltieri, m., buckley, c. l., & bruineberg, j. (2020). predictions in the eye of the beholder: an active inference account of watt governors. in j. bongard, j. l. l. hebert-dufrsne, r. dasari, & l. soros (eds.), proceedings of alife 2020: the 2020 conference on artificial life (pp. 121–129). mit press. https://doi.org/10.1162/isal_a_00288 barandiaran, x. e., di paolo, e. a., & rohde, m. (2009). defining agency: individuality, normativity, asymmetry, and spatiotemporality in action. adaptive behavior, 17(5), 367–386. https://doi.org/10.1177/1059712309343819 beer, r. d. (2015). characterizing autopoiesis in the game of life. artificial life, 21(1), 1–19. https://doi.org/10.1162/artl_a_00143 beer, r. d. (2020). bittorio revisited: structural coupling in the game of life. adaptive behavior, 28(4), 197–212. https://doi.org/10.1177/1059712319859907 beltrán, i., herculano-houzel, s., sinervo, b., & whiting, m. j. (2021). are ectotherm brains vulnerable to global warming? trends in ecology & evolution, 36(8), 691–699. https://doi.org/10.1016/j.tree.2021.04.009 biehl, m., pollock, f. a., & kanai, r. (2021). a technical critique of some parts of the free energy principle. entropy, 23(3), 293. https://doi.org/10.3390/e23030293 bitbol, m., & luisi, p. l. (2004). autopoiesis with or without cognition: defining life at its edge. journal of the royal society interface, 1(1), 99–107. https://doi.org/10.1098/rsif.2004.0012 di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1146/annurev-psych-010418-102836 https://doi.org/10.1016/j.neunet.2019.03.001 https://doi.org/10.1016/j.plrev.2021.11.001 https://doi.org/10.1007/s11229-016-1288-5 https://doi.org/10.1111/j.1756-8765.2012.01211.x https://doi.org/10.1162/isal_a_00288 https://doi.org/10.1177/1059712309343819 https://doi.org/10.1162/artl_a_00143 https://doi.org/10.1177/1059712319859907 https://doi.org/10.1016/j.tree.2021.04.009 https://doi.org/10.3390/e23030293 https://doi.org/10.1098/rsif.2004.0012 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 36 bourgine, p., & stewart, j. (2004). autopoiesis and cognition. artificial life, 10(3), 327–345. https://doi.org/10.1162/106454 6041255557 bruineberg, j., dolega, k., dewhurst, j., & baltieri, m. (2021). the emperor's new markov blankets. behavioral and brain sciences, 1–63. https://doi.org/10.1017/s0140525x21002351 bruineberg, j., & hesp, c. (2018). beyond blanket terms: challenges for the explanatory value of variational (neuro-) ethology. physics of life reviews, 24, 37–39. https://doi.org/10.1016/j.plrev.2017.11.015 bruineberg, j., kiverstein, j., & rietveld, e. (2018). the anticipating brain is not a scientist: the free-energy principle from an ecological-enactive perspective. synthese, 195(6), 2417–2444. https://doi.org/10.1007/s11229-016-1239-1 byrge, l., sporns, o., & smith, l. b. (2014). developmental process emerges from extended brain–body–behavior networks. trends in cognitive sciences, 18(8), 395–403. https://doi.org/10.1016/j.tics.2014.04.010 chemero, a. (2009). radical embodied cognitive science. mit press. clark, a. (2015). radical predictive processing. the southern journal of philosophy, 53, 3–27. https://doi.org/10.1111/sjp.12120 clark, a. (2017). how to knit your own markov blanket: resisting the second law with metamorphic minds. in t. metzinger. & w. wiese (eds.), philosophy and predictive processing (pp. 1–19). frankfurt am main: mind group. https://doi.org/ 10.15502/9783958573031 clavel vázquez, m. j. (2020). a match made in heaven: predictive approaches to (an unorthodox) sensorimotor enactivism. phenomenology and the cognitive sciences, 19(4), 653–684. https://doi.org/10.1007/s11097-019-09647-0 colombo, m., & palacios, p. (2021). non-equilibrium thermodynamics and the free energy principle in biology. biology & philosophy, 36(5). https://doi.org/10.1007/s10539-021-09818-x colombo, m., & wright, c. (2021). first principles in the life sciences: the free-energy principle, organicism, and mechanism. synthese, 198(s14), 3463–3488. https://doi.org/10.1007/s11229-018-01932-w constant, a., clark, a., & friston, k. j. (2021). representation wars: enacting an armistice through active inference. frontiers in psychology, 11, 598733. https://doi.org/10.3389/fpsyg.2020.598733 costa, l. d., friston, k., heins, c., & pavliotis, g. a. (2021). bayesian mechanics for stationary processes. proceedings of the royal society a: mathematical, physical and engineering sciences, 477(2256), 20210518. https://doi.org/10.1098/rspa.2 021.0518 de jaegher, h. (2021). loving and knowing: reflections for an engaged epistemology. phenomenology and the cognitive sciences, 20(5), 847–870. https://doi.org/10.1007/s11097-019-09634-5 de jaegher, h., & di paolo, e. a. (2007). participatory sense-making. phenomenology and the cognitive sciences, 6(4), 485– 507. https://doi.org/10.1007/s11097-007-9076-9 de jaegher, h., di paolo, e. a., & adolphs, r. (2016). what does the interactive brain hypothesis mean for social neuroscience? a dialogue. philosophical transactions of the royal society b: biological sciences, 371(1693), 20150379. https://doi.org/10.1098/rstb.2015.0379 de jaegher, h., di paolo, e. a., & gallagher, s. (2010). can social interaction constitute social cognition? trends in cognitive sciences, 14(10), 441–447. https://doi.org/10.1016/j.tics.2010.06.009 desjardins, e. (2011). historicity and experimental evolution. biology & philosophy, 26(3), 339–364. https://doi.org/10.100 7/s10539-011-9256-4 desjardins, e. (2015). historicity and ecological restoration. biology & philosophy, 30(1), 77–98. https://doi.org/10.1007/s1 0539-014-9467-6 di paolo, e. a. (2001). artificial life and historical processes. in j. kelemen & p. sosík (eds.), advances in artificial life (pp. 649–658). springer berlin heidelberg. https://doi.org/10.1007/3-540-44811-x_75 di paolo, e. a. (2005). autopoiesis, adaptivity, teleology, agency. phenomenology and the cognitive sciences, 4(4), 429–452. https://doi.org/10.1007/s11097-005-9002-y di paolo, e. a. (2009). extended life. topoi, 28(1), 9–21. https://doi.org/10.1007/s11245-008-9042-3 di paolo, e. a. (2014). the worldly constituents of perceptual presence. frontiers in psychology, 5. https://doi.org/10.3389/ fpsyg.2014.00450 di paolo, e. a. (2018). the enactive conception of life. in a. newen, s. gallagher, & l. de bruin (eds.), the oxford handbook of cognition: embodied, embedded, enactive and extended (pp. 71–94). oxford: oxford university press. di paolo, e. a. (2021). enactive becoming. phenomenology and the cognitive sciences, 20(5), 783–809. https://doi.org/10.100 7/s11097-019-09654-1 di paolo, e. a., buhrmann, t., & barandiaran, x. (2017). sensorimotor life: an enactive proposal. oxford university press. di paolo, e. a., cuffari, e. c., & de jaegher, h. (2018). linguistic bodies: the continuity between life and language. mit press. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1162/1064546041255557 https://doi.org/10.1162/1064546041255557 https://doi.org/10.1017/s0140525x21002351 https://doi.org/10.1016/j.plrev.2017.11.015 https://doi.org/10.1007/s11229-016-1239-1 https://doi.org/10.1016/j.tics.2014.04.010 https://doi.org/10.1111/sjp.12120 https://doi.org/10.15502/9783958573031 https://doi.org/10.15502/9783958573031 https://doi.org/10.1007/s11097-019-09647-0 https://doi.org/10.1007/s10539-021-09818-x https://doi.org/10.1007/s11229-018-01932-w https://doi.org/10.3389/fpsyg.2020.598733 https://doi.org/10.1098/rspa.2021.0518 https://doi.org/10.1098/rspa.2021.0518 https://doi.org/10.1007/s11097-019-09634-5 https://doi.org/10.1007/s11097-007-9076-9 https://doi.org/10.1098/rstb.2015.0379 https://doi.org/10.1016/j.tics.2010.06.009 https://doi.org/10.1007/s10539-011-9256-4 https://doi.org/10.1007/s10539-011-9256-4 https://doi.org/10.1007/s10539-014-9467-6 https://doi.org/10.1007/s10539-014-9467-6 https://doi.org/10.1007/3-540-44811-x_75 https://doi.org/10.1007/s11097-005-9002-y https://doi.org/10.1007/s11245-008-9042-3 https://doi.org/10.3389/fpsyg.2014.00450 https://doi.org/10.3389/fpsyg.2014.00450 https://doi.org/10.1007/s11097-019-09654-1 https://doi.org/10.1007/s11097-019-09654-1 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 37 di paolo, e. a., & de jaegher, h. (2021). enactive ethics: difference becoming participation. topoi. https://doi.org/10.1007/ s11245-021-09766-x di paolo, e. a., heras-escribano, m., chemero, a., & mcgann, m. (eds.). (2021). enaction and ecological psychology: convergences and complementarities. frontiers media sa. https://doi.org/10.3389/978-2-88966-431-3 di paolo, e. a., & thompson, e. (2014). the enactive approach. in l. shapiro (ed.), the routledge handbook of embodied cognition (pp. 68–78). london: routledge. fischer, k. w., & bidell, t. r. (2006). dynamic development of action, thought, and emotion. in w. damon & r. m. lerner (eds.), handbook of child psychology, vol. 1: theoretical models of human development (pp. 313–399). new york: wiley. fisher, a. j., medaglia, j. d., & jeronimus, b. f. (2018). lack of group-to-individual generalizability is a threat to human subjects research. proceedings of the national academy of sciences, 115(27), e6106–e6115. https://doi.org/10.1073/pn as.1711978115 freeman, w. j. (1999). consciousness, intentionality and causality. journal of consciousness studies, 6(11–12), 143–172. https://www.ingentaconnect.com/content/imp/jcs/1999/00000006/f0020011/995 friston, k. (2013). life as we know it. journal of the royal society interface, 10(86), 20130475. https://doi.org/10.1098/rsif.2 013.0475 friston, k. (2012). a free energy principle for biological systems. entropy, 14(11), 2100–2121. https://doi.org/10.3390/e141 12100 friston, k. (2019). a free energy principle for a particular physics. https://arxiv.org/abs/1906.10184 friston, k. j. (2011). embodied inference: or “i think therefore i am, if i am what i think.” in w. tschacher & c. bergomi (eds.), the implications of embodiment: cognition and communication (pp. 89–125). exeter: imprint academic. friston, k. j., costa, l. d., & parr, t. (2021). some interesting observations on the free energy principle. entropy, 23(8), 1076. https://doi.org/10.3390/e23081076 friston, k. j., parr, t., yufik, y., sajid, n., price, c. j., & holmes, e. (2020). generative models, linguistic communication and active inference. neuroscience & biobehavioral reviews, 118, 42–64. https://doi.org/10.1016/j.neubiorev.2020.07.005 friston, k. j., rosch, r., parr, t., price, c., & bowman, h. (2018). deep temporal models and active inference. neuroscience & biobehavioral reviews, 90, 486–501. https://doi.org/10.1016/j.neubiorev.2018.04.004 friston, k., kilner, j., & harrison, l. (2006). a free energy principle for the brain. journal of physiology-paris, 100(1-3), 70–87. https://doi.org/10.1016/j.jphysparis.2006.10.001 fuchs, t. (2017). ecology of the brain: the phenomenology and biology of the embodied mind. oxford university press. gould, s. j. (2002). the structure of evolutionary theory. harvard university press. hohwy, j. (2013). the predictive mind. oxford university press. hutto, d., & myin, e. (2013). radicalizing enactivism: basic minds without content. mit press. kirchhoff, m. d. (2018). autopoiesis, free energy, and the life–mind continuity thesis. synthese, 195(6), 2519–2540. https://doi.org/10.1007/s11229-016-1100-6 kirchhoff, m. d., & froese, t. (2017). where there is life there is mind: in support of a strong life-mind continuity thesis. entropy, 19(4), 169. https://doi.org/10.3390/e19040169 kirchhoff, m. d., & kiverstein, j. (2021). how to determine the boundaries of the mind: a markov blanket proposal. synthese, 198(5), 4791–4810. https://doi.org/10.1007/s11229-019-02370-y kirchhoff, m., parr, t., palacios, e., friston, k., & kiverstein, j. (2018). the markov blankets of life: autonomy, active inference and the free energy principle. journal of the royal society interface, 15(138), 20170792. https://doi.org/10.1 098/rsif.2017.0792 kiverstein, j. (2020). free energy and the self: an ecological–enactive interpretation. topoi, 39(3), 559–574. https://doi.org/ 10.1007/s11245-018-9561-5 korbak, t. (2021). computational enactivism under the free energy principle. synthese, 198(3), 2743–2763. https://doi.org/ 10.1007/s11229-019-02243-4 kostrubiec, v., zanone, p.-g., fuchs, a., & kelso, j. a. s. (2012). beyond the blank slate: routes to learning new coordination patterns depend on the intrinsic dynamics of the learner—experimental evidence and theoretical model. frontiers in human neuroscience, 6, 222. https://doi.org/10.3389/fnhum.2012.00222 kuhn, t. (1962). the structure of scientific revolutions. university of chicago press. landecker, h. (2011). food as exposure: nutritional epigenetics and the new metabolism. biosocieties, 6(2), 167–194. https://doi.org/10.1057/biosoc.2011.1 litwin, p., & miłkowski, m. (2020). unification by fiat: arrested development of predictive processing. cognitive science, 44(7), e12867. https://doi.org/10.1111/cogs.12867 di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11245-021-09766-x https://doi.org/10.1007/s11245-021-09766-x https://doi.org/10.3389/978-2-88966-431-3 https://doi.org/10.1073/pnas.1711978115 https://doi.org/10.1073/pnas.1711978115 https://www.ingentaconnect.com/content/imp/jcs/1999/00000006/f0020011/995 https://doi.org/10.1098/rsif.2013.0475 https://doi.org/10.1098/rsif.2013.0475 https://doi.org/10.3390/e14112100 https://doi.org/10.3390/e14112100 https://arxiv.org/abs/1906.10184 https://doi.org/10.3390/e23081076 https://doi.org/10.1016/j.neubiorev.2020.07.005 https://doi.org/10.1016/j.neubiorev.2018.04.004 https://doi.org/10.1016/j.jphysparis.2006.10.001 https://doi.org/10.1007/s11229-016-1100-6 https://doi.org/10.3390/e19040169 https://doi.org/10.1007/s11229-019-02370-y https://doi.org/10.1098/rsif.2017.0792 https://doi.org/10.1098/rsif.2017.0792 https://doi.org/10.1007/s11245-018-9561-5 https://doi.org/10.1007/s11245-018-9561-5 https://doi.org/10.1007/s11229-019-02243-4 https://doi.org/10.1007/s11229-019-02243-4 https://doi.org/10.3389/fnhum.2012.00222 https://doi.org/10.1057/biosoc.2011.1 https://doi.org/10.1111/cogs.12867 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org ezequiel a. di paolo, evan thompson, and randall d. beer 38 longo, g. (2018). how future depends on past and rare events in systems of life. foundations of science, 23(3), 443–474. https://doi.org/10.1007/s10699-017-9535-x longo, g., & montévil, m. (2014). perspectives on organisms: biological time, symmetries and singularities. springer. maturana, h. r., & varela, f. j. (1980). autopoiesis and cognition: the realization of the living. d. reidel publishing company. maturana, h. r., & varela, f. j. (1987). the tree of knowledge. shambhala. medaglia, j. d., ramanathan, d. m., venkatesan, u. m., & hillary, f. g. (2011). the challenge of non-ergodicity in network neuroscience. network: computation in neural systems, 22(1–4), 148–153. https://doi.org/10.3109/09638237.2011.639604 molenaar, p. c. m. (2004). a manifesto on psychology as idiographic science: bringing the person back into scientific psychology, this time forever. measurement: interdisciplinary research & perspective, 2(4), 201–218. https://doi.org/10 .1207/s15366359mea0204_1 molenaar, p. c. m., & nesselroade, j. r. (2014). new trends in the inductive use of relational developmental systems theory: ergodicity, nonstationarity, and heterogeneity. in p. c. molenaar, r. m. lerner, & k. m. newel (eds.), handbook of developmental systems and methodology (pp. 442–462). new york: wiley. montévil, m. (2020). historicity at the heart of biology. theory in biosciences. https://doi.org/10.1007/s12064-020-00320-8 nicolis, g., & prigogine, i. (1977). self-organization in nonequilibrium systems: from dissipative structures to order through fluctuations. new york, ny: john wiley & sons. orth, d., kamp, j. van der, memmert, d., & savelsbergh, g. j. p. (2017). creative motor actions as emerging from movement variability. frontiers in psychology, 8, 1903. https://doi.org/10.3389/fpsyg.2017.01903 oyama, s., griffiths, p. e., & gray, r. d. (eds.). (2001). cycles of contingency: developmental systems and evolution. mit press. palacios, e. r., razi, a., parr, t., kirchhoff, m., & friston, k. (2020). on markov blankets and hierarchical self-organisation. journal of theoretical biology, 486, 110089. https://doi.org/10.1016/j.jtbi.2019.110089 parr, t., costa, l. d., & friston, k. (2019). markov blankets, information geometry and stochastic thermodynamics. philosophical transactions of the royal society a: mathematical, physical and engineering sciences, 378(2164), 20190159. https://doi.org/10.1098/rsta.2019.0159 pessoa, l. (2014). understanding brain networks and brain organization. physics of life reviews, 11(3), 400–435. https: //doi.org/10.1016/j.plrev.2014.03.005 pezzulo, g., rigoli, f., & friston, k. (2015). active inference, homeostatic regulation and adaptive behavioural control. progress in neurobiology, 134, 17–35. https://doi.org/10.1016/j.pneurobio.2015.09.001 raja, v., valluri, d., baggs, e., chemero, a., & anderson, m. l. (2021). the markov blanket trick: on the scope of the free energy principle and active inference. physics of life reviews, 39, 49–72. https://doi.org/10.1016/j.plrev.2021.09.001 ramstead, m. j. d., badcock, p. b., & friston, k. j. (2018). answering schrödinger's question: a free-energy formulation. physics of life reviews, 24, 1–16. https://doi.org/10.1016/j.plrev.2017.09.001 ramstead, m. j. d., kirchhoff, m. d., constant, a., & friston, k. j. (2021). multiscale integration: beyond internalism and externalism. synthese, 198(s1), 41–70. https://doi.org/10.1007/s11229-019-02115-x rath, d., amlinger, l., rath, a., & lundgren, m. (2015). the crispr-cas immune system: biology, mechanisms and applications. biochimie, 117, 119–128. https://doi.org/10.1016/j.biochi.2015.03.025 rose, l. t., rouhani, p., & fischer, k. w. (2013). the science of the individual. mind, brain, and education, 7(3), 152–158. https://doi.org/10.1111/mbe.12021 seth, a. k. (2014). a predictive processing theory of sensorimotor contingencies: explaining the puzzle of perceptual presence and its absence in synesthesia. cognitive neuroscience, 5(2), 97–118. https://doi.org/10.1080/17588928.2013. 877880 soucy, s. m., huang, j., & gogarten, j. p. (2015). horizontal gene transfer: building the web of life. nature reviews genetics, 16(8), 472–482. https://doi.org/10.1038/nrg3962 speelman, c. p., & mcgann, m. (2016). editorial: challenges to mean-based analysis in psychology: the contrast between individual people and general science. frontiers in psychology, 7, 1234. https://doi.org/10.3389/fpsyg.2016.01234 stewart, j., gapenne, o., & di paolo, e. a. (eds.). (2010). enaction: a new paradigm for cognitive science. mit press. thelen, e. (2005). dynamic systems theory and the complexity of change. psychoanalytic dialogues, 15(2), 255–283. https://doi.org/10.1080/10481881509348831 thelen, e., corbetta, d., & spencer, j. p. (1996). development of reaching during the first year: role of movement speed. journal of experimental psychology: human perception and performance, 22(5), 1059–1076. https://doi.org/10.1037/00 96-1523.22.5.1059 thompson, e. (2007). mind in life: biology, phenomenology and the sciences of mind. cambridge, ma: harvard university press. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s10699-017-9535-x https://doi.org/10.3109/09638237.2011.639604 https://doi.org/10.1207/s15366359mea0204_1 https://doi.org/10.1207/s15366359mea0204_1 https://doi.org/10.1007/s12064-020-00320-8 https://doi.org/10.3389/fpsyg.2017.01903 https://doi.org/10.1016/j.jtbi.2019.110089 https://doi.org/10.1098/rsta.2019.0159 https://doi.org/10.1016/j.plrev.2014.03.005 https://doi.org/10.1016/j.plrev.2014.03.005 https://doi.org/10.1016/j.pneurobio.2015.09.001 https://doi.org/10.1016/j.plrev.2021.09.001 https://doi.org/10.1016/j.plrev.2017.09.001 https://doi.org/10.1007/s11229-019-02115-x https://doi.org/10.1016/j.biochi.2015.03.025 https://doi.org/10.1111/mbe.12021 https://doi.org/10.1080/17588928.2013.877880 https://doi.org/10.1080/17588928.2013.877880 https://doi.org/10.1038/nrg3962 https://doi.org/10.3389/fpsyg.2016.01234 https://doi.org/10.1080/10481881509348831 https://doi.org/10.1037/0096-1523.22.5.1059 https://doi.org/10.1037/0096-1523.22.5.1059 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org laying down a forking path 39 thompson, e. (2018). review of daniel d. hutto and erik myin, evolving enactivism: basic minds meet content. notre dame philosophical reviews. https://ndpr.nd.edu/reviews/evolving-enactivism-basic-minds-meet-content/ thompson, e., & stapleton, m. (2009). making sense of sense-making: reflections on enactive and extended mind theories. topoi, 28(1), 23–30. https://doi.org/10.1007/s11245-008-9043-2 turner, j. s. (2000). the extended organism: the physiology of animal-built structures. harvard university press. van es, t. (2021). living models or life modelled? on the use of models in the free energy principle. adaptive behavior, 29(3), 315–329. https://doi.org/10.1177/1059712320918678 van es, t., & kirchhoff, m. d. (2021). between pebbles and organisms: weaving autonomy into the markov blanket. synthese, 199(3-4), 6623–6644. https://doi.org/10.1007/s11229-021-03084-w varela, f. j. (2011). preface to the second edition of “de máquinas y seres vivos autopoiesis: la organización de lo vivo” by francisco varela. systems research and behavioral science, 28(6), 601–617. https://doi.org/10.1002/sres.1122 varela, f. j. (1979). principles of biological autonomy. north holland. varela, f. j. (1991a). making it concrete: before, during and after breakdowns. revue internationale de psychopathologie, 4(6), 601–617. https://associationpsychanalytiquedefrance.org/publications/revues/n4-1991/ varela, f. j. (1991b). organism: a meshwork of selfless selves. in a. i. tauber (ed.), organism and the origins of self. boston studies in the philosophy of science vol. 129 (pp. 79–107). dordrecht: springer. https://doi.org/10.1007/978-94-0113406-4_5 varela, f. j., thompson, e., & rosch, e. (1991). the embodied mind: cognitive science and human experience. mit press. varela, f., lachaux, j.-p., rodriguez, e., & martinerie, j. (2001). the brainweb: phase synchronization and large-scale integration. nature reviews neuroscience, 2(4), 229–239. https://doi.org/10.1038/35067550 veissière, s. p. l., constant, a., ramstead, m. j. d., friston, k. j., & kirmayer, l. j. (2019). thinking through other minds: a variational approach to cognition and culture. behavioral and brain sciences, 43, e90. https://doi.org/10.1017/s01405 25x19001213 villalobos, m., & ward, d. (2015). living systems: autonomy, autopoiesis and enaction. philosophy & technology, 28(2), 225–239. https://doi.org/10.1007/s13347-014-0154-y vygotsky, l. (2012). thought and language. revised and expanded edition. mit press. waddington, c. h. (1957). the strategy of the genes: a discussion of some aspects of theoretical biology. george allen; unwin. walsh, k. s., mcgovern, d. p., clark, a., & o’connell, r. g. (2020). evaluating the neurophysiological evidence for predictive processing as a model of perception. annals of the new york academy of sciences, 1464(1), 242–268. https://doi.org/10 .1111/nyas.14321 weron, a., burnecki, k., akin, e. j., solé, l., balcerek, m., tamkun, m. m., & krapf, d. (2017). ergodicity breaking on the neuronal surface emerges from random switching between diffusive states. scientific reports, 7(1), 5404. https: //doi.org/10.1038/s41598-017-05911-y wiese, w., & friston, k. j. (2021). examining the continuity between life and mind: is there a continuity between autopoietic intentionality and representationality? philosophies, 6(1), 18. https://doi.org/10.3390/philosophies6010018 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. di paolo, e. a., thompson, e., & beer, r. d. (2022). laying down a forking path: tensions between enaction and the free energy principle. philosophy and the mind sciences, 3, 2. https://doi.org/10.33735/phimisci.2022.9187 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://ndpr.nd.edu/reviews/evolving-enactivism-basic-minds-meet-content/ https://doi.org/10.1007/s11245-008-9043-2 https://doi.org/10.1177/1059712320918678 https://doi.org/10.1007/s11229-021-03084-w https://doi.org/10.1002/sres.1122 https://associationpsychanalytiquedefrance.org/publications/revues/n4-1991/ https://doi.org/10.1007/978-94-011-3406-4_5 https://doi.org/10.1007/978-94-011-3406-4_5 https://doi.org/10.1038/35067550 https://doi.org/10.1017/s0140525x19001213 https://doi.org/10.1017/s0140525x19001213 https://doi.org/10.1007/s13347-014-0154-y https://doi.org/10.1111/nyas.14321 https://doi.org/10.1111/nyas.14321 https://doi.org/10.1038/s41598-017-05911-y https://doi.org/10.1038/s41598-017-05911-y https://doi.org/10.3390/philosophies6010018 https://doi.org/10.33735/phimisci.2022.9187 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction an enactive fep? misreadings of autopoiesis and enaction autopoiesis boundaries points of tension between the two approaches non-equilibrium steady states versus history mutual enablement and co-constitution of internal and external domains conclusions how does the psychiatrist know? how does the psychiatrist know? adrian kinda,b,c (adrian.kind@ovgu.de) abstract an important question in the philosophy of psychiatry is: what is the proper method of psychiatric diagnostic reasoning? let us call this the methodological question. in this paper, i criticize the answer that proponents of phenomenological psychiatry have given to this question and present an alternative. i argue that their answer fails to meet several adequacy conditions for a theory of psychiatric diagnostic reasoning. i then show how my own answer to the methodological question – the model-based account of diagnostic reasoning – meets these adequacy conditions. i conclude that the model-based account of diagnostic reasoning is preferable to the phenomenological account. keywords diagnostic reasoning ∙ epistemology of psychiatry ∙ model-based reasoning ∙ phenomenological psychiatry this article is part of a special issue on “models and mechanisms in philosophy of psychiatry,” edited by lena kästner and henrik walter. 1 introduction imagine you are worrying about someone you know because they suffer from severe mental distress. who should they see to find out what their problem is? most people might suggest consulting a psychiatrist. this seems reasonable, as psychiatrists undergo theoretical and practical training meant to ensure they are and remain experts in matters of psychiatric diagnosis. given their training, it seems fair prima facie to assume that psychiatrists deserve this epistemic authority in determining someone’s psychopathological status.1 but even if one grants epistemic ahumboldt-universität zu berlin, faculty of philosophy, berlin school of mind and brain. botto von guericke universität magdeburg, faculty of humanities, institute of philosophy. ccharité universitätsmedizin, research division mind and brain. 1i take the expert status of the psychiatrist to remain unharmed by a commonsensical mild skepticism towards current psychiatric diagnostics. this skepticism claims that there is room to improve kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-5798-8513 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 2 authority to psychiatrists, questions nevertheless arise. one of these questions concerns how psychiatrists, with all their epistemic authority, arrive at a diagnosis. this question is what i will call the methodological question. the methodological question: what is the method of psychiatric diagnostic reasoning? the methodological question presumes that there is, indeed, some method of psychiatric diagnostic reasoning to describe. how plausible this presumption is depends on what one takes to be a method. following goldman (2000), i assume a method is a learned belief-forming procedure that shapes inquiry. such procedures are specifiable on different levels of generality, ranging from general procedures like deduction or induction, to more specific ones like the philosophical method of phenomenological reduction or the scientific method of dna sequencing. on this understanding of method, it appears prima facie plausible that, given the theoretical and practical training psychiatrists receive in diagnostics, there is some sort of methodology at work behind clinical psychiatric diagnostics that can be explicated. the methodological question is not a clear-cut normative or descriptive question. to provide an adequate answer descriptive as well as normative adequacy conditions should be considered. an answer to the methodological question should be descriptively adequate, meaning that it be consistent with what we know about how well-trained psychiatrists engage in diagnostic reasoning in clinical diagnostics. but above that, explicating what should be considered the method behind these practices should also be constrained by widely shared normative commitments within the psychiatric profession. a proposal not considering such professional normative standards might present an answer that matches aspects of psychiatric practice but implies that diagnostic experts ignore their own explicit standards in practice. prima facie, this seems less plausible and charitable than to assume that the normative standards that are regularly discussed in the literature and stressed in clinical training also affect diagnostic practice. as a result, both descriptive accuracy and normative commitments of the accuracy of diagnoses but does not doubt the expertise of the psychiatrist on account of this fact. raising such doubts based on this fact would have absurd consequences: most scientists should not be considered experts in their field, since many scientific methods are recognized as imperfect, and a lot of scientific vocabulary is left open to revision as research progresses. expertise, as we usually speak about it, appears to be determined not in relation to an ideal system of theoretical reasoning that already gets everything completely right, but in relation to the skillful use of the currently established theoretical and practical resources for inquiry; these resources may, of course, be improved. only the acceptance of strong skepticism towards psychiatric diagnostics would be a problem for recognizing the epistemic authority of psychiatrists. such skepticism claims that psychiatric diagnostics is a through and through invalid activity that gets nothing right about patients. this view is often associated with rosenhan (1973) and szasz (1960). addressing strong skepticism is beyond the scope of this paper, but see spitzer (1975), guze (1992), and cahalan (2019). kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 3 clinical practice provide relevant constraints for answering the methodological question. the methodological question is not only a fair question to ask, but also a significant one deserving philosophical attention. indeed, addressing the methodological question is a vital matter for philosophy of psychiatry, and for several reasons. answering the methodological question is necessary for gaining a systematic, comprehensive understanding of the epistemology of psychiatry, which includes not only the epistemology of psychiatric sciences but also that of clinical practice. without an answer to the methodological question, it is impossible to address more specific epistemological and ethical questions. for example, without a proper understanding of the diagnostic, inferential practices of the clinical psychiatrist, it is impossible to evaluate the epistemic strengths and weaknesses of the ways in which psychiatrists justify their diagnoses. or, to give another example, lacking an understanding of the method of psychiatric clinical diagnosis would make it impossible to explicate its internal standards so that it could be determined when diagnostic efforts violate these norms and constitute malpractice.2 while the importance of the methodological question for medicine in general is well recognized, asking this question specifically for psychiatry is a more recent development. in the past, work on the methodological question concerning medical diagnostic reasoning in general has been done by cognitive psychologists and medical education scientists, with only few notable exceptions among philosophers (e.g., sober, 1979).3 only more recently has the methodological question as put in this paper – with focus on diagnostic reasoning in clinical psychiatry – came to greater attention. this time with philosophers of psychiatry as driving forces of the debate (e.g., banner & thornton, 2007; cooper, 2007; fuchs, 2010; gupta et al., 2019; murphy, 2006; parnas et al., 2013; reznek, 1998). as a contribution to this development, this paper presents a new answer to the methodological question and shows its benefits over a competing account: the 2by “internal standards”, i mean the epistemic norms that psychiatrists ought to follow to arrive at permissible diagnostic conclusions. the conclusions are considered permissible insofar as they are justified by the standards of the expert community of clinicians. being “objectively” wellgrounded or reliable does not guarantee the acceptability of a diagnosis. the justification of a diagnosis is always also tied to matters of responsibility. if a psychiatrist’s diagnosis is wrong but complies with the internal standards, she will not be considered culpable because she was justified in making this diagnosis. if, on the other hand, a psychiatrist guesses a diagnosis and thereby violates the internal standards of psychiatric diagnostics, she will be guilty of diagnostic malpractice. for a similar take on justification, see pollock (1999, p. 125) and adam & littlejohn (2021, pp. 320–321). 3some landmark cognitive science papers about clinical reasoning are elstein et al. (2014); nurcombe & fitzhenry-coor (1982); beck & bergman (1986); schmidt et al. (1990); magnani (1992); charlin et al. (2000); croskerry (2009); and fernando et al. (2013). good overviews with different emphases can be found in gruppen & frohna (2002); norman (2005); patel et al. (2012), and higgs et al. (2019). kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 4 phenomenological account of psychiatric diagnostic reasoning.4 to begin, section 2 presents the phenomenological account. next, section 3 points out the difficulties faced by this account in meeting several normative requirements and empirical adequacy conditions that belong to a plausible account of psychiatric diagnostic reasoning. after raising these criticisms, section 4 introduces my explanation, the model-based account of psychiatric diagnostic reasoning, and evaluates my account against the challenges brought up against the phenomenological account to show how the model-based account deals better with those problems and is therefore the preferable theory. 2 the phenomenological account the phenomenological account is an answer to the methodological question. it is defended by philosophers of psychiatry standing in the tradition of phenomenology stemming from husserl (husserl & ströker, 1900/2013), first applied to psychiatry by the philosopher-psychiatrists jaspers (1913/1973).5 among the defenders of the phenomenological account, we find some of the currently most influential exponents influenced by this line of tradition: thomas fuchs (2010), josef parnas, dan zahavi, and louis a. sass (2013). for simplicity, i will refer to these authors as phenomenologists. i shall add briefly that what i present here as the phenomenological account is of course not all that this tradition of philosophy of psychiatry has to offer but will only focus on their epistemic proposal that can be taken to be their response to the methodological question. beyond this aspect, phenomenological psychiatry has a rich tradition of discussing, e.g., the psychiatrist-patient relationship and its implication for care, as well as the relevance of the overall existential situation of patients. all this will not be part of what i discuss as the phenomenological account. here i am only interested in the phenomenologists’ proposal concerning how psychiatrists arrive at their diagnostic conclusions. accordingly, my later presented criticism and proposed alternative account will only concern this aspect of phenomenological psychiatry. i remain agnostic regarding other topics discussed in phenomenological psychiatry. to reconstruct the phenomenologists’ answer to the methodological question, i begin by presenting some quotes indicating their understanding of diagnostic reasoning in psychiatry. the most general statement of how psychiatrists engage in diagnostic reasoning according to the phenomenologists is provided by fuchs (2010, p. 271), who claims that 4i focus on one alternative proposal here due to limitations of space. a more detailed version of my proposal, as well as a comprehensive discussion of other alternative accounts will be provided in future work. 5for details on husserl’s influence on jasper see wiggins & schwartz (1997). kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 5 experienced clinicians do not diagnose and practice by ticking off the diagnostic criteria of the manuals. they work with the prototypal approach to diagnosis […] that help[s] to grasp the essence of a phenomenon as an organizing and meaningful ‘gestalt’ over particular details. while the phenomenologists provide no details on how this process of the direct recognition of a clinical gestalt is taking place, they do flesh out why they believe that diagnostics must directly address the whole gestalt of a disorder. they do so by discussing the nature of a disorder’s clinical gestalt and the epistemic constraints it puts on attempts to recognize a patient’s psychopathology. on the nature of the clinical gestalt, parnas et al. (2013) claim that it is “not a simple aggregate; the ‘whole is more than the sum of its parts.’ this unity emerges from the relations between component features and is influenced by the whole (part-whole relations)” (2013, p. 275), whereas these components are symptoms of these mental disorders. with regards to the epistemic access to these symptoms and the clinical gestalt of the disorder, they remark that symptoms of a disorder serve to elaborate “[a]spects of a gestalt [that] […] may be focused on in diagnosis or research; but one must remember that these aspects are interdependent in a mutually constitutive and implicative manner” (2013, p. 275). so that “[w]hat, defines a given individual experience/expression as a specific symptom or sign, […] articulates itself from within an experiential expressive whole [of the gestalt]” (2013, p. 275). how should we understand these remarks in relation to the methodological question as we have posed it here? let me begin again with fuchs’general point. according to him, the psychopathological feature recognized first and attributed to a patient is the whole gestalt of a disorder, existing over and above any of its details. that means that the psychiatrist does not first discern symptoms and signs, but rather directly recognizes a disorder based on its prototypical gestalt as it shows up in the patient’s behaviour and reports. this is the first step of the diagnostic reasoning process: the psychiatrist directly recognizes the disorder (e.g., a major depression). however, not fuchs nor parnas et al. give details about how this is meant to happen. one interpretation of what the phenomenologists have in mind, suggested by fuchs’ and parnas’ et al.’s use of the term’s “prototype” and “gestalt,” is that psychiatrists engage in a form of pattern recognition, namely prototype processing allowing them to recognise the clinical gestalt of a disorder. the prototype theory of pattern recognition in cognitive psychology is a model of pattern recognition, according to which different prototypes of objects are memorized by the system so that in the process of pattern recognition, outside simulation only needs to be compared with the prototype, and the sense to objects comes from the matching between input information and prototype. once outside simulating information matches best with a certain prototype in the brain, the information can be ranged in the category of that prototype and recognized (pi et al., 2008, p. 435). kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 6 an essential feature of this kind of pattern recognition is that it only contains top-down processing but no bottom-up processing (see pi et al., 2008, p. 436), which means that recognizing the relevant object begins with the matched prototype itself. given that phenomenologists emphasize the notion of gestalt, it seems that this prototype occurring in the psychiatrist’s mind and leading to ascribing a diagnosis has gestalt quality. according to ehrenfels – one of the founders of gestalt psychology – that means it has a positive content of presentation bound up in consciousness with the presence of complexes of mutually separable (i.e., independently presentable) elements. that complex of presentations which is necessary for the existence of a given gestalt quality we call the foundation of that quality (ehrenfels 1890, in smith & ehrenfels, 1988, p. 93). keeping in mind the description of the process of prototype recognition, which entails an immediate matching between information input and prototypes, and the understanding of these prototypes as having gestalt quality (i.e., being necessarily complexes of elements), the following picture emerges: according to phenomenologists, the psychiatrist directly perceives the disorders as complex or gestalt in the patient after being confronted with diagnostic information without further explicit cognitive efforts. this interpretation of fuchs’ general idea also matches with parnas’ et al.’s elaborations. as we saw, they claim (in their first quotation) that while the disorder and its symptoms are ontologically mutually constitutive, the clinical gestalt nevertheless enjoys epistemic primacy. as we saw in their last quote, they seem to believe that signs and symptoms are epistemically secondary insofar as the psychiatrist determines them after identifying the disorder. an interpretation following from their statement that only when the gestalt of the disorder is recognized a psychiatrist can proceed to identify the symptoms and signs of that disorder in the patient. the gestalt must be recognized first, as only the clinical gestalt of the psychopathology allows for a symptom or sign to “articulate itself” and therefore become epistemically accessible to the psychiatrist. in sum, the phenomenological proposal to answer the methodological question is that the method of psychiatric diagnostic reasoning is an automated form of prototype-based pattern recognition. this form of prototype-based pattern recognition leads psychiatrists to form cognitively unmediated (i.e., without explicit inferential reasoning) assumptions about the presence of a disorder in a patient occurring to the psychiatrist as a gestalt quality of their perception of this patient. any details of the psychopathological state of the patients are thereby epistemically secondary.6 particular features of a disorder can only be accessible and become relevant to the psychiatrist if the disorder is already recognized. the proposal 6to interpret the phenomenologists’ proposal as the idea that psychiatrists can directly access patients’ overall psychopathological mental condition by a quasi-perceptual process fits with other views held by authors from the phenomenological tradition. zahavi (2019), for example, defends a similar position not in regard to psychopathological mental condition but for our overall interperkind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 7 of the phenomenologists therefore has the feature of being epistemically top-down. it is a disorder first, symptoms second account to psychiatric diagnostic reasoning. 3 problems for the phenomenological account the phenomenological account faces problems that make it an implausible answer to the methodological question. some of these problems stem from empirical research about diagnostic reasoning that the phenomenological account does not accommodate, but that an empirically informed philosophical theory of psychiatric diagnostic reasoning should be constrained by. other problems stem from some absurd consequences of the phenomenological account. both sets of problems shall be presented here, beginning with those emerging from empirical research. cognitive psychologists and medical education researchers investigating the modes of cognition employed by diagnosing clinicians identify two types of processing: bottom-up pattern recognition (e.g., coderre et al., 2003; groves et al., 2003) and analytic reasoning (e.g., croskerry, 2009). these types of processing have become widely recognized, and so a theory of diagnostic reasoning that wishes to draw on empirical research should therefore make sure that these results cohere with their proposal. however, neither cognitive process has a place in the phenomenological account. although one might suspect that gestalt recognition could be understood as pattern recognition, it cannot – at least not in the sense of the term usually employed in medical cognition research. in the context of that research, pattern recognition is generally understood as a highly automatic, cue-based, feature-outcome associating process. the cues are usually assumed to be signs and symptoms on the basis of which the pattern – the illness or disorder – is recognized (e.g., loveday et al., 2013). this understanding of pattern recognition, however, does not square with the phenomenological account. pattern recognition, as understood by medical cognition researchers, treats the symptoms and signs as cues that must be individuated before the pattern itself is recognized. pattern recognition therefore requires following a symptom first, disorder second approach. the phenomenological account turns this process upside down, since it maintains that the disorder is recognized by a clinician prior to the elements making up the patter.7 sonal access. in his view, also our everyday knowledge about each other’s minds, e.g., if someone is angry, is won in a direct quasi-perceptual manner without cognitive mediating processes. 7one might wonder if there is an inversion in the process of disorder-symptoms recognition or if one could not say instead that what is described by bottom-up pattern recognition is an implicit part of what happens in the process described by the phenomenologists. in my view, this would be too much of a stretch. as we saw in the passages quoted earlier, the epistemic primacy of recognizing the whole over its parts is central for the phenomenologist. we should take them seriously in this claim. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 8 analytic reasoning also plays no role in the phenomenological account. analytic reasoning involves the explicit and careful consideration of the patient’s presentation, identifying symptoms given certain background knowledge, and carefully weighing what diagnostic options are most plausible based on the available evidence. while pattern recognition is often used in simple diagnostic tasks (e.g., diagnosing a flu), analytic reasoning is commonly employed when medical experts face complex or ambiguous diagnostic scenarios (see croskerry, 2009). such complexities and ambiguities often appear in psychiatric cases. on the phenomenological account, by contrast, the disorder is first recognized as a whole gestalt, and symptoms are only individuated after the gestalt of the disorder has been recognized in the patient. it therefore seems that analytic reasoning plays no part in the actual diagnostic reasoning process that identifies a disorder. if such reasoning is exercised at all, it would provide only a circular form of post-hoc justification for the diagnostic intuitions by which the clinician recognized the patient’s disorder in the first place as it is this initial diagnosis based on that the confirming symptoms would be recognized and not on independent grounds. as the phenomenologists claimed earlier: symptoms are only epistemically individuated in the context of the previously recognized gestalt. if psychiatrists really diagnosed in the manner described by the phenomenological account, it seems that they would not engage in analytic diagnostic reasoning.8 the phenomenological account fails to consider not only empirical research on the cognitive processes of practicing clinicians, but also critical diagnostic reasoning. critical diagnostic reasoning is a technique in which clinical professionals are trained and expected to practice, no matter their specialization (e.g., harjai & tiwari, 2009; mamede et al., 2007). engaging in critical diagnostic reasoning 8defenders of the phenomenological account may claim that such empirical constraints generated from medical cognition and education science do not mean anything to psychiatry. the argument might go as follows: since psychiatry is a special medical discipline and psychiatric diagnosis is unique, its diagnostic reasoning is also different from that belonging to any other medical field. a theory that explains psychiatric diagnostic reasoning consequently should not be constrained by the empirical findings of researchers who investigated diagnostic reasoning across different branches of medicine. by making this move, however, defenders of the phenomenological account try to free themselves of the need to show that psychiatry is special insofar as it does not employ pattern recognition or analytic reasoning. but the burden of proof remains on them, since medical education and cognition research has so persuasively established these two modes of cognition as belonging to diagnostic reasoning across all branches of medicine. if the universality were not widely recognized, we should see considerable debate among researchers in that those more particular concerned with diagnostic reasoning in psychiatry speak up. given the lack of debate, it seems fair to assume that psychiatry is not in this respect distinct from internal medicine, neurology, or dermatology. consequently, we should assume that the constraints discussed here apply to psychiatry and other branches of medicine until there is good reason to think differently. however, there might be other things crucial to diagnostic reasoning in psychiatry, less critical in different branches of medicine, making psychiatry unique because it has something additional, more specific about it. this possibility does not undermine the relevance of the plausibly shared constraints. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 9 means critically examining one’s own or another’s diagnostic judgments in order to avoid making mistakes in diagnostics due to biases or other errors in reasoning. questions like “why exactly should i draw this diagnostic conclusion?”, “what could be an alternative explanation?”, “did i consider all available and potentially relevant information?” are questions typically asked when engaging in this kind of reasoning. critical diagnostic reasoning can take place intrapersonally (by critically evaluating one’s own diagnostic judgments) or interpersonally (by evaluating the diagnoses of others, as a clinician who supervises or works on a team might). to engage in critical reasoning about the justification of one’s diagnosis in a noncircular way however, analytical diagnostic reasoning is a prerequisite. the problem here for the phenomenological account is that if a psychiatrist were to diagnose in the manner it prescribes, this intraand interpersonal critical diagnostic reasoning would be impossible, or at least unnecessary. intrapersonal critical reasoning is not required, since considerations of a more plausible alternative diagnosis, given the symptoms and signs of disease, could not be found. it is the initial diagnosis that determines what signs and symptoms the clinician will be able to individuate in the patient. it follows that any attempt to evaluate one’s own diagnosis will, by the logic of the phenomenological account, lead to a necessarily self-confirming result. to get out of this vicious circle the identification of signs and symptoms have to be archived analytically as described above. there is a related problem when it comes to interpersonal critical evaluation. part of the critical interpersonal discussion of diagnoses involves explaining why one has given a diagnosis to other clinical experts. these experts suggest potential alternatives in order that they may eventually come to an agreement on the best diagnostic decision. if, however, all symptoms and signs that the clinicians recognize depend on their initial diagnoses, pointing out other symptoms or signs to them would be hopeless because they would not be able to individuate those symptoms or signs independently of their original diagnosis. interpersonal disagreements about diagnoses would become unresolvable and farcical, since any one participant could never rationally convince the other participants who disagree with him. this is because it would be impossible for both sides to recognize the symptoms and signs that could serve as counterevidence to their own diagnostic proposal, given that those symptoms and signs would not fit the gestalt they recognized. the symptoms and signs individuated by each side in the disagreement would, at least in principle, be epistemically inaccessible to the other. 4 the model-based account of diagnostic reasoning it seems desirable to develop an alternative theory of psychiatric diagnostic reasoning that does not face the same problems as the phenomenological account. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 10 my proposed alternative is what i call the model-based account to psychiatric diagnostic reasoning.9 the basic idea behind this proposal is that we can best understand the diagnostic activities of psychiatrists as a kind of modelling process. it is a modelling process not in the sense that all cognition may be a form of modelling on some level of description, as psychologists and philosophers have claimed (e.g., hohwy, 2013; johnson-laird, 1983). the point rather is that the diagnostic reasoning of psychiatrists is a modelling process akin to the epistemic activity of modelling in the (applied) sciences. however, in order to make this claim intelligible and interesting, an understanding of modelling must be established that is sufficiently specific so that its application to psychiatric diagnostics can offer an informative answer to the methodological question and such application must be shown to be plausible. to specify my understanding of modelling and apply it to psychiatric diagnostic reasoning i will (in section 4.1) spell out the general modelling i have in mind and provide and discuss the specifics kind of modelling i claim to be present in psychiatric diagnostic reasoning, namely qualitative and compositional diagnostic modelling. i will then (in section 4.2) present the core elements of clinical psychiatric diagnostics and argue (in section 4.3) that the kind of modelling i discuss provides a good blueprint for these diagnostic efforts. finally (in section 4.4), i will show that my model-based account is superior to the phenomenological account insofar as it holds up better against the challenges. 4.1 modelling my general understanding of modelling takes up a currently prominent proposal in philosophy of science made by godfrey-smith (2006) and weisberg (2013, 2007).10 their basic idea is that modelling follows a certain strategy of theorizing, called the indirect strategy of representation, that can be contrasted with another approach to theorizing, called the abstract direct strategy of representation. the difference between these two approaches can be stated as follows. following the abstract direct strategy, means to begin with a detailed investigation of the representational target at the beginning of the theorizing process and follow a data-driven approach: first, collect large amounts of data about the system of interest before beginning to theorize. then, based on the analysis of 9the idea that diagnostics in general or psychiatric diagnostics in particular might be some sort of modeling is not entirely new but has not yet been spelled out in detail. upshur & colak (2003) mentioned this option briefly in regards to general medical reasoning, and murphy (2006, pp. 205– 206, and 265-266) develops it only briefly, and in a much more abstract and substantively different way than i do here. discussing murphy´s approach is beyond the scope of this paper. 10this proposal was developed by them from their own case studies on modeling in science (especially evolution and population biology), as well as from lessons from previous debates on modeling in the philosophy of science (especially physics and economics) e.g., by giere (1988, 1999), cartwright & nordby (1983), wimsatt (1987), and boumans (1999). kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 11 that data, determine what are the properties of the system and how certain properties of the system account for other properties. then, based on these findings, a theoretical structure is developed, encompassing the findings to arrive at an overall theoretical structure meant to represent the properties of the target systems as faithfully as possible. with the indirect strategy of representation, by contrast, theorists do not begin by directly investigating the representational target. instead, they set up a theoretical structure to represent the target system with little previous knowledge about it, analyse the dynamics and predictions of this structure, and only then turn to the actual real-world system. then they test whether the developed structure suits the representational interests of the modeler well enough to be used as a model of the target. these interests include, for example, representing relevant dynamics or allowing predictions of interest about the behaviour of the actual system. the core difference between this strategy and the abstract direct strategy consists in the indirect way by which the process arrives at a representational structure for a given target. instead of directly addressing the target in order to set up a theoretical structure, a theorist first comes up with a theoretical structure and only then turns to the task of evaluating the fit of the theoretical structure to the real-world target. it is this later strategy that, according to the case studies of godfrey-smith and weisberg, is the modelling strategy followed in (applied) science. now that the core difference is clear, let me focus on the details of the indirect strategy of representation. it follows a three-step procedure. these three steps of the procedure may be referred to simply as the construal of the model, its analysis, and finally, the model-world comparison. understanding the details of each step will be important later when considering whether diagnostic reasoning involves this kind of modelling. in the construal step, the theorist sets up the theoretical structure that is intended to serve as a model of the target system. this structure might be quantitative, as is most often the case in scientific models, but it may also be qualitative. mathematical formulas, graphical presentations, or a set of interrelated propositions expressed by sentences – all these theoretical structures can form the basis of models. by setting up such structures, the theorist determines the scope of the model (i.e., what aspects of the empirical system are to be targeted by the model) and assigns the elements of the structure to specific aspects of the empirical systems that have been included within the model’s scope, or what aspects of the world the model shall be a model about. the choice of how to specify a model’s structure is governed by the modeler’s expectations of what theoretical structure would best target those aspects of the real-world system they are interested in modelling. to come up with such a model structure, modelers can draw on different sources. sometimes modelers come up with a new model structure based on their intuitions and past experiences in modelling; this has led philosophers of science to talk about the “art” and the “know-how” of modelling (godfrey-smith, 2006; morrison, 1999). but, more often, background knowledge and past experikind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 12 ences with modelling play a chief role in setting up a model structure. in such cases models can be based on background theories11, the cross-contextual reuse of model structures from other disciplines12, or the reuse of model structures to model other targets within the same discipline.13 a number of these sources may be drawn upon in order to puzzle together a useful model structure.14 the second step in the indirect strategy of representation is the analysis of the model. here the modeler investigates the dynamics of the model structure. which changes in what parameters of the model would lead to what kind of changes in other parameters? what predictions does the model make? this step of the process is independent of the final step, the model-world comparison, since it is for now only concerned with discovering regularities within the model itself, not with the real-world phenomenon. what is discovered in this analysis might then lead the model to be useful in later epistemic and practical applications. in the last step, the model-world comparison, the results of the analysis are compared to the real-world system in order either to see how well the model simulations match the behaviour of the system, or at least to predict occurrences in it. after comparing the model with the real-world target, it is decided whether this model should be accepted or rejected based on fidelity criteria – criteria stipulated by the modeler concerning how good the model must fit the real-world target’s behaviour in order to be deemed acceptable. these criteria usually include a certain error tolerance. if the model meets the fidelity criteria, it will be accepted. if not, the model may be rejected and alternative models might be tested; or, alternatively, the existing structure of the model may be revised. given this general procedure of modelling, let me specify the instances of modelling that provide the relevant blueprint for understanding psychiatric diagnostics, namely: qualitative modelling and compositional diagnostic modelling. it is necessary to consider diagnostic reasoning as qualitative modelling in order to make my proposal realistic. for although most cases of modelling in the 11for a detailed discussion of such cases, see morrison’s (2007) insightful discussion of the relationship between theories and models, and giere’s (1999) famous showcase of classical mechanics and how it can inform modeling of real-world systems like the pendulum. 12a classic example of the reuse of theoretical model structures across disciplines are the lotkavolterra equations. the equations consist of a pair of first-order nonlinear differential equations which have been used as model structures in different scientific contexts. the equations were first used to model autocatalytic chemical reactions (lotka, 1910), later to model predator-prey interactions in ecology (lotka, 1925; volterra, 1926), and have even been applied in economics to model economic fluctuations (goodwin, 1967). more examples of model structure reuse across contexts can be found in knuuttila & loettgers (2016). 13as an example for this kind of model reuse, godfrey-smith (2009) proposes an occasion in which scientists develop a detailed model of how one particular trait became selected in the evolutionary process, and then apply the same structure they used to model the selection of this trait to model another trait’s selection. 14boumans (1999) spelled out some of these puzzle cases in the context of modeling in economics, where different sources – e.g., economic models, phenomenological laws, and assumed economic “laws” – are integrated into business-cycle models. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 13 (applied) sciences are quantitative, it would be implausible to expect the same for diagnostic reasoning. case studies used to explicate diagnostic reasoning for pedagogical purposes in psychiatry (see, e.g., wright et al., 2017) characterize diagnostic reasoning as verbal, with the occasional use of folk-mathematical talk of likeliness used to guide diagnostic inferences. these characteristics of diagnostic reasoning are confirmed by the findings of medical cognition research whose studies involve speaking out loud the protocols of medical experts (pinnock et al., 2015).15 diagnostic reasoning in medicine seems to take place qualitatively, not quantitatively. psychiatrists do not calculate their diagnoses; they think about them verbally, and in qualitative terms. accordingly, if diagnostic reasoning in psychiatry is a form of modelling, it should be expected to be a form of qualitative modelling.16 what, then, are qualitative models, and how do they differ from quantitative ones? the core difference between quantitative and qualitative models lies in the number of parameters and the discrete nature of the values these parameters and their relationships take. since qualitative models are often used in contexts in which the resulting models are meant to stay cognitively tractable, the number of parameters is often limited to a cognitively manageable maximum, and the values of these parameters, rather than being continuous parameters, are discrete in nature and in most cases also very limited in number – such as “present,” “absent,” or “neutral.” the same goes for the relationships between these elements that likewise take the form of qualitative values such as “increases,” “decreases,” or “irrelevant,” rather than indicating the influences between parameters in quantitative terms (forbus, 2008). because the values in the model structure are discrete, a qualitative model can be understood as symbolic and presented in various formats, including box and arrow diagrams, logical formulas, or a collection of interrelated, linguistically specified propositions. for my purposes this last format, which philosophers have called propositional models (thomson-jones, 2012), is of particular interest. given that most diagnostic reasoning takes place in terms of verbal reasoning, as demonstrated in speak out loud protocols, it is natural that a cognitively realistic inter15to produce speak out loud protocols, clinicians are presented with a diagnostic task and asked to say aloud what goes through their minds while they engage in diagnostic reasoning. 16the literature on qualitative modeling shows that it is not only a plausible approach to understanding expert reasoning in the context of diagnostic reasoning. it has also helped cognitive psychologists in their efforts to understand how we think about physical systems in everyday life, or how other technical experts think about the systems they handle, build and repair. it has also helped artificial intelligence researchers develop expert systems that support these experts in their tasks by mimicking their reasoning via qualitative models. the most intensely researched field related to qualitative models has been “qualitative physics,” which tries to understand the nature and use of qualitative models by engineers and other technical experts to think about artifacts they develop, use, or repair, in addition to the insights offered by quantitative models (see, e.g., bobrow, 1984; falkenhainer & forbus, 1991; weld & de kleer, 2013). beyond its application in qualitative physics, qualitative modeling has been used to understand expert reasoning in economics and engineering (farley, 1987) and, more recently, as a tool to develop highly idealized (but tractable) working models of complex marine ecosystems (reum et al., 2015). kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 14 pretation of the mental efforts in diagnostic reasoning would treat those mental efforts as a form of propositional, qualitative modelling. let me now turn from the characteristics of qualitative model structures to the act of developing and using qualitative models. the most important difference here is that qualitative modelling intentionally represents real-world systems in terms of discrete qualitative states, even though the corresponding aspect of the real-world system may not be discrete but continuous. though we know that water temperature can be described continuously, e.g., by the celsius temperature of the water, a qualitative model may only contain three values for the parameter of water temperature: cold, warm, and hot. this discretized type of representation is closer to everyday human reasoning about system features such as quantities, motion, space, time, causation, or frequencies. such everyday reasoning does not bother with differential equation and covariation analysis, nor with any other formal quantitative means of representation and modelling (forbus, 2008).17 given the importance and pervasiveness of discretized parameters, qualitative modelling involves a higher degree of idealization than quantitative modelling.18 the high degree of idealization in qualitative modelling usually prevents it from being the first choice of modelers who intend to develop a model that maximizes representational fidelity – the degree to which the model truly represents processes in the modelled system. however, in a context where representational fidelity is not a central requirement or not achievable, qualitative modelling can be highly beneficial. the main benefit of these models is, as already mentioned, that they allow cognitively tractable representations of the system that might be sufficient given the modeler’s interests and the model’s practical application. a qualitative 17note here that this understanding of qualitative theorizing differs from the understanding found in weisberg (2004), as well as from the discussion of qualitative theorizing in chemistry, e.g., in the work of hoffmann (1998). those qualitative models share many features with the account presented here, since weisberg also discusses, e.g., a high degree of idealization, and a typically restricted number of variables. but the qualitative models discussed here are not considered numerical, while weisberg explicitly states that, on his view, the difference of “qualitative and quantitative models is not about the use of numbers; both types of models can be numerical” (weisberg, 2004, p. 1071). 18to be precise, the limitations of parameters introduces a high degree aristotelian idealization (batterman, 2002): the act of stripping away the real-world system by limiting the scope of aspects of the real-world system considered in it. a discrete set of parameter values, on the other hand, leads to what is called galilean idealization (mcmullin, 1985): the act of intentionally distorting the representation of aspects of the real-world system by simplifying them. in principle, qualitative models could have infinitely many elements and discrete qualitative states so that the degree of idealization could be decreased. however, in practice this would undermine one of the main reasons why someone may choose this way of modeling: its computational simplicity and cognitive tractability. accordingly, when we talk about expert reasoning in terms of qualitative models, it seems right to expect them to be simpler given the limited computational power of humans. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 15 model of a system might also be preferred when attempts to provide a quantitative model fail owing to a lack of suitable quantitative data about the target system.19 let me now turn to diagnostic compositional modelling. as mentioned earlier, the approximated representations of real-world system behaviours in models can serve various epistemic and pragmatic purposes. they can, for example, be useful for predicting system behaviour, simulating aspects of its processes, or guiding interventions. models can also help indicate the presence of specific types of dysfunctions or irregularities in a modelled system. models set up for this latter purpose are the result of what is called diagnostic modelling.20 the standard approach to diagnostic modelling follows the assumption that the diagnosis of a systems error shall be identified by “the comparison of the observed system behaviour and the one which can be predicted with the use of the knowledge about the system model,” which itself requires “knowledge about the system model allowing for the prediction of its normal correct behaviour […] a model which can be used to simulate the normal work of the system in the case of lack of any faults” (falkenhainer & forbus, 1991, p. 140). but often our theoretical knowledge about systems is limited, and consequently our models based on those theoretical assumptions will be incomplete. nonetheless, it is still possible to use modelling to come to some diagnostic conclusions about real-world systems. this requires, however, deviating from the above-described standard approach. qualitative diagnostic modelling that starts with a less than fully-fledged and fine-grained model of the target system that it intends to diagnose is compositional diagnostic modelling (falkenhainer & forbus, 1991, p. 140). 19what has been said so far should make clear the principal idea of qualitative modeling rather than its boundaries: qualitative models can be more complex and ai researchers and mathematicians have worked out frameworks to give more technically rigorous representations of qualitative modeling through qualitative algebra (see forbus, 2008). due to the limited space here, i will not discuss any of these specific formal frameworks for qualitative models, but will instead stay with the understanding provided so far. what follows is therefore only handwaving in the direction of the relevant work in work in ai and mathematics. for concrete examples of proposals for formalized qualitative models of complex systems, one may look at the examples of two-valued models used to diagnose malfunctions in aircraft engines (abbott et al., 1988) or photocopiers (bell et al., 1994). a framework for employing three-valued formalizations based on positive (+), negative (-), or zero (0) values (on an ordinal scale) in order to model physical systems of varying levels of complexity can be found in de kleer & brown (1984). moreover, proposals have been made for the formalization of monotonic relationships (e.g., if a goes up, b goes up) between model elements and for compositional relationships (e.g., if a goes up, b goes up iff c goes down), as well as for how change over time can be considered in a time series of qualitative models. for discussion, see forbus (1984). 20historically this form of modeling was of special interest to cognitive scientists trying to understand how an expert reasoner (often an engineer) diagnoses failures in artificial systems. this approach to understanding their diagnostic reasoning started with the first formal proposals by reiter (1987). the approach is relatively young in comparison to attempts to apply hypothetical deductive, inductive, or abductive reasoning (ligęza & górny, 2017) theories to it. its more formal presentation has seen several modifications and extensions since reiter’s work (see, e.g., falkenhainer & forbus, 1991; hamscher, 1992/1995; weld & de kleer, 2013). kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 16 compositional modelling is “a strategy for organizing and reasoning about models of physical phenomena that addresses the following problem: given an artifact description and a query, produce a model of the artifact that is commensurate with the needs of the query” (falkenhainer & forbus, 1991, p. 96). applied to diagnostic modelling, the needs of the query are to look at some real-world system’s outputs and decide whether these outputs qualify as system errors, and, if so, what kind of error it is, based on models of sub-processes in the system that, as far as we know, can be responsible for the error. this does not necessarily require a full-fledged and exact theory of all interactions in the system, but leaves room for some idealization of the responsible processes, abstracting away aspects of the system less relevant for answering the particular diagnostic question it must answer. as falkenhainer and forbus put it, “answering questions about the through-put of a pump in a power plant requires neither consideration of the plant’s turbines and condensers, nor use of a quantum mechanical model of the pump’s parts and fluid” (falkenhainer & forbus, 1991, p. 96). let me make this procedure more concrete. instead of starting from complete, detailed models of systems, compositional modelling takes existing theories, models, or assumed principles (called domain theories) and uses them to develop models for the relevant aspects of the real-world system that the modeler is interested in. models set up this way are called “model fragments.” in the case of compositional diagnostic modelling, this would mean coming up with model fragments for specific dysfunctional or error outputs of the system. this description allows the modeler to capture reliable indicators in the modelled system, which are helpful for deciding whether a system’s output should be interpreted as a dysfunction of a certain type. this decision presupposes a taxonomy of errors in the system that is mapped onto the models – e.g., if it is an error of type a or type b. when the fragmented diagnostic models are found to apply to the system’s behaviour, all these applied models are considered together as a set of models for the aspects of the system’s behaviour that the modeler is interested in. the resulting set of models is called the scenario model (falkenhainer & forbus, 1991, p. 96).21 21the reader may recall the “puzzling” approach to model construal mentioned on p. 10. while there is, indeed, a similarity insofar as different theoretical sources may be used to set up a model structure, there is also an important difference. in the cases of puzzling, a whole model is constructed and tested in order to address all aspects of the system in one model. in this way the model can be evaluated for its performance, for example, in making predictions or simulating the system. but the scenario model does not have this quality. the scenario model is only constructed after its elemental models are tested and deemed capable of predictions or simulations. this feature of the model elements does not, however, carry over to the scenario model. to assume this would be to commit a mereological fallacy. indeed, the scenario model is the product of post-hoc theorizing that only brings together the previously gained information. but it itself does not add any epistemic value in terms of prediction or simulation to the contribution of the model fragments. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 17 4.2 clinical psychiatric diagnostics having completed my discussion of modelling, it is time to consider the diagnostic practices of clinical psychiatrists that i propose to be best understood as such modelling. for this i will focus on the core practices of clinical diagnostics: the mental status examination, the psychiatric interview, and, in certain contexts, cognitive and biological testing. while most clinicians agree that all these practices are central to clinical diagnostics, some might object that other widely used approaches are omitted – especially questionnaires and structured diagnostic interviews. the reason i do not discuss these is that i do not count them as constitutive core practices of proper, contemporary psychiatric diagnosis. discussing them is therefore unnecessary for my purposes.22 however, i will discuss in turn these other, truly constitutive practices of clinical diagnosis. two means of diagnostic information gathering are carried out in a face-to-face examination of the patient: the mental status examination and the psychiatric interview. the purpose of the mental status examination is to evaluate the different domains of cognitive functioning, such as perception, memory, thinking, affect, time-orientation, and thought-order, for the sake of identifying potentially psychopathological anomalies. this is done by the psychiatrist observing the patient’s behaviour as well as listening to the patient’s self-reports in response to concrete questions about his current experiences (casey & brendan, 2019; trzepacz & baker, 1993). the general idea behind the psychiatric interview is to seek a broader scope of self-report-based information about the current and past psychological and social functioning of the patient. this involves, for example, gathering information about employment, friendships and other relationships, housing situation, forensic history, substance abuse, sex drive, eating behaviour, sleeping habits, as well as other background information about family history, education, previous medical problems, and so forth (boland et al., 2022; poole & higgo, 2017). 22by proper and contemporary clinical psychiatric diagnosis, i mean clinical diagnostics carried out in psychiatry today and insofar as it is done so in accordance with the standards of the community of clinical professionals, as articulated by the most relevant expert communities within that group (e.g., the american psychiatric association). when i say that some diagnostic practices of contemporary proper psychiatric diagnostics are constitutive and others are not, i mean that they are necessary and sufficient for a procedure to count as a proper contemporary clinical diagnostic procedure. this way i try to address what is at the heart of diagnostic practice rather than spending time discussing what is (at least currently) accidental to it. by taking this approach i draw on tyler burge, who considers the explication of the constitutive conditions of a phenomenon as a matter of presenting the necessary and or sufficient conditions for something’s being what it is: “a constitutive question concerns conditions on something’s being what it is, in the most basic way. something cannot fail to be what it is, in this way, and be that something. constitutive conditions are necessary or sufficient conditions for something’s being what it is in this basic way. to be constitutive, the conditions must be capable of grounding ideal explanations of something’s nature, or basic way of being” (burge, 2010, xv). kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 18 cognitive testing is employed by a psychiatrist in order to evaluate the cognitive performance of patients, whereas biological testing is used to evaluate the presence or absence of biological markers that suggest the presence of disorders. cognitive tests are structured examinations consisting of questions posed to the patients (e.g., “what day is today?”) and cognitive-behavioural tasks to be executed (e.g., “please remember and repeat the following words,” “pick up the pen with your right hand and draw this clock”). the outcomes are scored and compared to cut-off criteria to decide whether anomalies are present. the cambridge cognitive to explain further how i distinguish what is constitutive and what not, let me say the following: employing questionnaires and structured interviews (as opposed to the discussed diagnostic means) is not necessary for comprehensive psychiatric assessment, nor is it sufficient to gather the diagnostic information required for a comprehensive diagnostic process. rather than counting as core diagnostic practices, questionnaires and structured diagnostic interviews are useful complements to those practices. since questionnaires or structured interviews are complements, including them in my presentation would be unnecessary given my focus on core procedures of psychiatric diagnostics. i do not claim that questionnaires or structured interviews are useless or irrelevant. questionnaires such as the beck depression inventory (beck et al., 1961) can be useful for screening, and structured interviews such the structured clinical interview for dsm-5 (first et al., 2016) can help attain lots of important diagnostic information. all i want to say is that a look at psychiatric practice shows that diagnostic information gathering by questionnaires and structured interviews plays a subordinate role in clinical diagnostics. they are practices done in the context of psychiatric diagnostics, but they do not individuate it. consider questionnaires. a responsible clinician would not make a categorical disorder diagnosis solely based on the answers to a questionnaire, nor can a diagnostic case formulation be provided based on them. however, drawing diagnostic conclusions without using questionnaires is not improper to diagnostic practice in psychiatry. questionnaires can contribute to a diagnosis, but only when interpreted in the context of an overall clinical impression, generated from what i consider to be the three core procedures. now consider structured interviews. at face value, they are not a necessary component of psychiatric diagnostics. we find them rarely used outside of research contexts, though there are some contexts in that they are used more often than in others – e.g. in autism diagnostics (tarbox et al., 2016). rather than being essential for proper clinical diagnostics, they are an essential tool to clinical research (aboraya, 2009; bruchmüller et al., 2011; mueller & segal, 2015). that these practices are not sufficient to make a diagnosis can also be shown along two lines. first, structured interviews do not yield the information necessary for providing a case formulation (discussed in the next section). a case formulation must be provided as part of the diagnostic proposal, and it needs to include, among other things, biographical and psychosocial data from patients that are not obtained in typical structured interviews, but only in the psychiatric interview. second, structured interviews usually ask questions that explicitly mention symptoms relevant for categorical diagnoses. such interviews thereby hope to attain answers that allow one to make a diagnosis based on those answers. however, as research suggests, experienced clinicians do not just consider the answers to these question in evaluating patients. they also consider a wide range of patient behaviors they observe in their contact with the patient (nakash & alegría, 2013). that sort of information is usually gathered instead by the mental status examination. such examinations are an important aspect of diagnostic practice for clinicians. research suggests, for example, that scid interviews carried out by laypeople who cannot perform the mental status examination have relatively low validity when measured against the diagnostic judgments of expert clinicians (nordgaard et al., 2012). in conclusion, using a structured interview does not make up for the psychiatric interview nor does it make superfluous a mental status examination. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 19 examination revised (roth et al., 1986) or the alzheimer disease assessment scale (adas) (hodges, 2018) are examples of such tests, and they are mostly relevant for diagnosing dementia and other neurodegenerative disorders. biological testing uses specific biological markers that indicate the presence or absence of specific conditions that in turn can inform (differential) diagnoses. such testing includes serological testing, genetic testing, as well as radiological examinations. here are some examples of how such testing informs diagnoses: liquor analysis can reveal levels of β-amyloid, total tau, and phospho-tau-181, which indicate the presence of irreversible forms of dementia (reitz & mayeux, 2014). genetic testing can show whether patients are carriers of ultra-high-risk genes for developing huntington (myers, 2004). neuroimaging data can be important for identifying strokes or major structural alterations of brain substances that may be responsible for cognitive and behavioural changes (first et al., 2018; power et al., 2016). neuroimaging data also allows for subtyping in prefrontal lobe dementia over alzheimer’s dementia, indicated by atrophy in the prefrontal lobe depicted in neuroimaging data (rohrer & rosen, 2013).23 again, such testing mainly supports the diagnoses of neurodegenerative diseases, but it can also be relevant for reaching differential diagnostic conclusions that render a psycho-behavioural condition into a non-psychiatric case. for example, if the patient is found to have a brain tumour or brain lesion, that can be assumed to cause their condition.24 these three domains of diagnostic evaluation – mental status examination, psychiatric interview, and cognitive and biological testing – make different contributions to the process of psychiatric diagnosis. this process itself can be divided into three steps: the screening procedure, the diagnostic in-depth evaluation, and the drawing of diagnostic conclusions. the mental status examination as well as the psychiatric interview contribute to both the screening procedure as well as the indepth evaluation. in the current state of psychiatry, the cognitive and biological testing, if they are used, only contribute to this second, evaluative step. but what exactly is the purpose of these two steps in the diagnostic process? the purpose of screening is to arrive at a list of patient complaints which, after be23i am aware that success in this domain of diagnostics is still limited. the method is not very good at differentiating between alzheimer’s disease and other forms of dementia (e.g., lewy body-, frontal temporal-, and vascular dementia) (maclin et al., 2019). moreover, the method has not yet been able to diagnose parkinson’s disease (he et al., 2018; miller & o’callaghan, 2015); huntington (silajdžić & björkqvist, 2018) or amyotrophic lateral sclerosis (verber & shaw, 2020). 24by talking about neurodegenerative diseases as psychiatric disorders, and by regarding psychological abnormalities caused by brain lesions as not being psychiatric disorders, i do not want to take a stance in the ongoing ontological debate as to whether mental disorders are brain disorders (e.g., boorse, 1977; graham, 2021; insel & cuthbert, 2015; insel & quirion, 2005; jefferson, 2020; miller, 2010; olbert & gala, 2015; papineau, 1994; schramme, 2013). nor do i want to take a stance in the related debate as to whether we should distinguish between neuropsychiatric or psychiatric disorders in clinical contexts (e.g., baker, 2002; david & nicholson, 2015; price et al., 2000). i instead adopt the current standard in psychiatry of regarding neurodegenerative disorders as part of its responsibility, but not the psychological conditions resulting from brain tumors or lesions. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 20 ing subject to more in-depth evaluation, can be classified as psychiatric symptoms and signs (or not). the list of complaints is formed by paying attention to prima facie obstructed aspects of the patient’s psychology and behaviour. these aspects appear sufficiently similar to psychopathological phenomena so as to warrant spending more time carefully examining whether they are, indeed, psychopathological symptoms or signs. if such complaints are discovered in the initial screening, an in-depth evaluation of the candidates usually follows in the same face-to-face interview. the clinician does this by asking further questions or conducting simple psycho-behavioural tests. if it is decided that cognitive and biological tests are necessary for the in-depth evaluation, this will, for practical reasons (e.g., getting material to take samples for the lab, or preparing the cognitive test) take place during another encounter.25 an example of one such in-depth evaluation will be discussed later. conducting an in-depth evaluation usually allows the clinician to draw two sorts of diagnostic conclusions. the first kind of conclusion concerns whether the enumerated complaints should be regarded as psychopathological symptoms or signs, or whether they instead indicate some sort of psycho-behavioural problem caused by a psychopathological condition. in the first case, it may be concluded that the complaints derive from a medical but non-psychopathological problem. or it may be that the psychological or behavioural complications turn out to be of no diagnostic relevance.26 if this is decided, the next step is to decide which diagnosis, if any, should be given to the evaluated patient. to do this is to make a diagnostic proposal. according to the standards of psychiatric expert associations like the american psychiatric association (2013), such a proposal shall consist of 25often the mental status examination and the psychiatric interview – and therefore the screening procedure and the bulk of the in-depth evaluation – are carried out by one clinician in a single encounter with the patient. in principle, it is also possible first to do a complete screening, and on another occasion to do an in-depth evaluation. that is, it is possible to do the mental status examination on one occasion and the psychiatric interview on another. but it is very handy to do both at the same time because many observations important for the mental status examination (e.g., regarding language production) can basically be made on the sideline of the psychiatric interview. the two are therefore often lumped together. 26why should two hypothetically similar token behaviors or mental states be typed as psychiatric symptoms or signs on one occasion, but as non-psychiatric ones on another? that is not due to some strong metaphysical distinction, but rather due to the special place that psychiatric symptoms and signs currently have in medical semiology (altable, 2012). in medicine symptoms are traditionally considered manifestations of a disease, or, to put it in more philosophical terms, they are representations of the presence of disease, and therefore of physiological alterations considered causally responsible for their presence. if a symptom or sign is caused by a disease condition that is not considered a mental disorder, it is, for the purpose of providing diagnoses of psychiatric disorders, not considered to be a psychiatric sign or symptom. this does not mean that, in the end, research might show that part of the causal pathways responsible for the occurrence of the symptoms is shared by a psychiatric disorder and a disease with similar psychological or behavioral symptoms. for an alternative view of medical semiotics in psychiatry, see borsboom (2017). kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 21 a diagnostic case formulation and a manual based diagnostic classification.27 both parts shall be discussed in turn. a case formulation is put forward based on the diagnostic information gathered in the diagnostic assessment and the in-depth evaluation. the formulation essentially presents the psychiatrist’s “analysis and integration of information” (gelder, 2000, p. 73). as such, it contains a careful clinical history of the patient, a list of the patient’s complaints gathered in the diagnostic process, and the psychiatrist’s decisions concerning which of the patient’s complaints should be interpreted as psychiatric symptoms and signs, and which should not. in addition to presenting a list of initially suspected and endorsed symptoms, the case formulation will present the information that contributed to the decision as to which of the complaints are deemed symptoms or signs. the case formulation thereby identifies the “social, psychological, and biological factors that may have contributed to developing a given mental disorder” (american psychiatric association, 2013, p. 19). by organizing diagnostic information this way and relating it to complaints, the case formulation intends to provide a structured presentation of diagnostic information that explains the patient’s complaints. the structure of the presentation allows a case formulation to present the justificatory grounds for attributing the psychopathologically relevant features to the patient.28 the other part of the diagnostic proposal is the diagnostic classification. one approach to diagnostic classification that has been around since the diagnostic statistic manual (dsm) iii (american psychiatric association, 1986) and the international classification of disease (icd) 10 (world health organization, 1992) – a classification still prevalent among the most recent editions of most diagnostic manuals – is based on the presence or absence of clusters of symptoms. these clusters consist in necessary criteria plus further diagnostic criteria from a fixed list which, in combination, are sufficient to diagnose a disorder. the diagnostic criteria for major depression (american psychiatric association, 2013, p. 160), for example, allow for 50 different combinations of signs of symptoms. diagnostic criteria can be either fulfilled (symptom present) or not fulfilled (symptom absent). 27note however that not only the apa requires the case formulation to be a proper part of psychiatric diagnostics in addition to the more widely discussed categorical diagnosis (outside the united states sometimes called clinical formulations). case formulations are widely recognized as diagnostic standard in psychiatric diagnostics as a look at proclamations and education guidelines of various expert societies show, that consider it a core competence in diagnostics, and part of good psychiatric practice. see for example american board of psychiatry & neurology (2011); royal college of psychiatrists (2013); royal australian and new zealand college of psychiatrists (2014); royal college of psychiatrists (2013); royal college of psychiatrists (2017); royal college of physicians and surgeons of canada (2019). 28while approaches differ in how exactly a case formulation should look, what has been said here appears to representative of the core idea assumed across the literature on case formulations. to compare, see e.g., meyer & turkat (1979); varghese & mellsop (1983); weerasekera (1996); mchugh & slavney (1998); butler (1998); mcwilliams (1999); eells (2010)); kuruvilla & kuruvilla (2010); chisolm & lyketsos (2012); goldman & greenberg (2015)); kirk & werbrook (2016); ryan (2020). kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 22 in some cases, further criteria such as temporal qualifications (e.g., the condition must be present for at least two weeks) or environmental factors (e.g., the experience of a life-threatening, dangerous, or significantly abusive circumstance for post-traumatic stress disorder) are explicitly mentioned. in addition to this long-time standard approach, the recent editions of the dsm and the upcoming icd-11 include a new approach. they include dimensional ratings as part of categorical diagnostics.29 the general idea behind dimensional ratings is to evaluate the presence of at least some psychopathological features on an ordinal severity rating scale rather than as simply being present or absent. dimensional ratings have been introduced as mandatory in the evaluation of diagnostic criteria for some mental disorder categorizations in the dsm-5 (e.g., autism spectrum disorder, intellectual disability) and as optional for others (e.g., primary psychotic disorder and personality disorders). dimensional ratings have also been made mandatory in some disorders categorized by icd-11 (e.g., autism spectrum disorder, personality disorder) and optional for others (e.g., primary psychotic disorders). while the basic idea is always the same, the use of dimensional diagnostics can have different faces. in some instances, dimensional rating systems are simply used as add-ons to the specification of present symptoms (e.g., whether the delusions present are clinically mild, moderate, or severe). in other cases, as in the personality disorder diagnostics in icd-11, significant changes accompanied the implementation of dimensional diagnostics. in the case of personality disorder diagnostics, for example, this change was a deflation of the rich personality disorder taxonomy present in icd-10 in favour of one general personality disorder category that is then further specified by severity (none, difficulty, mild, moderate, or severe) in light of dimensional ratings of the patient’s selfand social functioning. although the the american psychiatric association does not explicitly make a statement on the relationship between the diagnostic case formulation and the diagnostic classification, it provides some remarks regarding what is necessary and sufficient to make a psychiatric diagnosis. in so doing, it introduces a notion of diagnostic “clinical judgment” that allows us to reconstruct the relationship in question. the state that for reaching a categorical diagnosis “[…] it is not sufficient to simply check off the symptoms in the diagnostic criteria to make a mental disorder diagnosis” (american psychiatric association, 2013, p. 19, my emphasis). for, in addition, the “relative severity and valence of individual criteria and their 29these changes were introduced following the increased interest in psychiatric research in the idea that at least some psychopathological features as occur on a spectrum. proposals along this line have been made already, for example, for personality disorders (trull & durrett, 2005) and psychosis (cuthbert & morris, 2021; esterberg & compton, 2009). they were adopted in one way or another by relevant research movements in the field, such as the national institutes of mental health rdoc-project (national institute of mental health, 2013), or the hitop research consortium (kotov et al., 2017). discussing the scientific and clinical motivations for a dimensional understanding of mental disorders is beyond the scope of my project. for discussion of these motivations, see, e.g., helzer & association (2008); krueger & bezdjian (2009); adam (2013); reed et al. (2019). kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 23 contribution to a diagnosis requires clinical judgment” (2013, p. 19). clinical diagnostic judgment, however, is a capacity that “requires clinical training” to acquire. once acquired, it enables a psychiatrist “to recognize when the combination of predisposing, precipitating, perpetuating, and protective factors has resulted in a psychopathological condition” (2013, p. 19). this description of diagnostic clinical judgment necessary for determining which patient complaints qualify as symptoms and signs and with what severity turns out to be the reasoning process by which the psychiatrist develops their case formulation. the case formulation sets down the results of the psychiatrist’s analysis of the diagnostic information. it records which complaints are psychiatric symptoms or signs. this identification of symptom and signs in a case formulation allows for a quicker application of the proposed diagnostic categories and the justification of their application. the profiles of categories, consisting in lists of signs, symptoms, and their severity, can be compared to the psychopathological conditions determined to be present by the diagnostic case formulation. this allows for a diagnostic category to be chosen. when presented alongside the categorical diagnosis, the case formulation makes transparent the reasons for which a specific categorical choice was made. 4.3 clinical psychiatric diagnostics as modelling after giving this brief overview of the core components and usual steps in the process of clinical diagnostics – both in accordance with the standards of the clinical expert community – the next step is to map these components and processes onto the account of modelling presented earlier. as discussed in section 4.1, modelling usually starts from relatively little initial information about the modelled system. this information does little more than inform the modeler about the specific features of the system that will be modelled. in the context of clinical diagnostics, this step is the initial screening procedure, since its task it is to inform the psychiatrist about the specific features of the patient, i.e., the system he encounters given his professional interest in clinical diagnostics. the patient complaints are, according to the aforementioned modelling framework, the systems features that will be modelled.30 based on the initially gathered information about the patient’s complaints, the psychiatrist begins to theorize about the potential psychopathological, general medical, or non-medical reasons that might lead to the presentation of those complaints. this theorizing is based on the psychiatrist’s background knowledge 30please note that “complaints,” as the term is used here, are not necessarily subjective. complaints may entail aspects of patients’ behavior or mental life coming up in the screening procedure, though the patient is not reporting suffering from it. this is the case if these aspects are potentially psychiatric-relevant symptoms. if e.g., in the screening phase of diagnostics, a patient says they only sleep three hours a night, but that is fine for them, sleep problems will none the less be added to the list of complaints. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 24 about psychopathology, medicine, and folk psychology. the psychiatrist then determines the conditions that should be present in the patient if the presented complaint were a certain psychopathological, medical, or non-medical condition. that condition might subsequently be evaluated in a closer examination of the patient. in the language of modelling, this means that the psychiatrist sets up multiple diagnostic models for the identified complaints. those models are based on domainspecific background knowledge about the potential conditions that might be responsible for the presented complaints. this process corresponds in the modelling process to the construal step. this has implications for the model’s dynamics: the absence or presence of a parameter, along with the degree of severity of that parameter, has consequences for other parameters in the model. in the application of this model to the real-world system, i.e., in the in-depth evaluation of patient complaints, the goodness of fit can be assessed by the psychiatrist. this corresponds to the analysis step in the modelling process. models set up this way in the process of diagnostic reasoning are usually qualitative ones. they present the psychiatrist’s thought as a set of propositions that express the model’s structure and implications. the models are qualitative, since they are non-numeric but symbolic, and with parameters that are usually discrete. they are symbolic because they appear in the form of linguistic thought. their parameters are discrete because their values are usually determined either on a presence/absence dichotomy scale, or an ordinal scale indicating the degree of expression of the trait in the patient (e.g., a three-step absent-mild-severe scale). setting up these models and becoming aware of their implications through analysis is often a simple task frequently practiced by highly trained diagnostic experts who know the relevant background theories by heart and have used such models in the past. to develop new theories or gain new insights might, however, demand extra effort. the next step of the diagnostic procedure is the in-depth evaluation, which corresponds to the model-world comparison step in the modelling process. for every potential psychopathological condition, the psychiatrist engages in an in-depth evaluation of the patient in order to assess which of their theorized reasons for the presence of the complaint best fits the overall presentation of the patient.31 this confirms that the clinical diagnostics is a form of diagnostic and compositional modelling. for every complaint suspected to be of potential psychopathological relevance, models are established for determining whether the complaint is a psychopathological condition (i.e., a psychiatric symptom or sign). also established are models indicating whether the complaint derives from a non-psychiatric medical condition, or does not qualify as a medically relevant condition at all. the best-fitting model will be selected while the other models will be rejected. the selected models indicate how the evaluated complaints are to be interpreted. if, 31it might not always be possible to make such decisions unambiguously, and this leads to the wellknown problem of diagnostic ambivalence. due to lack of space, i cannot discuss this problem here. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 25 for example, a model is selected whose structure is provided by a theory about a certain psychiatric symptom, the condition will be determined to be an instance of that symptom. these selected models indicating the correct diagnostic evaluations of the complaints do not constitute an overall model of the patient as a whole system, nor do they constitute a model of the relationships between the different psychopathological complaints. instead, they are model fragments which, in accordance with what has been said about compositional modelling earlier, are brought together in what has been called a scenario model. this brings me to the last step of the diagnostic procedure: the drawing of diagnostic conclusions. when it comes to drawing such conclusions, psychiatrists do two things: (i) they develop a scenario model based on the model elements presented verbally in the diagnostic case formulation, and (ii) they make a categorical disorder diagnosis based on the symptoms and signs indicated by the accepted models. in the first step, the psychiatrist sets up a case formulation, whether it be only for themselves or to present to other clinicians. in the case formulation they bring together all diagnostic information relevant for the evaluation of the complaint, and they do so in a way showing how that information supports the chosen evaluation. sometimes other plausible diagnostic options might also be mentioned, and the model needed for that alternative option is spelled out and shown to fit less well than the one the psychiatrist accepted in the end. based on this evaluation, the psychiatrist then extracts the list of psychiatric symptoms and signs and, if relevant, the severity of those symptoms and signs based on the scenario model. the psychiatrist matches this list with the taxonomy of the relevant diagnostic manual used in order to provide a categorical disorder diagnosis based on the defined criteria. this brings an end to the diagnostic process, which turns out to be a model-based inference to the presence of one or more psychiatric disorders. this last step of the process does not necessarily require the psychiatrist to look up diagnostic categories in books. in fact, this last step often takes place rapidly for experienced clinicians who know by heart the criteria of most if not all the disorders they regularly encounter in their field. to make all this a bit more tangible, let me provide a brief example of what is perhaps the most elusive part of this procedure: the process from screening to the selection of the adequate model element or, in other words, the diagnostic classification of a specific psychopathological condition. consider a patient who shows an atypical kind of formal language production behaviour (as opposed to the atypical production of language content) that stands out to the clinician as potentially indicating a psychiatric condition. to illustrate how the abnormality may appear, consider the following table: noticeable normal psychiatrist: good morning, mr. x. what can i do for you? psychiatrist: good morning, mr. x. what can i do for you? kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 26 noticeable normal patient: you can help me. patient: good morning, doctor. i came to you because i have some problems that i think i need help with… psychiatrist: and i will try my best to do so. can you tell me something about the reason you reached out for help? psychiatrist: and i will try my best to do so. can you tell me something about the reason you reach out for help? patient: yes patient: well, thanks. i feel sad and empty, and i do not know what i should do about it. it started… what can be recognized in the patient is a presentation of reluctant speech – a presentation which, on closer examination, may turn out to be “poverty of speech,” a form of alogia. alogia is a psychiatric symptom which can take the form of poverty of language with respect to the quantity of speech. it is considered present in various mental and neuro-psychiatric disorders such as dementia, schizophrenia, severe depression, or schizotypal personality disorders (american board of psychiatry & neurology, 2011, p. 817). whether the patient indeed suffers from this condition will depend on a closer evaluation that must take place in order to decide whether the objective complaint of reluctant speech is, indeed, best diagnosed as alogia. to do so, the psychiatrist must consider different models that might be applied to the patient and evaluated. let us look at some of the theoretical considerations and the models that can be set up based on them that, if they apply, would lead to different evaluations. a first consideration might be that the behaviour of the patient is not psychopathologically relevant but part of their usual premorbid behaviour. as discussed in the psychiatric literature, sometimes people who are found to be unusually pedantic in their speech may evoke the impression of poverty of speech. this is a problem that has been observed in the case of administrators, politicians, scientists, and (guess what) philosophers (andreasen, 2016). if this were the case here, the complaint would not seem to qualify for the diagnosis of alogia. another reason the patient may show reluctant speech is that they feel uncomfortable or have anxiety in the interview situation. the patient may consequently choose his words carefully and sparingly, for example, because they fear the psychiatrist may “take them in.” again, if this would be the case, the complaint would not seem to qualify as a case of alogia. let me now come to a strain of theorizing that can lead to a model whose successful application would support a diagnosis of alogia. the chief cognitive impairment behind alogia in psychiatric cases involves an impairment of control retrieval – part of the executive functioning that enables the retrieval of information from memory. alogia occurs when the information is not automatically rekind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 27 trieved, or when there is more than one potential piece of information matching the search profile (docherty et al., 2011; doughty & done, 2009; wagner et al., 2001). if a test of speech production, conducted on a cooperative patient, shows patterns indicating this kind of cognitive impairment, diagnosis of alogia seems warranted. in order to evaluate these different options and decide whether the complaint should be evaluated as a psychiatric symptom, the psychiatrist has to come up with a more precise idea as to what diagnostically attained information would support which of these assumptions. that is, the psychiatrist needs to set up models that can be evaluated during the in-depth evaluation. this will usually lead the psychiatrist to consider a set of interrelated propositions for each of his different diagnostic hypotheses. the psychiatrists may for example test the hypothesis that the patient’s reluctant speech is not alogia, but simply a feature of her premorbid style of communication. she might be philosopher used to providing information as densely as possible – not only in papers for analysis, but also in verbal communication. for this purpose the psychiatrist may set up and evaluate the following set of propositions: i) the patient can elaborate their answers if asked to. ii) the patient recognizes that their answers are unusually short and can justify their manner of speaking by explaining their motivation (e.g., wanting to save the doctor’s time, or wanting to be as precise as possible). iii) the patient can report that her way of presenting information is nothing that has developed recently, but is rather her normal way of conveying information. if possible, the psychiatrist also speaks to relatives, friends, or other medical professionals in order to verify this last statement so that: iv) people who know the patient report that the patient always tended to speak this way. to evaluate the hypothesis that the patient’s reluctant speech might be due to a lack of motivation or mistrust in speaking to the psychiatrist, the psychiatrist will seek further information about the patient in the in depth-evaluation interview. this information would suggest the following: i) the patient has little to no intrinsic motivation to speak with the > psychiatrist. ii) if the patient comes to appreciate the potential help that treatment > could offer, her speech pattern changes. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 28 and if it is possible to speak to relatives, or if personal contacts of the patient can be contacted, the psychiatrist may reach out to them and evaluate: iii) if relatives or other personal contacts who have recently seen her > report either that she has not shown laconic speech previously or > that it is typical for her to behave that way in certain > circumstances. finally, let us come to the propositional model which the psychiatrist may set up and which suggests a positive evaluation of the presence of alogia as a psychiatric symptom. this will require saying a bit more about how the psychiatrist evaluates this condition based on their psychopathological insights. as already mentioned, alogia is considered a condition deriving from a disorder of the cognitive function of control retrieval – a part of the executive function that enables the retrieval of information from memory when either that information is not automatically retrieved or when there is more than one potential piece of information matching the search profile. this cognitive function can be tested with verbal fluency tasks. such tasks require subjects to follow a production rule in voicing words. they may be required, for example, to say words beginning with a certain letter (testing word-letter fluency) or falling into a category, such as animals (testing word fluency). more specifically, when tested for alogia, an individual would be asked to produce lexical items for a certain span of time. if an individual suffers from a cognitive impairment of control retrieval, there is an increased mean response latency between each reported word when asked to produce words of a category. if the individual does not show this deficit, that suggests the impairment of other language-related cognitive functions that, in principle, could also lead to the clinical presentation. these other impairments include disorganized semantic memory – which would lead to poorer performance in category fluency relative to letter fluency – or context processing – which should lead to a decrease in the proportion of correctly reported semantically-related words (docherty et al., 2011). if verbal fluency testing of the patient meets this prediction, it may be reasonably concluded that the patient’s complaints are an instance of the symptom of alogia. in accordance with these insights into the underlying psychology of alogia, the psychiatrist will construe a model consisting of a simple proposition that can be evaluated during an in-depth evaluation in cognitive testing, as described above: i) the patient has an impaired capacity for controlled retrieval leading to the phenomenological presence of reluctant speech.32 32besides this etiology of alogia, it is also possible for patients to suffer from a phenomenologically similar condition due to traumatic brain injury (tbi). this option has to be evaluated as well. there is, therefore, need for an additional model requiring in-depth evaluation as part of the psychiatric interview and the mental status examination. further evidence may be needed from biological testing – in this case, by computed tomography (ct) or magnetic resonance imaging (mri). these tests would potentially support the case for a brain lesion that, if present, kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 29 when all these models are evaluated against the information collected during the in-depth evaluation, the psychiatrist must decide which model best applies to the patient’s case. the psychiatrist thereby makes a decision as to whether the patient’s complaint should be evaluated as a psychopathologically relevant symptom. with this decision, in combination with all other similarly achieved decisions about the other complaints of the patient, the psychiatrist will have a selection of model elements at hand. each model element indicates whether a particular complaint should be evaluated as a psychopathological symptom or not, and, if yes, which one. these results will then be used to provide a case formulation for the scenario model in which, to continue with our example, the description of the patient will contain a sentence like: “when first meeting the patient, he showed a significant abnormality in verbal behaviour: a noticeable reluctance of speech. for example, when the patient was asked […] they answered […]” later in the case formulation, the formulation will come back to potential complaints like this one and state how it has been evaluated based on the accepted model for the patient’s condition. for example, the case formulation may state: “based on a cognitive evaluation (verbal fluency tasks), it is most plausible to assume that the patient’s reluctant speech qualifies as psychiatrically relevant alogia.” the psychiatrist may also spell out other diagnostic hypotheses and why they failed to apply. the case formulation might state: “the speech behaviours do not appear to be a feature of the patient’s premorbid personality, nor to result from reservations regarding the use of psychiatric service.” such a statement would be followed by the negatively evaluated propositions comprising the particular diagnostic model. the case formulation will ultimately provide a list of psychiatric symptoms or signs that then can be used to decide which diagnostic categories would apply to the patient’s case, if any. since the rest of the case has not been spelled out, in this simplified example it would remain open whether the patient may fulfil criteria supporting any of the following diagnoses for which alogia is a potential feature, namely: dementia, schizophrenia, severe depression, or schizotypal personality disorders (american psychiatric association, 2013, p. 817). can but need not show up in neuroimaging data (national academies of sciences, engineering, and medicine, 2019). to consider this option, the psychiatrist would evaluate whether: • the patient recently took some sort of blow to the head (e.g., by falling or having an accident). • if the patient did suffer some such blow, if the patient lost consciousness or has any loss of memory of events immediately before or after the accident. • if there were any alterations of mental state at the time of the accident (e.g., feeling dazed, disoriented, or confused). • if lesions that indicate traumatic brain damage can be seen in a neuroimaging test. if this model fits, it might turn out to be more plausible to consider the present language impairment not as a symptom of a psychiatric disorder, but rather the psychological symptom of a brain disorder that should not be considered when diagnosing a psychiatric disorder like schizophrenia. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 30 4.4 assessing the model-based account of psychiatric diagnostic reasoning how good does this model-based account of psychiatric diagnostic reasoning perform against the phenomenological account? to answer this question, let me once more bring up the four problems that beset the phenomenological account. first, pattern recognition plays a significant role in diagnostic reasoning, but it has no place in the phenomenological account. second, analytic reasoning also has an important place in diagnostic reasoning, but it is not to be found in the phenomenological account. third, the phenomenological account leaves no room for meaningful intrapersonal critical diagnostic reasoning. fourth, the phenomenological account also denies the need for the critical interpersonal diagnostic reasoning we find, for example, in expert teams. does my model-based account acknowledge or accommodate these features of diagnostic reasoning? it does. it uses pattern recognition in the inference from symptoms to disorder, analytic reasoning for model choice, and analytic reasoning to enable interpersonal and intrapersonal conflict. in the diagnostic process described by the model-based account, pattern recognition occurs in the last step of the diagnostic procedure. it involves the inference from the present symptoms to the disorder diagnosis. based on the identified symptoms of the patient, a well-trained psychiatrist who knows the nosology of their field by heart can make rapid inferences to the right categorization of the disorder. the inference is a form of pattern recognition that takes the identified symptoms as cues for identifying a pattern assigned to a disorder category. in this way the model-based account includes pattern recognition for experienced clinicians. it does so, moreover, in the same way pattern recognition in diagnostic reasoning is usually described. my account is, in this respect, in line with common sense medical thinking and medical education research. analytic reasoning also plays a significant role in the model-based account of diagnostic reasoning. analytic reasoning in the process of diagnosis involves weighing the evidence carefully in order to determine which diagnostic conclusions are most plausible given all available evidence. such reasoning clearly takes place in the model-based account, and in particular in the step involving the evaluation of different potential models that may apply to patients. here again, the model-based account is in keeping with common sense thinking and medical education research. the model-based account also regards critical diagnostic reasoning as an intraand interpersonal process. recall that critical diagnostic reasoning is the activity by which the clinician actively questions their own diagnostic conclusions (or that of another) and makes transparent to themselves (or others) why they made the particular diagnostic decision they did. in contrast to the phenomenological account, the model-based account acknowledges that critical diagnostic reasoning takes place. psychiatrists can reconsider for themselves why they have made the diagnostic decision regarding a certain symptom or sign. they can do this by explikind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 31 cating not only all the models they have considered and tested during the in-depth evaluation, but also why they think the model they prefer best suits the patient’s case. in further contrast to the phenomenological account, it may be possible that by going through one’s own reasoning, the psychiatrist may discover an error. maybe there is another model they forgot to evaluate? maybe there is new evidence that suggests a change in the models considered? or maybe the patient’s reports actually support the proposition of a model that had been refuted? such considerations can then lead to re-evaluating a diagnostic conclusion about a symptom or sign. as a result, it can also lead to a modified diagnosis. the same goes for the intersubjective level at which clinical experts engage in critical discussions about one another’s diagnostic conclusions. by discussing the case formulation and the reasoning standing behind it, clinicians critically reflect on the models their colleagues have adopted and tested. they may argue, for example, that a model is inadequate given more recent evidence, or that a model that should have been tested was not considered. or maybe by reporting what a patient did or said, a colleague might be able to argue in favour of a proposition that was initially rejected. such discussions may similarly lead to the re-evaluation of some complaints of the patient and may, as a result, change the diagnosis. in conclusion, it appears that the model-based account can handle all the problems that beset the phenomenological account of psychiatric diagnostic reasoning. the model-based account also appears to capture more closely the actual process of psychiatric diagnoses, as opposed to the rather vague statements by phenomenologists about how psychiatrists in practice arrive at their conclusions. 5 conclusion the starting point of this paper was what i called the methodological question: what is the method of psychiatric diagnostic reasoning? i briefly discussed the relevance of this question. after that, i introduced one of the more widely recognized answers that have been given to this question: the phenomenological approach. i then pointed out four problems, some of them empirical, others theoretical, that this account faces and that motivated the need for an alternative answer to the methodological question. i then introduced and illustrated one such alternative: the model-based account of psychiatric diagnostic reasoning. this model drew an analogy between clinical psychiatric diagnostics and modelling – more precisely: qualitative, diagnostic, compositional modelling. after spelling out this account in some detail, i came back to the initially presented problems faced by the phenomenologist account, arguing that the same problems do not plague the model-based account. in conclusion, it appears that the model-based account of psychiatric diagnostic reasoning is a preferable answer to the methodological question. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 32 acknowledgments i want to thank all members of the research training group extrospections for their helpful comments on this work. i especially thank sascha benjamin fink, laura kaltwasser, chiara caporuscio, anna daniels, and katharina bögl for repeated discussions on earlier drafts of this paper. i thank trevor griffith for his support in proofreading. finally, i thank isabel grahn for her open ear whenever i struggled with my manuscript. funded by the deutsche forschungsgemeinschaft (dfg, german research foundation) 337619223 / rtg2386. references abbott, k. h., schutte, p. c., palmer, m. t., & ricks, w. r. (1988). faultfinder: a diagnostic expert system with graceful degradation for onboard aircraft applications. https://ntrs.nasa.gov/citations/19880018634 aboraya, a. (2009). use of structured interviews by psychiatrists in real clinical settings: results of an open-question survey. psychiatry (edgmont (pa.: township)), 6(6), 24–28. adam, d. (2013). mental health: on the spectrum. nature, 496(7446), 416–418. https://doi.org/10.1038/496416a altable, c. r. (2012). logic structure of clinical judgment and its relation to medical and psychiatric semiology. psychopathology, 45(6), 344–351. https://doi.org/10.1159/000337968 american board of psychiatry & neurology. (2011). american board of psychiatry & neurology: neurology core competencies outline. american board of psychiatry; neurology inc. https://www.abpn.com/wp-content/uploads/2015/02/2011_c ore_n_mree.pdf american psychiatric association. (1986). diagnostic and statistical manual of mental disorders: dsm-3 (3rd ed). american psychiatric association. (2013). diagnostic and statistical manual of mental disorders: dsm-5 (5th ed). https://doi.org/10.1176/appi.books.9780890425596 andreasen, n. c. (2016). thought disorder. in s. h. fatemi & p. j. clayton (eds.), the medical basis of psychiatry (pp. 497–505). springer new york. https://doi.org/10.1007/978-1-4939-2528-5_25 baker, m. g. (2002). the wall between neurology and psychiatry. bmj, 324(7352), 1468–1469. https://doi.org/10.1136/bmj. 324.7352.1468 banner, n. f., & thornton, t. (2007). the new philosophy of psychiatry: its (recent) past, present and future: a review of the oxford university press series international perspectives in philosophy and psychiatry. philosophy, ethics, and humanities in medicine, 2(1), 9. https://doi.org/10.1186/1747-5341-2-9 batterman, r. w. (2002). asymptotics and the role of minimal models. the british journal for the philosophy of science, 53(1), 21–38. http://www.jstor.org/stable/3541639 beck, a. l., & bergman, d. a. (1986). using structured medical information to improve studentsʼ problem-solving performance: academic medicine, 61(9), 749–756. https://doi.org/10.1097/00001888-198609000-00019 beck, a. t., ward, c. h., mendelson, m., mock, j., & erbaugh, j. (1961). an inventory for measuring depression. archives of general psychiatry, 4, 561–571. https://doi.org/10.1001/archpsyc.1961.01710120031004 bell, d., bobrow, d., falkenhainer, b., fromherz, m., saraswat, v., & shirley, m. (1994). rapper: the copier modelling project. proceedings of the eighth international workshop on qualitative reasoning about physical systems. https://www.qrg.northwestern.edu/papers/files/qr-workshops/qr94/bell_1994_rapper.pdf bobrow, d. g. (1984). qualitative reasoning about physical systems: an introduction. artificial intelligence, 24(1-3), 1–5. https://doi.org/10.1016/0004-3702(84)90036-5 boland, r. j., verduin, m. l., ruiz, p., shah, a., & sadock, b. j. (eds.). (2022). kaplan & sadock’s synopsis of psychiatry 12th edition (twelfth edition). wolters kluwer. boorse, c. (1977). health as a theoretical concept. philosophy of science, 44(4), 542–573. https://doi.org/10.1086/288768 borsboom, d. (2017). a network theory of mental disorders. world psychiatry: official journal of the world psychiatric association (wpa), 16(1), 5–13. https://doi.org/10.1002/wps.20375 boumans, m. (1999). built-in justification. in m. s. morgan & m. morrison (eds.), models as mediators (1st ed., pp. 66–96). cambridge university press. https://doi.org/10.1017/cbo9780511660108.005 bruch, m., & bond, f. w. (eds.). (1998). beyond diagnosis: case formulation approaches in cbt. wiley. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://ntrs.nasa.gov/citations/19880018634 https://www.ncbi.nlm.nih.gov/pmc/articles/pmc2720839 https://www.ncbi.nlm.nih.gov/pmc/articles/pmc2720839 https://doi.org/10.1038/496416a https://doi.org/10.1159/000337968 https://www.abpn.com/wp-content/uploads/2015/02/2011_core_n_mree.pdf https://www.abpn.com/wp-content/uploads/2015/02/2011_core_n_mree.pdf https://doi.org/10.1176/appi.books.9780890425596 https://doi.org/10.1007/978-1-4939-2528-5_25 https://doi.org/10.1136/bmj.324.7352.1468 https://doi.org/10.1136/bmj.324.7352.1468 https://doi.org/10.1186/1747-5341-2-9 http://www.jstor.org/stable/3541639 https://doi.org/10.1097/00001888-198609000-00019 https://doi.org/10.1001/archpsyc.1961.01710120031004 https://www.qrg.northwestern.edu/papers/files/qr-workshops/qr94/bell_1994_rapper.pdf https://doi.org/10.1016/0004-3702(84)90036-5 https://doi.org/10.1086/288768 https://doi.org/10.1002/wps.20375 https://doi.org/10.1017/cbo9780511660108.005 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 33 bruchmüller, k., margraf, j., suppiger, a., & schneider, s. (2011). popular or unpopular? therapists’ use of structured interviews and their estimation of patient acceptance. behavior therapy, 42(4), 634–643. https://doi.org/10.1016/j.be th.2011.02.003 burge, t. (2010). origins of objectivity. oxford university press. https://doi.org/10.1093/acprof:oso/9780199581405.001.0001 butler, g. (1998). clinical formulation. in comprehensive clinical psychology (pp. 1–24). elsevier. https://doi.org/10.1016/ b0080-4270(73)00186-3 cahalan, s. (2019). the great pretender: the undercover mission that changed our understanding of madness (first edition). grand central publishing. carter, j. a., & littlejohn, c. (2021). this is epistemology: an introduction. wiley-blackwell. cartwright, n., & nordby, j. (1983). how approximations take us away from theory and towards truth. pacific philosophical quarterly, 64(3), 273–280. https://doi.org/10.1111/j.1468-0114.1983.tb00200.x casey, p. r., & kelly, b. (2019). fish’s clinical psychopathology: signs and syptoms in psychiatry (4th ed). cambridge university press. charlin, b., tardif, j., & boshuizen, h. p. a. (2000). scripts and medical diagnostic knowledge: theory and applications for clinical reasoning instruction and research. academic medicine, 75(2), 182–190. https://doi.org/10.1097/00001888200002000-00020 chisolm, m. s., & lyketsos, c. g. (2012). systematic psychiatric evaluation: a step-by-step guide to applying the perspectives of psychiatry. johns hopkins university press. coderre, s., mandin, h., harasym, p. h., & fick, g. h. (2003). diagnostic reasoning strategies and diagnostic success. medical education, 37(8), 695–703. https://doi.org/10.1046/j.1365-2923.2003.01577.x cooper, r. v. (2007). psychiatry and philosophy of science. mcgill-queen’s university press. croskerry, p. (2009). a universal model of diagnostic reasoning. academic medicine: journal of the association of american medical colleges, 84(8), 1022–1028. https://doi.org/10.1097/acm.0b013e3181ace703 cuthbert, b. n., & morris, s. e. (2021). evolving concepts of the schizophrenia spectrum: a research domain criteria perspective. frontiers in psychiatry, 12, 641319. https://doi.org/10.3389/fpsyt.2021.641319 david, a. s., & nicholson, t. (2015). are neurological and psychiatric disorders different? british journal of psychiatry, 207(5), 373–374. https://doi.org/10.1192/bjp.bp.114.158550 de kleer, j., & brown, j. s. (1984). a qualitative physics based on confluences. artificial intelligence, 24(1-3), 7–83. https://doi.org/10.1016/0004-3702(84)90037-7 division of clinical psychology. (2001). the core purpose and philosophy of the profession. the british psychological society. https://www.liverpool.ac.uk/media/livacuk/doctorateinclinicalpsychologyprogramme/docs/core_purpose_and_phi lospohy_of_the_profession.pdf docherty, a. r., berenbaum, h., & kerns, j. g. (2011). alogia and formal thought disorder: differential patterns of verbal fluency task performance. journal of psychiatric research, 45(10), 1352–1357. https://doi.org/10.1016/j.jpsychires.201 1.04.004 doughty, o. j., & done, d. j. (2009). is semantic memory impaired in schizophrenia? a systematic review and meta-analysis of 91 studies. cognitive neuropsychiatry, 14(6), 473–509. https://doi.org/10.1080/13546800903073291 eells, t. d. (2010). history and current status of psychotherapy case formulation. in t. d. eells (ed.), handbook of psychotherapy case formulation (2. ed., paperback ed, pp. 3–33). guilford press. elstein, a. s., allal, l., gordon, m., hilliard, j., jordan, r. d., kagan, n., loupe, m. j., shulman, l. s., & sprafka, s. a. (2014). medical problem solving: an analysis of clinical reasoning. https://0-doi-org.pugwash.lib.warwick.ac.uk/10.4159/harv ard.9780674189089 esterberg, m. l., & compton, m. t. (2009). the psychosis continuum and categorical versus dimensional diagnostic approaches. current psychiatry reports, 11(3), 179–184. https://doi.org/10.1007/s11920-009-0028-7 falkenhainer, b., & forbus, k. d. (1991). compsitional modeling: finding the right model for the job. artificial intelligence, 51(1-3), 95–143. https://doi.org/10.1016/0004-3702(91)90109-w farley, a. m. (1987). qualitative modeling of economic systems. ifac proceedings volumes, 20(10), 41–44. https://doi.org/10.1016/s1474-6670(17)59029-1 fernando, i., cohen, m., & henskens, f. (2013). a systematic approach to clinical reasoning in psychiatry. australasian psychiatry, 21(3), 224–230. https://doi.org/10.1177/1039856213486209 first, m. b., drevets, w. c., carter, c., dickstein, d. p., kasoff, l., kim, k. l., mcconathy, j., rauch, s., saad, z. s., savitz, j., seymour, k. e., sheline, y. i., & zubieta, j.-k. (2018). clinical applications of neuroimaging in psychiatric disorders. the american journal of psychiatry, 175(9), 915–916. https://doi.org/10.1176/appi.ajp.2018.1750701 first, m. b., williams, j. b. w., spitzer, r. l., & karg, r. s. (2016). user’s guide for the scid-5-cv structured clinical interview for dsm-5 disorders: clinician version. american psychiatric association publishing. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.beth.2011.02.003 https://doi.org/10.1016/j.beth.2011.02.003 https://doi.org/10.1093/acprof:oso/9780199581405.001.0001 https://doi.org/10.1016/b0080-4270(73)00186-3 https://doi.org/10.1016/b0080-4270(73)00186-3 https://doi.org/10.1111/j.1468-0114.1983.tb00200.x https://doi.org/10.1097/00001888-200002000-00020 https://doi.org/10.1097/00001888-200002000-00020 https://doi.org/10.1046/j.1365-2923.2003.01577.x https://doi.org/10.1097/acm.0b013e3181ace703 https://doi.org/10.3389/fpsyt.2021.641319 https://doi.org/10.1192/bjp.bp.114.158550 https://doi.org/10.1016/0004-3702(84)90037-7 https://www.liverpool.ac.uk/media/livacuk/doctorateinclinicalpsychologyprogramme/docs/core_purpose_and_philospohy_of_the_profession.pdf https://www.liverpool.ac.uk/media/livacuk/doctorateinclinicalpsychologyprogramme/docs/core_purpose_and_philospohy_of_the_profession.pdf https://doi.org/10.1016/j.jpsychires.2011.04.004 https://doi.org/10.1016/j.jpsychires.2011.04.004 https://doi.org/10.1080/13546800903073291 https://0-doi-org.pugwash.lib.warwick.ac.uk/10.4159/harvard.9780674189089 https://0-doi-org.pugwash.lib.warwick.ac.uk/10.4159/harvard.9780674189089 https://doi.org/10.1007/s11920-009-0028-7 https://doi.org/10.1016/0004-3702(91)90109-w https://doi.org/10.1016/s1474-6670(17)59029-1 https://doi.org/10.1177/1039856213486209 https://doi.org/10.1176/appi.ajp.2018.1750701 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 34 forbus, k. d. (1984). qualitative process theory. artificial intelligence, 24(1-3), 85–168. https://doi.org/10.1016/00043702(84)90038-9 forbus, k. d. (2008). chapter 9 qualitative modeling. in foundations of artificial intelligence (vol. 3, pp. 361–393). elsevier. https://doi.org/10.1016/s1574-6526(07)03009-x fuchs, t. (2010). subjectivity and intersubjectivity in psychiatric diagnosis. psychopathology, 43(4), 268–274. https://doi.org/10.1159/000315126 gelder, m. g. (ed.). (2000). new oxford textbook of psychiatry (1nd ed). oxford university press. giere, r. n. (1988). explaining science: a cognitive approach (8. [repr.]). univ. of chicago pr. giere, r. n. (1999). using models to represent reality. in l. magnani, n. j. nersessian, & p. thagard (eds.), model-based reasoning in scientific discovery (pp. 41–57). springer us. https://doi.org/10.1007/978-1-4615-4813-3_3 godfrey-smith, p. (2006). the strategy of model-based science. biology & philosophy, 21(5), 725–740. https://doi.org/10.100 7/s10539-006-9054-6 godfrey-smith, p. (2009). models and fictions in science. philosophical studies, 143(1), 101–116. https://doi.org/10.1007/s1 1098-008-9313-2 godwin, r. m. (1967). a growth cycle. in c. h. feinstein (ed.), socialism, capitalism and economic growth: essays presented to maurice dobb (pp. 54–59). cambridge university press. goldman, a. (2000). can science know when you’re conscious? journal of consciousness studies, 7(5), 3–22. goldman, r. n., & greenberg, l. s. (2015). case formulation in emotion-focused therapy: co-creating clinical maps for change. american psychological association. https://doi.org/10.1037/14523-000 goodwin, r. m. (1967). a growth cycle. in c. h. feinstein (ed.), socialism, capitalism and economic growth (pp. 54–59). cambridge university press. graham, g. (2021). the disordered mind (third edition). routledge, taylor & francis group. groves, m., o’rourke, p., & alexander, h. (2003). the clinical reasoning characteristics of diagnostic experts. medical teacher, 25(3), 308–313. https://doi.org/10.1080/0142159031000100427 gruppen, l. d., & frohna, a. z. (2002). clinical reasoning. in g. r. norman, c. p. m. van der vleuten, d. i. newble, d. h. j. m. dolmans, k. v. mann, a. rothman, & l. curry (eds.), international handbook of research in medical education (vol. 7, pp. 205–230). springer netherlands. https://doi.org/10.1007/978-94-010-0462-6_8 gupta, m., potter, n., & goyer, s. (2019). diagnostic reasoning in psychiatry: acknowledging an explicit role for intersubjective knowing. philosophy, psychiatry, & psychology, 26(1), 49–64. https://doi.org/10.1353/ppp.2019.0003 guze, s. b. (1992). why psychiatry is a branch of medicine. oxford university press. hamscher, w. (ed.). (1995). readings in model-based diagnosis (4. print). morgan kaufmann publ. (original work published 1992) harjai, p. k., & tiwari, r. (2009). model of critical diagnostic reasoning: achieving expert clinician performance. nursing education perspectives, 30(5), 305–311. https://journals.lww.com/neponline/abstract/2009/09000/model_of_critica l_diagnostic_reasoning__achieving.10.aspx he, r., yan, x., guo, j., xu, q., tang, b., & sun, q. (2018). recent advances in biomarkers for parkinson’s disease. frontiers in aging neuroscience, 10, 305. https://doi.org/10.3389/fnagi.2018.00305 helzer, j. e., & association, a. p. (eds.). (2008). dimensional approaches in diagnostic classification: refining the research agenda for dsm-v (1st ed). american psychiatric association. higgs, j., jensen, g. m., loftus, s., & christensen, n. (eds.). (2019). clinical reasoning in the health professions (fourth edition). elsevier. hodges, j. r. (2018). cognitive assessment for clinicians (third edition). oxford university press. hoffmann, r. (1998). qualitative thinking in the age of modern computational chemistry—or what lionel salem knows. journal of molecular structure: theochem, 424(1-2), 1–6. https://doi.org/10.1016/s0166-1280(97)00219-4 hohwy, j. (2013). the predictive mind (first edition). oxford university press. husserl, e., & ströker, e. (2013). logische untersuchungen (2. aufl.). meiner. (original work published 1900) insel, t. r., & cuthbert, b. n. (2015). brain disorders? precisely. science, 348(6234), 499–500. https://doi.org/10.1126/scienc e.aab2358 insel, t. r., & quirion, r. (2005). psychiatry as a clinical neuroscience discipline. jama, 294(17), 2221–2224. https://doi.org/10.1001/jama.294.17.2221 jaspers, k. (1973). allgemeine psychopathologie (9., unveränd. aufl). springer. (original work published 1913) jefferson, a. (2020). what does it take to be a brain disorder? synthese, 197(1), 249–262. https://doi.org/10.1007/s11229018-1784-x kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/0004-3702(84)90038-9 https://doi.org/10.1016/0004-3702(84)90038-9 https://doi.org/10.1016/s1574-6526(07)03009-x https://doi.org/10.1159/000315126 https://doi.org/10.1007/978-1-4615-4813-3_3 https://doi.org/10.1007/s10539-006-9054-6 https://doi.org/10.1007/s10539-006-9054-6 https://doi.org/10.1007/s11098-008-9313-2 https://doi.org/10.1007/s11098-008-9313-2 https://doi.org/10.1037/14523-000 https://doi.org/10.1080/0142159031000100427 https://doi.org/10.1007/978-94-010-0462-6_8 https://doi.org/10.1353/ppp.2019.0003 https://journals.lww.com/neponline/abstract/2009/09000/model_of_critical_diagnostic_reasoning__achieving.10.aspx https://journals.lww.com/neponline/abstract/2009/09000/model_of_critical_diagnostic_reasoning__achieving.10.aspx https://doi.org/10.3389/fnagi.2018.00305 https://doi.org/10.1016/s0166-1280(97)00219-4 https://doi.org/10.1126/science.aab2358 https://doi.org/10.1126/science.aab2358 https://doi.org/10.1001/jama.294.17.2221 https://doi.org/10.1007/s11229-018-1784-x https://doi.org/10.1007/s11229-018-1784-x https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 35 johnson-laird, p. n. (1983). mental models: towards a cognitive science of language, inference, and consciousness. harvard university press. johnstone, l., & dallos, r. (eds.). (2014). formulation in psychology and psychotherapy: making sense of people’s problems (second edition). routledge. kirk, k. h., & werbrook, d. (2016). assessment and formulation. in h. kennerley, j. kirk, & d. westbrook (eds.), an introduction to cognitive behaviour therapy: skills and applications (3rd edition, pp. 67–109). sage publications ltd. knuuttila, t., & loettgers, a. (2016). model templates within and between disciplines: from magnets to gases – and socioeconomic systems. european journal for philosophy of science, 6(3), 377–400. https://doi.org/10.1007/s13194-0160145-1 kotov, r., krueger, r. f., watson, d., achenbach, t. m., althoff, r. r., bagby, r. m., brown, t. a., carpenter, w. t., caspi, a., clark, l. a., eaton, n. r., forbes, m. k., forbush, k. t., goldberg, d., hasin, d., hyman, s. e., ivanova, m. y., lynam, d. r., markon, k., … zimmerman, m. (2017). the hierarchical taxonomy of psychopathology (hitop): a dimensional alternative to traditional nosologies. journal of abnormal psychology, 126(4), 454–477. https://doi.org/10.1037/abn000 0258 krueger, r. f., & bezdjian, s. (2009). enhancing research and treatment of mental disorders with dimensional concepts: toward dsm-v and icd-11. world psychiatry, 8(1), 3–6. https://doi.org/10.1002/j.2051-5545.2009.tb00197.x kuruvilla, k., & kuruvilla, a. (2010). diagnostic formulation. indian journal of psychiatry, 52(1), 78. https://doi.org/10.410 3/0019-5545.58905 ligęza, a., & górny, b. (2017). model-based diagnosis. in l. magnani & t. bertolotti (eds.), springer handbook of model-based science (pp. 435–461). springer international publishing. https://doi.org/10.1007/978-3-319-30526-4_20 lotka, a. j. (1910). contribution to the theory of periodic reactions. the journal of physical chemistry, 14(3), 271–274. https://doi.org/10.1021/j150111a004 lotka, a. j. (1925). elements of mathematical biology (williams and wilkins.). dover. loveday, t., wiggins, m., festa, m., schell, d., & twigg, d. (2013). pattern recognition as an indicator of diagnostic expertise. in p. latorre carmona, j. s. sánchez, & a. l. n. fred (eds.), pattern recognition applications and methods (vol. 204, pp. 1–11). springer berlin heidelberg. https://doi.org/10.1007/978-3-642-36530-0_1 maclin, j. m. a., wang, t., & xiao, s. (2019b). biomarkers for the diagnosis of alzheimer’s disease, dementia lewy body, frontotemporal dementia and vascular dementia. general psychiatry, 32(1), e100054. https://doi.org/10.1136/gpsych2019-100054 magnani, l. (1992b). abductive reasoning: philosophical and educational perspectives in medicine. in d. a. evans & v. l. patel (eds.), advanced models of cognition for medical training and practice (pp. 21–44). springer. https://doi.org/10.1 007/978-3-662-02833-9_2 mamede, s., schmidt, h. g., & rikers, r. (2007). diagnostic errors and reflective practice in medicine. journal of evaluation in clinical practice, 13(1), 138–145. https://doi.org/10.1111/j.1365-2753.2006.00638.x mandin, h., jones, a., woloschuk, w., & harasym, p. (1997). helping students learn to think like experts when solving clinical problems: academic medicine, 72(3), 173–179. https://doi.org/10.1097/00001888-199703000-00009 mchugh, p. r., & slavney, p. r. (1998). the perspectives of psychiatry. johns hopkins university press. mcmullin, e. (1985). galilean idealization. studies in history and philosophy of science part a, 16(3), 247–273. https://doi.org/10.1016/0039-3681(85)90003-2 mcwilliams, n. (1999). psychoanalytic case formulation. guilford press. merleau-ponty, m. (2018). phenomenology of perception (c. smith, trans.). franklin classics. meyer, v., & turkat, i. d. (1979). behavioral analysis of clinical cases. journal of behavioral assessment, 1(4), 259–270. https://doi.org/10.1007/bf01321368 miller, d. b., & o’callaghan, j. p. (2015). biomarkers of parkinson’s disease: present and future. metabolism, 64(3), s40–s46. https://doi.org/10.1016/j.metabol.2014.10.030 miller, g. a. (2010). mistreating psychology in the decades of the brain. perspectives on psychological science, 5(6), 716–743. https://doi.org/10.1177/1745691610388774 morgan, m. s., & morrison, m. (eds.). (1999). models as mediators: perspectives on natural and social science (1st ed.). cambridge university press. https://doi.org/10.1017/cbo9780511660108 morrison, m. (2007). where have all the theories gone? philosophy of science, 74(2), 195–228. https://doi.org/10.1086/520778 morrison, m. (1999). models as autonomous agents. in m. s. morgan & m. morrison (eds.), models as mediators (1st ed., pp. 38–65). cambridge university press. https://doi.org/10.1017/cbo9780511660108.004 mueller, a. e., & segal, d. l. (2015). structured versus semistructured versus unstructured interviews. in r. l. cautin & s. o. lilienfeld (eds.), the encyclopedia of clinical psychology (pp. 1–7). john wiley & sons, inc. https://doi.org/10.1002/ 9781118625392.wbecp069 kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s13194-016-0145-1 https://doi.org/10.1007/s13194-016-0145-1 https://doi.org/10.1037/abn0000258 https://doi.org/10.1037/abn0000258 https://doi.org/10.1002/j.2051-5545.2009.tb00197.x https://doi.org/10.4103/0019-5545.58905 https://doi.org/10.4103/0019-5545.58905 https://doi.org/10.1007/978-3-319-30526-4_20 https://doi.org/10.1021/j150111a004 https://doi.org/10.1007/978-3-642-36530-0_1 https://doi.org/10.1136/gpsych-2019-100054 https://doi.org/10.1136/gpsych-2019-100054 https://doi.org/10.1007/978-3-662-02833-9_2 https://doi.org/10.1007/978-3-662-02833-9_2 https://doi.org/10.1111/j.1365-2753.2006.00638.x https://doi.org/10.1097/00001888-199703000-00009 https://doi.org/10.1016/0039-3681(85)90003-2 https://doi.org/10.1007/bf01321368 https://doi.org/10.1016/j.metabol.2014.10.030 https://doi.org/10.1177/1745691610388774 https://doi.org/10.1017/cbo9780511660108 https://doi.org/10.1086/520778 https://doi.org/10.1017/cbo9780511660108.004 https://doi.org/10.1002/9781118625392.wbecp069 https://doi.org/10.1002/9781118625392.wbecp069 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 36 murphy, d. (ed.). (2006). psychiatry in the scientific image. mit press. myers, r. h. (2004). huntington’s disease genetics. neurorx, 1(2), 255–262. https://doi.org/10.1602/neurorx.1.2.255 nakash, o., & alegría, m. (2013). examination of the role of implicit clinical judgments during the mental health intake. qualitative health research, 23(5), 645–654. https://doi.org/10.1177/1049732312471732 national academies of sciences, engineering, and medicine. (2019). diagnosis and assessment of traumatic brain injury. in evaluation of the disability determination process for traumatic brain injury in veterans (pp. 23–56). national academies press. https://doi.org/10.17226/25317 national institute of mental health. (2013). research domain criteria (rdoc). https://www.nimh.nih.gov/research/resear ch-funded-by-nimh/rdoc/nimh-research-domain-criteria-rdoc nordgaard, j., revsbech, r., saebye, d., & parnas, j. (2012). assessing the diagnostic validity of a structured psychiatric interview in a first-admission hospital sample. world psychiatry, 11(3), 181–185. https://doi.org/10.1002/j.20515545.2012.tb00128.x norman, g. (2005). research in clinical reasoning: past history and current trends. medical education, 39(4), 418–427. https://doi.org/10.1111/j.1365-2929.2005.02127.x nurcombe, b., & fitzhenry-coor, i. (1982). how do psychiatrists think⁇ clinical reasoning in the psychiatric interview: a research and education project. australian & new zealand journal of psychiatry, 16(1), 13–24. https://doi.org/10.310 9/00048678209159466 olbert, c. m., & gala, g. j. (2015). supervenience and psychiatry: are mental disorders brain disorders? journal of theoretical and philosophical psychology, 35(4), 203–219. https://doi.org/10.1037/teo0000023 papineau, d. (1994). mental disorder, illness and biological disfunction. royal institute of philosophy supplement, 37, 73–82. https://doi.org/10.1017/s135824610000998x parnas, j., sass, l. a., & zahavi, d. (2013). rediscovering psychopathology: the epistemology and phenomenology of the psychiatric object. schizophrenia bulletin, 39(2), 270–277. https://doi.org/10.1093/schbul/sbs153 patel, v. l., arocha, j. f., & zhang, j. (2012). medical reasoning and thinking. oxford university press. https://doi.org/10.1 093/oxfordhb/9780199734689.013.0037 pi, y., liao, w., liu, m., & lu, j. (2008). theory of cognitive pattern recognition. in p.-y. yin (ed.), pattern recognition techniques, technology and applications. intech. https://doi.org/10.5772/6251 pinnock, r., young, l., spence, f., henning, m., & hazell, w. (2015). can think aloud be used to teach and assess clinical reasoning in graduate medical education? journal of graduate medical education, 7(3), 334–337. https://doi.org/10.4 300/jgme-d-14-00601.1 pollock, j. l., & cruz, j. (1999). contemporary theories of knowledge (2nd ed). rowman & littlefield publishers. poole, r., & higgo, r. (2017). psychiatric interviewing and assessment (2nd ed.). cambridge university press. https://doi.org/10.1017/9781316676554 power, b. d., nguyen, t., hayhow, b., & looi, j. (2016). neuroimaging in psychiatry: an update on neuroimaging in the clinical setting. australasian psychiatry, 24(2), 157–163. https://doi.org/10.1177/1039856215618525 price, b. h., adams, r. d., & coyle, j. t. (2000). neurology and psychiatry: closing the great divide. neurology, 54(1), 8–8. https://doi.org/10.1212/wnl.54.1.8 rao, v., spiro, j. r., schretlen, d. j., & cascella, n. g. (2007). apathy syndrome after traumatic brain injury compared with deficits in schizophrenia. psychosomatics, 48(3), 217–222. https://doi.org/10.1176/appi.psy.48.3.217 readings in qualitative reasoning about physical systems. (1990). elsevier. https://doi.org/10.1016/c2013-0-08298-9 reed, g. m., first, m. b., kogan, c. s., hyman, s. e., gureje, o., gaebel, w., maj, m., stein, d. j., maercker, a., tyrer, p., claudino, a., garralda, e., salvador‐carulla, l., ray, r., saunders, j. b., dua, t., poznyak, v., medina‐mora, m. e., pike, k. m., … saxena, s. (2019). innovations and changes in the icd‐11 classification of mental, behavioural and neurodevelopmental disorders. world psychiatry, 18(1), 3–19. https://doi.org/10.1002/wps.20611 reiter, r. (1987). a theory of diagnosis from first principles. artificial intelligence, 32(1), 57–95. https://doi.org/10.1016/00043702(87)90062-2 reitz, c., & mayeux, r. (2014). alzheimer disease: epidemiology, diagnostic criteria, risk factors and biomarkers. biochemical pharmacology, 88(4), 640–651. https://doi.org/10.1016/j.bcp.2013.12.024 rejón altable, c. (2012). logic structure of clinical judgment and its relation to medical and psychiatric semiology. psychopathology, 45(6), 344–351. https://doi.org/10.1159/000337968 reum, j. c. p., mcdonald, p. s., ferriss, b. e., farrell, d. m., harvey, c. j., & levin, p. s. (2015). qualitative network models in support of ecosystem approaches to bivalve aquaculture. ices journal of marine science, 72(8), 2278–2288. https://doi.org/10.1093/icesjms/fsv119 reznek, l. (1998). on the epistemology of mental illness. history and philosophy of the life sciences, 20(2), 215–232. http://www.jstor.org/stable/23332102 kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1602/neurorx.1.2.255 https://doi.org/10.1177/1049732312471732 https://doi.org/10.17226/25317 https://www.nimh.nih.gov/research/research-funded-by-nimh/rdoc/nimh-research-domain-criteria-rdoc https://www.nimh.nih.gov/research/research-funded-by-nimh/rdoc/nimh-research-domain-criteria-rdoc https://doi.org/10.1002/j.2051-5545.2012.tb00128.x https://doi.org/10.1002/j.2051-5545.2012.tb00128.x https://doi.org/10.1111/j.1365-2929.2005.02127.x https://doi.org/10.3109/00048678209159466 https://doi.org/10.3109/00048678209159466 https://doi.org/10.1037/teo0000023 https://doi.org/10.1017/s135824610000998x https://doi.org/10.1093/schbul/sbs153 https://doi.org/10.1093/oxfordhb/9780199734689.013.0037 https://doi.org/10.1093/oxfordhb/9780199734689.013.0037 https://doi.org/10.5772/6251 https://doi.org/10.4300/jgme-d-14-00601.1 https://doi.org/10.4300/jgme-d-14-00601.1 https://doi.org/10.1017/9781316676554 https://doi.org/10.1177/1039856215618525 https://doi.org/10.1212/wnl.54.1.8 https://doi.org/10.1176/appi.psy.48.3.217 https://doi.org/10.1016/c2013-0-08298-9 https://doi.org/10.1002/wps.20611 https://doi.org/10.1016/0004-3702(87)90062-2 https://doi.org/10.1016/0004-3702(87)90062-2 https://doi.org/10.1016/j.bcp.2013.12.024 https://doi.org/10.1159/000337968 https://doi.org/10.1093/icesjms/fsv119 http://www.jstor.org/stable/23332102 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org how does the psychiatrist know? 37 rohrer, j. d., & rosen, h. j. (2013). neuroimaging in frontotemporal dementia. international review of psychiatry, 25(2), 221–229. https://doi.org/10.3109/09540261.2013.778822 rosenhan, d. l. (1973). on being sane in insane places. science, 179(4070), 250–258. https://doi.org/10.1126/science.179.40 70.250 roth, m., tym, e., mountjoy, c. q., huppert, f. a., hendrie, h., verma, s., & goddard, r. (1986). camdex: a standardised instrument for the diagnosis of mental disorder in the elderly with special reference to the early detection of dementia. british journal of psychiatry, 149(6), 698–709. https://doi.org/10.1192/bjp.149.6.698 royal australian and new zealand college of psychiatrists. (2014). clinical examinations: formulation guidelines for trainees (formulation guidelines 0304). https://psychiatry-training.wiki.otago.ac.nz/images/6/6c/ranzcp-formulationsguide.pdf royal college of physicians and surgeons of canada. (2019). objectives of training in the specialty of psychiatry. version 2.0. http://www.royalcollege.ca/rcsite/education-strategy-accreditation/pgme-training-programs/accreditationresidencyprogram royal college of psychiatrists. (2017). using formulation in general psychiatric care: good practice. https://www.yumpu.co m/en/document/view/56758061/using-formulation-in-general-psychiatric-care-good-practice royal college of psychiatrists. (2013). a competency based curriculum for specialist core training in psychiatry: core training in psychiatry ct1–ct3. https://www.rcpsych.ac.uk/docs/default-source/training/curricula-and-guidance/curriculacore-psychiatry-curriculum-april-2018.pdf?sfvrsn=881b63ca_2 ryan, p. (ed.). (2020). enhancing clinical case formulation: theoretical and practical approaches for mental health practitioners (1st ed.). routledge. schmidt, h. g., norman, g. r., & boshuizen, h. p. (1990). a cognitive perspective on medical expertise: theory and implication [published erratum appears in acad med 1992 apr;67(4):287]. academic medicine, 65(10), 611–621. https://doi.org/10.1097/00001888-199010000-00001 schramme, t. (2013). on the autonomy of the concept of disease in psychiatry. frontiers in psychology, 4. https://doi.org/ 10.3389/fpsyg.2013.00457 silajdžić, e., & björkqvist, m. (2018). a critical evaluation of wet biomarkers for huntington’s disease: current status and ways forward. journal of huntington’s disease, 7(2), 109–135. https://doi.org/10.3233/jhd-170273 smith, b., & ehrenfels, c. (eds.). (1988). foundations of gestalt theory. philosophia verlag. sober, e. (1979). the art and science of clinical judgment:: an informational approach. in h. t. engelhardt, s. f. spicker, h. t. engelhardt, s. f. spicker, & b. towers (eds.), clinical judgment: a critical appraisal (vol. 6, pp. 29–44). springer netherlands. https://doi.org/10.1007/978-94-009-9399-0_3 spitzer, r. l. (1975). on pseudoscience in science, logic in remission, and psychiatric diagnosis: a critique of rosenhan’s “on being sane in insane places”. journal of abnormal psychology, 84(5), 442–452. https://doi.org/10.1037/h0077124 szasz, t. s. (1960). the myth of mental illness. american psychologist, 15(2), 113–118. https://doi.org/10.1037/h0046535 tarbox, j., la cava, s., & hoang, k. (2016). types of assessment. in j. l. matson (ed.), handbook of assessment and diagnosis of autism spectrum disorder (pp. 11–26). springer international publishing. https://doi.org/10.1007/978-3-319-27171-2_2 thomson-jones, m. (2012). modeling without mathematics. philosophy of science, 79(5), 761–772. https://doi.org/10.1086/ 667876 trull, t. j., & durrett, c. a. (2005). categorical and dimensional models of personality disorder. annual review of clinical psychology, 1(1), 355–380. https://doi.org/10.1146/annurev.clinpsy.1.102803.144009 trzepacz, p. t., & baker, r. w. (1993). the psychiatric mental status examination. oxford university press. upshur, r. e. g., & colak, e. (2003). arguments and evidence. theoretical medicine and bioethics, 24(4), 283–299. https://doi.org/10.1023/a:1026006801902 varghese, f. t., & mellsop, g. w. (1983). formulating a psychiatric case. bulletin of the royal college of psychiatrists, 7(3), 48–50. https://doi.org/10.1192/s0140078900008002 verber, n., & shaw, p. j. (2020). biomarkers in amyotrophic lateral sclerosis: a review of new developments. current opinion in neurology, 33(5), 662–668. https://doi.org/10.1097/wco.0000000000000854 volterra, v. (1926). variazioni e fluttuazioni del numero d’individui in specie animali conviventi. società anonima tipografica “leonardo da vinci”. wagner, a. d., paré-blagoev, e. j., clark, j., & poldrack, r. a. (2001). recovering meaning. neuron, 31(2), 329–338. https://doi.org/10.1016/s0896-6273(01)00359-2 weerasekera, p. (1996). multiperspective case formulation: a step towards treatment integration (original ed). krieger pub. co. weisberg, m. (2013). simulation and similarity: using models to understand the world. oxford university press. https://doi.org/10.1093/acprof:oso/9780199933662.001.0001 kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.3109/09540261.2013.778822 https://doi.org/10.1126/science.179.4070.250 https://doi.org/10.1126/science.179.4070.250 https://doi.org/10.1192/bjp.149.6.698 https://psychiatry-training.wiki.otago.ac.nz/images/6/6c/ranzcp-formulations-guide.pdf https://psychiatry-training.wiki.otago.ac.nz/images/6/6c/ranzcp-formulations-guide.pdf http://www.royalcollege.ca/rcsite/education-strategy-accreditation/pgme-training-programs/accreditation-residency-%20program http://www.royalcollege.ca/rcsite/education-strategy-accreditation/pgme-training-programs/accreditation-residency-%20program https://www.yumpu.com/en/document/view/56758061/using-formulation-in-general-psychiatric-care-good-practice https://www.yumpu.com/en/document/view/56758061/using-formulation-in-general-psychiatric-care-good-practice https://www.rcpsych.ac.uk/docs/default-source/training/curricula-and-guidance/curricula-core-psychiatry-curriculum-april-2018.pdf?sfvrsn=881b63ca_2 https://www.rcpsych.ac.uk/docs/default-source/training/curricula-and-guidance/curricula-core-psychiatry-curriculum-april-2018.pdf?sfvrsn=881b63ca_2 https://doi.org/10.1097/00001888-199010000-00001 https://doi.org/10.3389/fpsyg.2013.00457 https://doi.org/10.3389/fpsyg.2013.00457 https://doi.org/10.3233/jhd-170273 https://doi.org/10.1007/978-94-009-9399-0_3 https://doi.org/10.1037/h0077124 https://doi.org/10.1037/h0046535 https://doi.org/10.1007/978-3-319-27171-2_2 https://doi.org/10.1086/667876 https://doi.org/10.1086/667876 https://doi.org/10.1146/annurev.clinpsy.1.102803.144009 https://doi.org/10.1023/a:1026006801902 https://doi.org/10.1192/s0140078900008002 https://doi.org/10.1097/wco.0000000000000854 https://doi.org/10.1016/s0896-6273(01)00359-2 https://doi.org/10.1093/acprof:oso/9780199933662.001.0001 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org adrian kind 38 weisberg, m. (2007). who is a modeler? the british journal for the philosophy of science, 58(2), 207–233. https://doi.org/10 .1093/bjps/axm011 weisberg, m. (2004). qualitative theory and chemical explanation. philosophy of science, 71(5), 1071–1081. https://doi.org/ 10.1086/428011 weld, d. s., & de kleer, j. (2013). readings in qualitative reasoning about physical systems. morgan kaufmann. wiggins, o. p., & schwartz, m. a. (1997). edmund husserl’s influence on karl jaspers’s phenomenology. philosophy, psychiatry, &amp; psychology, 4(1), 15–36. https://doi.org/10.1353/ppp.1997.0011 wimsatt, w. c. (1987). false models as means to truer theories. in m. h. nitecki & a. hoffman (eds.), neutral models in biology (pp. 23–55). oxford university press. world health organization. (1992). the icd-10 classification of mental and behavioural disorders: clinical descriptions and diagnostic guidelines. wright, b., dave, s., & dogra, n. (2017). 100 cases in psychiatry (second edition). crc press/taylor & francis group. zahavi, d. (2019). phenomenology: the basics (original edition). routledge, taylor & francis group. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. kind, a. (2023). how does the psychiatrist know? philosophy and the mind sciences, 4, 6. https://doi.org/10.33735/phimisci.2023.9391 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1093/bjps/axm011 https://doi.org/10.1093/bjps/axm011 https://doi.org/10.1086/428011 https://doi.org/10.1086/428011 https://doi.org/10.1353/ppp.1997.0011 https://doi.org/10.33735/phimisci.2023.9391 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction the phenomenological account problems for the phenomenological account the model-based account of diagnostic reasoning modelling clinical psychiatric diagnostics clinical psychiatric diagnostics as modelling assessing the model-based account of psychiatric diagnostic reasoning conclusion predictive processing as a systematic basis for identifying the neural correlates of consciousness. predictive processing as a systematic basis for identifying the neural correlates of consciousness jakob hohwya (jakob.hohwy@monash.edu) anil sethb (a.k.seth@sussex.ac.uk) abstract the search for the neural correlates of consciousness is in need of a systematic, principled foundation that can endow putative neural correlates with greater predictive and explanatory value. here, we propose the predictive processing framework for brain function as a promising candidate for providing this systematic foundation. the proposal is motivated by that framework’s ability to address three general challenges to identifying the neural correlates of consciousness, and to satisfy two constraints common to many theories of consciousness. implementing the search for neural correlates of consciousness through the lens of predictive processing delivers strong potential for predictive and explanatory value through detailed, systematic mappings between neural substrates and phenomenological structure. we conclude that the predictive processing framework, precisely because it at the outset is not itself a theory of consciousness, has significant potential for advancing the neuroscience of consciousness. keywords active inference ∙ consciousness ∙ neural correlates ∙ prediction error minimization this article is part of a special issue on “the neural correlates of consciousness”, edited by sascha benjamin fink. 1 introduction the search for the neural correlates of consciousness (nccs) has now been underway for several decades (crick & koch, 1990b). several interesting and suggestive acognition & philosophy lab, monash university bsackler center for consciousness science and school of engineering and informatics, university of sussex; azrieli program in brain, mind, and consciousness toronto hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-3906-3060 https://orcid.org/0000-0002-1421-6051 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 2 patterns of brain activity have been revealed but no unequivocal ‘correlate’ has emerged. alongside these developments, a plethora of theories of consciousness have been formulated. many of these theories are conceptually intriguing, and some have empirical evidence in their favour, but the field of consciousness science is some distance from consensus on a theory of consciousness (for overviews, see bayne et al., 2009); the lack of consensus has sparked funding bodies to encourage adversarial collaborations between proponents of different theories (reardon, 2019). as with nccs, this lack of consensus can be attributed partly to continuing debate over the sense in which an adequate theory should ‘explain’ ‘consciousness’. there is a further distinction between scientific and metaphysical theories of consciousness; the former typically at least tacitly assumes materialism and aims at explanation and mechanistic approaches to consciousness, whereas the latter is concerned with the ultimate nature of consciousness rather than with specifics about neuronal mechanisms. here, we propose that it will be fruitful for the field to capitalize on a general account of brain function, which can serve as a systematic basis for identifying nccs, and which may as a consequence increase the predictive and explanatory power of the science of consciousness. we suggest, in particular, that a predictive processing (pp) approach brings many advantages to the scientific investigation of consciousness, primarily through encouraging and operationalizing closer links between phenomenological properties of conscious experience and mechanistic properties of underlying neural substrates. perhaps counter-intuitively, we suggest these advantages accrue in part because the pp approach is not, in itself, a theory of consciousness. 2 the need for a systematic ncc to identify the neural correlates of consciousness, it is commonly accepted that one needs to find the minimally sufficient conditions for the occurrence of a global conscious state or level (such as being conscious rather than unconscious, or being in the minimally conscious state rather than the awake state) or content (such as consciously experiencing a face rather than a house). there exist a number of different methods and typical paradigms for revealing such correlates. for example, binocular rivalry and masking paradigms have been extensively used in studies of conscious content, while for the global conscious state the natural or artificial manipulations of (respectively) sleep and anaesthesia provide productive approaches. these ncc methods are generally, at least implicitly, considered relatively theory neutral – they are not immediately proprietary to any particular scientific theory of consciousness (and proponents of most metaphysical theories, even dualists, can agree that these methods implement a valid scientific endeavor even if, by their lights, they cannot reveal the ultimate truth about consciousness). there is discussion about whether states can be conscious even if not accessed and reportable, and if that is the case, whether the neural correlates of such states can be identihohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 3 fied by paradigms that require behavioural report (block, 1995). there has been progress in extending ncc methods across conditions where introspective or behavioural report are not required and such ‘no report’ paradigms may begin to address this discussion (tsuchiya et al., 2015). while ncc methods may initially appear to be theory-neutral, different theories may gravitate towards certain types of paradigms and proposed neural correlates. for example, theories that deny the possibility of unaccessed conscious experience will tend to gravitate to paradigms that capitalise on reportability. after paradigm-defining research in the early 1990s, the ncc methodology became a focus of the science of consciousness (metzinger, 2000), and has received several reviews and commentaries since, reporting on significant progress (aru et al., 2012; fink, 2016; hohwy, 2009; klein & barron, 2020; koch et al., 2016; miller, 2007; seth, 2009; tononi & koch, 2008). an influential early discussion was given in chalmers (chalmers, 2000), which pointed to a crucial and so far underappreciated distinction between arbitrary and systematic nccs. an arbitrary ncc is revealed when some neural activity in some region of the organism’s brain (as accessed with some neuroimaging or neurophysiological method) reliably correlates with a given conscious state or content. such a correlate is ‘arbitrary’ because mere activation on its own does not help us extrapolate, predict, understand or explain much about the mechanisms underlying consciousness. the fact that region x ‘lights up’ when consciously rather than unconsciously perceiving gabor patches does not, by virtue of this fact alone, allow us to predict under what other stimulus conditions x might also be activated, or which regions might be activated for other types of stimuli. nor does it throw much light on the particular phenomenological properties that characterize such a conscious perception, nor what mechanisms underlie conscious perception generally. even with extensive brain mapping along these lines, the result would be an explanatorily inert descriptive list of areas or patterns of activity. arbitrary nccs are not likely to do much to let us understand the brain basis of consciousness, or to use such an understanding to help with consciousness-related conditions, such as disorders of consciousness (fernández-espejo & owen, 2013), split brains (de haan et al., 2020), or islands of consciousness (bayne et al., 2020). as a terminological note, we take ‘arbitrary’ in the sense that we would have no inkling why the pattern of activity would be like this and not some other way, given some particular phenomenology; such a fully arbitrary ncc is one extreme kind of ncc. we recognize that talk of ‘arbitrary’ nccs may not be entirely felicitous since the ncc in question may have a reliable correlation with phenomenal states that is law-like and exceptionless – and in this sense be ‘non-arbitrary’. in some respects, a better distinction could be between ‘non-explanatory’ and ‘explanatory’ correlates, as in (seth, 2009). note however that the arbitrariness we (following chalmers, 2000) refer to is not in the correlation, reliable or not, with phenomenal states, but rather in the question why it is this rather than another correlated pattern of neural activity that is the focus of the ncc search. being hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 4 ‘systematic’ then means being systematically guided by theoretical considerations of some sort. most actual instances of ncc research will not be fully arbitrary, as they naturally engage to some degree interpretations of the pattern or location of activation, mainly using that pattern or location to inspire a theoretical approach to consciousness. examples of such theory building can be found in the idea that frontoparietal activity is not arbitrary in the sense that it accounts for the global broadcast and access that seems predictive and explanatory at least for access consciousness (mashour et al., 2020), or in the interpretation of the role of ‘posterior hubs’, associated with integrated information theory approaches (koch et al., 2016). in light of such concerns, chalmers notes that if we could “find a neural system whose states do not just arbitrarily correlate with the phenomenal [contents and] states in question, but which vary along a corresponding systematic dimension, then the ncc in question will have much greater potential explanatory and predictive power”. we agree with this reasoning and label this the systematicity requirement for the ncc. (chalmers suggested as an example of systematicity correlations of conscious content with ‘neural representational systems’ (chalmers, 2000: 11); for critical discussion, see noë & thompson (2004) and responses; in this paper, we will consider a more inclusive notion than just representational systems). to unpack the systematicity requirement, we must first determine with respect to what there should be greater explanatory and predictive power. the answer to this question is that explanatory and predictive power should be with respect to both brain activity and phenomenology. a systematic approach will be useful for the ncc search only if its core elements can eventually be used to predict when and how conscious phenomena arise, explain why conscious phenomena have the structure and form they have, and potentially even control through interventions how conscious experience unfolds. we propose that such a systematic approach will best be developed by embedding the proposed correlates of the studied conscious phenomena within a larger understanding of brain function, one which is not itself primarily tied to consciousness. this will help move beyond mere arbitrary correlates by ensuring that the relevant brain activity can be identified and interpreted as more than just “the activity, whatever it is, that correlates with this or that conscious phenomenon”. in other words, it would enable us to look at the correlated brain activity and project systematically from that activity to predict related phenomenological states, and provide unifying explanations of their occurrence. we can then say that, to discharge the systematicity requirement, a larger understanding, theory or, as we shall phrase it, a framework for overall brain function must be found for the ncc. this framework must perform the task of providing explanatory and predictive power for certain patterns of brain activity and the structure and form of their correlated phenomenology. note, it is not necessary for such a framework to itself propose an explanation of consciousness. in fact, it hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 5 may serve the search for a systematic ncc better if this larger framework does not – at least not initially – contain a theory of consciousness. as we will discuss later, parts of the framework may nevertheless coalesce into a theory of consciousness if it turns out to be successful in contributing to a systematic ncc. (we do not argue that discharging the systematicity requirement necessarily requires a framework that is not a theory of consciousness. for example, integrated information theory is a theory of consciousness that proposes systematic relationships between neural mechanisms and phenomenological properties (tononi et al., 2016). note that iit is not generally considered as part of the ncc programme.) naturally, it will be a tremendous challenge to fully cash out this notion of a framework for overall brain function with the requisite explanatory and predictive power. we believe, as we shall discuss later, that there is cause for optimism. for now, an analogy might be useful, that the systematicity requirement is not specific to consciousness science. if we are interested in physiological phenotypic traits then a basic genetic theory of what governs changes to the dna sequence will not be useful on its own, even if that theory is in fact true. rather, to gain explanatory and predictive power with respect to phenotype we need to appeal to epigenetics. epigenetics places genes in context and allows us to build up mechanistic accounts of how genes, cells, bodies, and environments interact over developmental trajectories to explain phenotypic traits. epigenetics is a theoretical approach in its own right and there is considerable work to be done in correlating epigenetic processes with phenomena like cancer. similarly, for a theory of brain function, there will be kinds of frameworks that are not optimal for generating explanatory and predictive power for consciousness, even if they might adequately describe (for example) neuronal and synaptic activities, while there will be frameworks at other levels of description that will have the right kind of explanatory and predictive ‘grain’ for consciousness. with the help of a suitable framework for brain function, the fact that conscious perception of a gabor patch correlates with brain activity x would have greater potential for predicting under what other conditions x would or wouldn’t be present, and under what conditions a similar conscious perception would or wouldn’t be present. critically, such a framework should be capable of providing an adequate language in which to account for phenomenological properties of consciousness (e.g., representational content, sensory presence, feature binding, spatiality or the lack of spatiality, illusion formation, sense of agency etc.), which constitute the explananda of any satisfactory theory of consciousness. that is, unpacking brain activity x within the framework should cast light on phenomenology. in this manner, it would be useful for the empirical search for nccs to be coupled to a framework, not for consciousness per se but rather a framework for brain function, with the right internal structure and level of description to have explanatory potential for phenomenology. such an approach contrasts with the ncc search as it has largely unfolded until now, which may be characterized as identifying relatively arbitrary and piecemeal correlates. hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 6 3 three challenges to the ncc search identifying the general framework for brain function for the systematic ncc will be helped by considering three additional and extensively discussed challenges associated with the ncc search. 3.1 the global states and local contents challenge the first challenge has to do with the widely acknowledged distinction between states and contents of consciousness (laureys, 2005; rosenthal, 1986). most ncc programs are devoted to content nccs, and fewer (though increasingly more) to finding the ncc of global states (sometimes called ‘levels’) of consciousness. the challenge here is that a satisfying account of consciousness may require treating states and contents together (bayne, 2007; hohwy, 2009). the challenge can best be introduced with an example. binocular rivalry is a classic ncc paradigm, where different stimuli are shown continuously to each eye, and conscious perception alternates between them. in a typical binocular rivalry ncc study, a participant may be exposed to rivalry stimuli and fmri data may be acquired, showing alternating regions of activity associated with the alternating conscious percepts (as in, e.g., the classic study by tong et al., 1998). the problem, stressed early on by searle (2000), is that the participant was already conscious when put in the scanner and so the brain activity cannot reveal why one percept is conscious rather than not, but merely what brain activity is associated with selection of the content into consciousness, in an already conscious creature. more generally, the ncc search has largely sidestepped the issue of the relation between global states and contents (exceptions include bachmann, 2012; aru et al., 2019; haun & tononi, 2019; noreika et al., 2017). much remains to be discovered about interactions between global states and contents. for example, it may be that insertion of some contents into some global states changes those states, or that certain states may change the contents inserted into them (e.g., rivalry stimuli may be processed very differently in different global states of consciousness such as delirium, psychedelic states, or absence seizures). indeed, the psychedelic state serves as a compelling example in which there seem to be changes in both state/level and content, which, further, seem to be dependent on each other (schartner et al., 2017). this challenge constrains the search for a systematic ncc in the following way: the systematic dimension under which neural activity is described should be able to have both global states and local content as factors. there should be dimensions of overall functioning – a ‘whole of system’ approach – as well as of individual conscious contents. hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 7 3.2 the screening off challenge: brain structure and phenomenology the second challenge concerns how to ‘screen off’ less explanatory neural correlates from the more explanatory correlates. importantly, the degree to which a candidate ncc is explanatory is not simply a matter of how tight or reliable its correlation is with conscious states or contents. a number of researchers, beginning with miller (2007), have argued that even with very tight, reliable correlates being identified, it may still be an open question which of these are the true constituents or substrates of the conscious states or contents (with the underlying assumption being that only the true constituents would facilitate predictive and explanatory power). if there are three causally chained neural states ns1, ns2, ns3 that tightly and reliably correlate with the conscious experience of seeing a face, it may be that ns1 is a mere causal precursor, and ns3 a causal consequent of the true (and presumably most explanatory) constituent ns2, even though the empirical correlation between all three and the conscious experience is much the same (graaf et al., 2012; miller, 2015). one response to the screening off challenge is to make use of techniques for causal inference (aru et al., 2012; hohwy, 2009; klein & barron, 2020; neisser, 2012; pearl, 2000; woodward, 2003). in particular, better methods for causal intervention should be able to advance the issues. for example, tms, tdcs, and especially optogenetics should allow fine-grained (‘surgical’) manipulation of distinct elements of the causal neural chains of events underlying conscious states and contents (see marshel et al. (2019) for an impressive optogenetic study along these lines). this holds most obviously for precursors, where interventions would lead to effects on consciousness. it also holds for downstream consequences, where interventions should fail to influence consciousness. a second and perhaps more important response lies in examining whether potential correlates (e.g., ns2) better account for the corresponding conscious content or state than precursors (ns1) or consequents (ns3). this requires the ability to distinguish potential correlates on their explanatory potential with (phenomenological) explanatory targets. for example, ns2 and ns1 may correlate equally tightly with a conscious experience of a horizontal gabor patch, but ns1 correlates better than ns2 with several orientations of the experienced patch. here, it is a question for philosophy of science to formulate a substantial notion of ‘explanatory potential’; this may be, for example, in terms of homology, causal/functional isomorphism, contrastive explanations, or mechanistic reduction (bechtel, 2007; craver, 2007; hohwy & frith, 2004; o’brien & opie, 1999). meeting the screening off challenge would therefore be facilitated if brain activity could be ‘carved at the joints’ in a way that both allowed meaningful and efficient interventions and which mapped to an equivalent partitioning of the explanatory targets, namely the phenomenological properties of experience. there is a considerable and challenging task in revealing this ‘architecture’ of conscious hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 8 phenomenology, and it seems likely that approaches to screening off at the level of brain activity will develop in tandem with understanding of the phenomenological explanatory targets. in particular, we may learn more about the building blocks of the neural system and the varieties of phenomenology as we learn to better intervene on both. a standard example serves to illustrate the difficulties inherent in carving brain activity at its joints such that it matches the structure of the phenomenological explanatory targets. an early dynamical ncc approach proposed that conscious contents depended on gamma-band synchrony among neural populations (crick & koch, 1990a, 1990b). this proposal gained conceptual plausibility through its association with the classic ‘binding problem’ in neuroscience of how spatially segregated processing of distinct object features become ‘bound’ together. despite some early empirical support it became rapidly clear that synchrony among neural populations does not always go along with conscious contents and/or binding (for review and discussion, see gold, 1999; engel & singer, 2001). to the contrary, an excess of neuronal synchrony reliably leads to the global loss of consciousness, as in absence epilepsy. in hindsight, the insufficiency of a simple association between neural synchrony and consciousness is hardly surprising. allowing a more global view of binding, conscious scenes are not well described as single unified perceptual objects; rather, conscious experiences normally comprise of a multimodal range of contents each of which retain a distinctive identity while being embedded within a single unified scene. this phenomenological observation may then be a good guide for how to carve brain activity at the joints, where neuronal synchrony may still play some role – though no longer as the ncc itself (tononi & edelman, 1998). 3.3 the action challenge we turn next to a challenge which is more prescriptive, about where consciousness science should go, than the previous challenges. recent years have seen an increased focus on the role of action (in a broad and inclusive sense) for the mind. a number of broadly ‘enactive’ approaches have been suggested, which highlight the important point that conscious perception is closely entwined with agency, intention, movement and behaviour (contributions to this debate range very widely from radical reconceptualizations of cognitive science, biological approaches to active sensing, and computational approaches, see, e.g., hurley, 1998; engel et al., 2016; hutto & myin, 2013; krakauer et al., 2017; noë, 2004; noë & o’regan, 2001; schroeder et al., 2010). indeed, some of the pressure to include action in consciousness science comes from the predictive processing literature itself, insofar as this literature rests on a notion of action (as we discuss later) and has already begun to address consciousness (parr et al., 2019), and so insisting on this challenge will somewhat preemptively help secure a good position for this framework. hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 9 our conscious experience is intimately coupled with our actions at both personal levels (for example, a voluntary decision to reach for a cup of tea) and subpersonal levels (for example, as we actively sample the world in visual search, by moving our eyes around a visual scene). though we might readily imagine an utterly actionless but conscious system, it seems misguided to suggest that action is an optional extra for any actual system that is conscious. notice, this is not just the idea that consciousness is for action, but rather that the conscious experience has the structure and trajectory it has in part because we are active systems. presently, neuroscientific theories of consciousness and ncc research that focus on conscious states and contents pay little attention to the role of action in consciousness. the natural exceptions are studies that explicitly focus on experiences of intention, volition and agency, but these studies are rarely taken to motivate general accounts of consciousness or brain function (see also fleming et al. (2015) for an intriguing link between action and metacognition). one possible reason for general neglect of action is that action and consciousness seem to be dissociable, at least on the face of it. for example, we can be conscious without overt action as in dreaming, and we can act without consciousness, as in reflexes. nevertheless, a systematic dimension to brain function that includes action as an integrated part would be desirable, allowing us to posit the necessity of action for consciousness. such proposals might include covert, mental actions (such as endogenous allocation of attention) as well as overt actions – and they might distinguish between actions being constitutively necessary (i.e., necessary at any given time for consciousness), compared to being merely diachronously necessary (i.e., necessary as part of the history of the organism’s interaction with its environment). summarizing these three challenges for the systematic ncc search, we would like to have a framework for interpreting brain activity that allows interactions between global and local states of consciousness, enables intervention at ‘joints’ which make sense both in terms of brain activity and phenomenology, and which allows an integrated role for action in consciousness. next, after considering some theoretically motivated constraints, we will propose that the predictive processing framework fits the bill. 4 theories of consciousness there are now many theories of consciousness vying for attention and empirical evidence. here is an incomplete list of some of the most discussed theories, with some exemplifying references: global neuronal workspace theory (baars, 1988; mashour et al., 2020) integrated information theory (tononi et al., 2016; tononi & edelman, 1998) recurrent processing theory (lamme, 2020, 2010) higher order thought theories & metacognitive theories (brown et al., hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 10 2019; fleming, 2020; lau, 2008; lau & rosenthal, 2011; rosenthal, 1997) radical plasticity thesis (cleeremans, 2011) virtual reality theories (metzinger, 2004; revonsuo, 2006) attention-based theories (graziano & webb, 2015; prinz, 2012) heterophenomenology (dennett, 1991) core consciousness theory (damasio, 2000) orchestrated objective reduction (hameroff & penrose, 2014) electromagnetic theory (mcfadden, 2020, 2002) some of these labels and groupings ignore substantial differences. for example, though versions of higher-order thought theory are closely linked to metacognitive theories they differ significantly in how they cash out the higher-order element; the two attention-based theories differ, though both appeal to attention, and similarly for the two virtual reality theories. there are also now several theories proposed which connect to the framework we shall consider below, namely predictive processing; we discuss these theories below. though there is some diversity in the explanatory targets and in metaphysical assumptions (for discussion of this kind of diversity, see e.g., seth, 2021), there is general agreement that a theory of consciousness should explain the underlying mechanism(s) for conscious contents and/or global states of consciousness. regarding metaphysical assumptions, there is some debate about whether theories of consciousness are or should be aimed at resolving the metaphysical mindbody problem (the “hard problem”, see also alternate versions such as the knowledge argument or the explanatory gap; descartes, 1641; chalmers, 1996; jackson, 1982; levine, 1983). while some theories explicitly seek to solve the hard problem (e.g., integrated information theory, orchestrated objective reduction, electromagnetic theory), others do not, or do so only implicitly or indirectly (e.g. recurrent processing theory, global neuronal workspace theory). to solve the hard problem, a theory must remove any open question about whether a system characterised in terms of that theory is in fact conscious. this is a tall order. it seems to us the best hope for making inroads on the metaphysical question is to take an indirect approach, and to extend this approach by searching for systematic neural correlates of consciousness (though we acknowledge of course that the hard problem may be too metaphysically hard for any scientific approach to be successful). on one level, this strategy will deliver greater predictive and explanatory power regarding factors that shape and modulate consciousness. such advances will have important practical consequences in (for example) neurology and psychiatry. equally important, a focus on systematic nccs recognizes that ‘consciousness’ is not a singular explanatory target. the multifaceted structure of conscious phenomenology instead provides a repertoire of explanatory targets. applying a general theory of brain function alongside a refined repertoire of explanatory targets (as opposed to implicitly assuming consciousness to be a singular phenomenon) should prove powerful in circumventing the hard problem of why conscious experience happens hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 11 at all (seth, 2016). in other words, a sensible mapping – achieved by satisfying the systematicity constraint – of brain states and phenomenology may make the hard problem less pressing, or perhaps even dissolve it altogether. as mentioned, later we examine extensively whether predictive processing can provide the theoretical framework needed to deliver a systematic ncc. this discussion will raise the issue whether pp is itself a theory of consciousness (and really belongs on the list above), or is on the way to becoming one. for now, we note that there have been several recent proposals for how to connect pp to consciousness. our point here is that pp is not first and foremost a theory of consciousness since, as a framework for brain function, it could exist perfectly well even in the absence of any discussion of consciousness. this fact about it will, as we shall see, encourage its consideration as the basis for a systematic ncc. 4.1 common themes in spite of the diversity of current theories of consciousness, the most prominent have at least two elements in common. if these theories are at least roughly on the right track, then it will be desirable if the underlying framework for a systematic ncc, which we seek, can give rise to these common elements. the first element is uncertainty reduction. many theories associate consciousness with a mechanism such that experiencing a certain content in a certain conscious state is related to uncertainty reduction, in an information-theoretical sense. for example, global neuronal workspace theory operates with the idea of contents being ignited into consciousness when an evidence accumulation threshold is reached (dehaene, 2011); versions of higher-order thought theory operate with criterion setting that optimizes a variable signal-to-noise ratio (lau, 2008), the same holds for the radical plasticity thesis, and metacognitive theories; integrated information theory formally defines conscious states in part in terms of the possibilities excluded by being in that state (oizumi et al., 2014); recurrent processing theory considers whether the mechanisms for consciousness have to do with integrating information and with learning, both of which are uncertainty reducing. the reasoning behind this insistence on uncertainty reduction often seems to be that a conscious system has a tendency to settle in one unified and highly informative representational state. of course, we are not here placing all these diverse theories on the same footing with respect to how they conceive such uncertainty reduction. different theories propose diverse and often incompatible mechanisms, operating at quite different spatiotemporal and even conceptual scales (for discussion, see doerig et al., 2020; graziano et al., 2020). the second element that many theories of consciousness have in common is top-down signalling. many theories of consciousness agree that there is little chance for consciousness to occur in a system that does not have top-down (‘backward’ or ‘inside-out’) signalling in addition to the bottom-up (‘forward’ or ‘outsidein’) flow from the sensory cortices. for integrated information theory, it turns out hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 12 that a system with only bottom-up connectivity has no φ (‘phi’, the quantitative measure of consciousness on this theory; (oizumi et al., 2014: 19); recurrent processing theory is based on the notion of backward, ‘recurrent’ processing; ncc studies in support of global neuronal workspace theory suggest that consciousness is ‘ignited’ partly by recurrent connectivity; higher-order thought and metacognitive theories all operate with an assumption that some higher-level process can ‘look down’ at lower-level states; and finally, it is hard to see how virtual reality theory would work without extensive top-down signalling in the construction of the internal virtual reality model that can arise in the absence of bottom-up signals. we also note that studies of the global state of consciousness in anaesthesia suggest that a global state of consciousness depends on top-down signalling in the brain (boly et al., 2011, 2012; pal et al., 2020). there are significant differences amongst the theories concerning how this topdown signalling should work, neurophysiologically and computationally. some theories (e.g., global neuronal workspace theory and versions of higher-order thought and metacognitive theories place the source of top-down signals more forward in the brain than others, such as recurrent processing theory). in spite of these differences, the agreement on the need for top-down signalling is striking, and does seem to be supported by the available evidence. given that most theories of consciousness allocate key roles to uncertainty reduction and top-down signalling, and assuming these theories are at least roughly on the right track, it would be desirable if the systematic framework for the ncc was able to accommodate these two elements in a central but relatively theory neutral role. as we shall see, this requirement also makes pp a good fit. 5 core properties of pp predictive processing is a theoretical framework that posits that the brain’s overall function is to minimize long-term average prediction error. according to pp, the brain harbors a generative model of its internal, bodily environment as well as its external, worldly environment and continually finesses this model through a prediction/prediction error dynamic. the predictions are compared to the actual input, and the brain’s job is to keep prediction error within expected bounds in the long term (friston, 2010, 2005; friston, parr, et al., 2017). the underlying objective for this pp activity is for the organism to act (internally and externally) to maintain itself in its expected (homeostatic) states. as we shall acknowledge further in the final section, the pp framework is by no means universally accepted and there is substantial, interesting and challenging debate about it. here, we seek to primarily develop its promising properties for the science of consciousness. predictive processing in the brain can occur in four basic ways, all analogous to familiar descriptions of scientific inference (for comprehensive reviews, see e.g., bogacz, 2017; buckley et al., 2017; hohwy, 2020a; yon et al., 2019). the central hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 13 principle for the operation of these four processes is that on average and over the long run, (statistical) surprise should be minimized: perception. the model can be revised in the light of prediction error, which will reduce the error over different timescales, at different levels of the model’s hierarchy, and improve the model fit. on short timescales prediction errors are minimized to find the optimal (or near-optimal) posterior (using approximations to bayesian inference). this is perceptual inference (friston, 2003). on longer timescales the parameters of generative models can change, corresponding to perceptual learning, which in turn finesses models to perform better in perceptual inference. this is analogous to how a scientist might update their hypothesis in the light of new data, hoping to eventually learn models that generate predictions encountering only little prediction error. action. the model can be kept fixed and used to generate predictions even though prediction error may transiently increase. action, or selective sampling of the sensory input, can then be used to minimize the prediction error. in particular, proprioceptive predictions are delivered to the classic reflex arcs, which engage the body plant until the expected sensory input is obtained. this is action, and is analogous to how a scientist may retain a hypothesis and control the environment for confounds until the expected evidence obtains (or, less optimistically, engages in repeating slight variations of an experiment until the desired result arrives). action is then simply prediction error minimization with a different direction of fit than perceptual inference. in the pp framework, this is active inference and involves inference of policies for action, under expectations that precise (unambiguous) policies best minimize future prediction error (friston et al., 2013; friston, fitzgerald, rigoli, schwartenbeck, o’doherty, et al., 2016). conceiving of active inference in this way allows for both instrumental actions (actions that directly fulfil perceptual predictions) and exploratory, epistemic actions that improve the capacity for future instrumental action (tschantz, seth, et al., 2020). importantly, actions can also be internal to the organism (‘intero-actions’), so that interoceptive predictions enslave autonomic reflexes in order to maintain physiological homeostasis (allen et al., 2019; seth, 2013; seth & tsakiris, 2018). active inference is indispensable to pp in any sufficiently rich understanding – a pp system without action will not be like any biological creature, certainly not like any system with a brain. attention. the dynamic exchange of predictions and prediction errors in pp can be modulated according to the expected precision of prediction error, relative to the prediction, such that prediction error minimization occurs on the basis of prediction error estimated to be trustworthy; this amounts to gain control (or gating), and functionally maps on to attention (feldman & friston, 2010; hohwy, 2012). this is analogous to assessment of variance in statistical inference, and to how a scientist over time will build up expectations about which types of measurements will be precise and which not. precision optimization refers to inference and learning for precision expectations. note that despite the common association of hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 14 attention with exteroceptive modalities, precision optimization is ubiquitous in pp and applies (for example) equally to interoceptive inference (petzschner et al., 2019). recognition, understanding, exploration. the model can be simplified (cf. complexity reduction) to enable prediction error minimization in the long run; this corresponds to bayesian model selection, where complexity is penalized. it is analogous to how a scientist may prefer simpler models in the long run even though a more complex model may better fit the current evidence. functionally, this may map on to aspects of recognition (rather than mere detection) of sensory causes and their interactions, within an understanding of the causal structure of the environment, where different models of what may have caused sensory input are weighed against each other. as mentioned above, in relation to epistemic action, an emphasis on minimizing future prediction error can lead naturally to notions of exploration, where prediction error is occasionally allowed to increase in order to enable new causal information to be discovered, opening new avenues for efficient future prediction error minimization (friston, fitzgerald, rigoli, schwartenbeck, o’doherty, et al., 2016; friston et al., 2015; tschantz, seth, et al., 2020). putting these elements together, we arrive at the powerful idea that the brain embodies a model of its environment (including the organism itself), and exerts control on this environment through statistical inference geared towards minimizing its own long-term prediction error. this ambitious framework has been the focus of much work across neuroscience, cognitive science, biology and philosophy (e.g., metzinger & wiese, 2017; hohwy, 2020a; mendonça et al., 2020). the point of minimizing prediction error in the long run average, is that this allows the organism to stay within its expected states – i.e., to stay alive. this gives a perspective on the whole organism as primarily concerned with maintaining homeostasis by repeatedly visiting (expected) states of its (internal and external) environment. the overall prediction error minimization of the organism is therefore put into the context of the specific environmental impact (on physiological homeostasis) the organism is exposed to. this foundational biological perspective underwrites the claim that the pp framework is not in itself a theory of consciousness. a key point about pp is that the mechanistic process of sensory prediction error minimization will (given certain assumptions such as normally distributed data and separable timescales) come to approximate bayesian inference on the causes of sensory signals. this provides another perspective on pp: as transforming a typically intractable inference problem (exact bayesian inference) into a tractable optimization problem. this perspective in turn licenses the view that a system that manages to minimize prediction error in the long run will tend to behave as if it were a representational system. this follows because bayesian inference is representational, in the sense of resting on statistical models of the causes of data. (already there is a link here to chalmers’ example of a systematic ncc, mentioned earlier, as being a neural representational system, though pp is much richer than just this). hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 15 this perspective highlights that the pp framework can be cast at different levels of abstraction which make different claims about the underlying mechanism. on ‘strong’ interpretations, the brain is in fact utilizing generative models and approximate bayesian inference to accomplish prediction error minimization. on ‘weaker’ interpretations, brains (and other systems) can be described as if they are utilizing such models and inferential processes, though the underlying physicalcausal mechanism may be different (cf. the watt governor; van gelder, 1995). here lie interesting questions about the necessary and sufficient conditions required to be vs. to have a model, what this implies for pp, and for the application of pp as a framework for expressing systematic nccs (baltieri et al., 2020; kiefer, 2017; seth & tsakiris, 2018; van es, 2020). note that for organisms living in a dynamically changing world, bayesian inference will have to anticipate changes in the world that may threaten their prior beliefs (especially those prior beliefs concerning homeostasis). that is, such systems cannot just accumulate evidence under the assumption that the causes they are sampling from remain constant, they must take volatility into account too (i.e., changes in statistical distributions over time). this implies that bayesian inference should be hierarchical: an adequate prediction-error minimizing system should find patterns in patterns and build corresponding expectations about expectations, be able to regulate the rate with which it learns, and it should be prepared to relinquish parts of the model that it expects to become obsolete (mathys et al., 2011, 2014). in addition, with an appeal to hierarchical inference there is an opportunity to ameliorate the problem of understanding where priors come from: in a hierarchical setting, posteriors at one level form (empirical) priors for the level below. notice that under pp, hierarchies should not be understood in overly idealized terms of neatly stacked levels passing messages only to their nearest neighbours. cortical hierarchies, though mathematically describable, will recapitulate the causal structure of the world it is modelling, with several leap-frogging message pathways among more distant levels (markov et al., 2013). these core properties of the overall pp framework can be implemented mechanistically through process theories for particular kinds of systems, such as brains. the best known process theory for perceptual inference is predictive coding, and gradient descent on variational free energy for active inference (friston, 2003; friston, fitzgerald, rigoli, schwartenbeck, & pezzulo, 2016). predictive coding is a concrete ‘message passing’ scheme where neuronal representations in higher levels of cortical hierarchies generate predictions of representations in lower levels. the top-down predictions are compared with representations at the lower level to form a prediction error (associated with the activity of superficial pyramidal cells). this mismatch signal is passed back up the hierarchy, to update higher level representations (associated with the activity of deep pyramidal cells). as a result of this recursive exchange of signals, prediction errors are suppressed (‘explained away’) at all hierarchical levels, to provide a hierarchical explanation for sensory inputs at the very lowest (sensory) level. this message passing process encompasses a hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 16 limited number of signals, consistent with pp: top-down predictions and expected precisions, endogenous precision modulation, and bottom-up precision-weighted prediction error and precision prediction errors (canonical microcircuits for this scheme have been proposed bastos et al., 2012; see also walsh et al., 2020; and friston, 2019b; see also teufel & fletcher, 2020 for possible roles for “bottom-up” signaling of predictions). minimization of variational free energy for active inference is less widely known but is beginning to be extensively studied too (friston, fitzgerald, rigoli, schwartenbeck, & pezzulo, 2016; parr & friston, 2018; tschantz, millidge, et al., 2020; tschantz, seth, et al., 2020). outstanding issues in this area of research concern the underlying neuronal architecture of continuous vs. discrete processing, the relative extent to which predictive coding or complexity reduction occurs at higher or deeper levels of the cortical hierarchy (friston, parr, et al., 2017), how best to formalize notions of ‘expected free energy’ (millidge et al., 2020), and – more generally – its relation to machine learning approaches such as reinforcement learning and ‘control as inference’ (tschantz, millidge, et al., 2020). 6 pp as a systematic ncc now we are in a position to propose pp as a promising candidate for satisfying the systematicity requirement for the ncc. recall that to meet this goal we need a general framework for brain function with sufficient explanatory and predictive power to account for the structure and form of its correlated phenomenological distinctions. we further identified three challenges to the ncc search (global vs. local, screening off, and action) as well as two common elements of theories of consciousness (uncertainty reduction and top-down signal flow), which should be used as selection criteria for this general framework. pp evidently fits this bill. here we show how this is the case assuming some ‘strong’ version of pp, that is, the existence of some explicit message passing scheme similar to that described for perceptual and active inference above. (this assumption is important because it opens the possibility that ‘weaker’ versions of pp will work less well to establish a systematic correlate for consciousness, in other words, pp will also be able to describe systems that are not conscious). pp holds promise for addressing the ‘global states and local contents’ challenge. pp is an account of the global state of creatures that act to remain in their expected states, and who in the course of this process install a probabilistic model of their internal and external environment. this perspective allows that local input, caused by the occurrence of particular environmental states, can be inserted into this global state, potentially leading to interactions between the global prediction error landscape and the local exposure to new causes. this perspective then connects specific contents to the global regulation of homeostatic and allostatic states (for discussion, see stephan et al., 2016; corcoran & hohwy, 2018). this connection illustrates how the machinery of pp can be used to systematically account for properties of consciousness, without being a theory of consciousness hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 17 itself. from this perspective, global states of consciousness like sleep and anesthesia may involve disruption of the top-down or bottom-up signaling that is needed to instantiate the process of prediction error minimization with respect to local exposure to sensory causes; sleep and dreaming may relate to how expected reductions in precision of sensory input lead to global reduced gain on prediction error (hobson & friston, 2012), and propofol-induced loss of consciousness seems to relate to disruption of top-down signaling (boly et al., 2012). pp holds promise also for the ‘screening off: basic brain states and phenomenology’ challenge about how to ‘screen off’ less explanatory neural correlates from the more explanatory correlates. to address this challenge, we highlighted the importance of having ways of intervening on causal chains in neural systems that in a reasonable sense ‘carve nature at the joints’, that is, are informed by a mechanistic framework for brain function, and which are formulated at a level that can capture phenomenological distinctions. pp meets the first part of this challenge since it posits a mechanism of prediction error minimization, replicated throughout levels of a cortical hierarchy. in other words, manipulations and interventions that are aimed at screening off for the purposes of the ncc, should be informed and interpreted by the process theories animating pp (we recognise that there remain many open theoretical and empirical questions about these process theories (friston, fitzgerald, rigoli, schwartenbeck, & pezzulo, 2016; heeger, 2017; van der helm, 2016)). consider, as an example, an attempt to screen off upstream and downstream correlates for conscious face perception. in this example, it makes sense to ask how areas implicated in face perception, such as the fusiform face area (ffa) and the occipital face area (ofa) stand to each other in terms of hierarchical inference and message passing of predictions and prediction errors. it could be that even if activity in ffa is more tightly coupled to face perception than activity in ofa, we cannot understand the systematic contribution of ffa to conscious perception without appreciating the role of prediction error signals arriving from another area (such as ofa). in other words, it may be that pp can make contributions to the search for systematic nccs by guiding the balancing of functional segregation (localization) and functional connectivity (integration) in formulating candidate nccs. in a concrete instance of this kind of approach, weilnhammer and colleagues (weilnhammer et al., 2020) employed a predictive coding model of binocular rivalry to provide evidence that prefrontal activity is not downstream to alternations in conscious perception. pp then also offers a powerful approach to the phenomenological part of this challenge, namely, to match neural substrates to the structure of conscious experience. certain aspects of pp readily lend themselves to matching with informal descriptions of perceptual phenomenology. for example, conscious experiences typically contain many objects set in a contextualized scene. the hierarchical nature of pp nicely accommodates this phenomenological observation, so that sensory signals are ‘explained away’ by predictive models combining many degrees of invariance. this admits predictions about sensory causes that have different hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 18 degrees of invariance with respect to exogenous environmental dynamics and sensory changes resulting from actions. this is a useful property in light of the range of invariances characterizing perceptual phenomenology (e.g., objects are perceived invariantly as being a certain way while at the same time we experience their changing position, lighting, and even presence to our sensory surfaces at all during occlusion). pp also readily accounts for multisensory integration through approximation to precision-weighted optimal bayesian inference (alais & burr, 2004; hohwy, 2013: ch. 5); it also speaks to unusual conscious experiences, for example in delusions and hallucinations, which may reflect sensory signals being inappropriately overwhelmed by perceptual priors (corlett et al., 2019; fletcher & frith, 2009; powers et al., 2017). perhaps most interestingly, the language of pp enables new questions (or reanimates old, neglected questions) about conscious phenomenology. for example, extending early claims that peripheral vision is ‘somehow statistical’ (lettvin, 1976), we could ask whether perceptual content in the visual periphery is constituted by precision estimations, instead of merely being shaped by them. or, starting from empirical observations rather than phenomenology, we could ask how the conceptual tools of pp account for intriguing findings such as the inability to distinguish between large variations in colour saturation for rapidly presented images (cohen & rubenstein, 2020). in other words, the language of pp enables new and productive mappings to be drawn between phenomenological descriptions and underlying neurocognitive processes. if pp can facilitate informative mapping of structured phenomenology with neural substrates screened off in an informed manner, then it can make significant headway on this challenge. pp can also help address the ‘action challenge’. this challenge asks for a flexible way to integrate action into an account of the structuring of conscious experience. pp does this first in a general way, by defining the organism in terms of how it acts to maintain itself in its expected states. within this general setting, pp can account for the specific role of action in perception through the notion of active inference where policies are inferred on the basis of the expected prediction error, given the selected policy. policies cover long-term goals such as going shopping, or choosing an education (or maintaining homeostatic and interoceptive states in the face of environmentally or internally caused fluctuations), and translate these goals through shorter-term policies all the way down to control states that can be executed by the agent’s body (such as limb or eye movement). the key move here is to recognize that goals (in a reinforcement learning sense) can be expressed as priors within pp, so that an organism will come to act in ways that fulfil its (biased) expectations about its future condition (friston, fitzgerald, rigoli, schwartenbeck, o’doherty, et al., 2016; tschantz, millidge, et al., 2020). the fact that both perception and action are inferential, albeit with different directions of fit, places them on equal footing and ensures they can meaningfully participate within the same, systematic ncc project. this can be exemplified in a computational model that explains one of the basic paradigms in consciousness hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 19 science, binocular rivalry, in terms of active inference where selective sampling, driven by beliefs about volatility, leads to perceptual alternations (parr et al., 2019). in addition, since, within pp, actions themselves emerge from fulfillment of (proprioceptive) predictions, the pp framework provides a means of accounting for the phenomenology of action – such as sense of agency; (pacherie, 2008) as well as the influence of action on other conscious contents (such as the fidgeting we experience when bored or excited (perrykkad & hohwy, 2020), or the role of actionrelated predictions in the shaping of conscious access (skora et al., 2020). pp is therefore well placed to systematically integrate an action component into the ncc, and thereby to afford an understanding of the key role for action in consciousness, rather than conceiving action as an optional extra to consciousness. in addition to the three challenges for a systematic ncc, we also identified two constraints, extracted from theories of consciousness, concerning uncertainty reduction and top-down signaling. pp can easily provide a systematic framework for the ncc that identifies uncertainty reduction as a core element. the reason is simple: prediction error minimization, construed as (implicit) minimization of surprisal, is equivalent to uncertainty reduction over the long-term average. prediction error minimization happens over multiple hierarchical, spatiotemporal scales, and through several interlinked processes (perceptual and active inference, precision optimization, complexity reduction); this richness affords some capacity to capture the different styles of uncertainty reduction that various theories of consciousness focus on. thus pp seems consistent with the following ideas: that (phenomenal) consciousness correlates with recurrent processing within perceptual cortices (as opposed to requiring the contribution of domain-general, higher-order, or ‘executive’ cortices) (cf. recurrent processing theory); that (access) consciousness requires ignition into a more anterior global workspace, perhaps once a prediction error minimization threshold has been crossed (cf. global neuronal workspace theory; this leaves an interesting open question for pp as to what determines this threshold, with one option being that it varies contextually with expected precision); that a conscious experience is the way it is, and not some other way, in part because of the repertoire of excluded alternatives (cf. integrated information theory; though much formal work still needs to be done to ascertain in which ways uncertainty reduction in pp and integrated information theory may align); or that metacognitive assessment of first-order representations requires prefrontal areas (cf. higher order thought and metacognitive theories). note, we are not claiming that all these theories reduce to pp. instead, we argue that several of them are consistent with different aspects of pp (at least in their theoretical motivation, if not in formal detail), and – further – that pp’s tools may be useful in broadening the conceptual appeal and empirical grip of some of these theories. pp is also well placed to underpin the second constraint: top-down signalling. this is because most prediction error minimization schemes, such as hierarchical predictive coding, are cashed out in ways that emphasise top-down message passhohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 20 ing. it may be that in some (e.g., single celled) pp organisms, there is no straightforward way to distinguish anything resembling top-down or bottom-up signalling. but it is reasonably clear that directionally-oriented message passing does occur in primates and many other animals, and the prevalence of top-down or ‘inside out’ connections in the brain is indeed one of the main motivations for pp in the first place (e.g., friston, 2003). while this basic functional architecture is undeniably prominent, many open questions remain, the resolution of which likely deliver a more complicated picture. for example, it may be that different message passing schemes are required for continuous and discrete processing (friston, parr, et al., 2017), and there may be a role for ‘bottom-up predictions’ (teufel & fletcher, 2020), perhaps in relation to ‘amortized’ inference in machine learning. from the perspective of developing a systematic ncc, these questions should be considered in relation to their relevance for phenomenological properties of consciousness. in short, we find that pp is an exceptionally promising candidate to satisfy the systematicity requirement in the search for candidate nccs. it is a generalpurpose mechanistic framework, which has plausible neurobiological underpinnings, and which is well equipped to address significant challenges facing the ncc project. moreover, it is able to marry the ncc concept to key considerations in some prominent theories of consciousness, and its constructs already show potential in accounting for fundamental aspects of conscious phenomenology. by interpreting ncc findings in terms of something that is not itself, in the first instance, a theory of consciousness, it therefore seems possible to increase the predictive and explanatory value of the ncc. of course, something that has great predictive and explanatory power with respect to x may itself be, or become, a theory of x. and we view it as indeed possible (even likely) that, in time, some aspects of pp will emerge as a theoretical explanation of consciousness. thus, pp could eventually supersede current theories of consciousness. for this to happen, a core challenge will be to distinguish those aspects of pp that apply particularly to conscious creatures, from those that don’t. in its broadest and most general formulation, in terms of the free energy principle (friston, 2019a), the theory applies to any system in a far-from-equilibrium steady state, which almost certainly includes non-conscious creatures – indeed it may apply literally to every thing. the strategy should therefore be to learn what is special about the way in which prediction error is minimized in creatures that we are relatively sure are conscious, such as primates (probably all mammals), and with this understanding to formulate a theory of at least some aspects of consciousness in terms of pp (wiese, 2018, 2020). there are various possible avenues here, but to give this prospect some substance, it may be that conscious selfhood is grounded in control-oriented active inference of physiological homeostasis – a “beast-machine” theory of consciousness (seth, 2021; seth & tsakiris, 2018). we return to this theme in the concluding remarks below. hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 21 7 predictive and explanatory power of pp as a systematic basis for ncc implementation of pp to arrive at a systematic ncc can happen on several fronts, with much relevant work already underway. we give a flavor of some of this work below. at the theoretical level, there are emerging basic formulations of predictive processing (or active inference, or the free energy principle, or adjacent or kindred formulations) directed at consciousness (or sentience, or self) (carhart-harris et al., 2014; chang et al., 2020; clark, 2019; friston, 2018; friston, wiese, et al., 2020; rudrauf et al., 2017; williford et al., 2018), including the emergence of the problem of consciousness itself (clark et al., 2019), or how pp might supersede, dissolve, or deflate the ‘hard problem’ of consciousness (clark, 2019; seth, 2016, 2021). there has been a significant amount of research mapping a large set of properties of pp to many different aspects of conscious phenomenology (clark, 2015; hohwy, 2013, 2020a; seth, 2019). existing theories of consciousness are being interpreted in the light of pp, such as heterophenomenology (dołęga & dewhurst, 2020), global neuronal workspace theory (hohwy, 2013; whyte, 2019; whyte & smith, 2020), versions of higher-order thought theory and metacognitive theories (hohwy, 2015; sandved smith et al., 2020), attention-based theories (marchi & hohwy, 2020), and integrated information theory (for related discussion, see kanai et al., 2019). moving closer to empirical implementation, pp is increasingly able to speak to key paradigms in consciousness science, such as binocular rivalry. an early contribution here proposed that if some representational content enters consciousness, then this content is captured in the perceptual hypothesis that is best at minimizing prediction error; or, equivalently, for content to be conscious it must be represented by the posterior with the most evidence (highest probability) (hohwy et al., 2008). note that this is only a starting point since it does not pretend to be a sufficient condition for consciousness (and even with that caveat, there are some additional questions about how active inference may contribute to this idea, as the action challenge highlights; for example, desired or survival-relevant states might influence the ‘winning’ posterior through increased prior precision (seth & tsakiris, 2018)). the idea that the best perceptual hypothesis determines conscious content has been pursued successfully with quantitative predictions and neuroimaging methods (weilnhammer et al., 2017), pointing to involvement of prefrontal cortex in the lead-up to perceptual transitions (weilnhammer et al., 2020). this position has since been re-visited in the computational model mentioned earlier, which integrates active inference and thereby displays the potential for pp to see conscious perception in its essential relation to action (parr et al., 2019). one aspect of pp that applies across many empirical contexts is the association of top-down signaling with predictions (or expectations) that enable and shape perceptual content. this sharply defined functional role contrasts with some other perspectives, like recurrent processing theory, in which the role of top-down sighohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 22 naling is left comparatively unspecified. following early demonstrations that the integrity of top-down signaling within visual cortex was necessary for conscious perception of motion (pascual-leone & walsh, 2001), there has been a productive upsurge in studies of how expectations shape perception, much of which has been directly inspired by versions of pp (lange et al., 2018; summerfield & egner, 2009). behavioural experiments have addressed the simple question of whether we (are more likely to) consciously see what we expect to see, or what violates our expectations. pinto and colleagues found evidence using continuous flash suppression that valid perceptual expectations accelerated conscious access (pinto et al., 2015); melloni and colleagues used a perceptual hysteresis design to show that valid expectations facilitated conscious perception and altered the latency of event-related potentials distinguishing seen from unseen stimuli (melloni et al., 2011); this focus extends to action, where yon and colleagues found that action sharpens the sensory representations of expected outcomes (yon et al., 2018), and where suzuki and colleagues found that valid sensorimotor predictions accelerated breakthrough of virtual objects from continuous flash suppression (suzuki et al., 2019; see also skora et al., 2020). a growing number of studies using fmri have provided evidence for how contextual expectations modulate activity in a content specific manner in early visual areas even in the absence of sensory input to those areas, and have begun unraveling how the interplay of predictions and prediction errors might take place across distinct laminae of the cortex (kok et al., 2019; muckli et al., 2015). in one recent study, expectations based on word context were found to both increase the perceptual legibility of individual letters, and to sharpen their representations within early visual cortex (heilbron et al., 2020). another study, using meg, was able to quantify the time course of the influence of probabilistic top-down priors in perception (aru et al., 2016; dijkstra et al., 2020). a further study systematically explored the dynamic interactions of predictions for conscious awareness (meijs et al., 2018). a number of studies have applied pp to the question of aberrant conscious perception, for example, in schizophrenia (cole et al., 2020; stuke et al., 2019) and autism (lawson et al., 2017; skewes et al., 2015). an important question in consciousness science is the relation between consciousness and attention. theories of the attention-based type suggest a tight relation (graziano et al., 2020; prinz, 2012), while some theoretical and empirical considerations suggest that consciousness and (at least top-down) attention can be dissociated (koch & tsuchiya, 2007). pp seems able to provide a middle ground allowing a tight connection between attention and consciousness together with elements of dissociation, via the multifaceted involvement of precision optimisation (hohwy, 2012). as mentioned, pp sees attention as precision optimisation, which is required for any perceptual and active inference in a changing world. this means that any processing leading to any conscious perception must include attention, at least in this pp-sense, to some degree. methodologically, pp has inspired studies of attention and expectation treated as orthogonal factors. for example, one influential study used pp to derive hyhohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 23 potheses about interactions between attention and expectation, and explored how these interactions would manifest for conscious perception at low visual processing areas of the brain (v1); (kok et al., 2012). dual-task paradigms can also be used to dissociate the effects of expectation from those of attention (e.g., sherman et al., 2015, in a study connecting prior expectations to metacognition). recent studies using frequency tagging in eeg have sought to use pp to disentangle the top-down and bottom-up pathways for expectation and attention (gordon et al., 2019). there is increasing focus on metacognitive approaches to consciousness. there is a clear sense in which pp involves a certain kind of metacognition, namely in terms of higher-level expected precisions. that is, the pp system is constantly ‘looking down’ at its own representations seeking to learn how their precisions (and more generally their sufficient statistics) change over time. many questions are still open about how this general pp element can speak to the role of different kinds of metacognition in consciousness (for deployment of active inference to metacognition, see sandved smith et al., 2020). an outstanding challenge here will be to formally integrate the framework of signal detection theory, in which most theories of metacognition are articulated (e.g., barrett et al., 2013) with pp (see, e.g., fitzgerald et al., 2014; fleming, 2020). as we look briefly to more subtle aspects of conscious experience, pp is surprisingly resourceful. conscious perception combines both invariant and variant elements, and is mostly accompanied by a sense of presence of the perceived objects and properties; this may be accounted for by the way states of affairs are represented throughout the cortical hierarchy, which exhibits a gradient from variance to invariance (clark, 2015; felleman & essen, 1991; hohwy, 2013), which appears to underwrite perceptual stability in the visual cortex (podvalny et al., 2017). a further contributor to presence could be the ability, in some pp systems, for counterfactual prediction error minimization (seth, 2014) allowing representational access to occluded properties. since pp applies equally to body-related perceptual inference (e.g., inference on the causes of proprioceptive and interoceptive signals), it is able to address issues about bodily self-awareness (allen et al., 2019; seth, 2013). the role of action for self-representation is also being explored with pp; in particular, sensory attenuation (manifested, e.g., in our inability to tickle ourselves) in action has received a new explanation in terms of precision optimization (brown et al., 2013; van doorn et al., 2015; vasser et al., 2019). this overall approach to self-related inference may be extended to models of the philosophically recalcitrant topic of the self (apps & tsakiris, 2014; hohwy & michael, 2017; limanowski & blankenburg, 2013); two recent collections have several relevant pp contributions on this topic (milliere & metzinger, 2020; tsakiris & preester, 2018). indeed, ambitious applications of pp have begun to account for the subjective experiences of embodied selfhood as arising from active inference of interoceptive states, where such inference is motivated by the control or regulation of the hidden causes of these states rather than by their accurate representation (seth, 2015b, 2015a; wiese, 2015). these ideas call on early cybernetic formulations hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 24 of predictive regulation of essential variables, like those of ross ashby (ashby, 1954), which provide an alternative context for the free energy principle. it is worth sounding a note of caution about interpreting evidence for (especially) altered experiences of embodiment in terms of bayesian inference on bodyrelated sensory signals. for example, a recent study by lush and colleagues has revealed a link between subjective reports of illusory ownership in the well-known ‘rubber hand illusion’ (rhi) and hypnotic suggestion (lush et al., 2020). in this study, individual differences in the strength of the rhi correlated tightly with individual differences in hypnotic suggestibility (which the authors refer to as trait ‘phenomenological control’). this result raises the possibility that reported experiences of embodiment of the rhi may be due, at least in part, to implicit imaginative suggestions arising from the experimental context, putting pressure on accounts of the rhi that are based primarily on multisensory integration. however, the pp view offers a potential solution to this confound too, by considering suggestion effects as top-down expectations of a distinctive kind (see, for example, martin & pacherie, 2019). as a final example, the sense of the present as temporally extended may be explained by pp-like internal simulation (grush, 2006); the sense of temporal flow of conscious experience – how one moment passes seamlessly into the next even when nothing much happens in the environment – may be explained by the dynamics of hierarchical bayes that encode expected change in the environment (hohwy et al., 2016; sherman et al., 2020; vogel et al., 2020, 2019); this process may in turn relate to the alternation of percepts in binocular rivalry, as discussed above. in summary, then, pp is already able to cash out significant predictive and explanatory value for the study of consciousness. this supports the proposal that pp is an excellent candidate for providing the foundational framework for a systematic ncc. it also encourages the prospect that, eventually, aspects of pp may themselves coalesce into a theory of consciousness in its own right. 8 challenges and concluding remarks we have proposed that the science of consciousness can move toward a more explanatory and predictive conception of the ncc by capitalising on the predictive processing (pp) framework. pp holds unique promise for addressing challenges that a successful ncc strategy must overcome, and for linking candidate nccs to key theoretical constraints shared among several prominent theories of consciousness. we are not aware of any other theoretical framework with this kind of potential for moving toward a systematic ncc. several areas of theoretical and empirical research are already demonstrating the potential of pp as such a basis for a systematic, rather than arbitrary, ncc. full exploitation of this potential will be facilitated by focused efforts to draw on the pp framework when formulating and testing for candidate nccs. altogether, this approach will deliver significantly enhanced predictive and explanatory value hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 25 for the ncc program, as well as for the science of consciousness at large. in the end, this approach may also change the extent to which the hard problem of consciousness is still seen, by some, as a fundamental limit on a materialist science of consciousness. in this discussion we have focused on the promising explanatory and unificatory aspects of pp. however, pp, although increasingly pervasive, is certainly not universally accepted and there are conflicting findings and perspectives, driving ongoing debate (e.g., alilović et al., 2019; cao, 2020; li & ma, 2020; litwin & miłkowski, 2020; rahnev & denison, 2018). for example, one recurring challenge to pp is that worry that an imperative to minimize sensory prediction errors will lead to agents that do nothing at all – the so-called ‘dark room problem’ (friston et al., 2012; sun & firestone, 2020a, 2020b). however, this challenge has already been effectively addressed by extensions of pp that consider the minimization of expected future prediction error, rather than just prediction error in the here and now (hohwy, 2020b; seth et al., 2020; van de cruys et al., 2020). overall, pp as a framework continues to evolve, with new formal and modelling approaches appearing at a steady pace, some of which are yet to find its way into the broader debates about pp (friston, parr, et al., 2017; friston, da costa, et al., 2020). these new approaches will plausibly help generate robust experimental results, and could also enable closer links between neural substrates and conscious phenomenology. we believe the incorporation of active inference into consciousness science will be one major step in this revision, and that an increased focus on interoceptive inference will bring us closer to conscious phenomenology, especially in relation to affect, emotion, and self (seth, 2021). likewise, it will be important to gain clarity on the distinction between continuous and discrete predictive processing, both in order to reveal neural substrates and message passing and to begin exploring the phenomenology of thinking and (propositional) reasoning. a foundational challenge to pp is the question whether the brain is actually engaged in approximate bayesian inference, or whether this is just an attractive framework within which to interpret behavior and brain dynamics. this relates to the question whether organisms have explicit internal models, implementing pp processes, or whether their body plant in a more abstract sense embodies an implicit model. the answers to these questions will help determine how pp can underwrite the search for nccs, and in what sense it will be of explanatory and predictive value to do so. these answers might also afford an opportunity to begin addressing a related challenge, namely why and how pp should underpin consciousness at all, if it is so general that it literally underpins every organism’s existence – including those that are not conscious (e.g., viruses)? a first approach to this challenge would be to acknowledge that even if pp (in its free energy formulation) is indeed of maximal scope, the process theories under pp apply to particular kinds of systems and thereby allow for differences in the way pp is implemented. such differences could in principle speak to the distinction between conscious and non-conscious creatures. in addition to the distinction between explicit and imhohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 26 plicit generative models, a further useful distinction could be between creatures with and without a capacity for explicit counterfactual modelling of expected prediction error, or with and without temporally deep generative models (friston, 2017). such differences may potentially also account for differences in global conscious states in conscious creatures – such as the difference between anaesthesia and normal wakefulness in humans. it remains to be seen if developments in these respects will help address important clinical and ethical issues for individuals and creatures at the edges of consciousness, such as disorders of consciousness, infants, and non-verbal animals. it also remains to be seen if the accumulation of these pp-driven advances in consciousness science will eventually lead to meaningful inroads on the hard problem of consciousness, or onto the meta-problem of why many people are convinced that there is a hard problem of consciousness (chalmers, 2018). it is precisely because pp is not itself a theory of consciousness that it holds unique promise for shedding new light on these perennially intractable issues. acknowledgments aks is grateful to the dr. mortimer and theresa sackler foundation, which supports the sackler centre for consciousness, and to the cifar program on brain, mind, and consciousness. jh and aks are supported by the australian research council discovery grant dp160102770. references alais, d., & burr, d. (2004). the ventriloquist effect results from near-optimal bimodal integration. current biology, 14(3), 257–262. https://doi.org/10.1016/j.cub.2004.01.029 alilović, j., timmermans, b., reteig, l. c., van gaal, s., & slagter, h. a. (2019). no evidence that predictions and attention modulate the first feedforward sweep of cortical information processing. cerebral cortex, 29(5), 2261–2278. https: //doi.org/10.1093/cercor/bhz038 allen, m., levy, a., parr, t., & friston, k. j. (2019). in the body’s eye: the computational anatomy of interoceptive inference. biorxiv, 603928. https://doi.org/10.1101/603928 apps, m. a. j., & tsakiris, m. (2014). the free-energy self: a predictive coding account of self-recognition. neuroscience & biobehavioral reviews, 41, 85–97. https://doi.org/10.1016/j.neubiorev.2013.01.029 aru, j., bachmann, t., singer, w., & melloni, l. (2012). distilling the neural correlates of consciousness. neuroscience & biobehavioral reviews, 36(2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.12.003 aru, j., rutiku, r., wibral, m., singer, w., & melloni, l. (2016). early effects of previous experience on conscious perception. neuroscience of consciousness, 2016(1). https://doi.org/10.1093/nc/niw004 aru, j., suzuki, m., rutiku, r., larkum, m. e., & bachmann, t. (2019). coupling the state and contents of consciousness. frontiers in systems neuroscience, 13. https://doi.org/10.3389/fnsys.2019.00043 ashby, w. r. (1954). design for a brain. wiley. baars, b. j. (1988). a cognitive theory of consciousness. cambridge university press. bachmann, t. (2012). how to begin to overcome the ambiguity present in differentiation between contents and levels of consciousness? frontiers in psychology, 3. https://doi.org/10.3389/fpsyg.2012.00082 baltieri, m., buckley, c. l., & bruineberg, j. (2020). predictions in the eye of the beholder: an active inference account of watt governors. artificial life conference proceedings, 32, 121–129. https://doi.org/10.1162/isal_a_00288 barrett, a. b., dienes, z., & seth, a. k. (2013). measures of metacognition on signal-detection theoretic models. psychological methods, 18(4), 535–552. https://doi.org/10.1037/a0033268 hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.cub.2004.01.029 https://doi.org/10.1093/cercor/bhz038 https://doi.org/10.1093/cercor/bhz038 https://doi.org/10.1101/603928 https://doi.org/10.1016/j.neubiorev.2013.01.029 https://doi.org/10.1016/j.neubiorev.2011.12.003 https://doi.org/10.1093/nc/niw004 https://doi.org/10.3389/fnsys.2019.00043 https://doi.org/10.3389/fpsyg.2012.00082 https://doi.org/10.1162/isal_a_00288 https://doi.org/10.1037/a0033268 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 27 bastos, a. m., usrey, w. m., adams, r. a., mangun, g. r., fries, p., & friston, k. j. (2012). canonical microcircuits for predictive coding. neuron, 76(4), 695–711. https://doi.org/10.1016/j.neuron.2012.10.038 bayne, t. (2007). conscious states and conscious creatures: explanation in the scientific study of consciousness. philosophical perspectives, 21(1), 1–22. https://doi.org/10.1111/j.1520-8583.2007.00118.x bayne, t., cleeremans, a., & wilken, p. (eds.). (2009). oxford companion to consciousness. oxford university press. bayne, t., seth, a. k., & massimini, m. (2020). are there islands of awareness? trends in neurosciences, 43(1), 6–16. https: //doi.org/10.1016/j.tins.2019.11.003 bechtel, w. (2007). mental mechanisms: philosophical perspectives on cognitive neuroscience. taylor & francis. block, n. (1995). on a confusion about a function of consciousness. behavioral and brain sciences, 18(2), 227–247. https: //doi.org/10.1017/s0140525x00038188 bogacz, r. (2017). a tutorial on the free-energy framework for modelling perception and learning. journal of mathematical psychology, 76, 198–211. https://doi.org/10.1016/j.jmp.2015.11.003 boly, m., garrido, m. i., gosseries, o., bruno, m.-a., boveroux, p., schnakers, c., massimini, m., litvak, v., laureys, s., & friston, k. j. (2011). preserved feedforward but impaired top-down processes in the vegetative state. science, 332(6031), 858–862. https://doi.org/10.1126/science.1202043 boly, m., moran, r., murphy, m., boveroux, p., bruno, m.-a., noirhomme, q., ledoux, d., bonhomme, v., brichant, j.-f., tononi, g., laureys, s., & friston, k. j. (2012). connectivity changes underlying spectral eeg changes during propofolinduced loss of consciousness. journal of neuroscience, 32(20), 7082–7090. https://doi.org/10.1523/jneurosci.376911.2012 brown, h., adams, r. a., parees, i., edwards, m., & friston, k. j. (2013). active inference, sensory attenuation and illusions. cognitive processing, 14(4), 411–427. https://doi.org/10.1007/s10339-013-0571-3 brown, r., lau, h., & ledoux, j. e. (2019). understanding the higher-order approach to consciousness. trends in cognitive sciences, 23(9), 754–768. https://doi.org/10.1016/j.tics.2019.06.009 buckley, c. l., kim, c. s., mcgregor, s., & seth, a. k. (2017). the free energy principle for action and perception: a mathematical review. journal of mathematical psychology, 81, 55–79. https://doi.org/10.1016/j.jmp.2017.09.004 cao, r. (2020). new labels for old ideas: predictive processing and the interpretation of neural signals. review of philosophy and psychology, 11(3), 517–546. https://doi.org/10.1007/s13164-020-00481-x carhart-harris, r. l., leech, r., hellyer, p. j., shanahan, m., feilding, a., tagliazucchi, e., chialvo, d. r., & nutt, d. (2014). the entropic brain: a theory of conscious states informed by neuroimaging research with psychedelic drugs. frontiers in human neuroscience, 8. https://doi.org/10.3389/fnhum.2014.00020 chalmers, d. j. (2000). what is a neural correlate of consciousness? in t. metzinger (ed.), neural correlates of consciousness (pp. 17–39). mit press. chalmers, d. j. (2018). the meta-problem of consciousness. journal of consciousness studies, 25(9-10), 6–61. chalmers, d. j. (1996). the conscious mind: in search of a fundamental theory. oxford university press. chang, a. y. c., biehl, m., yu, y., & kanai, r. (2020). information closure theory of consciousness. frontiers in psychology, 11. https://doi.org/10.3389/fpsyg.2020.01504 clark, a. (2015). surfing uncertainty: prediction, action, and the embodied mind. oxford university press. clark, a. (2019). consciousness as generative entanglement. journal of philosophy, 116(12), 645–662. https://doi.org/10.584 0/jphil20191161241 clark, a., friston, k. j., & wilkinson, s. (2019). bayesing qualia: consciousness as inference, not raw datum. journal of consciousness studies, 26(9-10), 19–33. cleeremans, a. (2011). the radical plasticity thesis: how the brain learns to be conscious. frontiers in psychology, 2. https://doi.org/10.3389/fpsyg.2011.00086 cohen, m. a., & rubenstein, j. (2020). how much color do we see in the blink of an eye? cognition, 200, 104268. https: //doi.org/10.1016/j.cognition.2020.104268 cole, d. m., diaconescu, a. o., pfeiffer, u. j., brodersen, k. h., mathys, c. d., julkowski, d., ruhrmann, s., schilbach, l., tittgemeyer, m., vogeley, k., & stephan, k. e. (2020). atypical processing of uncertainty in individuals at risk for psychosis. neuroimage: clinical, 26, 102239. https://doi.org/10.1016/j.nicl.2020.102239 corcoran, a. w., & hohwy, j. (2018). allostasis, interoception, and the free energy principle: feeling our way forward. in m. tsakiris & h. de preester (eds.), the interoceptive basis of the mind (pp. 272–292). oxford university press. corlett, p. r., horga, g., fletcher, p. c., alderson-day, b., schmack, k., & powers, a. r. (2019). hallucinations and strong priors. trends in cognitive sciences, 23(2), 114–127. https://doi.org/10.1016/j.tics.2018.12.001 craver, c. f. (2007). explaining the brain: mechanisms and the mosaic unity of neuroscience. oxford university press. hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neuron.2012.10.038 https://doi.org/10.1111/j.1520-8583.2007.00118.x https://doi.org/10.1016/j.tins.2019.11.003 https://doi.org/10.1016/j.tins.2019.11.003 https://doi.org/10.1017/s0140525x00038188 https://doi.org/10.1017/s0140525x00038188 https://doi.org/10.1016/j.jmp.2015.11.003 https://doi.org/10.1126/science.1202043 https://doi.org/10.1523/jneurosci.3769-11.2012 https://doi.org/10.1523/jneurosci.3769-11.2012 https://doi.org/10.1007/s10339-013-0571-3 https://doi.org/10.1016/j.tics.2019.06.009 https://doi.org/10.1016/j.jmp.2017.09.004 https://doi.org/10.1007/s13164-020-00481-x https://doi.org/10.3389/fnhum.2014.00020 https://doi.org/10.3389/fpsyg.2020.01504 https://doi.org/10.5840/jphil20191161241 https://doi.org/10.5840/jphil20191161241 https://doi.org/10.3389/fpsyg.2011.00086 https://doi.org/10.1016/j.cognition.2020.104268 https://doi.org/10.1016/j.cognition.2020.104268 https://doi.org/10.1016/j.nicl.2020.102239 https://doi.org/10.1016/j.tics.2018.12.001 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 28 crick, f., & koch, c. (1990a). some reflections on visual awareness. sympos quant biol, 953–962. https://doi.org/10.1101/ sqb.1990.055.01.089 crick, f., & koch, c. (1990b). towards a neurobiological theory of consciousness. seminars in the neurosciences, 2, 263–275. damasio, a. (2000). the feeling of what happens: body and emotion in the making of consciousness. mariner books. de haan, e. h. f., corballis, p. m., hillyard, s. a., marzi, c. a., seth, a., lamme, v. a. f., volz, l., fabri, m., schechter, e., bayne, t., corballis, m., & pinto, y. (2020). split-brain: what we know now and why this is important for understanding consciousness. neuropsychology review, 30(2), 224–233. https://doi.org/10.1007/s11065-020-09439-3 dehaene, s. (2011). conscious and nonconscious processes: distinct forms of evidence accumulation? biological physics. progress in mathematical physics, 60. https://doi.org/10.1007/978-3-0346-0428-4_7 dennett, d. c. (1991). consciousness explained. little, brown & company. descartes, r. (1641). descartes: meditations on first philosophy: with selections from the objections and replies (j. cottingham, ed.). cambridge university press. dijkstra, n., ambrogioni, l., vidaurre, d., & gerven, m. van. (2020). neural dynamics of perceptual inference and its reversal during imagery. elife, 9, e53588. https://doi.org/10.7554/elife.53588 doerig, a., schurger, a., & herzog, m. h. (2020). hard criteria for empirical theories of consciousness. cognitive neuroscience, 1–22. https://doi.org/10.1080/17588928.2020.1772214 dołęga, k., & dewhurst, j. e. (2020). fame in the predictive brain: a deflationary approach to explaining consciousness in the prediction error minimization framework. synthese. https://doi.org/10.1007/s11229-020-02548-9 engel, a. k., friston, k. j., & kragic, d. (eds.). (2016). the pragmatic turn: toward action-oriented views in cognitive science. the mit press. engel, a. k., & singer, w. (2001). temporal binding and the neural correlates of sensory awareness. trends in cognitive sciences, 5(1), 16–25. https://doi.org/10.1016/s1364-6613(00)01568-0 feldman, h., & friston, k. j. (2010). attention, uncertainty, and free-energy. frontiers in human neuroscience, 4. https: //doi.org/10.3389/fnhum.2010.00215 felleman, d. j., & essen, d. c. v. (1991). distributed hierarchical processing in the primate cerebral cortex. cereb cortex, 1–47. https://doi.org/10.1093/cercor/1.1.1 fernández-espejo, d., & owen, a. m. (2013). detecting awareness after severe brain injury. nature reviews neuroscience, 14(11), 801–809. https://doi.org/10.1038/nrn3608 fink, s. b. (2016). a deeper look at the “neural correlate of consciousness”. frontiers in psychology, 7. https://doi.org/10.338 9/fpsyg.2016.01044 fitzgerald, t. h. b., schwartenbeck, p., moutoussis, m., dolan, r. j., & friston, k. j. (2014). active inference, evidence accumulation, and the urn task. neural computation, 27(2), 306–328. https://doi.org/10.1162/neco_a_00699 fleming, s. m. (2020). awareness as inference in a higher-order state space. neuroscience of consciousness, 2020(1). https: //doi.org/10.1093/nc/niz020 fleming, s. m., maniscalco, b., ko, y., amendi, n., ro, t., & lau, h. (2015). action-specific disruption of perceptual confidence: psychological science. https://doi.org/10.1177/0956797614557697 fletcher, p. c., & frith, c. d. (2009). perceiving is believing: a bayesian approach to explaining the positive symptoms of schizophrenia. nature reviews neuroscience, 10(1), 48–58. https://doi.org/10.1038/nrn2536 friston, k. j. (2017). the mathematics of mind-time. aeon. https://aeon.co/essays/consciousness-is-not-a-thing-but-aprocess-of-inference friston, k. j. (2018). am i self-conscious? (or does self-organization entail self-consciousness?). frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018.00579 friston, k. j. (2010). the free-energy principle: a unified brain theory? nature reviews neuroscience, 11(2), 127–138. https: //doi.org/10.1038/nrn2787 friston, k. j. (2005). a theory of cortical responses. philosophical transactions of the royal society b: biological sciences, 360(1456), 815–836. https://doi.org/10.1098/rstb.2005.1622 friston, k. j. (2019a). a free energy principle for a particular physics. arxiv. https://doi.org/arxiv:%201906.10184 friston, k. j. (2019b). waves of prediction. plos biology, 17(10), e3000426. https://doi.org/10.1371/journal.pbio.3000426 friston, k. j. (2003). learning and inference in the brain. neural networks, 16(9), 1325–1352. https://doi.org/10.1016/j.neun et.2003.06.005 friston, k. j., da costa, l., hafner, d., hesp, c., & parr, t. (2020). sophisticated inference. arxiv. https://doi.org/arxiv: %202006.04120 friston, k. j., fitzgerald, t., rigoli, f., schwartenbeck, p., o’doherty, j., & pezzulo, g. (2016). active inference and learning. neuroscience & biobehavioral reviews, 68, 862–879. https://doi.org/10.1016/j.neubiorev.2016.06.022 hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1101/sqb.1990.055.01.089 https://doi.org/10.1101/sqb.1990.055.01.089 https://doi.org/10.1007/s11065-020-09439-3 https://doi.org/10.1007/978-3-0346-0428-4_7 https://doi.org/10.7554/elife.53588 https://doi.org/10.1080/17588928.2020.1772214 https://doi.org/10.1007/s11229-020-02548-9 https://doi.org/10.1016/s1364-6613(00)01568-0 https://doi.org/10.3389/fnhum.2010.00215 https://doi.org/10.3389/fnhum.2010.00215 https://doi.org/10.1093/cercor/1.1.1 https://doi.org/10.1038/nrn3608 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.3389/fpsyg.2016.01044 https://doi.org/10.1162/neco_a_00699 https://doi.org/10.1093/nc/niz020 https://doi.org/10.1093/nc/niz020 https://doi.org/10.1177/0956797614557697 https://doi.org/10.1038/nrn2536 https://aeon.co/essays/consciousness-is-not-a-thing-but-a-process-of-inference https://aeon.co/essays/consciousness-is-not-a-thing-but-a-process-of-inference https://doi.org/10.3389/fpsyg.2018.00579 https://doi.org/10.1038/nrn2787 https://doi.org/10.1038/nrn2787 https://doi.org/10.1098/rstb.2005.1622 https://doi.org/arxiv:%201906.10184 https://doi.org/10.1371/journal.pbio.3000426 https://doi.org/10.1016/j.neunet.2003.06.005 https://doi.org/10.1016/j.neunet.2003.06.005 https://doi.org/arxiv:%202006.04120 https://doi.org/arxiv:%202006.04120 https://doi.org/10.1016/j.neubiorev.2016.06.022 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 29 friston, k. j., fitzgerald, t., rigoli, f., schwartenbeck, p., & pezzulo, g. (2016). active inference: a process theory. neural computation, 29(1), 1–49. https://doi.org/10.1162/neco_a_00912 friston, k. j., parr, t., & vries, b. de. (2017). the graphical brain: belief propagation and active inference. network neuroscience, 1(4), 381–414. https://doi.org/10.1162/netn_a_00018 friston, k. j., rigoli, f., ognibene, d., mathys, c., fitzgerald, t., & pezzulo, g. (2015). active inference and epistemic value. cognitive neuroscience, 6(4), 187–214. https://doi.org/10.1080/17588928.2015.1020053 friston, k. j., schwartenbeck, p., fitzgerald, t., moutoussis, m., behrens, t., & dolan, r. j. (2013). the anatomy of choice: active inference and agency. frontiers in human neuroscience, 7. https://doi.org/10.3389/fnhum.2013.00598 friston, k. j., thornton, c., & clark, a. (2012). free-energy minimization and the dark-room problem. frontiers in psychology, 3. https://doi.org/10.3389/fpsyg.2012.00130 friston, k. j., wiese, w., & hobson, j. a. (2020). sentience and the origins of consciousness: from cartesian duality to markovian monism. entropy, 22(5), 516. https://doi.org/10.3390/e22050516 gold, i. (1999). does 40-hz oscillation play a role in visual consciousness? consciousness and cognition, 8(2), 186–195. https://doi.org/10.1006/ccog.1999.0399 gordon, n., tsuchiya, n., koenig-robert, r., & hohwy, j. (2019). expectation and attention increase the integration of top-down and bottom-up signals in perception through different pathways. plos biology, 17(4), e3000233. https: //doi.org/10.1371/journal.pbio.3000233 graaf, t. a. de, hsieh, p.-j., & sack, a. t. (2012). the “correlates” in neural correlates of consciousness. neuroscience & biobehavioral reviews, 36(1), 191–197. https://doi.org/10.1016/j.neubiorev.2011.05.012 graziano, m. s. a., guterstam, a., bio, b. j., & wilterson, a. i. (2020). toward a standard model of consciousness: reconciling the attention schema, global workspace, higher-order thought, and illusionist theories. cognitive neuropsychology, 37(3-4), 155–172. https://doi.org/10.1080/02643294.2019.1670630 graziano, m. s. a., & webb, t. w. (2015). the attention schema theory: a mechanistic account of subjective awareness. frontiers in psychology, 6. https://doi.org/10.3389/fpsyg.2015.00500 grush, r. (2006). how to, and how not to, bridge computational cognitive neuroscience and husserlian phenomenology of time consciousness. synthese, 153(3), 417–450. https://doi.org/10.1007/s11229-006-9100-6 hameroff, s., & penrose, r. (2014). consciousness in the universe: a review of the “orch or” theory. physics of life reviews, 11(1), 39–78. https://doi.org/10.1016/j.plrev.2013.08.002 haun, a., & tononi, g. (2019). why does space feel the way it does? towards a principled account of spatial experience. entropy, 21(12), 1160. https://doi.org/10.3390/e21121160 heeger, d. j. (2017). theory of cortical function. proceedings of the national academy of sciences, 114(8), 1773–1782. https: //doi.org/10.1073/pnas.1619788114 heilbron, m., richter, d., ekman, m., hagoort, p., & lange, f. p. de. (2020). word contexts enhance the neural representation of individual letters in early visual cortex. nature communications, 11(1), 321. https://doi.org/10.1038/s41467-01913996-4 hobson, j. a., & friston, k. j. (2012). waking and dreaming consciousness: neurobiological and functional considerations. progress in neurobiology, 98(1), 82–98. https://doi.org/10.1016/j.pneurobio.2012.05.003 hohwy, j. (2012). attention and conscious perception in the hypothesis testing brain. frontiers in psychology, 3. https: //doi.org/10.3389/fpsyg.2012.00096 hohwy, j. (2013). the predictive mind. oxford university press. hohwy, j. (2015). prediction error minimization, mental and developmental disorder, and statistical theories of consciousness. in disturbed consciousness: new essays on psychopathology and theories of consciousness (pp. 293–324). mit press. hohwy, j. (2020a). new directions in predictive processing. mind & language, 35(2), 209–223. https://doi.org/10.1111/mila .12281 hohwy, j. (2020b). self-supervision, normativity and the free energy principle. synthese. https://doi.org/10.1007/s11229020-02622-2 hohwy, j. (2009). the neural correlates of consciousness: new experimental approaches needed? consciousness and cognition, 18(2), 428–438. https://doi.org/10.1016/j.concog.2009.02.006 hohwy, j., & frith, c. (2004). can neuroscience explain consciousness? journal of consciousness studies, 11(7-8), 180–198. hohwy, j., & michael, j. (2017). why does any body have a self? in f. de vignemont & a. j. t. alsmith (eds.), the body and the self, revisited (pp. 363–391). mit press. hohwy, j., paton, b., & palmer, c. (2016). distrusting the present. phenomenology and the cognitive sciences, 15(3), 315–335. https://doi.org/10.1007/s11097-015-9439-6 hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1162/neco_a_00912 https://doi.org/10.1162/netn_a_00018 https://doi.org/10.1080/17588928.2015.1020053 https://doi.org/10.3389/fnhum.2013.00598 https://doi.org/10.3389/fpsyg.2012.00130 https://doi.org/10.3390/e22050516 https://doi.org/10.1006/ccog.1999.0399 https://doi.org/10.1371/journal.pbio.3000233 https://doi.org/10.1371/journal.pbio.3000233 https://doi.org/10.1016/j.neubiorev.2011.05.012 https://doi.org/10.1080/02643294.2019.1670630 https://doi.org/10.3389/fpsyg.2015.00500 https://doi.org/10.1007/s11229-006-9100-6 https://doi.org/10.1016/j.plrev.2013.08.002 https://doi.org/10.3390/e21121160 https://doi.org/10.1073/pnas.1619788114 https://doi.org/10.1073/pnas.1619788114 https://doi.org/10.1038/s41467-019-13996-4 https://doi.org/10.1038/s41467-019-13996-4 https://doi.org/10.1016/j.pneurobio.2012.05.003 https://doi.org/10.3389/fpsyg.2012.00096 https://doi.org/10.3389/fpsyg.2012.00096 https://doi.org/10.1111/mila.12281 https://doi.org/10.1111/mila.12281 https://doi.org/10.1007/s11229-020-02622-2 https://doi.org/10.1007/s11229-020-02622-2 https://doi.org/10.1016/j.concog.2009.02.006 https://doi.org/10.1007/s11097-015-9439-6 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 30 hohwy, j., roepstorff, a., & friston, k. j. (2008). predictive coding explains binocular rivalry: an epistemological review. cognition, 108(3), 687–701. https://doi.org/10.1016/j.cognition.2008.05.010 hurley, s. l. (1998). consciousness in action. harvard university press. hutto, d. d., & myin, e. (2013). radicalizing enactivism: basic minds without content. cambridge, mass.: mit press. jackson, f. (1982). epiphenomenal qualia. the philosophical quarterly 32, 32(127), 127–136. https://doi.org/10.2307/2960077 kanai, r., chang, a., yu, y., magrans de abril, i., biehl, m., & guttenberg, n. (2019). information generation as a functional basis of consciousness. neuroscience of consciousness, 2019(1). https://doi.org/10.1093/nc/niz016 kiefer, a. b. (2017). literal perceptual inference. in t. k. metzinger & w. wiese (eds.), philosophy and predictive processing. mind group. klein, c., & barron, a. b. (2020). how experimental neuroscientists can fix the hard problem of consciousness. neuroscience of consciousness, 2020(1). https://doi.org/10.1093/nc/niaa009 koch, c., massimini, m., boly, m., & tononi, g. (2016). neural correlates of consciousness: progress and problems. nature reviews neuroscience, 17(5), 307–321. https://doi.org/10.1038/nrn.2016.22 koch, c., & tsuchiya, n. (2007). attention and consciousness: two distinct brain processes. trends in cognitive sciences, 11(1), 16–22. https://doi.org/10.1016/j.tics.2006.10.012 kok, p., rahnev, d., jehee, j. f. m., lau, h. c., & lange, f. p. de. (2012). attention reverses the effect of prediction in silencing sensory signals. cerebral cortex, 22(9), 2197–2206. https://doi.org/10.1093/cercor/bhr310 kok, p., rait, l. i., & turk-browne, n. b. (2019). content-based dissociation of hippocampal involvement in prediction. journal of cognitive neuroscience, 32(3), 527–545. https://doi.org/10.1162/jocn_a_01509 krakauer, j. w., ghazanfar, a. a., gomez-marin, a., maciver, m. a., & poeppel, d. (2017). neuroscience needs behavior: correcting a reductionist bias. neuron, 93(3), 480–490. https://doi.org/10.1016/j.neuron.2016.12.041 lamme, v. a. f. (2020). visual functions generating conscious seeing. frontiers in psychology, 11. https://doi.org/10.3389/fp syg.2020.00083 lamme, v. a. f. (2010). how neuroscience will change our view on consciousness. cognitive neuroscience, 1(3), 204–220. https://doi.org/10.1080/17588921003731586 lange, f. p. de, heilbron, m., & kok, p. (2018). how do expectations shape perception? trends in cognitive sciences, 22(9), 764–779. https://doi.org/10.1016/j.tics.2018.06.002 lau, h. c. (2008). a higher order bayesian decision theory of consciousness. elsevier, 168, 35–48. https://doi.org/10.1016/ s0079-6123(07)68004-2 lau, h., & rosenthal, d. (2011). empirical support for higher-order theories of conscious awareness. trends in cognitive sciences, 15(8), 365–373. https://doi.org/10.1016/j.tics.2011.05.009 laureys, s. (2005). the neural correlate of (un)awareness: lessons from the vegetative state. trends in cognitive sciences, 9(12), 556–559. https://doi.org/10.1016/j.tics.2005.10.010 lawson, r. p., mathys, c., & rees, g. (2017). adults with autism overestimate the volatility of the sensory environment. nature neuroscience, 20(9), 1293–1299. https://doi.org/10.1038/nn.4615 lettvin, j. y. (1976). on seeing sidelong. the sciences, 16(4), 10–20. https://doi.org/10.1002/j.2326-1951.1976.tb01231.x levine, j. (1983). materialism and qualia: the explanatory gap. pacific philosophical quarterly, 64(4), 354–361. https: //doi.org/10.1111/j.1468-0114.1983.tb00207.x li, h.-h., & ma, w. j. (2020). confidence reports in decision-making with multiple alternatives violate the bayesian confidence hypothesis. nature communications, 11(1), 2004. https://doi.org/10.1038/s41467-020-15581-6 limanowski, j., & blankenburg, f. (2013). minimal self-models and the free energy principle. frontiers in human neuroscience, 7. https://doi.org/10.3389/fnhum.2013.00547 litwin, p., & miłkowski, m. (2020). unification by fiat: arrested development of predictive processing. cognitive science, 44(7), e12867. https://doi.org/10.1111/cogs.12867 lush, p., botan, v., scott, r. b., seth, a. k., ward, j., & dienes, z. (2020). trait phenomenological control predicts experience of mirror synaesthesia and the rubber hand illusion. nature communications, 11(1), 4853. https://doi.org/10.1038/s4 1467-020-18591-6 marchi, f., & hohwy, j. (2020). the intermediate scope of consciousness in the predictive mind. erkenntnis. https://doi.or g/10.1007/s10670-020-00222-7 markov, n. t., ercsey-ravasz, m., essen, d. c. v., knoblauch, k., toroczkai, z., & kennedy, h. (2013). cortical high-density counterstream architectures. science, 342(6158). https://doi.org/10.1126/science.1238406 marshel, j. h., kim, y. s., machado, t. a., quirin, s., benson, b., kadmon, j., raja, c., chibukhchyan, a., ramakrishnan, c., inoue, m., shane, j. c., mcknight, d. j., yoshizawa, s., kato, h. e., ganguli, s., & deisseroth, k. (2019). cortical layer–specific critical dynamics triggering perception. science, 365(6453). https://doi.org/10.1126/science.aaw5202 hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.cognition.2008.05.010 https://doi.org/10.2307/2960077 https://doi.org/10.1093/nc/niz016 https://doi.org/10.1093/nc/niaa009 https://doi.org/10.1038/nrn.2016.22 https://doi.org/10.1016/j.tics.2006.10.012 https://doi.org/10.1093/cercor/bhr310 https://doi.org/10.1162/jocn_a_01509 https://doi.org/10.1016/j.neuron.2016.12.041 https://doi.org/10.3389/fpsyg.2020.00083 https://doi.org/10.3389/fpsyg.2020.00083 https://doi.org/10.1080/17588921003731586 https://doi.org/10.1016/j.tics.2018.06.002 https://doi.org/10.1016/s0079-6123(07)68004-2 https://doi.org/10.1016/s0079-6123(07)68004-2 https://doi.org/10.1016/j.tics.2011.05.009 https://doi.org/10.1016/j.tics.2005.10.010 https://doi.org/10.1038/nn.4615 https://doi.org/10.1002/j.2326-1951.1976.tb01231.x https://doi.org/10.1111/j.1468-0114.1983.tb00207.x https://doi.org/10.1111/j.1468-0114.1983.tb00207.x https://doi.org/10.1038/s41467-020-15581-6 https://doi.org/10.3389/fnhum.2013.00547 https://doi.org/10.1111/cogs.12867 https://doi.org/10.1038/s41467-020-18591-6 https://doi.org/10.1038/s41467-020-18591-6 https://doi.org/10.1007/s10670-020-00222-7 https://doi.org/10.1007/s10670-020-00222-7 https://doi.org/10.1126/science.1238406 https://doi.org/10.1126/science.aaw5202 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 31 martin, j.-r., & pacherie, e. (2019). alterations of agency in hypnosis: a new predictive coding model. psychological review, 126(1), 133–152. https://doi.org/10.1037/rev0000134 mashour, g. a., roelfsema, p., changeux, j.-p., & dehaene, s. (2020). conscious processing and the global neuronal workspace hypothesis. neuron, 105(5), 776–798. https://doi.org/10.1016/j.neuron.2020.01.026 mathys, c., daunizeau, j., friston, k. j., & stephan, k. e. (2011). a bayesian foundation for individual learning under uncertainty. frontiers in human neuroscience, 5. https://doi.org/10.3389/fnhum.2011.00039 mathys, c. d., lomakina, e. i., daunizeau, j., iglesias, s., brodersen, k. h., friston, k. j., & stephan, k. e. (2014). uncertainty in perception and the hierarchical gaussian filter. frontiers in human neuroscience, 8. https://doi.org/10.3389/fnhum. 2014.00825 mcfadden, j. (2020). integrating information in the brain’s em field: the cemi field theory of consciousness. neuroscience of consciousness, 2020(1). https://doi.org/10.1093/nc/niaa016 mcfadden, j. (2002). the conscious electromagnetic information (cemi) field theory: the hard problem made easy? journal of consciousness studies, 9(8), 45–60. meijs, e. l., slagter, h. a., lange, f. p. de, & gaal, s. van. (2018). dynamic interactions between top–down expectations and conscious awareness. journal of neuroscience, 38(9), 2318–2327. https://doi.org/10.1523/jneurosci.1952-17.2017 melloni, l., schwiedrzik, c. m., müller, n., rodriguez, e., & singer, w. (2011). expectations change the signatures and timing of electrophysiological correlates of perceptual awareness. journal of neuroscience, 31(4), 1386–1396. https: //doi.org/10.1523/jneurosci.4570-10.2011 mendonça, d., curado, m., & gouveia, s. (2020). the philosophy and science of predictive processing. bloomsbury publishing. metzinger, t. (2004). being no one. mit press. metzinger, t. (ed.). (2000). neural correlates of consciousness: empirical and conceptual questions. mit press. metzinger, t., & wiese, w. (eds.). (2017). philosophy and predictive processing. mind group. https://predictive-mind.net/p apers miller, s. m. (2015). the constitution of phenomenal consciousness. john benjamins. miller, s. m. (2007). on the correlation/constitution distinction problem (and other hard problems) in the scientific study of consciousness. acta neuropsychiatrica, 19(3), 159–176. https://doi.org/10.1111/j.1601-5215.2007.00207.x millidge, b., tschantz, a., & buckley, c. l. (2020). whence the expected free energy? https://doi.org/arxiv:2004.08128v3 milliere, r., & metzinger, t. (2020). radical disruptions of self-consciousness: philosophy and the mind sciences, 1(i), 1–13. https://doi.org/10.33735/phimisci.2020.i.50 muckli, l., de martino, f., vizioli, l., petro, l. s., smith, f. w., ugurbil, k., goebel, r., & yacoub, e. (2015). contextual feedback to superficial layers of v1. current biology, 25(20), 2690–2695. https://doi.org/10.1016/j.cub.2015.08.057 neisser, j. (2012). neural correlates of consciousness reconsidered. consciousness and cognition, 21(2), 681–690. https: //doi.org/10.1016/j.concog.2011.03.012 noë, a. (2004). action in perception. mit press. noë, a., & o’regan, j. k. (2001). a sensorimotor account of vision and visual consciousness. behavioral and brain sciences, 24(5), 939–973. noë, a., & thompson, e. (2004). are there neural correlates of consciousness? journal of consciousness studies, 11(1), 3–28. noreika, v., canales-johnson, a., harrison, w. j., johnson, a., arnatkevičiūtė, a., koh, j., chennu, s., & bekinschtein, t. a. (2017). wakefulness state modulates conscious access: suppression of auditory detection in the transition to sleep. biorxiv, 155705. https://doi.org/10.1101/155705 o’brien, g., & opie, j. (1999). a connectionist theory of phenomenal experience. behavioral and brain sciences, 22, 127–148. oizumi, m., albantakis, l., & tononi, g. (2014). from the phenomenology to the mechanisms of consciousness: integrated information theory 3.0. plos computational biology, 10(5), e1003588. https://doi.org/10.1371/journal.pcbi.1003588 pacherie, e. (2008). the phenomenology of action: a conceptual framework. cognition, 107(1), 179–217. https://doi.org/10 .1016/j.cognition.2007.09.003 pal, d., li, d., dean, j. g., brito, m. a., liu, t., fryzel, a. m., hudetz, a. g., & mashour, g. a. (2020). level of consciousness is dissociable from electroencephalographic measures of cortical connectivity, slow oscillations, and complexity. journal of neuroscience, 40(3), 605–618. https://doi.org/10.1523/jneurosci.1910-19.2019 parr, t., corcoran, a. w., friston, k. j., & hohwy, j. (2019). perceptual awareness and active inference. neuroscience of consciousness, 2019(1). https://doi.org/10.1093/nc/niz012 parr, t., & friston, k. j. (2018). the anatomy of inference: generative models and brain structure. frontiers in computational neuroscience, 12. https://doi.org/10.3389/fncom.2018.00090 pascual-leone, a., & walsh, v. (2001). fast backprojections from the motion to the primary visual area necessary for visual awareness. science, 292(5516), 510–512. https://doi.org/10.1126/science.1057099 hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1037/rev0000134 https://doi.org/10.1016/j.neuron.2020.01.026 https://doi.org/10.3389/fnhum.2011.00039 https://doi.org/10.3389/fnhum.2014.00825 https://doi.org/10.3389/fnhum.2014.00825 https://doi.org/10.1093/nc/niaa016 https://doi.org/10.1523/jneurosci.1952-17.2017 https://doi.org/10.1523/jneurosci.4570-10.2011 https://doi.org/10.1523/jneurosci.4570-10.2011 https://predictive-mind.net/papers https://predictive-mind.net/papers https://doi.org/10.1111/j.1601-5215.2007.00207.x https://doi.org/arxiv:2004.08128v3 https://doi.org/10.33735/phimisci.2020.i.50 https://doi.org/10.1016/j.cub.2015.08.057 https://doi.org/10.1016/j.concog.2011.03.012 https://doi.org/10.1016/j.concog.2011.03.012 https://doi.org/10.1101/155705 https://doi.org/10.1371/journal.pcbi.1003588 https://doi.org/10.1016/j.cognition.2007.09.003 https://doi.org/10.1016/j.cognition.2007.09.003 https://doi.org/10.1523/jneurosci.1910-19.2019 https://doi.org/10.1093/nc/niz012 https://doi.org/10.3389/fncom.2018.00090 https://doi.org/10.1126/science.1057099 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 32 pearl, j. (2000). causality: models, reasoning and inference. cambridge university press. perrykkad, k., & hohwy, j. (2020). fidgeting as self-evidencing: a predictive processing account of non-goal-directed action. new ideas in psychology, 56, 100750. https://doi.org/10.1016/j.newideapsych.2019.100750 petzschner, f. h., weber, l. a., wellstein, k. v., paolini, g., do, c. t., & stephan, k. e. (2019). focus of attention modulates the heartbeat evoked potential. neuroimage, 186, 595–606. https://doi.org/10.1016/j.neuroimage.2018.11.037 pinto, y., gaal, s. van, lange, f. p. de, lamme, v. a. f., & seth, a. k. (2015). expectations accelerate entry of visual stimuli into awareness. journal of vision, 15(8), 13–13. https://doi.org/10.1167/15.8.13 podvalny, e., yeagle, e., mégevand, p., sarid, n., harel, m., chechik, g., mehta, a. d., & malach, r. (2017). invariant temporal dynamics underlie perceptual stability in human visual cortex. current biology, 27(2), 155–165. https: //doi.org/10.1016/j.cub.2016.11.024 powers, a. r., mathys, c., & corlett, p. r. (2017). pavlovian conditioning–induced hallucinations result from overweighting of perceptual priors. science, 357(6351), 596–600. https://doi.org/10.1126/science.aan3458 prinz, j. j. (2012). the conscious brain: how attention engenders experience. oxford university press. rahnev, d., & denison, r. n. (2018). behavior is sensible but not globally optimal: seeking common ground in the optimality debate. behavioral and brain sciences, 41. https://doi.org/10.1017/s0140525x18002121 reardon, s. (2019). “outlandish” competition seeks the brain’s source of consciousness. science. https://www.sciencemag .org/news/2019/10/outlandish-competition-seeks-brain-s-source-consciousness revonsuo, a. (2006). inner presence: consciousness as a biological phenomenon. mit press. rosenthal, d. m. (1986). two concepts of consciousness. philosophical studies, 49(3), 329–359. rosenthal, d. m. (1997). a theory of consciousness. in n. block, o. j. flanagan, & g. guzeldere (eds.), the nature of consciousness (pp. 729–753). mit press. rudrauf, d., bennequin, d., granic, i., landini, g., friston, k. j., & williford, k. (2017). a mathematical model of embodied consciousness. journal of theoretical biology, 428, 106–131. https://doi.org/10.1016/j.jtbi.2017.05.032 sandved smith, l., hesp, c., lutz, a., mattout, j., friston, k. j., & ramstead, m. (2020). towards a formal neurophenomenology of metacognition: modelling meta-awareness, mental action, and attentional control with deep active inference. in psyarxiv. https://doi.org/10.31234/osf.io/5jh3c schartner, m. m., carhart-harris, r. l., barrett, a. b., seth, a. k., & muthukumaraswamy, s. d. (2017). increased spontaneous meg signal diversity for psychoactive doses of ketamine, lsd and psilocybin. scientific reports, 7(1), 46421. https: //doi.org/10.1038/srep46421 schroeder, c. e., wilson, d. a., radman, t., scharfman, h., & lakatos, p. (2010). dynamics of active sensing and perceptual selection. current opinion in neurobiology, 20(2), 172–176. https://doi.org/10.1016/j.conb.2010.02.010 searle, j. r. (2000). consciousness. annual review of neuroscience, 23(1), 557–578. https://doi.org/10.1146/annurev.neuro. 23.1.557 seth, a. (2009). explanatory correlates of consciousness: theoretical and computational challenges. cognitive computation, 1(1), 50–63. https://doi.org/10.1007/s12559-009-9007-x seth, a. k. (2016). the real problem. aeon. https://aeon.co/essays/the-hard-problem-of-consciousness-is-a-distractionfrom-the-real-one seth, a. k. (2021). being you. faber & faber. seth, a. k. (2015a). inference to the best prediction. in t. k. metzinger & j. m. windt (eds.), open mind. mind group. seth, a. k. (2015b). the cybernetic bayesian brain: from interoceptive inference to sensorimotor contingencies. in t. k. metzinger & j. m. windt (eds.), open mind. mind group. seth, a. k. (2014). a predictive processing theory of sensorimotor contingencies: explaining the puzzle of perceptual presence and its absence in synesthesia. cognitive neuroscience, 5(2), 97–118. https://doi.org/10.1080/17588928.2013. 877880 seth, a. k. (2019). from unconscious inference to the beholder’s share: predictive perception and human experience. european review, 27(3), 378–410. https://doi.org/10.1017/s1062798719000061 seth, a. k. (2013). interoceptive inference, emotion, and the embodied self. trends in cognitive sciences, 17(11), 565–573. https://doi.org/10.1016/j.tics.2013.09.007 seth, a. k., millidge, b., buckley, c. l., & tschantz, a. (2020). curious inferences: reply to sun and firestone on the dark room problem. trends in cognitive sciences, 24(9), 681–683. https://doi.org/10.1016/j.tics.2020.05.011 seth, a. k., & tsakiris, m. (2018). being a beast machine: the somatic basis of selfhood. trends in cognitive sciences, 22(11), 969–981. https://doi.org/10.1016/j.tics.2018.08.008 sherman, m. t., fountas, z., seth, a. k., & roseboom, w. (2020). accumulation of salient perceptual events predicts subjective time. biorxiv, 2020.01.09.900423. https://doi.org/10.1101/2020.01.09.900423 hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.newideapsych.2019.100750 https://doi.org/10.1016/j.neuroimage.2018.11.037 https://doi.org/10.1167/15.8.13 https://doi.org/10.1016/j.cub.2016.11.024 https://doi.org/10.1016/j.cub.2016.11.024 https://doi.org/10.1126/science.aan3458 https://doi.org/10.1017/s0140525x18002121 https://www.sciencemag.org/news/2019/10/outlandish-competition-seeks-brain-s-source-consciousness https://www.sciencemag.org/news/2019/10/outlandish-competition-seeks-brain-s-source-consciousness https://doi.org/10.1016/j.jtbi.2017.05.032 https://doi.org/10.31234/osf.io/5jh3c https://doi.org/10.1038/srep46421 https://doi.org/10.1038/srep46421 https://doi.org/10.1016/j.conb.2010.02.010 https://doi.org/10.1146/annurev.neuro.23.1.557 https://doi.org/10.1146/annurev.neuro.23.1.557 https://doi.org/10.1007/s12559-009-9007-x https://aeon.co/essays/the-hard-problem-of-consciousness-is-a-distraction-from-the-real-one https://aeon.co/essays/the-hard-problem-of-consciousness-is-a-distraction-from-the-real-one https://doi.org/10.1080/17588928.2013.877880 https://doi.org/10.1080/17588928.2013.877880 https://doi.org/10.1017/s1062798719000061 https://doi.org/10.1016/j.tics.2013.09.007 https://doi.org/10.1016/j.tics.2020.05.011 https://doi.org/10.1016/j.tics.2018.08.008 https://doi.org/10.1101/2020.01.09.900423 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org predictive processing and the neural correlates of consciousness 33 sherman, m. t., seth, a. k., barrett, a. b., & kanai, r. (2015). prior expectations facilitate metacognition for perceptual decision. consciousness and cognition, 35, 53–65. https://doi.org/10.1016/j.concog.2015.04.015 skewes, j. c., jegindø, e.-m., & gebauer, l. (2015). perceptual inference and autistic traits. autism, 19(11). https://doi.org/ 10.1177/1362361313519872 skora, l., seth, a., & scott, r. b. (2020). sensorimotor predictions shape reported conscious visual experience in a breaking continuous flash suppression task. in psyarxiv. https://doi.org/10.31234/osf.io/hwkdj stephan, k. e., manjaly, z. m., mathys, c. d., weber, l. a. e., paliwal, s., gard, t., tittgemeyer, m., fleming, s. m., haker, h., seth, a. k., & petzschner, f. h. (2016). allostatic self-efficacy: a metacognitive theory of dyshomeostasis-induced fatigue and depression. frontiers in human neuroscience, 10. https://doi.org/10.3389/fnhum.2016.00550 stuke, h., weilnhammer, v. a., sterzer, p., & schmack, k. (2019). delusion proneness is linked to a reduced usage of prior beliefs in perceptual decisions. schizophrenia bulletin, 45(1), 80–86. https://doi.org/10.1093/schbul/sbx189 summerfield, c., & egner, t. (2009). expectation (and attention) in visual cognition. trends in cognitive sciences, 13(9), 403–409. https://doi.org/10.1016/j.tics.2009.06.003 sun, z., & firestone, c. (2020a). the dark room problem. trends in cognitive sciences, 24(5), 346–348. https://doi.org/10.101 6/j.tics.2020.02.006 sun, z., & firestone, c. (2020b). optimism and pessimism in the predictive brain. trends in cognitive sciences, 24(9), 683–685. https://doi.org/10.1016/j.tics.2020.06.001 suzuki, k., schwartzman, d. j., augusto, r., & seth, a. k. (2019). sensorimotor contingency modulates breakthrough of virtual 3d objects during a breaking continuous flash suppression paradigm. cognition, 187, 95–107. https://doi.org/ 10.1016/j.cognition.2019.03.003 teufel, c., & fletcher, p. c. (2020). forms of prediction in the nervous system. nature reviews neuroscience, 21(4), 231–242. https://doi.org/10.1038/s41583-020-0275-5 tong, f., nakayama, k., vaughan, j. t., & kanwisher, n. (1998). binocular rivalry and visual awareness in human extrastriate cortex. neuron, 21(4), 753–759. https://doi.org/10.1016/s0896-6273(00)80592-9 tononi, g., boly, m., massimini, m., & koch, c. (2016). integrated information theory: from consciousness to its physical substrate. nature reviews neuroscience, 17(7), 450–461. https://doi.org/10.1038/nrn.2016.44 tononi, g., & edelman, g. m. (1998). consciousness and complexity. science, 282(5395), 1846–1851. https://doi.org/10.112 6/science.282.5395.1846 tononi, g., & koch, c. (2008). the neural correlates of consciousness. annals of the new york academy of sciences, 1124(1), 239–261. https://doi.org/10.1196/annals.1440.004 tsakiris, m., & preester, h. d. (2018). the interoceptive mind: from homeostasis to awareness. oxford university press. tschantz, a., millidge, b., seth, a. k., & buckley, c. l. (2020). reinforcement learning through active inference. https: //arxiv.org/abs/2002.12636 tschantz, a., seth, a. k., & buckley, c. l. (2020). learning action-oriented models through active inference. plos computational biology, 16(4), e1007805. https://doi.org/10.1371/journal.pcbi.1007805 tsuchiya, n., wilke, m., frässle, s., & lamme, v. a. f. (2015). no-report paradigms: extracting the true neural correlates of consciousness. trends in cognitive sciences, 19(12), 757–770. https://doi.org/10.1016/j.tics.2015.10.002 van de cruys, s., friston, k. j., & clark, a. (2020). controlled optimism: reply to sun and firestone on the dark room problem. trends in cognitive sciences, 24(9), 680–681. https://doi.org/10.1016/j.tics.2020.05.012 van der helm, p. a. (2016). structural coding versus free-energy predictive coding. psychonomic bulletin & review, 23(3), 663–677. https://doi.org/10.3758/s13423-015-0938-9 van doorn, g., paton, b., howell, j., & hohwy, j. (2015). attenuated self-tickle sensation even under trajectory perturbation. consciousness and cognition, 36, 147–153. https://doi.org/10.1016/j.concog.2015.06.016 van es, t. (2020). living models or life modelled? on the use of models in the free energy principle. adaptive behavior. https://doi.org/10.1177/1059712320918678 van gelder, t. (1995). what might cognition be, if not computation? the journal of philosophy, 92(7), 345–381. https: //doi.org/10.2307/2941061 vasser, m., vuillaume, l., cleeremans, a., & aru, j. (2019). waving goodbye to contrast: self-generated hand movements attenuate visual sensitivity. neuroscience of consciousness, 2019(1). https://doi.org/10.1093/nc/niy013 vogel, d. h. v., beeker, t., haidl, t., kupke, c., heinze, m., & vogeley, k. (2019). disturbed time experience during and after psychosis. schizophrenia research: cognition, 17, 100136. https://doi.org/10.1016/j.scog.2019.100136 vogel, d. h. v., falter-wagner, c. m., schoofs, t., krämer, k., kupke, c., & vogeley, k. (2020). flow and structure of time experience – concept, empirical validation and implications for psychopathology. phenomenology and the cognitive sciences, 19(2), 235–258. https://doi.org/10.1007/s11097-018-9573-z hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.concog.2015.04.015 https://doi.org/10.1177/1362361313519872 https://doi.org/10.1177/1362361313519872 https://doi.org/10.31234/osf.io/hwkdj https://doi.org/10.3389/fnhum.2016.00550 https://doi.org/10.1093/schbul/sbx189 https://doi.org/10.1016/j.tics.2009.06.003 https://doi.org/10.1016/j.tics.2020.02.006 https://doi.org/10.1016/j.tics.2020.02.006 https://doi.org/10.1016/j.tics.2020.06.001 https://doi.org/10.1016/j.cognition.2019.03.003 https://doi.org/10.1016/j.cognition.2019.03.003 https://doi.org/10.1038/s41583-020-0275-5 https://doi.org/10.1016/s0896-6273(00)80592-9 https://doi.org/10.1038/nrn.2016.44 https://doi.org/10.1126/science.282.5395.1846 https://doi.org/10.1126/science.282.5395.1846 https://doi.org/10.1196/annals.1440.004 https://arxiv.org/abs/2002.12636 https://arxiv.org/abs/2002.12636 https://doi.org/10.1371/journal.pcbi.1007805 https://doi.org/10.1016/j.tics.2015.10.002 https://doi.org/10.1016/j.tics.2020.05.012 https://doi.org/10.3758/s13423-015-0938-9 https://doi.org/10.1016/j.concog.2015.06.016 https://doi.org/10.1177/1059712320918678 https://doi.org/10.2307/2941061 https://doi.org/10.2307/2941061 https://doi.org/10.1093/nc/niy013 https://doi.org/10.1016/j.scog.2019.100136 https://doi.org/10.1007/s11097-018-9573-z https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org jakob hohwy and anil seth 34 walsh, k. s., mcgovern, d. p., clark, a., & o’connell, r. g. (2020). evaluating the neurophysiological evidence for predictive processing as a model of perception. annals of the new york academy of sciences, 1464(1), 242–268. https://doi.org/10 .1111/nyas.14321 weilnhammer, v., fritsch, m., chikermane, m., eckert, a.-l., kanthak, k., stuke, h., kaminski, j., & sterzer, p. (2020). evidence for an active role of inferior frontal cortex in conscious experience. biorxiv. https://doi.org/10.1101/2020.0 5.28.114645 weilnhammer, v., stuke, h., hesselmann, g., sterzer, p., & schmack, k. (2017). a predictive coding account of bistable perception – a model-based fmri study. plos computational biology, 13(5), e1005536. https://doi.org/10.1371/journa l.pcbi.1005536 whyte, c. j. (2019). integrating the global neuronal workspace into the framework of predictive processing: towards a working hypothesis. consciousness and cognition, 73, 102763. https://doi.org/10.1016/j.concog.2019.102763 whyte, c. j., & smith, r. (2020). the predictive global neuronal workspace: a formal active inference model of visual consciousness. biorxiv. https://doi.org/10.1101/2020.02.11.944611 wiese, w. (2018). toward a mature science of consciousness. frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018 .00693 wiese, w. (2015). perceptual presence in the kuhnian-popperian bayesian brain. in t. k. metzinger & j. m. windt (eds.), open mind. mind group. wiese, w. (2020). the science of consciousness does not need another theory, it needs a minimal unifying model. neuroscience of consciousness, 2020(1). https://doi.org/10.1093/nc/niaa013 williford, k., bennequin, d., friston, k. j., & rudrauf, d. (2018). the projective consciousness model and phenomenal selfhood. frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018.02571 woodward, j. (2003). making things happen. oxford university press. yon, d., gilbert, s. j., lange, f. p. de, & press, c. (2018). action sharpens sensory representations of expected outcomes. nature communications, 9(1), 4288. https://doi.org/10.1038/s41467-018-06752-7 yon, d., lange, f. p. de, & press, c. (2019). the predictive brain as a stubborn scientist. trends in cognitive sciences, 23(1), 6–8. https://doi.org/10.1016/j.tics.2018.10.003 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. hohwy, j., & seth, a. (2020). predictive processing as a systematic basis for identifying the neural correlates of consciousness. philosophy and the mind sciences, 1(ii), 3. https://doi.org/10.33735/phimisci.2020.ii.64 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1111/nyas.14321 https://doi.org/10.1111/nyas.14321 https://doi.org/10.1101/2020.05.28.114645 https://doi.org/10.1101/2020.05.28.114645 https://doi.org/10.1371/journal.pcbi.1005536 https://doi.org/10.1371/journal.pcbi.1005536 https://doi.org/10.1016/j.concog.2019.102763 https://doi.org/10.1101/2020.02.11.944611 https://doi.org/10.3389/fpsyg.2018.00693 https://doi.org/10.3389/fpsyg.2018.00693 https://doi.org/10.1093/nc/niaa013 https://doi.org/10.3389/fpsyg.2018.02571 https://doi.org/10.1038/s41467-018-06752-7 https://doi.org/10.1016/j.tics.2018.10.003 https://doi.org/10.33735/phimisci.2020.ii.64 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction the need for a systematic ncc three challenges to the ncc search the global states and local contents challenge the screening off challenge: brain structure and phenomenology the action challenge theories of consciousness common themes core properties of pp pp as a systematic ncc predictive and explanatory power of pp as a systematic basis for ncc challenges and concluding remarks the varieties of selflessness the varieties of selflessness raphaël millièrea (raphael.milliere@philosophy.ox.ac.uk) abstract many authors argue that conscious experience involves a sense of self or self-consciousness. according to the strongest version of this claim, there can be no selfless states of consciousness, namely states of consciousness that lack self-consciousness altogether. disagreements about this claim are likely to remain merely verbal as long as the target notion of self-consciousness is not adequately specified. after distinguishing six notions of self-consciousness commonly discussed in the literature, i argue that none of the corresponding features is necessary for consciousness, because there are states of consciousness in which each of them is plausibly missing. such states can be said to be at least partially selfless, since they lack at least one of the ways in which one could be self-conscious. furthermore, i argue that there is also preliminary empirical evidence that some states of consciousness lack all of these six putative forms of self-consciousness. such states might be totally selfless, insofar as they lack all the ways in which one could be self-conscious. i conclude by addressing four objections to the possibility and reportability of totally selfless states of consciousness. keywords dreamless sleep ∙ drug-induced states ∙ global states of consciousness ∙ meditation ∙ psychopathology ∙ self-consciousness ∙ selflessness this article is part of a special issue on “radical disruptions of self-consciousness”, edited by thomas metzinger and raphaël millière. 1 introduction many authors in philosophy, psychology and neuroscience are sympathetic to the claim that conscious experience involves a sense of self. in first approximation, a subject’s overall experience involves a sense of self at t if and only if (a) the subject is self-conscious at t, and (b) the subject’s being self-conscious at t makes a constitutive contribution to her overall phenomenology at t.1 auniversity of oxford 1while the clause (b) may seem somewhat redundant, it is intended to highlight the fact that the notion of “sense of self” or “self-consciousness” under consideration is a phenomenal one – rather millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0001-6965-6073 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 2 the claim that conscious experience involves a sense of self comes in two varieties, depending on the alleged prevalence of the target phenomenon. according to the necessity claim (nc), self-consciousness is necessary for consciousness in general.2 according to the typicality claim (tc), self-consciousness is merely pervasive in ordinary experience.3 these two claims can be formulated more precisely as follows: nc for any subject s, if s is conscious at t, then (a) s is self-conscious at t, and (b) s’s being self-conscious at t makes a constitutive contribution to s’s overall phenomenology at t. tc for any healthy, neurotypical, adult human subject s, in any ordinary condition in which s is conscious, (a) s is self-conscious, and (b) s’s being selfconscious in that condition makes a constitutive contribution to s’s overall phenomenology.4 in this article, i focus mostly on the strongest of these claims: nc. proponents of nc deny that there can be selfless states of consciousness, namely states of consciousness that lack self-consciousness entirely. in section 2, i suggest that disagreements about the prevalence of self-consciousness in conscious experience are difficult to arbitrate unless the target notion of self-consciousness is adequately specified. accordingly, i distinguish six notions of self-consciousness commonly than a dispositional notion such as the capacity to use a concept of self. the inclusion of the qualifier “constitutive” is intended to highlight the fact that the relevant contribution to the subject’s phenomenology is not merely causal, since a non-conscious mental state could indirectly cause a change in a subject’s phenomenology (e.g. a non-conscious desire for ice cream could cause a subject to feel hungry, thereby indirectly causing a change in the subject’s phenomenology). 2for example: “if ‘self-consciousness’ is taken to mean ‘consciousness with a sense of self’, then all human consciousness is necessarily covered by the term – there is just no other kind of consciousness as far as i can see” (damasio, 1999, p. 19); “if a mental state is conscious, that mental state will involve a non-reflective self-awareness” (gallagher, 2010, p. 567); “self-consciousness is a constitutive feature of phenomenal consciousness” (zahavi, 2014, p. 17); “[s]elf-awareness is an integral function of all conscious experiences” (lou, changeux, & rosenstand, 2017, p. 767); “i believe that the self always shows up in every human experience” (duncan, 2019); “is self-consciousness necessary for consciousness? the answer is yes.” (friston, 2018, p. 1). 3for example: “[phenomenally] conscious states often seem to have a ‘me-ishness’ about them; the phenomenal content often represents the state as a ‘state of me’ ” (block, 1995, p. 235); “there is an elusive sense of self-presence or self-manifestation inherent in even a simple conscious experience of the blue sky. it is less clear to me, however, that this feature […] is constitutive of phenomenology” (kriegel, 2009, p. 177); “self-experience [is] fundamental to the daily experience of all normal human beings […]. it’s a particular way of experiencing oneself that comes to every normal human being in early childhood” (strawson, 2011, p. 8); “normal experience displays […] me-ness [i.e., a phenomenal awareness of oneself]” (guillot, 2017, p. 46). 4i purposefully leave open the specific meanings of “healthy”, “neurotypical” and “ordinary condition”, since different proponents of tc might have slightly different views about these constraints. in general, the claim is intended to apply broadly to conscious experiences occurring in the sober wakeful state, for adult individuals who do not suffer from psychopathologies or neuropathologies. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 3 discussed in the literature; each of these notions, as well as every possible combination of these notions, yields a distinct version of nc that can be assessed more precisely than the generic formulation of the claim. in section 3, i argue that none of these six putative forms of self-consciousness is necessary for consciousness in general, because there are cases in which each of them is plausibly missing. finally, in section 4, i discuss further evidence that some states of consciousness lack any form of self-consciousness – including evidence about drug-induced states that has remained largely overlooked – thus undermining all versions of nc. i conclude by discussing four objections to the possibility and reportability of totally selfless states of consciousness. 2 getting clearer on self-consciousness as claims about phenomenology, both nc and tc should in principle be supported by introspection.5 indeed, proponents of these claims often rely on introspective judgements (as evidenced by some of the quotes provided in thnotes 2 and 3). to some, it is simply manifest that conscious experience involves selfconsciousness. nonetheless, many proponents of those claims cautiously emphasize the elusiveness of the target phenomenon. thus, the relevant notion of the sense of self is deemed “deep and intangible” (chalmers, 1996, p. 10), “implicit” (gallagher & zahavi, 2019) or “elusive” (kriegel, 2009, p. 177). moreover, there has been no shortage of skepticism regarding the introspective accuracy of nc and tc since hume’s infamous denial that anything like a sense of self can be found in experience.6 introspective disagreements are famously difficult to arbitrate, as they often lead to seemingly intractable disputes in which it is hard for both camps to find common ground. given the polysemy of the notions of “sense of self” and “selfconsciousness”, it is not implausible that disagreements about nc and tc might hinge, at least in part, on terminological variation (bayne & spener, 2010). indeed, the two sides of the disagreements might lack a common understanding of what they are supposed to notice in their experience, although they could perhaps come to an agreement if the target feature were specified in more detail. consider, for example, prinz’s skeptical stance on the possibility of experiencing the self in isolation: among the various phenomenal qualities that make up an experience, there is none that can be characterized as an experience of the self […]. 5this does mean, of course, that such claims can be vindicated by introspection alone. as a necessity claim, nc cannot be conclusively defended on the sole basis of reports about phenomenology. however, both nc and tc should be at least consistent with introspection, and it should only take one convincing counter-example to invalidate either claim. 6hume, a treatise of human nature, book 1, section 6. there is, of course, some disagreement on the meaning of hume’s comments on the self (see below for neo-humean positions). millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 4 there is no phenomenal i. if i wait for myself to appear in experience, i will never arrive.7 (prinz, 2012, pp. 214–240) on this neo-humean view, it is impossible to find a distinctive experience of the self among the various phenomenal qualities that show up in one’s stream of consciousness. in other words, there is no such thing as a sui generis qualitative property of selfhood that can be noticed in every conscious experience. as it turns out, there is a rather widespread agreement that this view is phenomenologically plausible.8 in fact, most proponents of nc and tc would agree that the sense of self they deem pervasive in most or all conscious experiences is not “a detachable self quale that one could introspect in isolation from any other content of consciousness” (zahavi & kriegel, 2016, p. 38). in other words, they would happily endorse the claim that all or most conscious states are “selfless” insofar as they lack this feature, although they would maintain that all or most conscious experiences involve self-consciousness in some other sense. thus, disagreements about the existence and prevalence of the sense of self in experience are likely to remain merely verbal as long as both sides fail to reach an agreement on the nature of the target phenomenon. one might hope that once nc and tc have been properly disambiguated, the resulting claims can be assessed on their own merit and the disagreement can be ultimately resolved. but what might self-consciousness be, if not a sui generis qualitative property of selfhood? self-consciousness is traditionally defined as consciousness of oneself as oneself (smith, 2017).9 this definition is intended to rule out cases in which one might accidentally be conscious of what happens to be oneself, such as when one catches a glimpse of oneself in the mirror but fails to recognize oneself (kaplan, 1989; perry, 1979). in other words, genuine self-consciousness has first-personal (de se) content, 7see howell (2006) and dainton (2008) for similar considerations. 8explicit denials of this interpretation of the humean view are rare, although there are some exceptions such as the following passage quoted by nathan (1997) (seemingly to endorse it): “i consider […] my consciousness and feeling of myself, that taste of myself, [to be] more distinctive than the taste of ale or alum, more distinctive than the smell of walnut-leaf or camphor” (hopkins, 1959, p. 123). see also duncan (2019): “i experience myself from the inside, so to speak, in roughly the same way that i experience my pains, thoughts, or emotions. the self is among those items that i can be aware of through introspection.” (i shall come back to the question of whether there is a ubiquitous qualitative property of selfhood in experience in subsection 4.4.) 9there is a long-standing disagreement about whether self-consciousness need be consciousness of oneself, if this is intended to imply a commitment to a realist view of the self. to the extent that instances of self-consciousness have correctness conditions, it might be the case that no such experience is veridical. for example, instances of self-consciousness might systematically misrepresent some existing entity as the self, by analogy with perceptual illusions (mcclelland, 2019); or represent a wholly non-existent entity as the self, by analogy with perceptual hallucinations (bayne, 2010; metzinger, 2003). in other words, “self-consciousness” need not be a success term. strictly speaking, then, one would have to define self-consciousness as consciousness of something as oneself (or simply consciousness as of oneself) to remain neutral on this debate. in what follows, i will keep glossing self-consciousness as consciousness of oneself (as oneself) for convenience, although i will assume a neutral stance on the metaphysics of selfhood. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 5 in so far as it is non-accidentally about oneself. in addition, the notion of selfconsciousness is broadly intended to include cases in which one is conscious of a property or part of oneself as one’s own.10 the general notion of consciousness of oneself as oneself (or consciousness of a property/part of oneself as one’s own) is a determinable of more determinate notions. there are a number of candidates in the literature that can arguably be boiled down to six main determinate notions:11 1. cognitive self-consciousness: thinking of oneself as oneself (garcíacarpintero, 2015; guillot, 2016) 2. spatial self-consciousness: being conscious of one’s location (with respect to one’s perceived environment) as one’s own (bermudez, 1998; schwenkler, 2014) 3. sense of bodily ownership: being conscious of one’s body as one’s own (de vignemont, 2018; martin, 1995) 4. sense of bodily agency: being conscious of one’s bodily actions as one’s own (bayne, 2008; haggard, 2017) 5. sense of mental ownership: being conscious of one’s mental states as one’s own (billon, 2017b; zahavi, 2005) 6. sense of mental agency: being conscious of one’s mental actions as one’s own (peacocke, 2006; proust, 2009) each of these determinate notions can be used to formulate distinct versions of nc and tc where they refer to the target feature whose pervasiveness in conscious experience is at issue. for example, the six notions yield the following six versions of nc: nc1 for any subject s, if s is conscious at t, then s’s phenomenology involves cognitive self-consciousness at t. nc2 for any subject s, if s is conscious at t, then s’s phenomenology involves spatial self-consciousness at t. nc3 for any subject s, if s is conscious at t, then s’s phenomenology involves a sense of bodily ownership at t. nc4 for any subject s, if s is conscious at t, then s’s phenomenology involves a sense of bodily agency at t. 10see chisholm (1969, p. 21). 11while i will focus on these six notions in this article, i will consider in subsection 4.4 whether additional notions of self-consciousness could be defined and whether doing so would threaten my overall argument. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 6 nc5 for any subject s, if s is conscious at t, then s’s phenomenology involves a sense of mental ownership at t. nc6 for any subject s, if s is conscious at t, then s’s phenomenology involves a sense of mental agency at t. moreover, insofar as all of these phenomenal features are taken to be determinates of which self-consciousness is a determinable, they can be combined within disjunctive claims that are also versions of nc and tc. thus, one might claim that the ordinary experience of healthy, neurotypical adults involves cognitive self-consciousness or spatial self-consciousness; and similarly for any other disjunction including two or more of the six phenomenal features listed above.12 in particular, there are maximally disjunctive versions of nc and tc, such as: ncdisj for any subject s, if s is conscious at t, then s’s phenomenology involves any of the phenomenal features (1–6) at t (including any combination of these features). presumably, the claims resulting from the distinction of determinate notions of self-consciousness are less likely to generate verbal disputes than the generic formulations of nc and tc, because they leave less room for equivocation about the nature of the target phenomenon. nonetheless, there is plenty of genuine disagreements on whether each of the phenomenal features (1–6) actually exists. it is worth dwelling briefly on the nature of such disagreements. it is fairly uncontroversial that we can entertain thoughts about ourselves, and judge that our spatial location, body, action or mental state is our own. but this does not entail that there is a distinctive experience of thinking about oneself as oneself, or of being aware of one’s location, body, action or mental state as one’s own. if the six features distinguished above are real and distinct forms of self-consciousness, they ought to come with their own phenomenology and be irreducible to one another. for example, if there is such a thing as a phenomenology of bodily ownership, it should not simply be a matter of thinking of one’s body as one’s own (de vignemont, 2018); likewise, if there is a phenomenology of bodily agency, it should not merely be a matter of thinking of one’s bodily actions as one’s own (bayne, 2008); and so on for all other relevant features. in what follows, i shall remain agnostic on the issue of whether any of the phenomenal features (1–6) are ever instantiated in conscious experience. instead, i will consider whether there is any compelling evidence that they are missing in at least some states of consciousness – independently of whether or not these states occur in the ordinary experience of healthy, neurotypical adults. thus, my focus will be on the empirical plausibility of various versions of nc. against nc1–nc6, i will 12if we limit the analysis to these six determinate notions, this yields a total of 64 (26) possible versions of nc and tc. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 7 argue that it is possible to find states of consciousness lacking each of the six putative forms of self-consciousness individually. such states of consciousness are at least partially selfless, insofar as they lack at least some form of self-consciousness. i will then argue that some states of consciousness lack all six putative forms of self-consciousness. such totally selfless states of consciousness should count as evidence against all versions of nc, including ncdisj. 3 partial selflessness let us define a partially selfless state of consciousness as a state that lacks at least one form of self-consciousness, but not all: psl a global state of consciousness of a subject s is partially selfless if and only if, while being in that state, s is not self-conscious in at least one of the ways in which s can be self-conscious, but does not lack all of the ways in which s can be self-conscious at the same time. by contrast, let us define a totally selfless state of consciousness as a state that lacks all forms of self-consciousness: tsl a global state of consciousness of a subject s is totally selfless if and only if, while being in that state, s is not self-conscious in any way in which s can be self-conscious (if s can be self-conscious at all). if none of the putative forms of self-consciousness (1–6) exist, and no other form of self-consciousness exists, it trivially follows that all states of consciousness are totally selfless. if only one of the putative forms of self-consciousness (1–6) exists, then any state of consciousness lacking this unique form of self-consciousness will be totally selfless. finally, if several of the putative forms of self-consciousness (1–6) exist, then there might be different types of partially selfless states that differ with respect to the form(s) of self-consciousness they are lacking. states of consciousness that credibly lack one of the putative forms of self-consciousness distinguished in the previous section should be considered to be at least partially selfless, while states of consciousness that lack all of them should be considered to be totally selfless. 3.1 consciousness without cognitive self-consciousness the question of whether there can be states of consciousness in which one does not consciously think of oneself as oneself is rather straightforward.13 indeed, it 13one some versions of the higher-order thought (hot) theory of consciousness, being in a conscious mental state m necessarily involves thinking that one is in m (rosenthal, 1986). however, hot theories are not committed to the claim that higher-order thoughts about first-order mental states should themselves be conscious thoughts. in fact, proponents of the theory readily acknowledge that “the hots are typically not conscious thoughts” (rosenthal, 1997, p. 745). millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 8 would be odd to claim that subjects are perpetually thinking of themselves as themselves whenever they are conscious at all. even the most narcissistic individuals presumably live through some conscious episodes in which they are not thinking about themselves. admittedly, we are often lost in thoughts about ourselves, such as the planning and anticipation of future personal goals (see d’argembeau, 2018, for a review). however, we clearly have a great many thoughts that are not about ourselves in any way. for example, try solving the following equation: 𝑥2 + 5 = 13 − 4 once you have determined the value of x, ask yourself whether you were thinking about yourself in any way while you were pondering the problem. i predict that for you and other readers the answer will be unanimously negative. in addition to the routine occurrence of thoughts that are not focused on oneself in ordinary experience, there are also global states of consciousness in which self-related thoughts appear to be inhibited, if not suppressed altogether. for example, it has been argued that the performance of attention-demanding activities can induce a state of flow, especially at expert level, characterized by extreme and undivided focus on the task at hand (csikszentmihalyi, 1990). the flow state scale (fss) was designed to measure different aspects of flow states across nine factors (jackson & marsh, 1996). in a sample of 747 respondents from a variety of competitive and noncompetitive sport activities, the mean score (on a five-point likert scale) for the fss factors concentration on the task at hand (including the item “my attention was focused entirely on what i was doing”) and loss of self-consciousness (including the item “i was not concerned with what others may have been thinking of me”) were 3.73 and 3.92 respectively (jackson & eklund, 2004). similar results were obtained with a sample of 236 music student performing for an examination, with mean scores of 3.53 and 3.12 for the same fss factors (wrigley & emmerson, 2011). these data suggest that skilled individuals focused on a demanding performance may reach a state in which they are entirely focused on task-related goals, and therefore are unlikely to have thoughts about themselves (as themselves).14 another condition in which cognitive self-consciousness might temporarily subside is meditation. many meditative practices, especially in the buddhist tradition, explicitly aim at the cessation of spontaneous thoughts and the focus of attention on sensory stimuli (fox & cahn, forthcoming). specifically, these practices cultivate meta-awareness of episodes of mind-wandering, by monitoring spontaneous thoughts in order to disengage from them and re-focus attention on a particular object (such as the breath or a mantra) or on a wider awareness of the 14see høffding & montero (2019), however, for a different perspective on flow states, according to which “the enactment of expertise requires conscious thought” and “one never disappears, as one is always there, exclusively occupied with the momentary task at hand” (p. 17). note that the authors do mention the possibility that conscious thoughts only emerge when something goes wrong during the task. (i am grateful to dan zahavi for bringing this article to my attention.) millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 9 present moment (hasenkamp, 2018). there is some evidence that once attention has been stabilized and the mind “quieted”, meditators can undergo conscious episodes entirely lacking in self-referential thoughts (millière, carhart-harris, roseman, trautwein, & berkovich-ohana, 2018). finally, it is plausible that the experience of dreaming is often free of cognitive self-consciousness. an analysis of 500 rapid eye movement (rem) dreams found that only 20% of dream reports included instances of thinking, and that such instances made up only 5% of the content of reported dreams (meier, 1993). these findings suggest that “different forms of thinking can be lacking in dreams” (windt, 2015a, p. 424), including thinking of oneself as oneself. 3.2 consciousness without spatial self-consciousness in section 2, i defined spatial self-consciousness as the subject’s consciousness of her location with respect to her perceived environment as her own location. some perceptual experiences – visual and auditory experiences in particular – represent spatial properties of perceived objects within an egocentric frame of reference (evans, 1982; klatzky, 1998). indeed, such experiences represent the location of environmental landmarks (i.e., objects, properties and events perceived in the subject’s environment) with respect to a point of origin, roughly centered on the subject’s head in the case of visual and auditory perception. for example, a particular object might be represented as being located on the right hand side at a certain distance from the point of origin of the subject’s visuospatial perspective (over there to the right of here). correlatively, the location of the point of origin of one’s visuospatial perspective is itself represented relatively to environmental landmarks. to give a simple example, if an object perceived in the subject’s environment is represented as being located at some distance to the right of the subject’s point of origin, then it seems that the subject’s point of origin will itself be represented as being located at some distance to the left of the relevant object. a number of authors have gone further in suggesting that in ordinary perceptual experience, the location of the point of origin of a subject’s visuospatial perspective with respect to her perceived environment is represented as as the subject’s own location (bermudez, 1998; bermúdez, 2001; cassam, 1997; hurley, 1998; noë, 2005; schwenkler, 2014). in other words, environmental landmarks are not simply perceived as located with respect to here, where here refers to the location of the point of origin of the relevant frame of reference (e.g., the visuospatial perspective); they are perceived as located with respect to me, where me refers to the subject as such. as cassam puts it, “in experiencing objects as spatially related to one, one literally experiences the bodily self as located in the perceived world” (cassam, 1997, p. 53). let us call an experience that represents the relative location of the point of origin of its spatial frame of reference as the subject’s own location a self-locating experience, and the relevant kind of spatial self-representation selflocating content. insofar as self-locating content contributes to the phenomenal millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 10 character of self-locating experiences, the latter can be said to constitute instances of self-consciousness – namely, instances of spatial self-consciousness. whether ordinary visual and auditory experiences always involve spatial selfconsciousness is a matter of controversy. since my focus here is on nc2 rather than on the corresponding version of tc, i shall leave aside ordinary visual and auditory experiences, and ask whether there is compelling empirical evidence that some (non-ordinary) states of consciousness lack spatial self-consciousness. a good starting point for finding evidence is to consider states of consciousness in which stimuli providing information about distant environmental landmarks are not available. this is the case when subjects are deprived of distal sensory modalities such as sight and audition, a condition known as sensory deprivation or perceptual isolation. one of the most effective ways of artificially cutting off distal perception in healthy subjects is the use of so-called flotation tanks (feinstein et al., 2018). flotation tanks contain a solution of water and epsom salts at skin temperature allowing the patient to float supine in darkness and silence. in such conditions, visual, auditory, tactile, proprioceptive, thermal and vestibular signals are extremely attenuated. it is plausible that individuals floating in a such a tank may undergo episodes in which they cannot perceive any environmental landmark. indeed, even the tactile stimulation coming from the contact with water is extremely attenuated by the lack of movement and the fact that the water is at skin temperature. thus, flotation tanks and similar sensory deprivation devices should yield, at least occasionally, experiences that lack self-locating content. although there is little evidence on the phenomenology of experiences induced by flotation tanks from controlled studies, the few qualitative reports available are consistent with this hypothesis: i […] just flew away into nothing […] and could not feel whether i was in the tank or where i was. (kjellgren, buhrkall, & norlander, 2010, p. 1252) [t]here was only me and i was lying there shut in somewhere and i did not know where i was or time or space […]. (kjellgren, lyden, & norlander, 2008, p. 644) i feel like i am losing time and space. (kjellgren et al., 2008, p. 646) the extent to which artificial sensory deprivation can completely remove spatial cues about one’s position with respect to one’s environment is debatable, because the subjects’ distal perceptual capacities are not directly impaired. by contrast, some individuals lack distal sensory modalities altogether as a result of genetic conditions, illness or accidental organ damage. for example, the genetic disorder known as usher syndrome can lead to deafblindness, the combination of hearing and vision loss (dammeyer, 2014). there is some evidence that individuals with deafblindness can have experiences lacking self-locating content, especially when tactile and vestibular stimuli are attenuated as in flotation tanks. thus, a deafblind subject with usher syndrome interviewed about her daily experience described millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 11 commonly recurring episodes in which she appears to completely lose awareness of her spatial location while taking a bath: doing nothing but relaxing in the bath i always found i became lost as in relation to where i was… it is a feeling of being there but not there… you are here but not here. you are just existing… [i]f only for a short time i do lose the spatial awareness of the bath… [s]lowly the whole world vanishes… i have no sense of being present in any environment, the disconnect from all is like unplugging from the world… it is like being in a void, a nothingness that is just outside our reality, and i slip into the void when i have a bath… your sense of where you are… just disappear[s] to nothingness, the world becomes nothing […]. it is as if the subconscious decides that location is now not needed and turns off the gps area of the mind. (millière, 2019, p. 266) if there is such a thing as experiencing one’s location in egocentric space as one’s own, it is plausible that such an experience cannot occur when sight and audition are altogether missing and other sensory modalities are strongly attenuated. while anecdotal reports from single subjects should be interpreted with caution, the above report supports this idea. aside from sensory deprivation and deafblindness, there is some evidence that expert meditators can undergo states of consciousness described in terms that strongly suggest they lack self-locating content. thus, a recent study reported that a subject with around 20,000 hours of meditation practice could reliably induce an experience described in the following terms: i don’t have any kind of sense of location […]. i have no idea where i am […]. i’m not there basically, just world, so there’s no real location. (ataria, dor-ziderman, & berkovich-ohana, 2015 sup. material) another study of long-term mindfulness meditation practitioners found that some of them were able to induce a state of “spacelessness” (berkovich-ohana, dorziderman, glicksohn, & goldstein, 2013), described in such terms as “i felt outside time and space” (p. 4) and “the center of space became endless, without a reference point in the middle” (p. 5). the evidence that meditation can induce states of consciousness lacking spatial self-consciousness remains rather scarce, and as such is unlikely to convince those who think that spatial self-consciousness is ubiquitous in conscious experience. nonetheless, the cases of sensory deprivation and especially deafblindness strongly suggest that one can be conscious without being conscious of one’s location with respect to one’s environment. given the dependence of spatial selfconsciousness on the processing of sensory stimuli that carry spatial information about the subject’s location, and given that the capacity for consciousness does not depend on the availability of such stimuli, there is no reason to think that spatial self-consciousness should be necessary for consciousness in general.15 15perhaps one could be in a state where one is not aware of one’s location with respect one’s millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 12 3.3 consciousness without a sense of bodily ownership the sense of bodily ownership refers to the “phenomenological quality that [a] body part appears to be part of one’s body” (martin, 1995, p. 269). some authors hold that ordinary conscious experience involves such phenomenology (de vignemont, 2018), while others deny that the relevant phenomenal feature exists (bermúdez, 2011). are there states of consciousness that uncontroversially lack this phenomenal feature, namely states that do not involve experiencing one’s body as one’s own? proponents of the claim that the sense of bodily ownership is pervasive in ordinary experience argue that certain psychopathologies such as somatoparaphrenia and depersonalization disorder are characterized by loss of the sense of bodily ownership, over specific limbs or over the whole body respectively (de vignemont, 2018). somatoparaphenia is a monothematic delusion in which patients report that one of their limbs belongs not to them but to someone else. some reports suggest that this delusional belief is rooted in phenomenology: my eyes and my feelings don’t agree, and i must believe my feelings. i know they [the affected limbs] look like mine, but i can feel they are not, and i can’t believe my eyes. (nielsen, 1938, p. 555) depersonalized individuals frequently report feeling disconnected from their bodies, or feeling as if their bodies were absent or dead. unlike somatoparaphrenic patients, however, they are not delusional insofar as they do not endorse the content of these abnormal feelings (billon, 2016). some reports from depersonalized individuals seem to be elegantly explained by the hypothesis that they lack a sense of bodily ownership over their bodies: my body didn’t feel like my own. i just felt like a floating mind in space with blunted, blurry thoughts. i found it crazy sometimes that i even had my own body, because i didn’t feel connected to it. i’d look down at my hands and they didn’t feel real. (bradshaw, 2016) reports from psychopathologies might not be sufficient to convince those who think that we cannot have bodily sensations without a sense of bodily ownership attached to them. there is, however, a more indirect way to argue that there can be states of consciousness lacking a sense of bodily ownership. insofar as the putative sense of bodily ownership refers to an experience of one’s body or body part as one’s own, it is parasitic upon bodily awareness – which refers to one’s experience of one’s body in general. if a subject had no awareness of her body, then it is difficult to see how she could have a sense of bodily ownership. environment, and yet still have bodily sensations with spatial content (e.g., an itch located in one’s ankle). however, it is unclear that such sensations would involve a form of self-consciousness if they do not involve a sense of bodily ownership (see the next section). furthermore, the deafblind subject quoted above also reported “you’re not there in your body anymore […]. the boundaries of my body vanish when [i am] cocooned in the water” (millière, 2019, p. 271). millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 13 there is converging empirical evidence suggesting that subjects may be conscious without having any kind of bodily experience. examples of such states may include certain kinds of dreams and certain kinds of out-of-body experiences (generally resulting from neuropathologies). in the general population, there is evidence that a small subset of dreams completely lack bodily sensations or the representation of a dream body (cicogna & bosinelli, 2001; laberge & degracia, 2000; occhionero & cicogna, 2011). for example, some subjects report that during their dreams they “experienced themselves as a disembodied point or freely moving center of awareness” (windt, 2010, p. 301). this suggests that dreams need not involve bodily awareness, or therefore a sense of bodily ownership. likewise, there is a subset of out-of-body experiences (obes), known as “asomatic”, in which subjects have the experience of being disembodied without having the illusion of seeing their own body from the outside (alvarado, 2000; metzinger, 2013b). a recent study with a large sample size of both healthy participants and vestibular disorder patients who reported obes suggests that only 10% to 30% of such experiences feature autoscopic episodes, namely the hallucination of one’s “own” body in external space, at some distance from the location from which one perceives the world (lopez & elzière, 2018, p. 6). this suggests that the majority of obes might not involve the visual representation of one’s body. thus, alvarado (2000) reports that some individuals who experience an obe “do not experience a body at all, describing themselves as ‘pure consciousness’ […] or as ‘balls of light’, ‘points in space’, or ‘clouds’ ” (p. 186). admittedly, as many as 59% of vestibular disorder patients who reported obes experienced vestibular sensations concurrently (lopez & elzière, 2018). however, it is dubious that mere vestibular sensations would be sufficient to sustain a sense of bodily ownership in the absence of bodily sensations located within a body schema. indeed, vestibular sensations such as dizziness or floating sensations are not clearly instances of awareness of one’s body or body parts. in any case, it is plausible that some asomatic obes do not involve concurrent vestibular sensations. overall, the evidence discussed in this section strongly supports the claim that some states of consciousness do not involve a phenomenology of bodily ownership, even if such phenomenology exists and is relatively prevalent in ordinary experience. 3.4 consciousness without a sense of bodily agency many authors argue that there is a distinctive phenomenology of voluntary action that goes beyond the somatosensory awareness of bodily movements (bayne, 2008; hohwy, 2007; horgan, tienson, & graham, 2003). assessing whether there can be states of consciousness lacking such phenomenology (if it exists) is rather straightforward. firstly, one should expect the sense of bodily agency to be attached to volunmillière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 14 tary bodily actions rather than to involuntary bodily movements as a general rule, although there might be some exceptions (i.e., illusions of bodily agency). there are a number of involuntary movements that should not normally be associated with a sense of agency. these include involuntary muscle twitches and reflex actions, but also passive movements resulting from the action of another agents or the force exerted by inanimate objects. secondly, there are countless conditions in which one’s body is immobile and one lacks an experience of bodily movement altogether. individuals in the lockedin state suffering from complete motor paralysis offer a particularly dramatic example of such a condition (murguialday et al., 2011). but there are also mundane situations in which one is not moving a muscle, and therefore is unlikely to have any sense of bodily agency – for example as one falls asleep (see blanke & metzinger, 2009). finally, the cases in which one lacks bodily experiences altogether (discused in subsection 3.3) should a fortiori lack any sense of bodily agency. for example, bodiless dreams should not involve a distinctive phenomenology of bodily action in the absence of a representation of a dream body. 3.5 consciousness without a sense of mental ownership the sense of mental ownership, also called sense of “mineness”, refers to the putative “sense that i am the one who is undergoing an experience” (gallagher, 2000, p. 15). clearly one can judge, at least in some circumstances, that a conscious mental state is one’s own – in the sense that one undergoes the relevant mental state. for example, i can judge that i am currently undergoing a visual experience of a computer screen, such that this token visual experience is a mental state of which i am the subject. to say that a subject has a sense of ownership over a (conscious) mental state is to say something more, namely that the subject is aware of that mental state as her own (where this awareness is not a matter of judgement, but a distinctive phenomenology of ownership). the idea that all or almost all conscious experiences involve a sense of ownership (over those very experiences) can seemingly be found in both the analytic and the phenomenological traditions: [a]ll subjective experience is self-conscious in the weak sense that there is something it is like for the subject to have that experience. this involves a sense that the experience is the subject’s experience, that it happens to her, occurs in her stream. (flanagan, 1992, p. 194) when i (in nonpathological standard cases) am aware of an occurrent pain, perception, or thought from the first-person perspective, the experience in question is given immediately, noninferentially and noncriterially as mine. (zahavi, 2005, p. 124) millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 15 however, the sense of mental ownership is perhaps the most controversial of the six notions of self-consciousness distinguished in section 2. firstly, some of the proponents of the idea that one is aware of one’s occurrent experiences as one’s own seem to understand it in a deflationary manner that amounts to little more than the claim that one’s experiences constitutively contribute to one’s overall phenomenology (zahavi, 2018). if this deflationary interpretation is correct, then there is after all no distinctive phenomenology of mental ownership above and beyond the phenomenal character of one’s occurrent experience.16 it is difficult to find authors who argue that there is a sense of mental ownership and explicitly refer to it as a distinctive feeling (in addition, as it were, to the phenomenal character of the mental state to which it supposedly applies).17 secondly, a number of objections have been raised against the very idea that there is a sense of mental ownership (see chadha, 2018; mcclelland, forthcoming). thirdly, even those who think that there is a phenomenology of mental ownership generally agree that it is not ubiquitous. thus, some suggest that we have a phenomenology of mental ownership only in very specific cases, when we introspect our occurrent experiences (howell & thompson, 2017), while others hold that the sense of mental ownership is missing at least in certain pathological cases. guillot (2017) and billon (2017b) both argue that depersonalized patients lack a phenomenology of mental ownership (regardless of whether they also lack a phenomenology of bodily ownership). consider the following reports quoted by billon: [i]t is not me who feels. i am not interested in what i appear to be feeling, it is somebody else who feels mechanically. (janet, 1908, p. 515, billon’s translation) i feel pains in my chest, but they seem to belong to someone else, not to me. (mayer-gross, 1935, p. 114) taken at face value, these reports suggest that subjects were not aware of the relevant experiences as their own. similarly, martin & pacherie (2013) suggest that 16i will come back to this deflationary notion in subsection 4.4. 17martin & pacherie (2013) come close when they suggest that we normally have a “feeling or sense that our thoughts belong to us” (p. 112), but they specifically restrict their account to thoughts. block (2007) also comes close when he writes that “awareness is a matter of having a state whose content is in some sense ‘presented’ to the self or having a state that is ‘for me’ or that comes with a sense of ownership or that has ‘me-ishness’ ” (p. 484), although the equivalence between “awareness” and the sense of ownership suggests that block might endorse the deflationary interpretation of the notion. gallagher (2000) does not draw an explicit distinction between the sense of bodily ownership and the sense of mental ownership, giving as a canonical example of the “sense of ownership”: “the sense that my body is moving regardless of whether the movement is voluntary or involuntary” (p. 15). this makes it difficult to understand his proposal, because being aware of one’s body as one’s own (through bodily experience) is different from being aware of one’s bodily experience as one’s own (see de vignemont, 2011). finally, nida-rümelin (2017, p. 66) argues that “being presented with something necessarily involves being […] aware of being the subject to whom something is presented.” millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 16 reports of thought insertion in schizophrenia are best explained by the lack of a sense of mental ownership over affected thoughts. here is a characteristic report quoted by martin and pacherie in support of their interpretation: [s]he said that sometimes it seemed to be her own thought “but i don’t get the feeling that it is”. she said her “own thoughts might say the same thing”, “but the feeling isn’t the same”, “the feeling is that it is somebody else’s.” (allison-bolger, 1999, #89) of course, these reports might be best explained by the occurrence of an additional feeling of alienation or estrangement from one’s experiences (in the case of depersonalization) or from one’s thoughts (in the case of thought insertion) that is missing in ordinary experience (billon & kriegel, 2015; parrott, 2017). but this alternative interpretation would remove the need to posit the existence of a sense of mental ownership in the first place, suggesting that neither depersonalization/thought insertion nor ordinary experience comes with such phenomenology.18 aside from psychopathologies, one may wonder whether more mundane states of consciousness such as nonlucid dreams and mind-wandering might offer compelling evidence that the sense of mental ownership, if it exists at all, is unlikely to be ubiquitous. nonlucid dreams are characterized by the fact that the dreamer is not aware of dreaming. presumably, such dreams do involve conscious experiences, in so far as there is something it is like to undergo nonlucid dreams. admittedly, one should not confuse the awareness that one is dreaming with the awareness of one’s dream experiences as one’s own; in principle, one could have the latter without the former. nevertheless, if the experience of nonlucid dreams came with a distinctive sense of mental ownership, one could expect this feature to guarantee some form of continuity and coherence in dreams; but dreams are often incongruous, vague and discontinuous (hobson, 1988). mind-wandering is another ordinary scenario in which the occurrence of a distinctive sense of ownership over one’s experience is implausible. as i mentioned in subsection 3.1, subjects are often unaware that they undergo episodes of mindwandering.19 the suggestion that a subject who is lost in thought and unaware of her mental drift would nonetheless have a distinctive sense of ownership over her wandering thoughts is particularly unintuitive. as windt (2017) puts it: “mindwandering episodes are owned in the basic sense of occurring in me, they have phenomenal character […] [b]ut this does not mean that they are also experienced […] as mine” (p. 1727). 18it is also possible that the relevant reports are best explained by the lack of some phenomenal feature other than the putative sense of mental ownership, such as the sense of bodily ownership for depersonalization (de vignemont, 2018) or the sense of mental agency for thought insertion (stephens, poland, & radden, 2000). while this hypothesis is compatible with the assumption that the sense of mental ownership is ubiquitous in both normal and pathological experiences, this claim requires additional motivation, which would be difficult to find if phenomenal contrast cases were unavailable. 19i do not mean to suggest that episodes of mind-wandering are not consciously experienced, but rather that subjects lack meta-awareness that they are undergoing such episodes. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 17 3.6 consciousness without a sense of mental agency while most discussions of agency focus on bodily actions, there is an increasing interest within philosophy of mind in mental actions (o’brien & soteriou, 2009). mental actions are generally taken to include various endogenous mental states such as task-related thinking, judging, calculating, deciding, actively remembering or actively imagining. insofar as these mental states are under voluntary control and involve some degree of cognitive effort, it has been suggested that they come with a sense of mental agency (peacocke, 2006; proust, 2009). according to a radical version of this view, consciousness always involves a sense of mental agency, because being conscious requires some form of mental activity: the mind of one who is conscious is necessarily a mind actively governing the movement of its own attention and thinking processes […]. in the final analysis it is because thinking is active and thinking is essential to consciousness that mental action is a necessary condition of consciousness. (o’shaughnessy, 2000, pp. 89–264) does this view have any plausibility? even if we assume that mental actions are normally associated with a sense of mental agency, there seem to be plenty of conscious episodes that do not involve mental activity. i have already discussed the case of schizophrenic thought insertion, which has traditionally been interpreted in terms of a lack of the putative sense of mental agency over affected thoughts, leading patients to attribute them to foreign agents (stephens et al., 2000). but the familiar phenomenon of unbidden thoughts – thoughts that seem to pop unsolicited into one’s mind – provides a more mundane example of cognitive processes lacking a sense of mental agency. while this phenomenon occurs in the general population, it is particularly common in patients diagnosed with obsessive– compulsive disorder (ocd), who struggle to keep intrusive thoughts at bay and report that their occurrence feels irresistible (billon, 2013). more generally, the landscape of conscious cognitive processes appears to be filled with regions of passivity. as i have previously emphasized, there is ample evidence that mind-wandering is a phenomenon that is not under voluntary control, and whose occurrence we might even fail to notice until the moment at which we snap back to reality. in light of the empirical literature on mind-wandering, it is perhaps no exaggeration to say that control of our conscious thought processes is the exception rather than the rule (metzinger, 2013a); see also figure 2 in metzinger (2015, p. 294).20 the same could be said about many conscious episodes that occur during nonlucid dreaming, which are often characterized by their passivity – by contrast with imagination (windt, 2017). finally, it seems unlikely that mental actions would occur during certain states of absorption, such as the experience of being enthralled by a captivating movie. 20although see seli et al. (2017) for evidence that some episodes of mind-wandering are started intentionally. (i am grateful to thomas metzinger for bringing this article to my attention.) millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 18 overall, it does not seem difficult to find examples of conscious episodes that are characterized by a lack of mental activity, and therefore should not involve any sense of mental agency. 3.7 the varieties of partial selflessness the empirical cases discussed in this section delineate many potential forms of partial selflessness, depending on how many of the six candidate notions of selfconsciousness outlined in section 2 really exist and are plausibly missing in each case given available evidence (table 1). conditions csc ssc sbo sba smo sma mind-wandering x ? ? ? ? x flow states x ? ? ? ? ? meditation x x ? ? ? ? nonlucid dreams x ? ? ? x x bodiless dreams x ? x x x x deafblindness ? x ? ? ? ? sensory deprivation ? x ? ? ? ? asomatic obes ? ? x x ? ? somatoparaphrenia ? ? x x ? ? drug-induced states ? ? x x ? ? depersonalization ? ? x x x ? total paralysis ? ? ? x ? ? thought insertion ? ? ? ? x x unbidden thoughts ? ? ? ? ? x table 1: the varieties of (partial) selflessness. (csc = cognitive self-consciousness, ssc = spatial self-consciousness, sbo = sense of bodily ownership, sba = sense of bodily agency, smo = sense of mental ownership, sma = sense of mental agency. red crosses denote the absence of the relevant feature.) as i have repeatedly emphasized, several of these putative forms of self-consciousness might not exist. rather than weighing directly on their existence, i have argued – against nc1– nc6 – that for each of these notions of self-consciousness, it is plausible in light of available evidence that (at least) some states of consciousness lack the corresponding phenomenal feature. if several of the putative forms of self-consciousness (1–6) exist, however, then there might be not only different types of partially selfless state that differ with respect to the feature they lack, but also totally selfless states: states that lack any form of self-consciousness. i will now consider this possibility in light of empirical evidence, and discuss potential objections to the very idea of totally selfless states of consciousness. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 19 4 total selflessness i have previously defined totally selfless states of consciousness as follows: tsl a global state of consciousness of a subject s is totally selfless if and only if, while being in that state, s is not self-conscious in any way in which s can be self-conscious (if s can be self-conscious at all). several authors have argued that totally selfless states of consciousness are at least conceptually possible, if not nomologically possible. thus, peacocke (2014) suggests that there could be subjects whose (conscious) mental states are at “degree 0” of self-representation, meaning that they do not involve any kind of selfrepresentation. although he notes that “there is nothing incoherent in this description of a subject at degree 0 who perceives the world but does not self-represent” (peacocke, 2014, p. 31), and entertains the fictional case of “spherical underwater creature” as an example, he falls short of endorsing the claim that any actual creature has degree 0 conscious mental states. likewise, dainton (2016) considers as a thought experiment the case of newly born deafblind puppy floating in outer space, arguing that such creature would have experiences lacking self-consciousness altogether. while he acknowledges that this kind of fictional scenario is “rather speculative”, he suggests that it there is no alternative to speculation when it comes to the possibility of totally selfless states, because such states are dramatically different from the wakeful states we are familiar with. in this section, i will argue that we need not resort to far-fetched thought experiments to ascertain whether totally selfless states are nomologically possible. indeed, i will suggest that there might be actual examples of such states in healthy, neurotypical adult humans – although perhaps not in ordinary conscious experience.21 i will specifically discuss available evidence regarding three candidates for such states: drug-induced ego dissolution, meditation and lucid dreamless sleep. after discussing each of these cases, i will address four lingering objections to the possibility and reportability of totally selfless states.22 4.1 drug-induced ego dissolution certain psychoactive drugs are known to produce dramatic short-term effects on self-consciousness, especially at higher doses. this phenomenon, called drug-induced ego dissolution (died), is typically described as a (reversible) loss 21whether these examples really are totally selfless states of consciousness depends, of course, on whether the typology of self-consciousness proposed in section 2 is exhaustive. on this issue, see my response to the third objection in section 4.4 below. 22i remain agnostic here on whether there might be several types of totally selfless state of consciousness, or a single type of such state – although i am personally inclined to believe that the former is true, as the putative examples of totally selfless states discussed below seem to vary with respect to phenomenology. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 20 of one’s sense of self and sense of self–world boundaries (letheby & gerrans, 2017; millière, 2017).23 although such effects had already been reported in early studies on mescaline24, lsd25 and psilocybin26, recent controlled studies and anecdotal evidence converge in suggesting that they can be induced by a broad range of psychoactive drugs, including serotonin 2a agonists (so-called classic psychedelics), nmda antagonists (so-called dissociative anesthetics) and kappa-opioid agonists.27 a careful analysis of reports of died suggests that this notion is a complex construct that refers to a family of experiences rather than a single type of experience. while some self-reported experiences of died are described as involving a loss of access to autobiographical semantic information (e.g., information about one’s personal identity and history), others are characterized by the loss of awareness of bodily sensations, and yet others seem to involve distinct impairments. if there are several forms of self-consciousness, then it should not be surprising that different types of experiences might be indiscriminately reported as experiences of “self-loss”, even though they are in fact characterized by the loss of distinct phenomenal features. in other words, died might be an umbrella term that covers different forms of partial selflessness (millière et al., 2018). instead of using the term indiscriminately, it is more judicious to assess which putative forms of self-consciousness might be present or absent in experiences induced by different drugs. intriguingly, some reports of died describe experiences that appear to lack not one or a few of the putative forms of self-consciousness outlined in section 2, but all six of them. one drug called 5-meo-dmt seems to be particularly likely to induce this kind of experience (barsuglia et al., 2018; uthaug et al., 2019). 5meo-dmt is a classic psychedelic naturally occurring in the toxin of the toad bufo alvarius as well as in a wide variety of plants. its effects are both very intense and very short-lived, lasting for 10 to 20 minutes when the substance is insufflated. let us examine reports of died in light of the distinction between six notions of self-consciousness, with a particular focus on reports describing the effects of 5-meo-dmt. i have already argued that large parts of our waking lives are not spent thinking about ourselves. this is all the more likely to be the case under the influence of drugs that profoundly impair cognitive processes. consider the following descriptions of the effects of 5-meo-dmt gathered from erowid, a curated database of reports of drug-induced states:28 23see also letheby (2020) and deane (2020), both in this special issue. 24beringer (1927; guttmann & maclay, 1937; mayer-gross & stein, 1926). 25anderson & rawnsley (1954; bercel, travis, olinger, dreikurs, & polos, 1956; klee, 1963; pahnke & richards, 1966; savage, 1955; sedman & kenna, 1964; von mering, morimoto, hyde, & rinkel, 1955). 26pahnke (1966; rümmele & gnirss, 1961). 27see millière (2017) for a review. 28see erowid.org/experiences (each report has a unique identification number). millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://erowid.org/experiences https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 21 i recall no thought process, i wasn’t thinking at all, just simply experiencing. (#17378) the world as i knew it is gone forever […]. there is no thought. there is no time. there is no world. (#90018) my brain seemed unable to comprehend the detail of complexity and i wondered if my vision was temporarily impaired. my ability to think coherently seemed to be similarly truncated. (#79324) it’s impossible to think while undergoing the experience. (#52311) my thoughts shattered into nothing. i lost sense of self, and my environment […]. [i]t was [an] absence of conscious thought. (#16507) darkness surrounds me entirely, yet i’m not unconscious. i’m aware of the darkness, but there isn’t any mental chatter. i wasn’t wondering where i was […]. i wasn’t wondering what had happened to anything, or anyone that were with me a moment ago. just silence and darkness. (#90196) these reports suggest that at the peak of experiences induced by 5-meo-dmt it becomes difficult to think at all. a fortiori, one should expect that most subjects are unable to think about themselves as themselves in that state. there is also evidence that the experience of space is thoroughly disrupted by 5-meo-dmt. a recent online survey on the effects of the drug with a sample of 515 participants found that the overwhelming majority of respondents scored positively on items of the mystical experience questionnaire corresponding to “loss of your usual sense of space” (95% of respondents), “loss of usual awareness of where you were” (88%) and “being in a realm with no space boundaries” (87%) (davis, barsuglia, lancelotta, grant, & renn, 2018). anecdotal reports from erowid also suggest that subjects often lack awareness of their location in space: [s]oon i am falling, falling, up, down, inward, outward, i’m not really sure. i am flung into a room with no features and no depth cues. (#5816) i was the universe, i was everywhere and nowhere, everything and nothing all at the same time. (#49690) any concept of “time” or relative “space” […] was carried off along with the rest of manifest existence […] moving my consciousness inward and outward in unimaginable magnitude, “eventually” into the yawning maw of nothingness […]. all sense of “i” [was] instantly gone. (#23769) i previously argued in subsections 3.3 and 3.4 that one cannot have a sense of bodily ownership or a sense of bodily agency if one lacks bodily experiences altogether. reports from 5-meo-dmt frequently describe a complete loss of bodily awareness. for example, the following reports seem to describe a lack of both spatial selfconsciousness and bodily awareness: millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 22 i was completely disconnected somatically, unable to locate or feel my body […] unable to locate myself – or anything else – anywhere in particular. i had no body, not even the slightest semblance of a dreambody or mental-body, and i had absolutely no sense of where i was. (masters, 2005) “i” was no more, blasted to a confetti cloud by an explosive force i could no longer locate in my head, because it had exploded that too, expanding to become all that there was. whatever this was, it was not a hallucination. a hallucination implies a reality and a point of reference and an entity to have it. none of those things remained […]. and then suddenly the devolution of everything into the nothingness of pure force reverses course. one by one, the elements of our universe begin to reconstitute themselves: the dimensions of time and space returned first, blessing my still-scattered confetti brain with the cozy coordinates of place; this is somewhere! and then i slipped back into my familiar “i” like an old pair of slippers and soon after felt something i recognized as my body begin to reassemble. (pollan, 2018) these reports and many others suggest that at the height of the 5-meo-dmt experience subjects may lose all awareness of their bodies. consequently, they should also lack a sense of bodily ownership and a sense of bodily agency. what about the sense of mental ownership and the sense of mental agency? it is easier to start with the latter: as we have seen, the putative sense of mental agency depends upon the occurrence of mental actions. during an experience as intense, chaotic and uncontrollable as that elicited by 5-meo-dmt, in which it becomes difficult even to think, there should not be much room for mental activity. subjects almost unanimously report on their complete passivity as they go through the experience as if riding a particularly terrifying rollercoaster. a common occurrence in the reports is that one has to “let go” of control over what is happening in one’s mind and “surrender” to the experience.29 thus, it is reasonable to assume that experiences induced by 5-meo-dmt rarely if ever involve a sense of mental agency. it is more difficult to assess whether these experiences involve a sense of mental ownership because the notion is delicate to grasp. are subjects under the influence of 5-meo-dmt aware of their experience as their own? presumably, being aware of one’s experience as one’s own involves having a sense that one is the subject of one’s experience. this seems to require the ability to distinguish between the experience one is having and oneself as its subject. reports of died frequently mention that such states precisely lack a sense that one is having the experience: 29one report from the erowid database contains this characteristic piece of advice: “i would, personally, advise any potential user of this drug to simply be psychologically prepared to let go, and surrender ego and control, and accept that, once its effects are in full motion, it is absolutely impossible to grasp any control of your own life and consciousness, feeling utterly helpless and in vulnerable awe. it will be overwhelming” (#82407). millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 23 it felt as if “i” did no longer exist. there was purely my sensory perception of my environment, but sensory input was not translated into needs, feelings, or acting by “me”. (report cited in millière et al., 2018, p. 10) online reports from erowid also emphasize this form of extreme absorption in the experience induced by 5-meo-dmt, precluding any awareness of oneself as the experience’s subject: there was no self, nothing to separate me from the sensation. (#81888) i had no idea which part of what was happening or what i was seeing was “me”. (#112103) i had previous experience with [other] hallucinogens […] and had always felt able to differentiate and mediate between the experience and the “i” as its subject […]. this time, however, the experience was relentless and psychologically brutalizing. (#49107) these reports cast doubt on the claim that the putative sense of mental ownership could endure in some of the experiences induced by 5-meo-dmt. overall, the evidence reviewed here supports the claim that died, and specifically the kind of experience elicited by 5-meo-dmt, can lack any of the six putative forms of self-consciousness discussed in this article. consequently, such experiences are good candidate examples of total selflessness as i have defined it.30 4.2 meditation some meditation practices aim specifically to induce an experience of total selflessness, in particular so-called non-dual awareness meditation and samādhi meditation. non-dual awareness meditation loosely refers to a family of practices that can be found in several eastern contemplative traditions, including dzoghen and mahāmudrā within tibetan buddhism, and advaita vedanta and kashimiri shaivism within hinduism (dunne, 2011; josipovic, 2010). these practices rely upon the assumption that ordinary conscious experience is structured by a dichotomy between subject and object, which can be momentarily dispelled by expert meditators. such practice is meant to reduce the phenomenal character of one’s experience to a bare minimum, and remove any form of intentionality; in this state, allegedly, “non-dual awareness appears as an empty cognizance, aware and present, but without any thoughts, emotions or perceptions, without a sense of body, space, orientation, time or the usual sense of self” (josipovic, 2019, p. 7). 30another possibility to consider is that all six kinds of self-consciousness might disappear in the relevant drug-induced states, but in succession rather than simultaneously, such that those states would always involve at least one form of self-consciousness. however, i think this interpretation is particularly implausible in light of available reports. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 24 samādhi is a meditative practice from the buddhist tradition that aims at inducing a state of deep absorption (shankman, 2008). the practice of samādhi includes several stages (or in pāli, jhāna), culminating in a “formless” or “objectless” state. the aim of this practice is to reach a kind of “lucid trance” (harvey, 2012, p. 330), a state of pure consciousness which involves a complete loss of self-consciousness: “attaining samādhi is to reach the silent state of pure consciousness where there is no phenomenological content and a loss of any sense of individual self or duality” (wahbeh, sagher, back, pundhir, & travis, 2018). rather than lacking phenomenal character altogether (which could be a contradiction in terms), experiences induced by samādhi might simply lack ordinary conscious contents, including the forms of self-consciousness that might pervade ordinary experience. canonical descriptions of altered states of consciousness from ancient contemplative traditions cannot be taken at face value without further scrutiny. nonetheless, reports from advanced meditation practitioners intriguingly suggest that it is possible to reach states of deep absorption that lack the six putative forms of self-consciousness discussed in section 2.31 thus, samādhi practice is said to induce experiences that should lack cognitive self-consciousness in so far as they are “free from discursive thought” (shankman, 2008, p. 44).32 in such experiences, “all limitations imposed by the concept of ‘space’ are given up” (patel, 2012, p. 352); “awareness of the body falls away, leaving only the experience of limitless space” (shankman, 2008, p. 50). the buddhist monk and expert meditator ajahn brahmavamso further describes the experience of samādhi meditation in the following terms: “the first jhāna is when the body and the five senses have disappeared […] in the second jhāna your will has disappeared. it is not just that you do not want to do anything. you cannot do anything […]. it is quite clear that in jhānas you cannot feel anything to do with the body” (shankman, 2008, pp. 169–173). these anecdotal reports are consistent with reports from controlled studies with advanced meditators describing experiences that lack spatial self-consciousness and bodily awareness (ataria et al., 2015; dor-ziderman, berkovich-ohana, glicksohn, & goldstein, 2013; millière et al., 2018). interestingly, a recent qualitative study of 59 meditation-related experiences described during semi-structured interviews found that three quarters of these experiences involved disruptions of self-consciousness, including “loss of sense of agency” (25%), “loss of sense of basic self” (25%), “change in sense of embodiment” (22%), “change in narrative self” (22%) and “loss of sense of ownership” (18%) (lindahl, fisher, cooper, rosen, & britton, 2017). specifically, the authors report that a number of meditation-related experiences involved “fundamental changes in [the practitioners’] sense of self […] such that they felt like they no longer existed at all” (lindahl et al., 2017, p. 20). admittedly, the hypothesis that some meditation practices can induce states of 31see also metzinger (2020, this issue). 32see also: “crucial to the content-less jhānas are the absence of discursive reasoning and thought, and the cessation of the activities of mind” (patel, 2012, p. 352). millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 25 consciousness entirely lacking self-consciousness remains tentative. it difficult to find detailed descriptions of such states, especially from controlled studies; and the theory-ladenness of traditional descriptions of meditative states might raise some concerns regarding their validity. nevertheless, existing reports warrant further investigation into the phenomenology of specific types of meditation, in particular non-dual awareness and samādhi. 4.3 lucid dreamless sleep it has recently been suggested that conscious episodes may occur during sleep outside of dreams, and that such episodes may lack any kind of self-consciousness (thompson, 2015; windt, 2015b; windt, nielsen, & thompson, 2016). in the context of this hypothesis, dreams are defined as immersive spatio-temporal hallucinations, namely non-veridical perceptual experiences that involve the sense of being present in a simulated environment (windt, 2010). thus, a conscious experience occurring during sleep would qualify as “dreamless” if it lacked the immersive character of dreaming, as well as any specific imagery. the idea that consciousness might persist during dreamless sleep can be traced back to classical discussions in indian philosophy, particularly in the advaita vedānta and yoga schools (thompson, 2015). the modern practice of yoga nidrā (or “yogic sleep”) is claimed to induce a conscious experience of “witnessing sleep” in the absence of dreams (parker, 2019). however, this claim has not been rigorously assessed in controlled studies (parker, 2017). while there is some evidence that lucid dreaming is more frequent in long-term meditators than in meditation-naïve individuals (baird, riedner, boly, davidson, & tononi, 2019), there is no evidence that meditators can remain conscious in sleep outside of dreams, other than anecdotal reports often associated with dubious parapsychological claims (bharati, 2014; although see alcaraz-sanchez, 2019). an interesting avenue of research for the investigation of consciousness during deep sleep is the study of so-called “white dreams” in the laboratory. white dreams refer to cases in which subjects waking from sleep report the feeling of having been dreaming prior to awakening, but are unable to recall any details about these putative dreams. as many as a third of post-awakening reports in serial-awakening studies describe white dreams (siclari, larocque, postle, & tononi, 2013). according to the standard interpretation of such cases, subjects do report the occurrence of contentful dreams whose contents they simply forgot upon awakening. however, it has been suggested that a subset of such reports might accurately describe conscious episodes occuring during dreamless sleep, without sensory or propositional content. on this interpretation, “subjects experience only a minimal form of conscious presence with no narrative structure, and no specific percepts, bodily sensations or thoughts occurring during such experiences” (fazekas, nemeth, & overgaard, 2019, p. 85). this kind of conscious episode might occur specifically during the transition from unconscious sleep to immersive dreaming (windt et al., 2016). millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 26 interestingly, there are many anecdotal reports from lucid dreamers suggesting that some episodes of lucidity during sleep can be completely devoid of imagery (bogzaran, 2003; hurd, 2008; magallon, 1987). these episodes are often reported to occur either at the onset of lucid dreaming or during the transition between the end of one contentful dream and the beginning of the next one. anecdotal reports often describe these lucid episodes as the experience of a “void”: i found myself in an empty void […]. i had no body and my consciousness was sort of floating there.33 the void as i’ve experienced it is a state beyond self-concept and selfawareness. it exists as a sense of pure awareness without any perceiver (as such), no sense of personal self or egoic point of focus […]. within the void state of a lucid dream, “i” ceased to exist as “myself” […] without any sense of “me” having the experience.34 whether or not such experiences should be considered to be (minimalistic) lucid dreams rather than lucid dreamless sleep is debatable, since some descriptions suggest that they retain the immersive character of dreams (see windt, 2015a, chapter 11). if subjects do feel “immersed” within an empty space with no visual features, one may also wonder whether they might retain some form of spatial self-consciousness during these episodes. while reported experiences of the “void” involve no depth cues and often lack bodily sensations, it is possible that subjects still have a sense of the spatial structure of their environment and their location within it. yet it seems that at least some of these reports describe conscious episodes that lack any kind of spatial content. windt has argued on the basis of similar reports that such experiences might be characterized by “pure subjective temporality”, meaning that their phenomenal character would be reduced to the mere phenomenology of “now-ness” and duration (windt, 2015b). to the extent that the relevant conscious episodes lack propositional thoughts, sensory content and spatial content, they should lack all the putative forms of self-consciousness distinguished in section 2. while anecdotal reports offer intriguing clues about the range of conscious experiences that can occur during sleep, it is important to note that the occurrence of lucid dreamless sleep has not yet been systematically investigated in the laboratory. future research using a serial-awakening paradigm with subjects who report frequent lucid episodes without imagery during sleep might shed further light on whether it is possible to remain conscious during dreamless sleep. 33report retrieved from https://www.reddit.com/r/luciddreaming/comments/4l1nr9/found_som e_kind_of_black_void_while_wilding. 34report retrieved from https://www.reddit.com/r/luciddreaming/comments/nuam2/trapped_i n_your_own_mind_and_the_void. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.reddit.com/r/luciddreaming/comments/4l1nr9/found_some_kind_of_black_void_while_wilding https://www.reddit.com/r/luciddreaming/comments/4l1nr9/found_some_kind_of_black_void_while_wilding https://www.reddit.com/r/luciddreaming/comments/nuam2/trapped_in_your_own_mind_and_the_void https://www.reddit.com/r/luciddreaming/comments/nuam2/trapped_in_your_own_mind_and_the_void https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 27 4.4 objections and replies i have discussed so far tentative evidence that totally selfless states of consciousness may actually occur under various conditions. given that the relevant reports remain anecdotal, however, they may not be sufficient to convince proponents of the claim that self-consciousness is necessary for consciousness. to address this concern, i will now consider four objections to the possibility and reportability of totally selfless states of consciousness. objection 1. there can be no experience of total selflessness, because there is no experience without a subject of experience. this objection appears to rest upon a category mistake. the claim that “there is no experience without a subject of experience” (frege, 1921/1956, p. 299, translation modified) pertains to the metaphysics of experience, and does not have implications for phenomenology: it states that every experience has an experiencing subject as its bearer. this claim does not entail that the subject of a conscious experience need be conscious of herself as herself. to say that experiences can be totally selfless, as i have previously defined this notion, is not to say that experiences can be “free-floating” and exist without a subject of experience. the relevant notion of selflessness pertains to the absence of self-consciousness – construed as a phenomenal notion – rather than to the absence of a subject of experience. objection 2. there can be no experience of total selflessness, because it would be an experience devoid of content, and no experience can be contentless. there is a lingering ambiguity about the notion of content in this objection. the notion of “phenomenal content” is occasionally used as a synonym for the phenomenal character of an experience, or what it is like to have that experience (mcginn, 1989, p. 66). it is trivially true that there can be no conscious experience without content in that sense, insofar as any conscious experience is such that there is something it is like to have it. consequently, a conscious episode lacking any kind of self-consciousness should still have phenomenal character. admittedly, some of the tentative examples of totally selfless states i have discussed are meant to involve extreme forms of absorption and sensory gating, resulting in a dramatically impoverished phenomenology compared to ordinary conscious experience. nonetheless, they must have a phenomenal character in order to be conscious states at all. there is nothing inconsistent in the idea of a totally selfless state that has phenomenal character, because consciousness does not constitutively involve consciousness of oneself as oneself. secondly, the notion of content commonly refers to intentional content rather than phenomenal character. whether or not a conscious experience can completely lack intentional content is a controversial question. it is true that the descriptions of some states induced by non-dual awareness meditation or samādhi, millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 28 for example, suggest that such states lack intentional content altogether. but even if this characterization were inaccurate – even if there could be no conscious experience without intentional content – it would not entail that the relevant states should involve a form of self-consciousness. there is no reason why the possibility of totally selfless states should rest upon the possibility of conscious states without intentional content. totally selfless states as i have defined them – states of consciousness lacking any form of selfconsciousness – need not be “minimal phenomenal experiences” defined as experiences lacking intentionality (metzinger, 2018; windt, 2015b). interestingly, some descriptions of drug-induced ego dissolution involve rich sensory phenomenology, including intense visual hallucinations that presumably have intentional content (millière, 2017). consider also this intriguing report of an acute psychotic episode involving perceptual content: consciousness gradually loses its coherence. one’s center gives way. the center cannot hold. the “me” becomes a haze, and the solid center from which one experiences reality breaks up like a bad radio signal. there is no longer a sturdy vantage point from which to look out […]. no core holds things together […]. random moments of time follow one another. sights, sounds, thoughts, and feelings don’t go together. (saks, 2007, p. 13) while this anecdotal report is too coarse-grained to assess whether it describes an episode that plausibly lacks any form of self-consciousness, it is least conceivable that a similar experience with intentional content could be totally selfless. objection 3. there is a form of self-consciousness that is more basic than any of the six notions of self-consciousness discussed in this article and is preserved in allegedly ”selfless” states. this is an important objection that deserves careful consideration. among authors who argue that there are several forms of self-consciousness, many consider that these features can be ordered such that there is a most minimal or most basic form of self-consciousness. one can find a number of different labels referring to such a notion in the literature, including “minimal self” (zahavi, 2010), “minimal phenomenal selfhood” (blanke & metzinger, 2009) and “basic self-awareness” (billon, 2017a). what it means for a form of self-consciousness to be more “minimal” or “basic” than any other is not always explicitly spelled out, although it generally means that such a feature is the most ubiquitous form of self-consciousness in conscious experience. some authors argue that the most minimal form of self-consciousness is one of the six features distinguished in section 2. this is the case for proponents of a version of the necessity claim (nc) with respect to one (but not all) of these features: if a form of self-consciousness is necessary for consciousness in general, then it is more basic than any other form of self-consciousness that may exist. some propomillière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 29 nents of the typicality claim (tc) also hold that one of the six relevant features is more pervasive the others. for example, metzinger (2013b) and windt (2015a) argue that spatial self-consciousness is the most minimal form of self-consciousness, because it remains present even when all the other forms of self-consciousness are missing – although spatial self-consciousness can itself go missing in totally selfless states of consciousness.35 others suggest, however, that the most minimal form of self-consciousness is not reducible to any of the six notions distinguished in section 2; instead, they argue that it is a distinct notion variously referred to as “pre-reflexive self-awareness” (zahavi, 1999, 2005, 2014) or “for-me-ness” (kriegel, 2009; zahavi & kriegel, 2016). this notion denotes the subject’s awareness of her occurrent experience rather the subject’s awareness of herself (as herself). to mark this distinction, this notion is occasionally referred to as a non-egological conception of self-consciousness (gurwitsch, 1941). zahavi and kriegel themselves acknowledge that for-me-ness should not be conflated with self-consciousness as it is typically defined: for some, to be self-conscious is to think of oneself as oneself, or to be aware of one’s states or features as one’s own […]. clearly, on such understandings of the term “self-consciousness”, it would be quite implausible to suggest that all phenomenal consciousness involves prereflective self-consciousness. it should be clear, however, that this is not how we understand the notion of an experiential for-me-ness.36 (zahavi & kriegel, 2016, p. 40) accordingly, the objection that every experience involves self-consciousness in the form of for-me-ness appears to stem, at least in part, from a verbal dispute. even if we grant the claim that pre-reflective self-consciousness or for-me-ness is constitutive of consciousness, it does not follow that totally selfless states of consciousness – as i have defined them – are impossible. a subject could be aware of her occurrent conscious experience (i.e., have for-me-ness), yet fail to be aware of herself as herself (guillot, 2017). such a subject would be in a totally selfless state, defined as a state lacking any form of self-consciousness sensu stricto, i.e. egological self-consciousness.37 perhaps one could object that there is yet another notion of self-consciousness, distinct both from for-me-ness and from the six notions outlined in section 2, that is instantiated even in alleged examples of totally selfless states.38 specifically, one 35for a criticism of the claim that spatial self-consciousness is always instantiated whenever any other form of self-consciousness is instantiated, see millière (2019). 36see also zahavi (2018, p. 706): “being aware of one’s experiences when they occur is [not] tantamount to being aware of oneself as a (secondary) object [of one’s experience]”. 37in my opinion, referring to the non-egological notion as “self-consciousness” is somewhat misleading, because this term naturally refers to the subject’s consciousness of herself – just as “time consciousness” and “bodily consciousness” typically refer to the subject’s consciousness of time and of her body respectively. 38due to lack of space, i have left out of my discussion notions that are occasionally presented millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 30 could argue that there is after all such a thing as a sui generis qualitative property of selfhood that is more basic than the specific features discussed so far and irreducible to any of them. i find this proposal implausible for two reasons. firstly, the notion of a sui generis qualitative property of selfhood seems utterly mysterious. as i pointed out at the start of this paper, there is a widespread agreement that no such feature can be found through introspection. part of the problem is that it is unclear what one should look for if one were attempting to detect the presence of a sui generis qualitative property of selfhood in one’s experience. secondly, even if it were possible to detect such a feature in ordinary conscious experience, there would be no reason to think that the tentative examples of selfless states of consciousness i have discussed instantiate it. indeed, many of the relevant reports are quite explicit about the fact that the described experiences lack any feature that could be described as a sense of self. consider, for example, these additional reports of experiences induced by 5-meo-dmt, retrieved from erowid: i was completely disassociated from the “real world” and [from] any sense of self. it was the most jarring feeling. (#107905) it is a complete annihilation of self […]. i was absolutely nothing but a sensory perceiver, stuck within the split seconds that were eternity. (#18198) it felt as if all of the atoms of the molecules that typically form my physical self simply dispersed, and even my sense of self, or ego, vanished […]. (#56384) i wasn’t me any longer. there was no me. there was no ego. (#27601) the interpretation of such anecdotal reports can be debated. nonetheless, it is prima facie implausible that the reported experiences did involve a sui generis qualitative property of selfhood to which subjects were oblivious. given that there is no obvious evidence for the claim that all experiences – including tentative examples of totally selfless states – involve a sui generis qualitative property of selfhood, the burden of proof is on proponents of this claim. as “social” or “interpersonal” forms of self-consciousness (e.g. zahavi, 2014). however, these notions often seem reducible to one of the six forms of self-consciousness i have distinguished. for example, the conscious thought that i am less proficient in chinese language than someone else is an instance of cognitive self-consciousness. one notable exception might be the category of so-called self-conscious emotions (such as shame, pride and guilt), which does not seem reducible to any of the six features discussed in this article (sznycer, 2019); but it is very plausible that such emotions are lacking from most states of consciousness. (thanks to dan zahavi for raising this point.) millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 31 objection 4. there can be no autobiographical memory of totally selfless states, therefore reports of such states are suspect. this line of objection can be found in the following passage from metzinger:39 autophenomenological reports given by human beings about selfless states […] will typically not impress philosophers much, because they contain an inherent logical fallacy: how can you coherently report about a selfless state of consciousness from your own, autobiographical memory? (metzinger, 2003, p. 566) let us unpack this remark. to begin with, it is not obvious that the relevant reports contain a logical fallacy. take the report quoted above: “there was no me. there was no ego.” this report would be self-contradictory if it could be paraphrased as “i had an experience without a subject of experience” or even “i had an experience that i did not have”. however, this is clearly not the intended meaning of the report. rather, by “me” and “ego”, the author of the report appears to refer within the context of the quote to a phenomenal feature, namely to some form of self-consciousness (or indeed, to any form of self-consciousness).40 this is more obvious in other reports quoted above, such as “my sense of self, or ego, vanished”. however, the primary target of metzinger’s remark is not the logical consistency of the relevant reports per se, but their status as recollections of past experiences. specifically, the comment suggests that reporting a past conscious experience as totally selfless on the basis of an autobiographical memory is a form of confabulation. this suggestion seems premised upon the assumption that a conscious episode must involve some form of self-consciousness in order to be stored and later retrieved as an autobiographical memory. while a systematic discussion of the relationship between memory and self-consciousness is beyond the scope of this paper, i will succinctly argue that this assumption is unfounded.41 first of all, there is no agreement on how to define autobiographical memory, or how it relates to the standard taxonomy of declarative memory. for some, autobiographical memory forms a subset of episodic memory, namely memory of particular events from one’s past (bermúdez, 2017), while others argue that autobiographical memory also includes a subset of semantic memory, namely general knowledge about oneself (conway, 2005). in any case, retrospective reports of totally selfless states present themselves as reports of episodic rather than semantic memories, insofar as they describe a specific conscious episode from the subject’s past. 39as far as i know, the gist of the objection dates back to much older debates in buddhist philosophy (chadha, 2019; robins, 2017). 40compare this with a subject reporting that “there was no time” during an intense meditation practice: clearly, such report would not be intended to mean that the relevant experience did not unfold in time, but rather that it did not include a phenomenology of duration. 41see bermúdez (2017) for an overview of the relationship between memory and self-consciousness. see also fink (2020, this issue). millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 32 however, there is a further disagreement as to whether all episodic memories are autobiographical.42 thus, bermúdez (2017) argues that remembering an event such as the decisive goal of a football match by recalling an image seen on television is an instance of non-autobiographical episodic memory, because it is not specifically about oneself and one’s past (although it originates in an episode of one’s personal history). along the same lines, one could attempt to fend off metzinger’s concern by arguing that reports of totally selfless states are recollections of episodic but not autobiographical memory. the success of this strategy depends on how one cashes out the criterion that autobiographical memory be specifically about oneself and one’s past. while putative memories of totally selfless states do originate in past conscious episodes, one might argue that they are not specifically about oneself in the relevant sense, because the recollected conscious episodes allegedly lacked any form of self-consciousness. regardless of how autobiographical memory is defined, the crux of metzinger’s concern seems to be the fact that putative memories of totally selfless states are reported in the first person, as memories of experiences that one lived through – even though such experiences supposedly did not involve any awareness of oneself. there are several ways to address this concern. first, it is worth noting that a number of reports of putative selfless states precisely avoid using the first-person pronoun, or explicitly point out its inadequacy:43 there existed no one, not even me […] so would it be proper to still speak of “i”, even as the notion of “i” seemed so palpably illusory? (erowid report #42740, psychoactive compound: psilocybin) everything is slipping away […] and then i’m gone […]. i use the word “i” after-the-fact. at the time, i was no longer aware of my separate existence […]. (#95305, psychoactive compound: ketamine) [k]eep in mind, at this point, [that] i wasn’t aware of my own existence, so “i” isn’t the best word to use. (#32064, psychoactive compound: 4-acetoxy-dipt) [t]he use of the pronoun “i” to describe the experience is completely inappropriate […]. (#61129, psychoactive compound: salvinorin a) while many reports do use the first-person pronoun, such use need not be interpreted as evidence that subjects were self-conscious during the recollected experience. rather, grammatical conventions dictate that using the first-person pronoun to report episodes of one’s past is more appropriate than using convoluted impersonal constructions. on most accounts of episodic memory, at least three distinct mental states are involved in the process that leads from the occurrence of a con42for example, hoerl (1999) endorses the claim that all episodic memories are autobiographical. 43billon also points out that “patients suffering from severe forms of depersonalization, can express reluctance to use the first-person and even, in delusional cases, avoid it altogether, referring to themselves in the third-person exclusively” billon (2017a, p. 1). millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 33 scious experience to the retrospective report of that experience: (a) the original experience that is reported upon, (b) the representation of that experience retrieved during recollection, and (c) the recollective judgement expressed by the report of the experience upon retrieval of its representation. importantly, the occurrence of the first-person pronoun in recollective judgements cannot count as evidence that the mnemonic representation itself involves first-person content. a further possibility to consider is that the mnemonic representation of putative selfless states might involve first-person content even though the original experience did not. this possibility is consistent with simulation theories of episodic memory, according to which mnemonic representations are (at least partially) constructed during recollection through a process that bears many similarities to imagination (michaelian, 2016). on such a view, the simulation of the past conscious episode during retrieval might introduce a form of self-representation that was not present in the original experience, simply because it is difficult in the ordinary wakeful state to imagine undergoing a conscious experience lacking selfconsciousness altogether. furthermore, it has been suggested that the construction of episodic memory involves the enrichment of episodic memory traces with semantic information (cheng, werning, & suddendorf, 2016). given the subjects’ knowledge that they themselves underwent the recollected episode, it is perhaps not surprising that the mnemonic representation of the episode resulting from semantic enrichment might contain some form of self-representation. on this interpretation, some reports of totally selfless states that prominently feature the first-person pronoun might be instances of misremembering rather than confabulation, where the main or only inaccuracy is the post-hoc introduction of self-representation within the mnemonic representation.44 importantly, this would not entail that the relevant reports cannot be used to inform research on totally selfless states. indeed, if the recollection of totally selfless states is systematically biased by structural features of episodic memory construction, then we might be able to abstract away from this systematic bias, and ignore the incongruity of first-person pronoun use in retrospective reports. in conclusion, it is far from obvious that experiences lacking self-consciousness cannot be remembered as experiences that one has undergone. if such experiences really occur, it is not implausible that they could be reported post hoc. therefore, we should not dismiss putative reports of such experiences as incoherent or unreliable in principle without further assessment of the accuracy of the recollection process in each individual case. 5 conclusion is self-consciousness ubiquitous in conscious experience? to answer this question without engaging in merely verbal disagreements, it is important to disambiguate 44see robins (2016) on the difference between remembering, confabulating and misremembering. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 34 different notions of self-consciousness. starting from the common definition of self-consciousness as consciousness of oneself as oneself, i distinguished six candidate phenomenal features that can qualify as instances of self-consciousness in that sense. i subsequently argued that none of these phenomenal features is necessary for consciousness in general, because there are examples of states of consciousness that credibly lack each of them. such states of consciousness are at least partially selfless, insofar as they lack at least one of the ways in which one could be conscious of oneself as oneself. showing that each of the six putative forms of self-consciousness can be individually missing in conscious experience is not sufficient to rule out all versions of claim that self-consciousness is necessary for consciousness (nc). indeed, there are a number of disjunctive versions of this claim according to which some form of self-consciousness or other is necessary for consciousness, including the maximally disjunctive claim ncdisj: ncdisj for any subject s, if s is conscious at t, then s’s phenomenology involves any of the phenomenal features (1–6) at t (including any combination of these features). against all versions of nc – including ncdisj – i suggested that some states of consciousness might be totally selfless, insofar as they do not involve any form of self-consciousness. not only are such states conceptually possible, but they also plausibly occur in some actual conditions. the most promising evidence for the existence of totally selfless states of consciousness arguably comes from reports of certain drug-induced states, although there is also very preliminary evidence that such states may occur during certain forms of meditation practice and during dreamless sleep. given that a significant part of the relevant evidence remains anecdotal, more empirical research is needed to assess the hypothesis that these states of consciousness really lack all forms of self-consciousness. in particular, reports obtained from questionnaires and semi-structured interviews in controlled conditions could offer more robust evidence regarding the phenomenology of the candidate selfless states. nonetheless, available evidence speaks to the plausibility of totally selfless states. furthermore, common objections against the possibility and reportability of such states are misguided. the upshot of this discussion is twofold. firstly, proponents of the existence of a sense of self in conscious experience should be explicit about the nature of the target phenomenon. specifically, they should determine whether the relevant notion of “sense of self” refers to one of the six putative forms of self-consciousness i have distinguished, to a disjunction between several of them and/or their possible combinations, or to an entirely distinct phenomenal feature. secondly, if their target is indeed one of these six notions or any disjunction between them, they should restrict the scope of their claim to the ordinary experience of healthy, neurotypical adult humans – i.e., to some version of the typicality claim (tc) –, because the corresponding versions of the necessity claim (nc) are conceptually and empirically implausible. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 35 acknowledgments i wish to thank the audience of the selfless minds workshop in frankfurt, as well as dan zahavi and two anonymous referees for their comments on a previous version of this article. thanks are also due to thomas metzinger, jennifer windt and sascha fink for their editorial comments. finally, i am grateful to martin davies for comments on previous related materials. references alcaraz-sanchez, a. (2019). conscious states during sleep. a philosophical and psychological exploration (master’s thesis). university of edinburgh. allison-bolger, v. y. (1999). collection of case histories. unpublished typescript. alvarado, c. s. (2000). out-of-body experiences. in e. cardeña, s. j. lynn, & s. krippner (eds.), varieties of anomalous experiences: examining the scientific evidence (pp. 183–218). washington, dc: american psychological association. anderson, e. w., & rawnsley, k. (1954). clinical studies of lysergic acid diethylamide. european neurology, 128(1-2), 38–55. https://doi.org/10.1159/000139775 ataria, y., dor-ziderman, y., & berkovich-ohana, a. (2015). how does it feel to lack a sense of boundaries? a case study of a long-term mindfulness meditator. consciousness and cognition, 37, 133–147. https://doi.org/10.1016/j.concog.2015.09.002 baird, b., riedner, b. a., boly, m., davidson, r. j., & tononi, g. (2019). increased lucid dream frequency in long-term meditators but not following mindfulness-based stress reduction training. psychology of consciousness: theory, research, and practice, 6(1), 40–54. https://doi.org/10.1037/cns0000176 barsuglia, j., davis, a. k., palmer, r., lancelotta, r., windham-herman, a.-m., peterson, k., et al. (2018). intensity of mystical experiences occasioned by 5-meo-dmt and comparison with a prior psilocybin study. frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018.02459 bayne, t. (2008). the phenomenology of agency. philosophy compass, 3(1), 182–202. https://doi.org/10.1111/j.1747-9991.2007.00122.x bayne, t. (2010). the unity of consciousness. oxford: oxford university press. bayne, t., & spener, m. (2010). introspective humility. philosophical issues, 20(1), 1–22. https://doi.org/10.1111/j.1533-6077.2010.00176.x bercel, n. a., travis, l. e., olinger, l. b., dreikurs, e., & polos, m. g. (1956). model psychoses induced by lsd-25 in normals: i. psychophysiological investigations, with special reference to the mechanism of the paranoid reaction. a.m.a. archives of neurology & psychiatry, 75(6), 588–611. https://doi.org/10.1001/archneurpsyc.1956.02330240026003 beringer, k. (1927). der meskalinrausch: seine geschichte und erscheinungsweise. berlin: springer. berkovich-ohana, a., dor-ziderman, y., glicksohn, j., & goldstein, a. (2013). alterations in the sense of time, space, and body in the mindfulness-trained brain: a neurophenomenologically-guided meg study. frontiers in psychology, 4. https://doi.org/10.3389/fpsyg.2013.00912 bermudez, j. l. (1998). the paradox of self-consciousness. cambridge, ma: mit press. bermúdez, j. l. (2001). the sources of self-consciousness. proceedings of the aristotelian society, 102(1), 87–107. https://doi.org/10.1111/j.0066-7372.2003.00044.x bermúdez, j. l. (2011). bodily awareness and self-consciousness. in s. gallagher (ed.), the oxford handbook of the self (pp. 157–179). oxford: oxford university press. bermúdez, j. l. (2017). memory and self-consciousness. in k. michaelian & bernecker (eds.), the routledge handbook of philosophy of memory (pp. 180–191). london: routledge. bharati, s. v. (2014). my experiments with yoga nidra. rishikesh, india: himalayan yoga publications trust. billon, a. (2013). does consciousness entail subjectivity? the puzzle of thought insertion. philosophical psychology, 26(2), 291–314. https://doi.org/10.1080/09515089.2011.625117 billon, a. (2016). making sense of the cotard syndrome: insights from the study of depersonalisation. mind & language, 31(3), 356–391. https://doi.org/10.1111/mila.12110 billon, a. (2017a). basic self-awareness. european journal of philosophy, 25(3), 732–763. https://doi.org/10.1111/ejop.12168 billon, a. (2017b). mineness first: three challenges to the recent theories of the sense of bodily ownership. in f. de vignemont & a. alsmith (eds.), the subject’s matter: self-consciousness and the body (pp. 189–216). cambridge, ma: mit press. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1159/000139775 https://doi.org/10.1016/j.concog.2015.09.002 https://doi.org/10.1037/cns0000176 https://doi.org/10.3389/fpsyg.2018.02459 https://doi.org/10.1111/j.1747-9991.2007.00122.x https://doi.org/10.1111/j.1533-6077.2010.00176.x https://doi.org/10.1001/archneurpsyc.1956.02330240026003 https://doi.org/10.3389/fpsyg.2013.00912 https://doi.org/10.1111/j.0066-7372.2003.00044.x https://doi.org/10.1080/09515089.2011.625117 https://doi.org/10.1111/mila.12110 https://doi.org/10.1111/ejop.12168 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 36 billon, a., & kriegel, u. (2015). jaspers’ dilemma: the psychopathological challenge to subjectivity theories of consciousness. in r. j. gennaro (ed.), disturbed consciousness: new essays on psychopathology and theories of consciousness (pp. 29–54). cambridge, ma: mit press. blanke, o., & metzinger, t. (2009). full-body illusions and minimal phenomenal selfhood. trends in cognitive sciences, 13(1), 7–13. https://doi.org/10.1016/j.tics.2008.10.003 block, n. (1995). on a confusion about a function of consciousness. brain and behavioral sciences, 18(2), 227–247. https://doi.org/10.1017/s0140525x00038188 block, n. (2007). consciousness, accessibility, and the mesh between psychology and neuroscience. the behavioral and brain sciences, 30(5-6), 481–499; discussion 499–548. https://doi.org/10.1017/s0140525x07002786 bogzaran, f. (2003). lucid art and hyperspace lucidity. dreaming, 13(1), 29–42. https://doi.org/10.1023/a:1022186217703 bradshaw, m. (2016). a return to self: depersonalization and how to overcome it. seattle, wa: amazon services international. cassam, q. (1997). self and world. oxford: oxford university press. chadha, m. (2018). no-self and the phenomenology of ownership. australasian journal of philosophy, 96(1), 14–27. https://doi.org/10.1080/00048402.2017.1307236 chadha, m. (2019). reconstructing memories, deconstructing the self. mind & language, 34(1), 121–138. https://doi.org/10.1111/mila.12204 chalmers, d. j. (1996). the conscious mind: in search of a fundamental theory. oxford: oxford university press. cheng, s., werning, m., & suddendorf, t. (2016). dissociating memory traces and scenario construction in mental time travel. neuroscience & biobehavioral reviews, 60, 82–89. https://doi.org/10.1016/j.neubiorev.2015.11.011 chisholm, r. m. (1969). on the observability of the self. philosophy and phenomenological research, 30(1), 7–21. https://doi.org/10.2307/2105917 cicogna, p., & bosinelli, m. (2001). consciousness during dreams. consciousness and cognition, 10(1), 26–41. https://doi.org/10.1006/ccog.2000.0471 conway, m. a. (2005). memory and the self. journal of memory and language, 53(4), 594–628. https://doi.org/10.1016/j.jml.2005.08.005 csikszentmihalyi, m. (1990). flow: the psychology of optimal experience. new york: harper collins. dainton, b. (2008). the phenomenal self. oxford: oxford university press. dainton, b. (2016). the sense of self. aristotelian society supplementary volume, 90(1), 113–143. https://doi.org/10.1093/arisup/akw007 damasio, a. (1999). the feeling of what happens: body and emotion in the making of consciousness. san diego, ca: houghton mifflin harcourt. dammeyer, j. (2014). deafblindness: a review of the literature. scandinavian journal of public health, 42(7), 554–562. https://doi.org/10.1177/1403494814544399 d’argembeau, a. (2018). mind-wandering and self-referential thought. in k. c. r. fox & k. christoff (eds.), the oxford handbook of spontaneous thought (pp. 181–191). oxford: oxford university press. davis, a. k., barsuglia, j. p., lancelotta, r., grant, r. m., & renn, e. (2018). the epidemiology of 5-methoxy-n,n-dimethyltryptamine (5-meo-dmt) use: benefits, consequences, patterns of use, subjective effects, and reasons for consumption. journal of psychopharmacology, 32(7), 779–792. https://doi.org/10.1177/0269881118769063 deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 de vignemont, f. (2011). bodily immunity to error. in s. prosser & f. recanati (eds.), immunity to error through misidentification: new essays (pp. 1–27). cambridge: cambridge university press. de vignemont, f. (2018). mind the body: an exploration of bodily self-awareness. oxford: oxford university press. dor-ziderman, y., berkovich-ohana, a., glicksohn, j., & goldstein, a. (2013). mindfulness-induced selflessness: a meg neurophenomenological study. frontiers in human neuroscience, 7. https://doi.org/10.3389/fnhum.2013.00582 duncan, m. (2019). the self shows up in experience. review of philosophy and psychology, 10(2), 299–318. https://doi.org/10.1007/s13164-017-0355-2 dunne, j. (2011). toward an understanding of non-dual mindfulness. contemporary buddhism, 12(1), 71–88. https://doi.org/10.1080/14639947.2011.564820 evans, g. (1982). the varieties of reference. oxford: oxford university press. fazekas, p., nemeth, g., & overgaard, m. (2019). white dreams are made of colours: what studying contentless dreams can teach about the neural basis of dreaming and conscious experiences. sleep medicine reviews, 43, 84–91. https://doi.org/10.1016/j.smrv.2018.10.005 millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.tics.2008.10.003 https://doi.org/10.1017/s0140525x00038188 https://doi.org/10.1017/s0140525x07002786 https://doi.org/10.1023/a:1022186217703 https://doi.org/10.1080/00048402.2017.1307236 https://doi.org/10.1111/mila.12204 https://doi.org/10.1016/j.neubiorev.2015.11.011 https://doi.org/10.2307/2105917 https://doi.org/10.1006/ccog.2000.0471 https://doi.org/10.1016/j.jml.2005.08.005 https://doi.org/10.1093/arisup/akw007 https://doi.org/10.1177/1403494814544399 https://doi.org/10.1177/0269881118769063 https://doi.org/10.33735/phimisci.2020.i.39 https://doi.org/10.3389/fnhum.2013.00582 https://doi.org/10.1007/s13164-017-0355-2 https://doi.org/10.1080/14639947.2011.564820 https://doi.org/10.1016/j.smrv.2018.10.005 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 37 feinstein, j. s., khalsa, s. s., yeh, h. w., wohlrab, c., simmons, w. k., stein, m. b., & paulus, m. p. (2018). examining the short-term anxiolytic and antidepressant effect of floatation-rest. plos one, 13(2), e0190292–e0190292. https://doi.org/10.1371/journal.pone.0190292 fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 flanagan, o. j. (1992). consciousness reconsidered. cambridge, ma: mit press. fox, k. c. r., & cahn, b. r. (forthcoming). meditation and the brain in health and disease. in m. farias, d. brazier, & m. lalljee (eds.), the oxford handbook of meditation. oxford: oxford university press. frege, g. (1956). the thought: a logical inquiry (a. m. quinton & m. quinton, trans.). mind, 65(259), 289–311. retrieved from https://www.jstor.org/stable/2251513 (original work published 1921) friston, k. (2018). am i self-conscious? ( or does self-organization entail self-consciousness?). frontiers in psychology, 9. https://doi.org/10.3389/fpsyg.2018.00579 gallagher, s. (2000). philosophical conceptions of the self: implications for cognitive science. trends in cognitive sciences, 4(1), 14–21. https://doi.org/10.1016/s1364-6613(99)01417-5 gallagher, s. (2010). defining consciousness : the importance of non-reflective self-awareness. pragmatics and cognition, 18(3), 561–569. gallagher, s., & zahavi, d. (2019). phenomenological approaches to self-consciousness. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (summer 2019). metaphysics research lab, stanford university. garcía-carpintero, m. (2015). de se thought. https://doi.org/10.1093/oxfordhb/9780199935314.013.61 guillot, m. (2016). thinking of oneself as the thinker: the concept of self and the phenomenology of intellection. philosophical explorations, 19(2), 138–160. https://doi.org/10.1080/13869795.2016.1176232 guillot, m. (2017). i me mine: on a confusion concerning the subjective character of experience. review of philosophy and psychology, 8(1), 23–53. https://doi.org/10.1007/s13164-016-0313-4 gurwitsch, a. (1941). a non-egological conception of consciousness. philosophy and phenomenological research, 1(3), 325–338. https://doi.org/10.2307/2102762 guttmann, e., & maclay, w. s. (1937). mescaline and depersonalization. journal of neurology and psychopathology, 16(63), 193–212. haggard, p. (2017). sense of agency in the human brain. nature reviews neuroscience, 18(4), 196–207. https://doi.org/10.1038/nrn.2017.14 harvey, p. (2012). an introduction to buddhism: teachings, history and practices (2nd edition). new york: cambridge university press. hasenkamp, w. (2018). catching the wandering mind: meditation as a window into spontaneous thought. in k. c. r. fox & k. christoff (eds.), the oxford handbook of spontaneous thought (pp. 539–553). oxford: oxford university press. hobson, j. a. (1988). the dreaming brain (1st edition). new york: basic books. hoerl, c. (1999). memory, amnesia and the past. mind & language, 14(2), 227–251. https://doi.org/10.1111/1468-0017.00111 hohwy, j. (2007). the sense of self in the phenomenology of agency and perception. psyche, 13(1), 1–20. hopkins, g. m. (1959). the sermons and devotional writings of gerard manley hopkins (c. devlin, ed.). oxford: clarendon press. horgan, t. e., tienson, j. l., & graham, g. (2003). the phenomenology of first-person agency. in s. walter & h.-d. heckmann (eds.), physicalism and mental causation (p. 323). exeter, uk: imprint academic. howell, r. j. (2006). self-knowledge and self-reference. philosophy and phenomenological research, 72(1), 44–70. https://doi.org/10.1111/j.1933-1592.2006.tb00490.x howell, r. j., & thompson, b. (2017). phenomenally mine: in search of the subjective character of consciousness. review of philosophy and psychology, 8(1), 103–127. https://doi.org/10.1007/s13164-016-0309-0 hurd, r. (2008). exploring the void in lucid dreaming. retrieved from https://dreamstudies.org/2010/05/13/exploring-the-void-in-lucid-dreaming/ hurley, s. l. (1998). consciousness in action. cambridge, ma: harvard university press. høffding, s., & montero, b. g. (2019). not being there: an analysis of expertise-induced amnesia. mind & language, 1–20. https://doi.org/10.1111/mila.12260 jackson, s. a., & eklund, r. c. (2004). the flow scales manual. morgantown, wv: fitness information technology. jackson, s. a., & marsh, h. w. (1996). development and validation of a scale to measure optimal experience: the flow state scale. journal of sport and exercise psychology, 18(1), 17–35. https://doi.org/10.1123/jsep.18.1.17 janet, p. (1908). le sentiment de dépersonnalisation. journal de psychologie normale et pathologique, 5, 514–516. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1371/journal.pone.0190292 https://doi.org/10.33735/phimisci.2020.i.40 https://www.jstor.org/stable/2251513 https://doi.org/10.3389/fpsyg.2018.00579 https://doi.org/10.1016/s1364-6613(99)01417-5 https://doi.org/10.1093/oxfordhb/9780199935314.013.61 https://doi.org/10.1080/13869795.2016.1176232 https://doi.org/10.1007/s13164-016-0313-4 https://doi.org/10.2307/2102762 https://doi.org/10.1038/nrn.2017.14 https://doi.org/10.1111/1468-0017.00111 https://doi.org/10.1111/j.1933-1592.2006.tb00490.x https://doi.org/10.1007/s13164-016-0309-0 https://dreamstudies.org/2010/05/13/exploring-the-void-in-lucid-dreaming/ https://doi.org/10.1111/mila.12260 https://doi.org/10.1123/jsep.18.1.17 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 38 josipovic, z. (2010). duality and nonduality in meditation research. consciousness and cognition, 19(4), 1119–1121. https://doi.org/10.1016/j.concog.2010.03.016 josipovic, z. (2019). nondual awareness: consciousness-as-such as non-representational reflexivity. progress in brain research, 244, 273–298. https://doi.org/10.1016/bs.pbr.2018.10.021 kaplan, d. (1989). demonstratives. in j. almog, j. perry, & h. wettstein (eds.), themes from kaplan (pp. 481–563). oxford: oxford university press. kjellgren, a., buhrkall, h., & norlander, t. (2010). psychotherapeutic treatment in combination with relaxation in a flotation tank: effects on ”burn-out syndrome”. the qualitative report, 15(5), 1243–1269. retrieved from https://nsuworks.nova.edu/tqr/vol15/iss5/12/ kjellgren, a., lyden, f., & norlander, t. (2008). sensory isolation in flotation tanks: altered states of consciousness and effects on well-being. the qualitative report, 13(4), 636–656. retrieved from https://nsuworks.nova.edu/tqr/vol13/iss4/6 klatzky, r. (1998). allocentric and egocentric spatial representations. spatial information theory. a theoretical basis for gis, cosit93, 716, 14–24. klee, g. d. (1963). lysergic acid diethylamide (lsd-25) and ego functions. archives of general psychiatry, 8(5), 461–474. https://doi.org/10.1001/archpsyc.1963.01720110037005 kriegel, u. (2009). subjective consciousness: a self-representational theory. oxford: oxford university press. laberge, s., & degracia, d. j. (2000). varieties of lucid dreaming experience. in r. g. kunzendorf & b. wallace (eds.), individual differences in conscious experience (pp. 269–307). amsterdam, netherlands: john benjamins publishing company. letheby, c. (2020). being for no-one: psychedelic experience and minimal subjectivity. philosophy and the mind sciences, 1(i), 5. https://doi.org/10.33735/phimisci.2020.i.47 letheby, c., & gerrans, p. (2017). self unbound: ego dissolution in psychedelic experience. neuroscience of consciousness, 2017(1), nix016. https://doi.org/10.1093/nc/nix016 lindahl, j. r., fisher, n. e., cooper, d. j., rosen, r. k., & britton, w. b. (2017). the varieties of contemplative experience: a mixed-methods study of meditation-related challenges in western buddhists. plos one, 12(5), e0176239. https://doi.org/10.1371/journal.pone.0176239 lopez, c., & elzière, m. (2018). out-of-body experience in vestibular disorders a prospective study of 210 patients with dizziness. cortex, 104, 193–206. https://doi.org/10.1016/j.cortex.2017.05.026 lou, h. c., changeux, j. p., & rosenstand, a. (2017). towards a cognitive neuroscience of self-awareness. neuroscience & biobehavioral reviews, 83, 765–773. https://doi.org/10.1016/j.neubiorev.2016.04.004 magallon, l. l. (1987). awake in the dark: imageless lucid dreaming. lucidity letter, 6(1). retrieved from https://journals.macewan.ca/lucidity/article/download/752/693 martin, j.-r., & pacherie, e. (2013). out of nowhere: thought insertion, ownership and context-integration. consciousness and cognition, 22(1), 111–122. https://doi.org/10.1016/j.concog.2012.11.012 martin, m. g. f. (1995). bodily awareness: a sense of ownership. in j. l. bermudez, a. j. marcel, & n. m. eilan (eds.), the body and the self (pp. 267–289). cambridge, ma: mit press. masters, r. a. (2005). darkness shining wild: an odyssey to the heart of hell & beyond: meditations on sanity, suffering, spirituality, and liberation. surrey, bc: tehmenos press. mayer-gross, w. (1935). on depersonalization. british journal of medical psychology, 15(2), 103–126. https://doi.org/10.1111/j.2044-8341.1935.tb01140.x mayer-gross, w., & stein, h. (1926). über einige abänderungen der sinnestätigkeit im meskalinrausch. zeitschrift für die gesamte neurologie und psychiatrie, 101(1), 354–386. https://doi.org/10.1007/bf02878343 mcclelland, t. (2019). against virtual selves. erkenntnis, 84, 21–40. https://doi.org/10.1007/s10670-017-9945-8 mcclelland, t. (forthcoming). four impediments to the case for mineness. in m. garcia-carpintero & m. guillot (eds.), the sense of mineness. oxford: oxford university press. mcginn, c. (1989). mental content. oxford: basil blackwell. meier, b. (1993). speech and thinking in dreams. in c. cavallero & d. foulkes (eds.), dreaming as cognition (pp. 58–76). hertfordshire, uk: harvester wheatsheaf. metzinger, t. (2003). being no one: the self-model theory of subjectivity. cambridge, ma: mit press. metzinger, t. (2013a). the myth of cognitive agency: subpersonal thinking as a cyclically recurring loss of mental autonomy. frontiers in psychology, 4, 931. https://doi.org/10.3389/fpsyg.2013.00931 metzinger, t. (2013b). why are dreams interesting for philosophers? the example of minimal phenomenal selfhood, plus an agenda for future research. frontiers in psychology, 4, 746. https://doi.org/10.3389/fpsyg.2013.00746 millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.concog.2010.03.016 https://doi.org/10.1016/bs.pbr.2018.10.021 https://nsuworks.nova.edu/tqr/vol15/iss5/12/ https://nsuworks.nova.edu/tqr/vol13/iss4/6 https://doi.org/10.1001/archpsyc.1963.01720110037005 https://doi.org/10.33735/phimisci.2020.i.47 https://doi.org/10.1093/nc/nix016 https://doi.org/10.1371/journal.pone.0176239 https://doi.org/10.1016/j.cortex.2017.05.026 https://doi.org/10.1016/j.neubiorev.2016.04.004 https://journals.macewan.ca/lucidity/article/download/752/693 https://doi.org/10.1016/j.concog.2012.11.012 https://doi.org/10.1111/j.2044-8341.1935.tb01140.x https://doi.org/10.1007/bf02878343 https://doi.org/10.1007/s10670-017-9945-8 https://doi.org/10.3389/fpsyg.2013.00931 https://doi.org/10.3389/fpsyg.2013.00746 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 39 metzinger, t. (2015). m-autonomy. journal of consciousness studies, 22(11-12), 270–302. retrieved from https://www.ingentaconnect.com/content/imp/jcs/2015/00000022/f0020011/art00013 metzinger, t. (2018). minimal phenomenal experience. mindrxiv. https://doi.org/10.17605/osf.io/5wyg7 metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of ”pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 michaelian, k. (2016). mental time travel: episodic memory and our knowledge of the personal past. cambridge, ma: mit press. millière, r. (2017). looking for the self: phenomenology, neurophysiology and philosophical significance of drug-induced ego dissolution. frontiers in human neuroscience, 11(245), 1–22. https://doi.org/10.3389/fnhum.2017.00245 millière, r. (2019). are there degrees of self-consciousness? journal of consciousness studies, 26(3-4), 252–276. retrieved from https://www.ingentaconnect.com/content/imp/jcs/2019/00000026/f0020003/art00012 millière, r., carhart-harris, r. l., roseman, l., trautwein, f.-m., & berkovich-ohana, a. (2018). psychedelics, meditation and self-consciousness. frontiers in psychology, 9(1475), 1–29. https://doi.org/10.3389/fpsyg.2018.01475 murguialday, a. r., hill, j., bensch, m., martens, s., halder, s., nijboer, f., et al. (2011). transition from the locked in to the completely locked-in state: a physiological analysis. clinical neurophysiology, 122(5), 925–933. https://doi.org/10.1016/j.clinph.2010.08.019 nathan, n. m. l. (1997). self and will. international journal of philosophical studies, 5(1), 81–94. https://doi.org/10.1080/09672559708570846 nida-rümelin, m. (2017). self-awareness. review of philosophy and psychology, 8(1), 55–82. https://doi.org/10.1007/s13164-016-0328-x nielsen, m. (1938). gerstmann syndrome: finger agnosia, agraphia, confusion of right and left and acalculia. comparison of this syndrome with disturbance of body scheme resulting from lesions of the right side of the brain. archives of neurology & psychiatry, 39(3), 536–560. https://doi.org/10.1001/archneurpsyc.1938.02270030114009 noë, a. (2005). action in perception. cambridge, ma: the mit press. o’brien, l., & soteriou, m. (eds.). (2009). mental actions. oxford: oxford university press. occhionero, m., & cicogna, p. c. (2011). autoscopic phenomena and one’s own body representation in dreams. consciousness and cognition, 20(4), 1009–1015. https://doi.org/10.1016/j.concog.2011.01.004 o’shaughnessy, b. (2000). consciousness and the world. oxford: oxford university press. pahnke, w. n. (1966). drugs and mysticism: an analysis of the relationship between psychedelic drugs and the mystical consciousness (phd thesis). harvard university, cambridge, ma. pahnke, w. n., & richards, w. a. (1966). implications of lsd and experimental mysticism. journal of religion and health, 5(3), 175–208. https://doi.org/10.1007/bf01532646 parker, s. (2017). yoga nidrā: an opportunity for collaboration to extend the science of sleep states. sleep and vigilance, 1(2), 57–63. https://doi.org/10.1007/s41782-017-0026-5 parker, s. (2019). training attention for conscious non-rem sleep: the yogic practice of yoga-nidrā and its implications for neuroscience research. in n. srinivasan (ed.), progress in brain research (pp. 255–272). elsevier. parrott, m. (2017). subjective misidentification and thought insertion. mind & language, 32(1), 39–64. https://doi.org/10.1111/mila.12132 patel, k. c. (2012). eastern traditions, consciousness, and spirituality. in l. j. miller (ed.), the oxford handbook of psychology and spirituality (pp. 343–357). oxford: oxford university press. peacocke, c. (2006). mental action and self-awareness. in j. d. cohen & b. p. mclaughlin (eds.), contemporary debates in the philosophy of mind. oxford: blackwell. peacocke, c. (2014). the mirror of the world: subjects, consciousness, and self-consciousness. oxford: oxford university press. perry, j. (1979). the problem of the essential indexical. noûs, 13(1), 3–21. https://doi.org/10.2307/2214792 pollan, m. (2018). how to change your mind: what the new science of psychedelics teaches us about consciousness, dying, addiction, depression, and transcendence. new york: penguin press. prinz, j. j. (2012). the conscious brain: how attention engenders experience. oxford and new york: oxford university press. proust, j. (2009). is there a sense of agency for thought? in l. o’brien & m. soteriou (eds.), mental actions (pp. 253–279). oxford: oxford university press. robins, s. (2017). in defense of vasubandhu’s approach to episodic phenomenology. australasian philosophical review, 1(4), 416–419. https://doi.org/10.1080/24740500.2017.1411148 robins, s. k. (2016). misremembering. philosophical psychology, 29(3), 432–447. https://doi.org/10.1080/09515089.2015.1113245 millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.ingentaconnect.com/content/imp/jcs/2015/00000022/f0020011/art00013 https://doi.org/10.17605/osf.io/5wyg7 https://doi.org/10.33735/phimisci.2020.i.46 https://doi.org/10.3389/fnhum.2017.00245 https://www.ingentaconnect.com/content/imp/jcs/2019/00000026/f0020003/art00012 https://doi.org/10.3389/fpsyg.2018.01475 https://doi.org/10.1016/j.clinph.2010.08.019 https://doi.org/10.1080/09672559708570846 https://doi.org/10.1007/s13164-016-0328-x https://doi.org/10.1001/archneurpsyc.1938.02270030114009 https://doi.org/10.1016/j.concog.2011.01.004 https://doi.org/10.1007/bf01532646 https://doi.org/10.1007/s41782-017-0026-5 https://doi.org/10.1111/mila.12132 https://doi.org/10.2307/2214792 https://doi.org/10.1080/24740500.2017.1411148 https://doi.org/10.1080/09515089.2015.1113245 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org raphaël millière 40 rosenthal, d. m. (1986). two concepts of consciousness. philosophical studies, 49(3), 329–359. https://doi.org/10.1007/bf00355521 rosenthal, d. m. (1997). a theory of consciousness. in n. block, o. j. flanagan, & g. guzeldere (eds.), the nature of consciousness (pp. 729–754). cambridge, ma: mit press. rümmele, w., & gnirss, f. (1961). untersuchungen mit psilocybin, einer psychotropen substanz aus psilocybe mexicana. schweizer archiv für neurologie, neurochirurgie und psychiatrie, 87, 365–385. saks, e. r. (2007). the center cannot hold: my journey through madness. new york: hachette books. savage, c. (1955). variations in ego feeling induced by d-lysergic acid diethylamide (lsd-25). psychoanalytic review, 42, 1–16. schwenkler, j. (2014). vision, self-location, and the phenomenology of the “ point of view”. noûs, 48(1), 137–155. https://doi.org/10.1111/j.1468-0068.2012.00871.x sedman, g., & kenna, j. c. (1964). the occurrence of depersonalization phenomena under lsd. european neurology, 147(3), 129–137. https://doi.org/10.1159/000128893 seli, p., ralph, b. c. w., risko, e. f., w. schooler, j., schacter, d. l., & smilek, d. (2017). intentionality and meta-awareness of mind wandering: are they one and the same, or distinct dimensions? psychonomic bulletin & review, 24(6), 1808–1818. https://doi.org/10.3758/s13423-017-1249-0 shankman, r. (2008). the experience of samadhi: an in-depth exploration of buddhist meditation. boston: shambhala. siclari, f., larocque, j. j., postle, b. r., & tononi, g. (2013). assessing sleep consciousness within subjects using a serial awakening paradigm. frontiers in psychology, 4, 542. https://doi.org/10.3389/fpsyg.2013.00542 smith, j. (2017). self-consciousness. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (fall 2017). metaphysics research lab, stanford university. stephens, g. l., poland, j., & radden, j. (2000). when self-consciousness breaks: alien voices and inserted thoughts. cambridge, ma: mit press. strawson, g. (2011). selves: an essay in revisionary metaphysics. oxford: oxford university press. sznycer, d. (2019). forms and functions of the self-conscious emotions. trends in cognitive sciences, 23(2), 143–157. https://doi.org/10.1016/j.tics.2018.11.007 thompson, e. (2015). dreamless sleep, the embodied mind, and consciousness. in t. k. metzinger & j. m. windt (eds.), open mind. https://doi.org/10.15502/9783958570351 uthaug, m. v., lancelotta, r., van oorsouw, k., kuypers, k. p. c., mason, n., rak, j., et al. (2019). a single inhalation of vapor from dried toad secretion containing 5-methoxy-n,n-dimethyltryptamine (5-meo-dmt) in a naturalistic setting is related to sustained enhancement of satisfaction with life, mindfulness-related capacities, and a decrement of psychopathological symptoms. psychopharmacology, 236, 2653–2666. https://doi.org/10.1007/s00213-019-05236-w von mering, o., morimoto, k., hyde, r. w., & rinkel, m. (1955). experimentally induced depersonalization. proceedings of the annual meeting of the american psychopathological association, 66–77. wahbeh, h., sagher, a., back, w., pundhir, p., & travis, f. (2018). a systematic review of transcendent states across meditation and contemplative traditions. explore, 14(1), 19–35. https://doi.org/10.1016/j.explore.2017.07.007 windt, j. m. (2010). the immersive spatiotemporal hallucination model of dreaming. phenomenology and the cognitive sciences, 9(2), 295–316. https://doi.org/10.1007/s11097-010-9163-1 windt, j. m. (2015a). dreaming: a conceptual framework for philosophy of mind and empirical research. cambridge, ma: mit press. windt, j. m. (2015b). just in time—dreamless sleep experience as pure subjective temporality. in t. k. metzinger & j. m. windt (eds.), open mind. frankfurt am main: mind group. windt, j. m. (2017). from indian philosophy to cognitive neuroscience: two empirical case studies for ganeri’s self. philosophical studies, 174(7), 1721–1733. https://doi.org/10.1007/s11098-016-0826-9 windt, j. m., nielsen, t., & thompson, e. (2016). does consciousness disappear in dreamless sleep? trends in cognitive sciences, 20(12), 871–882. https://doi.org/10.1016/j.tics.2016.09.006 wrigley, w. j., & emmerson, s. b. (2011). the experience of the flow state in live music performance. psychology of music, 41(3), 292–305. https://doi.org/10.1177/0305735611425903 zahavi, d. (1999). self-awareness and alterity: a phenomenological investigation. evanston, il: northwestern university press. zahavi, d. (2005). subjectivity and selfhood: investigating the first-person perspective. cambridge, ma: mit press. zahavi, d. (2010). minimal self and narrative self. a distinction in need of refinement. in t. fuchs, h. c. sattel, & p. henningsen (eds.), the embodied self: dimensions, coherence, and disorders (pp. 3–11). stuttgart: schattauer. zahavi, d. (2014). self and other: exploring subjectivity, empathy, and shame. oxford: oxford university press. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/bf00355521 https://doi.org/10.1111/j.1468-0068.2012.00871.x https://doi.org/10.1159/000128893 https://doi.org/10.3758/s13423-017-1249-0 https://doi.org/10.3389/fpsyg.2013.00542 https://doi.org/10.1016/j.tics.2018.11.007 https://doi.org/10.15502/9783958570351 https://doi.org/10.1007/s00213-019-05236-w https://doi.org/10.1016/j.explore.2017.07.007 https://doi.org/10.1007/s11097-010-9163-1 https://doi.org/10.1007/s11098-016-0826-9 https://doi.org/10.1016/j.tics.2016.09.006 https://doi.org/10.1177/0305735611425903 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the varieties of selflessness 41 zahavi, d. (2018). consciousness, self-consciousness, selfhood: a reply to some critics. review of philosophy and psychology, 9(3), 703–718. https://doi.org/10.1007/s13164-018-0403-6 zahavi, d., & kriegel, u. (2016). for-me-ness: what it is and what it is not. in d. o. dahlstrom, a. elpidorou, & w. hopp (eds.), philosophy of mind and phenomenology: conceptual and empirical approaches (pp. 36–53). london: routledge. open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s13164-018-0403-6 https://doi.org/10.33735/phimisci.2020.i.48 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction getting clearer on self-consciousness partial selflessness consciousness without cognitive self-consciousness consciousness without spatial self-consciousness consciousness without a sense of bodily ownership consciousness without a sense of bodily agency consciousness without a sense of mental ownership consciousness without a sense of mental agency the varieties of partial selflessness total selflessness drug-induced ego dissolution meditation lucid dreamless sleep objections and replies objection 1. there can be no experience of total selflessness, because there is no experience without a subject of experience. objection 2. there can be no experience of total selflessness, because it would be an experience devoid of content, and no experience can be contentless. objection 3. there is a form of self-consciousness that is more basic than any of the six notions of self-consciousness discussed in this article and is preserved in allegedly "selfless" states. objection 4. there can be no autobiographical memory of totally selfless states, therefore reports of such states are suspect. conclusion minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of ``pure'' consciousness minimal phenomenal experience meditation, tonic alertness, and the phenomenology of “pure” consciousness thomas metzingera (metzinger@uni-mainz.de) abstract this is the first in a series of instalments aiming at a minimal model explanation for conscious experience, taking the phenomenal character of “pure consciousness” or “pure awareness” in meditation as its entry point. it develops the concept of “minimal phenomenal experience” (mpe) as a candidate for the simplest form of consciousness, substantiating it by extracting six semantic constraints from the existing literature and using sixteen phenomenological case-studies to incrementally flesh out the new working concept. one empirical hypothesis is that the phenomenological prototype of “pure awareness”, to which all such reports refer, really is the content of a predictive model, namely, a bayesian representation of tonic alertness. on a more abstract conceptual level, it can be described as a model of an unpartitioned epistemic space. keywords consciousness as such ∙ dharmakāya ∙ empty cognizance ∙ minimal model explanation ∙ minimal phenomenal experience ∙ minimal phenomenal selfhood ∙ pure awareness ∙ pure consciousness ∙ rigpa ∙ sākṣin ∙ samadhi ∙ self-luminosity ∙ tonic alertness ∙ transparency ∙ turiya ∙ wakefulness ∙ ye shes this article is part of a special issue on “radical disruptions of self-consciousness”, edited by thomas metzinger and raphaël millière. this paper is the first in a series of publications trying to develop a new approach to the problem of consciousness. in section 1 i will give a short introduction to the general idea, the strategy of explanation by minimal models, and list some of my epistemic goals. section 2 looks at the phenomenology of “pure consciousness” by using a selection of case studies and developing a set of six semantic constraints for a finer-grained concept of “minimal phenomenal experience”. section 3 provides ajohannes gutenberg university, mainz metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0003-2514-7191 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 2 some conceptual differentiation, distinguishing between the phenomenal character of “wakefulness” and the functional property of tonic alertness. the last section presents a new hypothesis about the experience of consciousness as such: it is a subsymbolic, bayesian, and aperspectival representation of tonic alertness. 1 introduction 1.1 from mps to mpe in 2009, olaf blanke and thomas metzinger asked two questions: what are the minimally sufficient conditions for the appearance of a phenomenal self, that is, the fundamental conscious experience of being someone? what are necessary conditions for self-consciousness in any type of system (blanke & metzinger, 2009, p. 7)? one of their findings was that bodily agency is a causally enabling but not a constitutive condition for phenomenal selfhood (blanke & metzinger, 2009, p. 13). blanke and metzinger then introduced the concept of “minimal phenomenal selfhood” (mps), the central defining features of which are (i) a globalized form of identification with the body as a whole (as opposed to mere ownership for body parts), (ii) spatiotemporal self-location, and (iii) a weak, geometrical first-person perspective (1pp). this paper begins to develop a related strategy for “minimal phenomenal experience” (mpe; this term was originally introduced in windt, 2015b). it will take as a new entry point not full-body illusions, but the specific phenomenology of “pure consciousness” in meditation. on the phenomenological level of analysis, part of the working hypothesis is that mpe lacks (i) mps and all other forms of egoic self-consciousness, (ii) time representation, and (iii) a spatial frame of reference. where for mps the global phenomenology of identifying with the body as a whole was the empirical starting point, here, i begin with all reports in which subjects claim that they have actually had an experience of consciousness as such. can a conscious system be exclusively aware of awareness itself? if yes, would there be a specific form of phenomenal character that is instantiated during such episodes? and if there really exists something like the simplest form of conscious experience – what are the minimally sufficient conditions for mpe to occur in neurotypical humans, and what would be necessary conditions for the experience of consciousness as such to appear in any type of system?1 1as a starting point, i have here mostly chosen reports about “pure awareness” or “pure consciousness” from contemplative practitioners, but in principle there may be many other relevant kinds of minimal phenomenal experience, for example as caused by subanesthetic doses of general anesthetics (russell, 2013; sanders, tononi, laureys, & sleigh, 2012) or as experienced by patients moving in and out of coma (giacino et al., 2002; laureys, owen, & schiff, 2004). my reason for selecting this highly specific subset of contemplative experiences is twofold. first, if we are interested in arriving at a minimal-model explanation for phenomenal consciousness, it is quite simply the most natural first step to closely investigate all those states in which subjects actually report experiencing the “essence” of consciousness or consciousness as such. second, what makes the metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 3 1.2 scientific explanation by minimal models one central motivation behind the current initiative is to test a fresh alternative to the now-classical research strategy of isolating the “neural correlate of consciousness” (ncc, metzinger, 2000). the new idea is to develop a “minimal model” of conscious experience. in the present context, developing a “minimal model” of conscious experience is something very benign and modest. for example, given the current state of consciousness science, a minimal model will not yet involve elaborate formal procedures like mathematically delimiting a universality class (batterman & rice, 2014). rather, it can be seen as a heuristic strategy of minimalist idealization: • constructing and investigating a model of conscious experience that includes only the core causal factors giving rise to the target phenomenon; • including only causal factors that make a difference to the actual occurrence and the essential phenomenal character itself; phenomenology of certain meditative states interesting from a scientific perspective is that the philosophical phenomenologies that typically come along with them claim a conjunction of minimal dimensionality (i.e., an “empty” form of conscious experience lacking internal structure and content) and maximal prototypicality (i.e., significant if often-unnoticed overlap with most other exemplars of the category of “consciousness”). i presuppose that the concept of phenomenal consciousness has not definitional but probabilistic structure. “probabilistic structure” means that a certain conscious experience falls under mpe if it satisfies a sufficient number of constraints encoded by mpe’s semantic constituents (margolis & laurence, 2014). in principle, there could exist multiple and loosely overlapping sets of sufficient conditions. ultimately, mpe is a cluster concept and “mpe” refers to a phenomenological prototype. this implies that there is no set of necessary and sufficient conditions that allows us to say whether a representational state is conscious or not. rather, “consciousness” refers to states, regions, or trajectories in a multidimensional space constituted by functional and contentrelated dimensions – it is a heterogeneous and graded construct, and membership comes in degrees (bayne, hohwy, & owen, 2016; for a similar approach to the concept of “mind wandering”, see seli, kane, metzinger, et al., 2018). the underlying dimensions which give a structure to the space of conscious experience create a complex pattern of family resemblances, and absolute orderings may not be possible in all cases. there will be degrees of prototypicality, because highly prototypical members have attributes that overlap with most other exemplars of the category, while low-prototypical members have little overlap. membership in the family of phenomenal states will therefore be graded, and quantitative assessments of prototypicality may actually help us understand why some phenomenological exemplars capture the intuitive “essence” of consciousness better than others. please note how this implies a radical departure from essentialist intuitions as might seem to be automatically implied by the two questions about necessary and sufficient questions in section 1.1: it may not be possible to isolate the simplest form of conscious experience in terms of one unique minimum of dimensionality for the relevant state-space (or an individual state displaying the minimal set of characteristic features), and for a given system multiple low-dimensional states may actually be possible or even co-exist. for example, in a 1,000,000-dimensional state space there might be five different kinds of three-dimensional states, each of them exhibiting a unique phenomenal character. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 4 • developing an idealized model of universal and repeatable features serving to gradually isolate the fundamental, explanatorily relevant, and structurally stable properties that underlie different forms of conscious experience. the opposite of a minimal model would be a maximal model (i.e., a model which would be “ideal” in a very different sense): a theory which would capture all the behaviour, both repeatable and unrepeatable, of a given system. for complex systems like the conscious human brain, this traditional modelling strategy is an unattainable goal; many of the fine details will actually get in the way of scientific explanation and detract from a deeper understanding of the research target (batterman, 2002, p. 22). however, if the phenomenon to be explained is a large-scale regularity that is largely independent of micro-scale details and features of individual elements, it can prove helpful to identify the minimal conditions necessary for the explanatory target to occur, later linking them back to the original phenomenon of interest (for three historically important case studies in ancestral heredity, the ideal gas law in physics, and evolutionary biology, see section 2 in rice, rohwer, & ariew, 2019; see also footnote 4). in the present context, there are two major aspects of the minimal-model strategy. the first is to describe the target phenomenon in an uncluttered way, abstracting away from everything that is not an essential feature of the core explanandum. if the systems of interest are conscious systems, three questions arise: (i) whether phenomenality as such has a distinct experiential character, (ii) whether this character ever occurs in isolation, and (iii) whether it can be described in a conceptually precise manner (this is the main goal of the current paper). the second aspect aims at an economical strategy for describing the network of causal factors giving rise to this specific explanandum – given that a precise description succeeds. here, constructing a minimal model would amount to eliminating superfluous details by extracting only the explanatorily relevant causal structure underlying the experience of phenomenality as such, thereby providing a medium-level functional analysis while still omitting most of the structural details of a fully mechanistic explanation (i.e., a “mechanism sketch”; cf. piccinini & craver, 2011). both aspects could come together if mathematically stable features of the physical dynamics characterizing the global neural correlate of mpe in humans could be directly mapped to corresponding abstract patterns within phenomenal dynamics independently of potentially unknowable and fine-grained neuroscientific details, thereby yielding what we are most interested in: universal and reliably repeatable properties instantiated by all forms of subjective experience. the first positive proposal of this paper is that individual episodes of mpe could be instantiations of such an abstract pattern, perhaps even of a universal or at least strongly prototypical feature. in terms of research heuristics, my second proposal is that it is simply natural to focus first on exactly those states (long identified by contemplative traditions) which are described as episodes of “pure awareness”, “empty cognizance”, or “consciousness as such”. prima facie and given the vastness metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 5 of our own phenomenal state space, this may be the most direct route to studying such an abstract pattern without any distracting additions. the strategy of “minimalist idealization” looks for the decisive explanatory and computational factors by first attempting to construct exactly solvable models (batterman, 2002, p. 37). therefore, minimal models do not aim at maximal fidelity or completeness; they are not yet mappings to fine-grained functional mechanisms because the accuracy conditions for an actual representation of causal pathways are not intended to be very strong (weisberg, 2012, chapter 6, section 6.1.2). however, explanation by minimal models may nevertheless involve a phase of deidealization at a later stage, turning them into a continuing research program eventually leading to a multilevel mechanistic explanation (piccinini & craver, 2011). this is one of my longer-term goals. but i also believe that in order to really solve the problem of consciousness we will ultimately need a theory of mpe – because only a minimal model can give us a deep scientific understanding of the essence of phenomenal experience. 1.3 sketching the epistemic goal of the mpe approach there are five factors that make the concept of “mpe” interesting. first, mpe could refer to the common phenomenological denominator that is always present whenever there is conscious experience at all, however rich or limited. second, a concerted attempt to isolate this hypothetical common denominator could lead to interdisciplinary unification in consciousness research by connecting different domains and experimental approaches in a new way. third, by aiming at a minimalmodel explanation we could arrive at an entirely new theoretical model of conscious experience itself, providing us with a fresh perspective and possibly even a new window onto the target. fourth, the mpe approach could lead to unification and considerable enrichment within analytical philosophy of mind itself, by elevating comparative and transcultural philosophy of mind to a new and more systematic level. and finally, reports about “pure consciousness” experiences are a neglected empirical phenomenon worth studying in its own right. from now on, i will use the phenomenological notions of “pure awareness” and “pure consciousness” interchangeably. “mpe” refers to the theoretical concept of minimal phenomenal experience as it could figure in a minimal-model explanation of consciousness. i will here limit myself to one plausibly homogeneous subclass of mpe: to a paradigmatic form of “pure consciousness” experience as it appears in some forms of contemplative practice (see footnote 1). if one accepts the relevance of the five reasons given above, then the most pressing epistemic goal consists in first getting the phenomenology right. the guiding phenomenological question is this: does pure awareness have a minimal, distinct, and unique phenomenal character sui generis? metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 6 we can also ask four types of slightly more detailed question: (1) is there a form of consciousness that does not instantiate minimal phenomenal selfhood (blanke & metzinger, 2009) or an epistemic agent model (metzinger, 2017, sec. 2.5)? → are selfhood (i.e., egoic self-consciousness, the experience of “being someone”) and perspectivalness (i.e., an individual and consciously experienced first-person perspective) necessary conditions for phenomenality? → if they are not, what is the epistemological status of reports about such episodes? (2) does “pure” consciousness really exist? → are there any concrete examples of phenomenality without intentional content? → are there types of conscious experience which do have content, but of which we would predict that they will be reported as “empty” or “contentless” forms of experience? (3) can we develop a minimal model for conscious experience as such? → is there a single kind of conscious experience that can count as the simplest instance of the target phenomenon? → fundamentality: is there a fundamental (and perhaps implicit) kind of phenomenal character sui generis, which can at times be made explicit and which underlies or “permeates” all other forms of phenomenal experience? (4) is there one maximally abstract phenomenal property? → is “pure awareness” a unique form of phenomenal character of which we could say that it forms a phenomenological kind of its own and that it subsumes all other forms? i will present preliminary answers to some of these questions in section 4. first, let us penetrate a little deeper into the phenomenological landscape surrounding mpe, using a set of fourteen case studies plus two canonical phenomenological descriptions from the relevant literature. 2 mpe on the phenomenological level of analysis one major research target for the current initiative is the development of a fully validated psychometric instrument for mpe. in this introductory paper i will take three steps. as a first, preparatory step for creating a finer-grained, more evidencebased description i will now treat mpe as an exclusively phenomenological notion, deliberately leaving aside all the metaphysical interpretations and theoretical background assumptions frequently offered along with experiential reports by metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 7 humankind’s contemplative traditions.2 second, i will attempt to semantically enrich the new concept by extracting six general constraints from both eastern and western philosophical traditions, sometimes mixing phenomenological analysis and exegesis. accordingly, we will arrive at a preliminary 6d-model of what has hitherto been called “pure” consciousness. the third step will have to consist in further increasing the dimensionality of our conceptual instruments, by asking actual meditators whether they have ever had the relevant kind of “pure awareness” experience and if so, how they would describe its phenomenal character (cf. footnote 1). the six constraints to be introduced in the following subsections are: • wakefulness (pc1) • low complexity (pc2) • self-luminosity (pc3) • introspective availability (pc4) • epistemicity (pc5) • transparency/opacity (pc6) 2.1 pure consciousness in contemplative phenomenology mpe is probably an experience every human being knows. nonetheless, most explicit reports come from contemplative traditions, and i will therefore treat it for the most part as a specific and very simple subset of meditative experiences. in his classic monograph philosophy and mysticism walter terence stace categorizes reports of “pure consciousness” as a specific, namely introvertive, type of mystical experience: there would be no mental content whatever but rather a complete emptiness, vacuum, void. one would suppose a priori that consciousness would then entirely lapse and one would fall asleep or become unconscious. but the introvertive mystics – thousands of them all over the world—unanimously assert that they have attained to this complete vacuum of particular mental contents, but that what then 2it must be clearly noted that the mpe approach as sketched out here operates under naturalistic background assumptions, and it does so without having given an independent argument to support these assumptions. second, in some relevant philosophical traditions there have been very different conceptual, epistemological, metaphysical, and soteriological interpretations of the relevant phenomenological reports than the one presented here, in particular by those scholarpractitioners of the past who extensively cultivated and subsequently underwent such experiences themselves (see, e.g., higgins, 2013, p. 10, footnote 4). the philosophical difficulty lies in correctly assessing the phenomenology of certainty and insight with respect to the states’ own metaphysical status that often goes along with such states, especially when they involve permanent alterations to conscious experience (see section 3.1 in metzinger & windt, 2015; metzinger, 2014). on this second point, an anonymous reviewer remarked that for at least some types of mpe experiences there is an important aspect of the phenomenological profile that should not be ignored without argument, namely their “noetic purport to be metaphysically ultimate”. for recent alternatives to naturalism, see for example albahari (2020); finnigan (2017). see also footnote 15. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 8 happens is quite different from a lapse into unconsciousness. on the contrary, what emerges is a state of pure consciousness – “pure” in the sense that it is not the consciousness of any empirical content. it has no content except itself. […] it is the bare unity of the manifold of consciousness from which the manifold itself has been obliterated. this seems analogous to saying that if from a whole or unity of many parts we could subtract all the parts, the empty whole or unity would be left. (stace, 1960, p. 86) similarly, in a well-known collection of texts, first published by oxford university press in 1990 under the title the problem of pure consciousness, editor robert forman defines “pure consciousness events” (pces)3 as a “wakeful though contentless (non-intentional) consciousness”:4 one neither thinks nor perceives any mental or sensory content. yet, despite this suspension of content, one emerges from such events confident that one had remained awake inside, fully conscious. this experience, which has been called the pure consciousness event, or pce, has been identified in virtually every tradition. […] the pure consciousness event may be defined as a wakeful but contentless (nonintentional) consciousness. (forman, 1998, pp. 185–186) 3brentyn ramm has recently introduced the closely related concept of a “pure awareness experience” or “pure awareness event” (pae), which allows him to distinguish between “objectless pure awareness experience” and “object-directed pure awareness experience”, cf. ramm (2019, p. 9). note how this then allows for two further distinctions, which are relevant and already well known in the contemplative literature: “subject-directed pure awareness experience” and “selfdirected pure awareness experience”, for examples see harding (2000, p. 9); kelly (2019, p. 100). the last phenomenological notion is directly related to the philosophical project of formulating a conceptually clear first-order reflexive model of mpe, cf., e.g., montague (2016, p. 63). ramm’s distinction itself points to the gradualist approach i pursue by introducing the low-complexity constraint below, which will hopefully allow for finer-grained phenomenological distinctions in the future. 4interestingly, the same author also endorses something strongly resembling the “minimal-model approach” introduced in section 1.2 above: when a biologist seeks to understand a complex phenomenon, one key strategy is to look to [sic] at it in its simplest form. probably the most famous is the humble bacterium e. coli. its simple gene structure has allowed us to understand much of the gene functioning of complex species. similarly, many biologists have turned to the “memory” of the simple sea slug to understand our own more kaleidoscopic memory. freud and durkheim both used totemism, which they construed as the simplest form of religion, to understand the complexities of religious life. the methodological principle is: to understand something complex turn to its simple forms. mystical experiences may represent just such a simple form of human consciousness. […] these pces, encounters with consciousness devoid of intentional content, may be just the least complex encounter with awareness per se that we students of consciousness seek. the pce may serve, in short, as the e. coli of consciousness studies. (forman, 1998, pp. 185–186) metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 9 from these classical sources, we can already derive the first two major phenomenological constraints (pcs): • wakefulness (pc1): the phenomenal character of tonic alertness (see section 3.1). • low complexity (pc2): often described as the complete absence of intentional content, in particular of high-level symbolic mental content (i.e., discursive, conceptual, or propositional thought), but also of sensorimotor or affective content. if we accept these first two constraints, then a theory of mpe will have to be a theory of contentless, restful alertness – a form of phenomenal experience that in its purest form does not instantiate any intentional properties. from a philosophical perspective, this conjunction of two necessary conditions is simple and straightforward. as a first characterization, it is certainly helpful. however, please note how even the very first classical quotations presented above introduce a whole range of conceptual ambiguities. in claiming that “[i]t has no content except itself”, walter stace (1960, p. 86) introduces reflexivity as an additional constraint for a conceptual prototype of pure consciousness; when he speaks of an “empty whole or unity”, one is confronted with the problem that on closer inspection the phenomenology of wholeness and the phenomenology of unity are very different (wiese, 2018). and if robert forman defines pces by making use of the notion of “(non-intentional) consciousness”, this immediately creates the need for a background theory about what “intentional” content really is, how many kinds of mental content there are, and how phenomenal and intentional properties are related. in particular, these conceptual ambiguities leave one specific possibility open: the absence of all empirical content could itself be an appearance, and what subjects (falsely) describe as a “contentless” phenomenal state could actually carry an abstract form of intentional content.5 if there is a content that can account for the reported experience of contentlessness, then representationalism remains tenable. this is the line i will be pursuing. from a very different perspective, here is how jonathan shear approaches the point: 5of course, as empirically plausible models of the human brain provide us with clear examples of non-conceptual abstraction, “abstractness” does not imply conceptual or propositional content. in addition, whenever phenomenological reports are the target of analysis, such reports can be doubted. as uriah kriegel puts the point: such a state of pure, contentless awareness is the state that certain meditative techniques are supposed to lead to. certainly the state in question has been described along these lines by some practitioners. it is nonetheless open to the theoretician to challenge the practitioner’s own description of the state she is in. (kriegel, 2009, p. 175, footnote 32) metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 10 experience of the deepest level, that of pure consciousness awake to its own nature in and by itself, is generally held to be especially important. the defining characteristic of this experience is the complete absence of all sounds, tastes, thoughts, feelings, images, and anything else that one can ever imagine. […] but they have in common the idea that it is possible for all empirical content to disappear, while one nevertheless remains awake. what then is the experience like? by all accounts it is not like anything. one can have it and remember it – one knows that one was not asleep. but one does not remember it as anything at all. it is just itself – unimaginable and indescribable. the experience itself is extraordinarily abstract. indeed, it is the logical ultimate of abstraction, since by all accounts it is what remains after everything that can possibly be removed from experience has been removed, while one nevertheless remains awake. it should be noted, however, that even if the natural response to this experience is to describe it as “contentless”, it is still appropriate to raise the question of whether it is in fact completely contentless. for while experiences […] might be so subtle and abstract that they naturally seem to be completely contentless, they might nevertheless actually have some, albeit very abstract, content. (shear, 2007, pp. 700, 702–703) if we are searching for a positive characterization of the target, an additional difficulty is that low complexity already introduces one exclusively negative element. and indeed, if one looks at the relevant literature, it is striking how many phenomenological characterizations of mpe episodes are of an exclusively negative kind. pces are described as: • non-sensory: mpe itself instantiates no perceptual qualities. this also includes “no-thingness”, i.e., the absence of the phenomenal property of “objecthood”. this means that there is no subjective experience of distinct multimodal objects as integrated from different sensory features and as segmented from a background or perceptual scene, being juxtaposed or “thrown against” a perceiving self; • non-motor: absolute stillness, no motion in space; • atemporal: an absence of temporal experience, no motion in time; • non-cognitive: non-symbolic and non-conceptual, no mind-wandering; • non-egoic: no self-location in time, no self-location in space, no quality of agency (either mental or bodily), no localized unit of identification (ui6), i.e., 6the “unit of identification” (ui) is whatever form of experiential content leads to phenomenologimetzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 11 mpe itself is not even characterized by mps, the minimal phenomenal sense of selfhood; • unbounded: there is no second, finite region to which attention could be directed, and there are no consciously experienced boundaries, limits, or a horizon; • aperspectival: no epistemic agent model (or eam7), no passive personallevel self-as-subject. hence, phenomenologically, the relevant phenomenal character seems not to be a first-person character: during its actual occurrence it is not tied to an individual first-person perspective (this feature is often called “non-duality” in buddhist or neo-advaitan literature). wakefulness is by far the most frequently mentioned kind of phenomenal character in descriptions of “pure consciousness”, and this observation is an important premise for the argument to be developed later. the second most frequent element is an absence of spatiotemporal self-location. “spatial self-location” means that one experiences oneself as located at a point or a region in a spatial frame of reference (typically, as embodied), while “temporal self-location” means that one is present in a temporal order, in a consciously experienced “now” (blanke & metzinger, 2009). “nowness” itself is a form of temporal experience, and this already leads to two conclusions that can guide future neuroscientific research into the underlying correlates: subjects experiencing the relevant states are not oriented in time and space, and the processing of temporal information has been transiently suspended. to demonstrate the first and second aspect, here are three case studies: case study #1. i experience pure consciousness as a state of unboundedness and total ease and deep relaxation. there are no thoughts, no feelings, or any other sensations like weight or temperature. i just know i am. there is no notion of time or space, but my mind is fully awake and perfectly clear. it is a very simple and natural state. (alexander, 1988, p. 3) cal reports of the type: “i am this!” in humans, typical uis are the body, the emotional self-model, and the sense of effort in attentional or cognitive agency (see next footnote). if a conscious system has no ui, it lacks the phenomenology of identification, and it has no sense of self. in the context of this special issue (and the mpe approach in particular), the philosophically interesting issue is whether there can be non-egoic uis, without a conscious sense of self. for more on uis, see metzinger (2018); for discussion, fasching (2008). 7an eam is a special layer in the human self-model. it is a conscious internal model of the self as actively selecting targets of knowledge, as a mental agent who stands in epistemic relations (like “perceiving”, “believing”, “knowing”) to the world and to itself (as in “controlling the focus of attention”, “reasoning about the world and oneself”, or “knowing that one knows”), and as an entity that has the capacity to actively create such relations of knowing. the eam is the origin of our consciously experienced first-person perspective. for more see metzinger (2017, sec. 2.5). metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 12 case study #2. […] a state of complete rest, full consciousness without content and unbounded in time and space. (severeide, 1990, p. 1570; quoted from bachmann, 2014, p. 52) case study #3. […] a couple of times per week i experience deep, unbounded silence, during which i am completely aware and awake, but no thoughts are present. there is no awareness of where i am, or the passage of time. i feel completely whole and at peace. (travis & pearson, 2000, p. 81) the third negative phenomenological descriptor which i want to highlight is “unboundedness”. it is present in all three case studies presented above, and it also frequently occurs in the literature. we must avoid a phenomenological non-sequitur at this point: to say that the space of pure consciousness is “boundless” does not imply that there is an explicit experience of infinite expansion or of large distances. rather, it means that there is no “other side beyond the boundary” to which attention could shift. to give an example, in visual awareness attention can shift from a red colour patch into an adjacent green colour patch, transgressing a chromatic boundary. for the phenomenology of tonic alertness there is no such boundary, because everything outside the space of wakefulness cannot be deliberately attended to. it is simply unconscious. prima facie this has the interesting consequence that the method of phenomenal contrast (siegel, 2007, p. 135) cannot be applied to the phenomenology of tonic alertness. it seems impossible to point to a pair of two experiences which differ only in terms of the relevant content; if there is no phenomenology of wakefulness, then there simply is no conscious experience altogether. in other words, if prototypical mpe as characterized above really is consciousness as such, then it makes no sense to contrast its specific phenomenal character with what could only be unconscious states devoid of such character. however, because for most of us the phenomenology of tonic alertness is normally fully transparent, this point will always remain strongly counterintuitive. in any case it would be false to interpret relevant reports of “unboundedness” as the phenomenology of an explicitly experienced and merely quantitative infinity, or as a concrete, endless expanse of some sort. it just means that a) the phenomenal character in question includes the potential for expansion, and b) that in the experience of pure awareness itself there is no such thing as a “beyond” – another consciously experienced finite region or realm “on the other side” of a boundary. this has a second implication. in the relevant phenomenological domain, there are no countable entities: the domain is an unstructured space which has neither centre nor periphery8 and which cannot be introspectively individuated to create indivisible units of experience – not even a single one. for our growing list of phenomenological descriptors for mpe, this has the interesting consequence of 8if this conclusion is valid, then describing inner awareness as peripheral ultimately rests on a misguided spatial metaphor, cf. kriegel (2009, p. 176). metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 13 adding “non-oneness” to “non-duality” (cf. mañjuśrīmitra, lipman, simmons, & norbu, 1987, p. 60). the first theoretical conclusion was that the method of phenomenal contrast may not be applicable to mpe. the second conclusion now is that the relevant phenomenal character refers to the conscious experience of an unstructured space without a centre or boundaries, which also lacks the phenomenal quality of singularity or numerical identity. against this background of predominantly negative characterizations there is a third conclusion that we can draw about wakefulness and low complexity: there will be variable degrees of constraint satisfaction. the concept of “mpe” really refers to a multidimensional space of possible conscious states, a space which may contain different points, regions, or trajectories (bayne et al., 2016). prima facie, judging from the currently available set of case studies, there clearly seems to be a primary dimension constituted by the phenomenal character we refer to by the folk-psychological term “wakefulness”, and it is this phenomenological dimension that should become the first target of empirical studies. “wakefulness as such” seems semantically related to similar phenomenological notions like “mental clarity” (but without mental content), “cognitive lucidity” (but without conscious cognition), or “bare awareness” (without object). what all of these have in common is that they refer to a kind of conscious experience that can clearly come in different strengths and degrees of intensity. “mpe” refers to their common phenomenological denominator. let me give two examples. first, it is plausible to assume different degrees of wakefulness phenomenologically characterizing such states. subjectively, there should be an introspectively detectable gradient of tonic alertness. for example, satisfaction of wakefulness could be higher during certain meditative or druginduced states, but lower during episodes of mental fatigue or sleep deprivation, or when we are close to sleep onset or fainting. to make this even clearer, let me briefly introduce a conceptual distinction which will be explained in detail in section 3.1. from now on, “wakefulness” exclusively refers to the phenomenal experience of tonic alertness, and not to the functional property of alertness as such or to the concrete physical property of cortical arousal. it is important never to conflate “wakefulness” as referring to a phenomenal property (that is, the conscious experience of sustained, cue-independent, tonic alertness as such) with “wakefulness” as a functional property (for example, as physically realized by the level of cortical arousal). not only is it logically possible to have functional wakefulness without conscious experience (i.e., “wakeful unawareness”), but the possibility is also empirically demonstrated by the existence of a wake/sleep cycle in some types of coma patients, for example in the minimally conscious state (mcs) or in unresponsive wakefulness syndrome (uws), tellingly termed wachkoma in german (laureys et al., 2010). in addition, it is empirically plausible that the conscious experience of wakefulness can exist in low-arousal states like nrem sleep, as shown by the existence of lucid dreamless sleep (dentico et al., 2016; ferrarelli et al., 2013; maruthai et al., 2016; mason et al., 1997; metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 14 mason & orme-johnson, 2010; metzinger, 2019, sec. 5; thompson, 2015a, 2015b; windt, 2015a; windt, nielsen, & thompson, 2016). hereafter, “wakefulness” therefore refers only to the graded subjective experience of tonic alertness. here is the second example. we can expect different degrees of absorption and content superimposition. for example, if the “absence of intentional content” demanded by low complexity refers only to a complete absence of mind-wandering and active high-level thinking, then episodes of clear, effortless mindfulness while still perceiving the physical body and visually observing the environment with an entirely quiet mind (as for example in buddhist walking meditation) already count as an mpe experience. a multimodal, egocentric frame of reference would still be preserved and superimposed onto mpe, like an abstract prior that has lost some but not all of its force. “mpe” then simply refers to the region of phenomenal state-space that we get by subtracting from the natural wake state the following attributes: a. the phenomenology of cognitive agency (as described in metzinger, 2013, 2017); b. the different types of spontaneous, task-unrelated thought (fox & christoff, 2018; seli, kane, smallwood, et al., 2018); c. all episodes in which perception is completely uncoupled from attention (mere “mind-blanking”; cf. kawagoe, onoda, & yamaguchi, 2019; ward & wegner, 2013). on the other hand, if “absence of intentional content” also refers to the content of perceptual, sensorimotor, interoceptive, or emotional states, then we make a stronger demand. if there is no sensory or other representational content whatsoever superimposed onto an mpe experience, then “mpe” refers to a much rarer kind of state. a full satisfaction of the low-complexity constraint means minimal complexity of reportable representational content on the level of conscious processing. therefore, the prototypical region in state-space will be high on the wakefulness dimension, and minimal in terms of content complexity. from now on, i will call this type a “full-absorption episode.”9 here are two case studies: 9please note how the phenomenological distinction between prototypical, highwakefulness/low-complexity states versus the variable spectrum of high-wakefulness/lowcomplexity states does not cleanly map onto walter stace’s conceptual distinction between “extrovertive mysticism” and “introvertive mysticism”. extrovertive mysticism exhibits four clusters of phenomenological features – fundamental unity, certainty about objective reference (i.e., william james’ “noetic quality”, cf. note 2), paradoxicality, and bliss/beatitude/joy – the last three of which do not seem to correspond to the ineffable contentlessness of full-absorption episodes (cf. stace, 1960, p. 67). introvertive mysticism is much closer to prototypical mpe, and here stace also includes all of the features we would today call ego-dissolution (cf. 1960, chapter 1, section 8), atemporality, and lack of spatial self-location stace (1960, p. 110) when he aptly describes the phenomenology of an “undifferentiated unity” which holds together the stream of metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 15 case study #4. after the content of consciousness had faded away, suddenly the experience of pure consciousness came. suddenly i felt that consciousness expanded, boundaries in time and space disappeared, and fullness, wholeness appeared, consciousness was completely without any content. i was completely awake, the awareness was there, very clear, but it was empty of any content, just pure consciousness in itself. when the experience was over i once again became aware of the sounds in the room, and the thoughts began wandering, […]. (severeide, 1990, p. 1570; quoted from bachmann, 2014, p. 52) case study #5. i had been meditating alone in my room all morning when someone knocked on my door. i heard the knock perfectly clearly, and upon hearing it i knew that, although there was no “waking up” before hearing the knock, for some indeterminate length of time prior to the knocking i had not been aware of anything in particular. i had been awake but with no content for my consciousness. […] the experience was […] utterly without content, […]. (forman, 1986, p. 55) beyond wakefulness, the content of a full-absorption episode cannot be reported, because the self-referential mechanisms of forming an autobiographical memory are suspended. a full-absorption episode is a radical disruption of self-consciousness, because mpe has not yet contracted into a first-person perspective. therefore, only the process of entering into it or of emerging out of it can be faithfully represented in the autobiographical self-model; as such the episode itself is not a part of the subject’s inner life narrative.10 on the other hand, it may be helpful to distinguish between concurrent ineffability and retrospective report, and we do know from amnesia and trauma research that an absence of reportable conscious experience need not imply an absence of memory. before moving on to the third constraint, let us observe how the first two items already allow us to differentiate between three relevant classes of phenomenal states. for simplicity, let us also assume that the wakefulness constraint is fully satisfied in all three cases: consciousness constituting the empirical ego (cf. stace, 1960, p. 87). yet, if one looks carefully at the impressive list of case studies walter stace presents, one cannot help but notice that the phenomenological landscape of these experiences really is much deeper, more complex, more idiosyncratic, more variable, and finer-grained than anything that could be captured by any simplistic, two-dimensional conceptual framework. 10for reasons of space, i have had to exclude a discussion of the three major methodological obstacles for the mpe approach (see footnote 27). for an extended discussion of the “problem of performative self-contradiction”, i recommend the contributions by sascha fink (2020) and raphaël millière (2020) in this special issue and section 4.5 in metzinger (2019). metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 16 • maximally satisfied low complexity plus maximal wakefulness would lead to what i have termed “full-absorption episodes”. in such episodes, the hypothetical phenomenal character of mpe would become a stand-alone feature; “pure awareness” would be the only aspect that could later be reported (see winter et al., 2020, for a recent single-case study). empirically plausible examples are “witness consciousness”11 during lucid dreamless sleep (for an introduction, see section 5 in metzinger, 2019) and deep meditative states during the practice of calm abiding and highly focused attention, for example in buddhist jhāna practice. • highly satisfied low complexity would lead to the prediction that at least some representational content remains active, while the stable experience of wakeful awareness presents itself as a continuous and explicit phenomenal quality. typical examples would be clear and relaxed states of restful alertness without any thought activity (e.g., during classical sitting meditation with closed eyes), or any natural form of mindfulness co-occurring with 11the notion of “witness-consciousness” refers to a term in classical advaita vedanta philosophy (sākṣin; sanskrit: सा ी). it is an example of a phenomenological notion with a long tradition in indian and tibetan philosophy, quite obviously bearing direct relevance to the present sketch of an interdisciplinary research program. the earliest appearance of the idea seems to be found in the atharva veda, in the muṇḍaka upaniṣad (chatterjee, 1982, p. 339), whereas the source of the philosophical concept sākṣin is obscure (fort, 1984, p. 287, n. 2; gupta, 1998, p. 19). witnessconsciousness is that which makes all knowledge possible, cannot itself become an object of knowledge, and is self-luminous. bina gupta gives the following epistemological characteristics: 1. the witness consciousness, although the basis of all knowing, is different from the object known. it is implied in every act of knowing. it is the ultimate subject; it can never become an object of knowledge. 2. it is the pure element of awareness in all knowing. it is one, immutable, indivisible reality. 3. it shines by its own light; it is self-luminous. 4. it is different from the empirical individual (jīva; sanskrit: जीव), who cognizes and enjoys. in other words, it is different from the empirical individual who is caught up in the triple states of waking, dreaming, and dreamless sleep. (gupta, 1998, p. 18) on a purely phenomenological reading that abstracts away from all metaphysical and epistemological aspects of the term “witness-consciousness”, we find something that strongly resembles mpe. it is a self-luminous form of bare awareness which can be superimposed by intentional objects during single “acts of knowing”, a non-egoic type of phenomenal character which has nothing to do with any phenomenal self-representation as a cognitive agent or an entity possessing subjective preferences; and, perhaps most importantly, it is something that could in principle be present in waking states, during dreaming, and in dreamless sleep, simply because it is different from any phenomenal self-model portraying only the “empirical individual”; cf. fort (1980). in short, witness-consciousness is an intrinsically non-reactive, non-egoic, and non-conceptual experience of knowing and wakefulness. one interesting reading is of the sākṣin as a “field” of consciousness, within which the distinction of subject, object, and the process of knowing arises as a “context for contents, or ‘space’ in which forms may appear” (fort, 1984, p. 278). for recent discussions and attempts at a definition, see albahari (2009, p. 66); fasching (2011, p. 211). metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 17 perceptual experience (e.g., during shamata-practice or “body scan”), or as a natural form of mindfulness co-occurring with bodily motion (e.g., during walking meditation [kinhin], the practice of tai-chi, or the feldenkrais method). • unsatisfied low complexity plus maximal wakefulness could serve to describe so-called “awakened” states. for example, following a global gestalt switch in which mpe has turned from an unrecognized background into the dominant and spontaneously present foreground of all experience, an explicit experience of wakefulness could appear as the medium or the centreand timeless epistemic space in which rich, variable, and complex phenomenal contents unfold over time. let us now add another dimension. in many classical texts, the phenomenology of wakefulness in meditation is described as functionally autonomous, for example as “self-generating ever-fresh awareness” (mañjuśrīmitra et al., 1987, xi) or as originary, naturally present, and non-transient “primordial knowing” (ye shes in tibetan; cf. higgins, 2012, p. 447, 2012, p. 99). this phenomenological feature is shared by self-luminosity, my third example of a phenomenological constraint that is frequently mentioned in a specific subset of the relevant literature (e.g., in tibetan buddhism). • self-luminosity (pc3): a phenomenal property instantiated during some mpe episodes, typically described as “radiance”, “brilliance”, or the “clear light” of primordial awareness. phenomenological descriptions of self-luminosity are a good example of cultural variance. as opposed to wakefulness and low complexity, selfluminosity is not frequently found in western phenomenological reports. it may therefore not be a necessary element of the common phenomenological denominator we are looking for and may not figure in the final theoretical concept of minimal phenomenal experience. in particular, it could be a metaphorical way of referring to wakefulness: some people compare this cognizance to a radiant “thing” that shines with a light like a “clear light”. it means a sense of being wakeful, a quality of being vividly wide awake, which is empty of any identity and naturally alert. (rinpoche, 2013b, p. 163) let us look at a classical example taken from the bardo-thödol, the “tibetan book of the dead” (8th century ce): case study #6. this brilliant emptiness is the radiant essence of your own awareness. it is beyond substance, beyond characteristics, beyond colour. […] the instant of your own presence is empty, yet it is not a nihilistic emptiness, but unimpeded radiance, brilliant and metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 18 vibrant. […] your own awareness, a vast luminous expanse, clarity inseparable from emptiness, is also the buddha of unchanging light, beyond birth and death. just to perceive this is enough. if you recognize this brilliant essence of your own awareness as buddha nature, then gazing into it is to abide in the state of enlightenment. (lingpa, 2014, pp. 14–15)12 functional autonomy and a self-revealing “clarity inseparable from awareness” are the first two features we can isolate. interestingly, there seems to be a third aspect of the relevant phenomenal character that can be described as a concrete, quasi-sensory phenomenology of self-luminosity (sometimes translated as eigenstrahlkraft in german). how can we make sense of it? i think there are three interrelated ways to do so: as the raw intensity of an internal stimulus source, which, via contemplative practice, has been made introspectively available; as functional autonomy; and as intrinsic epistemicity. first, i would propose that the relevant phenomenology of quasi-perceptual concreteness consists in the fact that the subjective experience of tonic alertness varies along a dimension of intensity, although it is clearly mode-neutral and therefore not associated with any interoceptive or exteroceptive sensory modality. something always radiates, but to different degrees. if this first interpretation is valid, then it follows that it should in principle be possible to introspectively trace back the radiance to its origin or hidden cause (buswell, 1991). second, not only does the subjective character of self-luminosity come in variable strengths and degrees, it is functionally autonomous and permeates all of phenomenal space. therefore, “self-luminosity” may actually be a visual analogy for the experienced functional autonomy of tonic alertness. it may be an attempt to point not so much to a distinct form of phenomenal character or representational content, but to a functional dimension structuring the space of conscious experience, which in turn allows for a relative ordering of global states (bayne et al., 2016, p. 410). it is also conceivable that an organism may possess an inner repre12if one takes an intellectually honest, analytical perspective, it is not easy to make sense of passages like these. prima facie, terms like “brilliant” and “radiant” refer to the sensory phenomenology of brightness, which is a necessary element of all visual experience. “your own” seems to refer to a phenomenology of ownership, that is, at least a minimal sense of selfhood. “vibrant” either refers to tactile phenomenology, or – perhaps like the terms “brilliant” and “radiant” – it is used in a purely metaphorical fashion. “perceiving” is standardly defined as a sensory object representation unfolding under a consciously experienced first-person perspective. similarly, if “gazing into” is interpreted as a goal-directed process, it must be an active form of visual perception involving an epistemic agent model (an eam, as introduced in footnote 7 above). the phenomenological configuration described is a highly complex one; it is certainly not “pure” in terms of an absence of intentional content or subject/object structure. rather, it seems to involve sensory qualities, an active phenomenal self, and a first-person perspective. it would be easy to present dozens of very similar descriptions referring to the phenomenology of self-luminous and “pure” awareness, mostly from the advaita and tibetan buddhist traditions, see kunsang, schmidt, & tweed (2012), rgyal-ba-g’yung-drung (2017), namgyal (2006), namgyal (2019), rinpoche & namgyal (2011). are such descriptions intelligible? metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 19 sentation of this dimension. self-luminosity may therefore be a way of talking about wakefulness: wakefulness is always there whenever there is conscious experience at all, autonomous and largely independent of the ever-changing kaleidoscope of surface phenomenology, but it comes in different degrees of intensity and somehow it “radiates.”13 third, self-luminosity – if read not as referring to some mysterious form of visual content, but as an epistemological metaphor – also bears a direct relation to epistemicity, arguably the most important constraint to be extracted in the following section. mpe is a necessary condition for conscious knowledge to be possible, and this includes self-knowledge. in the context of a visual metaphor, for something to be “self-luminous” means that it possesses a quality of intrinsic epistemicity, because it autonomously makes itself knowable. think of a single burning candle in an unobstructed, dark room. it not only illuminates the room, but also endows itself with visibility. it is not something that needs to be discovered or gradually constructed with epistemic effort, in an agentive, top-down manner. if we take the visual metaphor seriously, mpe must be continuously self-revealing (strawson, 2017) and transcendental (higgins, 2013, p. 133) at the same time – a self-intimating internal process that first creates the possibility of knowledge. 2.2 consciousness as such in western philosophy of mind let us now extract three further semantic constraints for a working concept of “pure consciousness”, but this time from western analytical philosophy. below are two passages from a now-classical paper by g. e. moore, titled “the refutation of idealism”. the second one is a canonical phenomenological description of the relevant kind of introspective attention: 13“self-luminosity”, under this second interpretation i am here offering, may therefore be best treated as a visual analogy for an amodal type of phenomenal character, as an attempt to express what it is like: a consciously experienced visual quality like “redness” cannot be analyzed merely as hue plus saturation since it also always varies along a third subjective dimension of “brightness”. phenomenally experienced wakefulness might be a like a domain-general form of brightness which comes in varying degrees of intensity. zero brightness means zero phenomenal experience. the subjectively experienced brightness of a red object represents sheer stimulus strength for the organism, that is, the luminance of a visually given physical object – and perhaps mpe literally turns out to be a mode-neutral form of brightness. for example, the internal “luminance” in question could be a dynamical property of some part of the neural body. attending to “luminous emptiness” might really be attending to the autonomous activity of this body part, and it might even be the case that, as a causal consequence of additional representational resources, the local body part begins to reflexively represent itself (for an excellent discussion see bachmann, 2014, p. 139). therefore, if we want to understand the phenomenal concreteness of mode-neutral and contentless wakefulness, then we have to ask: what, for bare wakefulness, is the non-sensory equivalent of the raw stimulus intensity caused by the physical property of luminance in a given visual target and which, in the domain of visual consciousness, leads to the experiential quality called “brightness”? is there something like “mode-neutral brightness”, and can it be “made reflexive” by introspective attention? what exactly is the stimulus source, its hidden cause? if we read “self-luminosity” as a visual analogy, then the question becomes: what exactly is “luminant” here, and is it perhaps a part of the neural body itself? metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 20 […] when we refer to introspection and try to discover what the sensation of blue is, it is very easy to suppose that we have before us only a single term. the term “blue” is easy enough to distinguish, but the other element which i have called “consciousness” – that which a sensation of blue has in common with a sensation of green – is extremely difficult to fix. […] and in general, that which makes the sensation of blue a mental fact seems to escape us; it seems, if i may use a metaphor, to be transparent – we look through it and see nothing but the blue; we may be convinced that there is something, but what it is no philosopher, i think, has yet clearly recognised. (moore, 1903, p. 446) case study #7. […] the moment we try to fix our attention upon consciousness and to see what, distinctly, it is, it seems to vanish: it seems as if we had before us a mere emptiness. when we try to introspect the sensation of blue, all we can see is the blue: the other element is as if it were diaphanous. yet it can be distinguished if we look attentively enough, and if we know that there is something to look for. (moore, 1903, p. 450) first, it is noteworthy that moore actually uses the term “emptiness”, which plays a great role in buddhist philosophy (cf. the preceding quotation). second, it is noteworthy that he also uses a visual metaphor to describe the mode-neutral phenomenal character in question. his paper is the locus classicus for the concept of “phenomenal transparency”, a property of some conscious states which he called “transparency” or “diaphanousness”. today, a standard definition of “phenomenal transparency” is that it essentially consists in only the content properties of a conscious mental representation being available for introspection, but not the fact that it also has non-intentional or “vehicle properties”. in other words, it is not experienced as a representation. introspectively, we can access its content, but not the content-formation process itself. as i have extensively argued elsewhere (metzinger, 2003a, 2003b, 2014), the empirically most plausible explanation for this fact is that, relative to the given temporal resolution of introspective attention, earlier processing stages are invisible to the system, because all competing interpretations have already been disambiguated. in philosophy, it is often assumed that transparency in the sense of the standard definition above is a property of all phenom enal states. but of course, the standard assumption is incomplete: opaque phenomenal representations also exist (whereas unconscious states are neither transparent nor opaque in this sense). the most notable phenomenological examples of opaque state-classes are consciously experienced thoughts: we experience them as mind-dependent and internally constructed, as mental representations that could be true or false. apart from that, some types of emotions, pseudo-hallucinations, and lucid dreams are also subjectively experienced as representational processes. they make the possibility of misrepresentation introspectively available, because the associated subjective confidence changes over time. a spectrum therefore exists between phenomenal opacity and phenomenal transparency, and any given metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 21 content can vary along this spectrum. phenomenally opaque processes sometimes appear to us as deliberately initiated cognitive or representational processes. however, they can also appear as automatic or spontaneously occurring, as lim ited or even global phenomenal simulations; they can draw us into a manifest daydream; and they frequently seem not to be under the experi ential subject’s control (for more, see metzinger, 2003a, 2003b, 2014; pagnoni, 2019, p. 308). moore presents a phenomenological argument against the transparency of consciousness as such. many authors have read moore as if he were talking about the “transparency of qualia.”14 however, the relevant point concerns not the transparency of awareness or qualia, but rather the fact that consciousness itself can be made phenomenally opaque. moore’s much deeper point is that conscious experience as such has a unique phenomenal character sui generis (namely, a secondorder “awareness-of”, a relational “signature of knowing”), and that this character can sometimes be detected by introspective attention. in fact, moore’s self-stated goal in introducing “transparency” is actually “to try to make the reader see it” (moore, 1903, p. 450). his important phenomenological claims are the following: • (m1) there is one most general phenomenal property, which is shared by all sensory qualities: → “consciousness” as such, pure awareness, phenomenality per se. • (m2) under standard conditions, this global property is “transparent” (moore, 1903, p. 446), a “mere emptiness” or “diaphanous”: → it is usually not explicitly experienced, but it can be. • (m3) this property is evasive: → it “seems to vanish” under attentional agency; i.e., if we actively try to “fix our attention” on it (moore, 1903, p. 450), thereby creating an epistemic agent model (see footnote 7). • (m4) this property can become phenomenally opaque, under two conditions: → we look “attentively enough”, → and we “know” that there is a possible object for introspective attention (moore, 1903, p. 450). • (m5) it is difficult to simultaneously direct and sustain introspective attention to the global property in question and to concrete perceptual qualities: → moore found no philosopher in the literature who was “able to hold it and blue before their minds and to compare them, in the same way in which they can compare blue and green” (moore, 1903, p. 450). 14for an excellent, well-researched, and substantial discussion containing further references, see hellie (2007). metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 22 • (m6) awareness is relational and consciousness is a second-order epistemic process: → having a sensation is an awareness of something, and consciousness is the knowledge that this awareness currently exists. in his own words: the true analysis of a sensation or idea is as follows. the element that is common to them all, and which i have called “consciousness”, really is consciousness. a sensation is, in reality, a case of “knowing” or “being aware of” or “experiencing” something. when we know that the sensation of blue exists, the fact we know is that there exists an awareness of blue. (moore, 1903, p. 449) • (m7) consciousness is meta-awareness: → the type of epistemic relation is identical: “awareness-of” is the same relation in sensation as it is in becoming aware of this awareness. once again, in moore’s own words: this relation is just that which we mean in every case by “knowing”. to have in your mind “knowledge” of blue, is not to have in your mind a “thing” or “image” of which blue is the content. to be aware of the sensation of blue is not to be aware of a mental image of a “thing,” of which “blue” and some other element are constituent parts in the same sense in which blue and glass are constituents of a blue bead. lt is to be aware of an awareness of blue; awareness being used, in both cases, in exactly the same sense. (moore, 1903, p. 449) given this analysis, one can now interpret moore as saying that a) an aperspectival and non-agentive form of meta-awareness is co-instantiated with all forms of conscious perceptual knowledge; that b) the second-order epistemic relation of an awareness-of the current existence of an awareness-of some specific perceptual phenomenal character is mostly unnoticed (i.e., like a mind-independent and transparent medium in which this character simply “appears”); but that c) we can make it phenomenally opaque by attending to it in the right way. from this interpretation we can extract three further phenomenological constraints to semantically enrich our six-dimensional working concept of mpe: • introspective availability (pc4): we can sometimes actively direct introspective attention to consciousness as such and we can distinguish possible states by the degree of actually ongoing access. • epistemicity (pc5): the phenomenal experience of knowing, which comes in degrees and can also be described as the subjective quality of confidence.15 15to avoid any misunderstandings at this point: the phenomenology of knowing can exist without metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 23 • transparency/opacity (pc6): like all other phenomenal representations, mpe can vary along a spectrum of opacity and transparency. constraints 4 – 6 can now be interestingly discussed in the context of the first three constraints presented in the previous section. here are some examples. 2.2.1 introspective availability introspective availability has been endorsed not only by g. e. moore, but by many eastern contemplative traditions. however, it is important to always clearly distinguish between the dispositional property of availability for introspective attention and actually ongoing introspective access to the relevant phenomenal character. “availability for introspective attention” is a functional property a given internal state may or may not have, whereas ongoing introspective access presupposes the capacity for active attentional control, at least at the beginning of an episode. this actually ongoing access may then change the overall state.16 at the same time, it is also clear that the phenomenology of ongoing introspective access can only be a transitory phase, and that it is not itself an element of the relevant conscious state, which is devoid of subject/object structure, mental agency, and any form of conceptual or propositional mental content. here is how the relevant transition is described from a tibetan buddhist perspective: case study #8. the moment you recognize, it is already seen. there is nothing extra remaining that you missed. this is not like space looking at itself, because space does not see anything. when your mind, which is cognizant, recognizes itself, you immediately see that there is no “thing” to see. it is already seen in the same moment. at that very moment there is no thought, because the present thought has naturally vanished. (rinpoche, 2013a, p. 79) in the following description of contemplative phenomenology, wakefulness and epistemicity are interpreted as aspects of a self-sustaining, functionally autonomous process that precedes introspective access: there being knowledge in the epistemological sense. pc5 is an exclusively phenomenological constraint, and it can be satisfied by “noetic” experiences (cf. footnote 2) as well as intuitions, hallucinations, illusions of control, complex delusional states, and all other forms of consciously experienced misrepresentation that are not experienced as misrepresentations. 16the concept of “introspective attention” is only indirectly related to classical theories about “metacognition” (flavell, 1979; metcalfe & shimamura, 1996). introspective attention is nonconceptual, a subsymbolic process that has nothing to do with “beliefs”, “thinking about thinking”, or declarative knowledge about one’s own states. rather, it is related to the optimization of precision expectations relative to internal stimuli, cf. wiese and metzinger (2017), pagnoni (2019). in terms of contemplative practice, the goal will be an effortlessly sustained, entirely non-conceptual and non-propositional state of meta-awareness. for a useful conceptual clarification, see dunne, thompson, and schooler (2019). metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 24 case study #9. empty cognizance is an original state that is never lost. it merely needs to be known. we need the knowledge of selfexisting empty cognizance – the knowing itself to be empty and cognizant. (rinpoche, 2013b, p. 114) there seem to be three elements here: first, a functionally autonomous state of bare subjective confidence, of “knowing without an object”; second, a transient episode of introspective attention directed at this state; and third, the emergence of a new phenomenal state which dynamically integrates both elements (see also rinpoche, 1995, pp. 106–129). if mpe were to be the first element – in terms of minimally complex, full-absorption states characterized by a full satisfaction of the low-complexity constraint (as discussed above) – then it would be a functionally autonomous kind of internal process that could in principle be actively introspected, but mostly is not. let us now assume the process in question would be the process by which the central nervous system activates itself, creating the property of tonic alertness (metzinger, 2019, sec. 6). if attended to, the original process could be “made reflexive” or “ignited” into a recurrent loop, thereby causing a new dynamical state to appear – plausibly yielding a distinct, new type of phenomenal character that we might provisionally call “self-aware wakefulness”. compare this classical phenomenological text from the 16th century: case study #10. […] it is an aware emptiness that defies any description of being such-and-such – it is inexpressible and yet it can be experienced. […] it is a wakefulness for which no words suffice. it is not a definable entity, but at the same time, it is a self-knowing aware emptiness that is clear, lucid, and awake. (namgyal, 2001, p. 29) this demonstrates an important difference. for moore, concrete introspective access is mere meta-representation that creates phenomenal opacity by making a global and previously transparent aspect introspectively available for a robust conscious self. however, contemplative traditions clearly favour not higher-order representation, but a first-order, reflexive theory of meta-awareness (ramm, 2019, p. 7).17 here, we find the notion of non-egoic reflexivity: a minimally complex, func17for this reason, many contemplative traditions (e.g. rdzogs chen) stress that the relevant state cannot be directly introspected because it is a self-constituting process which is functionally independent of all possible mental efforts a meditator could ever make, and because any such effort would a) contract the overall experience into an introspective first-person perspective (the “meditative self”) and b) immediately impose a subject/object prior onto it. in western analytical philosophy of mind, see for example michelle montague for the point that awareness of awareness cannot be transformed into direct introspective awareness and that it cannot be directly introspected; see montague (2017, p. 22), and for the notion of a “ ‘self-constituting’ representational relation”, see montague (2016, p. 59). in 2009 uriah kriegel first developed this idea into the “attention-shift model of introspection”, which points out that some phenomenological elements like “peripheral awareness” may well be manifest without being introspectable. part of the metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 25 tionally autonomous baseline state that dynamically represents itself on a subsymbolic level (josipovic, 2019, sec. 3). perhaps we could metaphorically describe the phenomenological correlate as “an emptiness that has awoken to itself”. on the other hand, the dispositional property of introspective availability itself can be interpreted as a second functional dimension (as opposed to a content-related one). nevertheless, it has to do with what bayne, hohwy, and owen (2016, p. 409) call a “gating” of the contents of consciousness. mpe itself may be a very simple, highly abstract, and unstructured type of phenomenal character, but there is variability along a dimension of actually ongoing introspective access. it might open the organism to other kinds of conscious contents, for example by making it sensitive to incoming perceptual stimuli or enabling orientation in time and space. adding introspective attention would transiently create a sense of effort, a phenomenal self, and an internal, self-directed first-person perspective. this could then be followed by brief periods of effortless meta-awareness (i.e., the non-egoic but reflexive form of “empty cognizance” described above) in which the first-person perspective briefly dissolves until the next train of apparently spontaneous, task-unrelated thought ensues. interestingly, this emerging picture is in full accordance with recent models of what actually happens during the process of mindfulness meditation (hasenkamp, 2018; hasenkamp, wilson-mendenhall, duncan, & barsalou, 2012). 2.2.2 epistemicity epistemicity can now be read as a prediction about the phenomenal character of more complex episodes in which introspective attention has been transiently added to a simple state of mpe. attentional agency creates the phenomenology of a knowing self, an epistemic agent model (metzinger, 2017; see also limanowski & friston, 2020, this issue), and such states would then still be characterized by an explicit subject/object structure. here, however, the additional idea is that there may be a non-agentive variety of “knowingness” in low-complexity mpe states lacking this structural feature but characterized by a more globalized and effortless form of meta-awareness. contemplative practice ultimately dissolves itself and reveals the phenomenology of subject/object structure to be merely “parasitic” (higgins, idea is that introspection does not generate an entirely new representational state, but involves a structural reorganization: “rather, it involves reorganizing the center/periphery structure of one’s overall experience by transforming one’s peripheral inner awareness of one’s current experience into a focal one” (kriegel, 2009, p. 183). see also giustina and kriegel (2017), kriegel (2012). while the idea that peripheral awareness is “inbuilt” (kriegel, 2020) is actually fully compatible with the metaphysics of the relevant traditional philosophical theories, it contradicts their phenomenology, which at many places explicitly negates the idea of a periphery (see footnote 8). it also overlooks the fact that the process kriegel imagines would create an explicit representational transition in the conscious self-model, namely by creating not only focalized structure, but also a new epistemic agent model at the other end of the internal “arrow of intentionality”, an “introspectively knowing self” forming the origin of the first-person perspective. see footnote 7 above and metzinger (2017, sec. 2.5). metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 26 2013, p. 124) on an entirely non-dual and more fundamental level of experience. after attentional agency ends, a distributed continuum of subjective confidence remains. the prediction would be that this should generate the phenomenal character of an all-pervading field of “distributed confidence” or “primordial knowing”. recall that moore introduced “awareness” as an epistemic process (m6, m7). in contemplative phenomenology, the specific phenomenal character of epistemicity i am here trying to isolate has sometimes been termed “the flash of knowing that gives awareness its quality” (mañjuśrīmitra et al., 1987, xxiii) and it is often described as a content-independent “cognizant quality”: case study #11. in that moment, we don’t see any concrete thing whatsoever. there is an immediate knowing that the essence is empty. there is something that cognizes that the mind is empty, and this cognizant quality is indivisible from the emptiness itself. at the moment that this is an actuality, you don’t need to do anything more. (rinpoche, 2013a, p. 89) mpe is not a mystical experience, but in research on mystical experience, this phenomenological feature has often been described as a unitary state exhibiting a “noetic” phenomenal quality, while still lacking any specific form of content (wainwright, 1981). in such states one would predict a phenomenal signature of knowing which a) has no object (i.e., a form of knowing or subjective confidence without content), and which b) has been globalized (i.e., which is not object-oriented or tied to a localized model of a knowing self any more). here is one phenomenological description from an advanced meditator: case study #12. “[…] there’s no personal point of view, it’s the world point of view, it’s like the world looking, not me looking, the world is looking.” (quoted from ataria, dor-ziderman, & berkovich-ohana, 2015, p. 141) compare this with the following, more complex phenomenological analysis: case study #13. the first characteristic is that it has no boundaries, no fence round it, no edges; it’s absolutely unlimited in all directions. the second characteristic is that it is absolutely clear, clean, empty of contamination. it is utterly simple, totally transparent, empty of everything but itself, empty even of itself, clearer than glass, cloudless, an infinite sky. the third characteristic is that it is also full of the world. because it’s empty, it’s full – full of the scene, whatever the scene is, absolutely united with it. the fourth characteristic is that it is awake, it’s aware, it’s conscious. and the fifth characteristic is that it is right where you are. (harding, 2000, p. 65)18 clearly, mpe is neither an internal model of a self-as-subject, nor a model of an object. philosophically, this creates a conceptual tension: if we want to follow 18for more on the relevance of douglas harding, see ramm (2019). metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 27 moore in interpreting the concept of “awareness” as an epistemic relation, then it becomes difficult to specify the subject-argument place. what exactly are the relata of this special relation? what is the known, and what is the knower? what is the “something that cognizes” in case study #11, and what do we make of the fifth characteristic in case study #13? it may turn out that the experience itself isn’t really contentless, but only described as such, and it may still be non-egoic in the sense that no knowing self (i.e., no epistemic agent model) is part of its content. but, under an epistemological analysis, who or what is the epistemic subject? i want to submit that the subject simply is the organism or conscious system as a whole, and that mpe could be an internal model of an epistemic space currently opened by it – and that the as-yet-unpartitioned nature of this space is what explains the abstract phenomenal character of epistemicity. call this the “esm”thesis (esm for “epistemic space model”): mpe is an integrated inner representation of a space of possible knowledge states or epistemic policies, defined by the organism’s set of epistemic capacities. consciousness would then appear whenever an information-processing system has a) created an internal epistemic space, and b) activated a model of this very epistemic space within it (metzinger, 2019, sec. 8.4.1). using moore’s terminology, we could say that typically this model is “diaphanous” or “transparent.”19 just like physical space, epistemic space creates an allocentric frame of reference, it is unbounded and non-centred, while of course its variable content is perspectival: being a simulation of an embodied behavioural space, it is egocentric, there are perceptual horizons, and so on. however, just as physical space cannot be touched, felt, or grasped, mpe as such cannot be “touched by attention” or “grasped by thought”, because this would only introduce subject/object structure into a phenomenal baseline state, in which we have already been immersed all along. in my view, the best phenomenological metaphor for “pure consciousness” or mpe, to be found in many places in the contemplative literature, is the epistemic-space metaphor: space is the example for mind essence, because space is unmade. but mind essence is not totally like space, in that space cannot think. space has no knowing. our mind is cognizant emptiness – empty like space, but with a natural knowing. that union of cognizance and emptiness is seen when recognizing. it is immediate, like the example i mentioned of pointing into mid-air. you do not have to wait to raise your arm for your finger to touch space – you are already touching space, all the time. you do not have to move your hand forward; the contact 19by definition, an epistemic space is a space of possibilities, it contains every possible epistemic scenario and every dynamic partitioning of itself that could ever take place, everything that could potentially be known and experienced by a given system. therefore, the internal model of such a space, of the esm as such, would be characterized by a very high degree of counterfactual invariance: whatever the system could know or experience would necessarily take place within it. here, “maximal counterfactual invariance” (for the esm itself) and “maximal counterfactual richness” (for its content) could be an excellent way to analyze the moorean concept of “transparency”. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 28 is already occurring and has been your entire life. all you have to do is recognize that it is taking place. it’s the same with mind essence. (rinpoche, 2013b, p. 112) if moore is right, then mpe actually is an epistemic relation (“awareness-of”). if mpe is successfully introspected (i.e., without – as moore and the tibetan rdzogs chen-scholars of the 8th to 14th centuries tell us – trying to “fix” attention on it in an effortful, agentive way), then it is natural to expect a distinct phenomenal character of knowing. however, in this very special case there would be no correlated epistemic object and all the other negative constraints listed in section 2 would be simultaneously satisfied as well. “epistemicity” therefore refers to the mere phenomenal signature of knowing, i.e., a functionally autonomous experience of subjective confidence or certitude, a representation of model evidence as such. mpe could then be the special phenomenal character of non-directed, non-agentive and aperspectival “knowingness” – but with or without a perceptual object, depending on the degree of absorption discussed above.20 an additional phenomenological aspect of pc5, the epistemicity constraint, is that mpe states are sometimes described as reflexive, and sometimes not. that is, while awareness as such has no external intentional object it is sometimes described as having itself as an object, as being self-reflexive in a way that is nonconceptual and which does not necessarily involve the experience of a substantial, separate self (mackenzie, 2008). if there were such an intrinsic and fundamental form of reflexivity, then one could imagine it either as co-existing with specific content or as a pure stand-alone phenomenon. for example, it would be conceivable that it co-exists with complex content during the daytime while normally remaining as entirely unrecognized by the subject of experience. at the same time one can now make sense of the idea of how it might occur in dreamless deep sleep, namely, as an“island of awareness” the content of which “is not shaped by sensory input from either the external world or the body and which cannot be expressed via motor output” (bayne, seth, & mas20this is interesting, because there also seems to be a graded but distinct phenomenal character of “certitude” or subjectively experienced confidence, which we can now analyse as a nonconceptual representation of the “degree of belief”, “plausibility”, “trustworthiness”, or “reliability” assigned to an inner or outer perception, or any other consciously experienced epistemic state. we could also term it the “phenomenology of validity” (or the “noetic quality” mentioned in footnote 2), and it is clearly an essential aspect of mpe. it can be found in complex hallucinations, various psychiatric syndromes, or during epileptic seizures as a result of cortical dysfunction; cf. picard (2013). what makes the phenomenal character of subjective confidence especially interesting from a philosophical point of view is a) that it relates directly to centuries of theorizing about consciousness as “knowing that one knows”, and b) that there are now unifying conceptual frameworks describing confidence as bayesian probability, bridging different fields of psychology and neuroscience and unifying disparate research perspectives. i will come back to this point in later publications, see also meyniel, sigman, & mainen (2015). epistemicity could be the way bayesian probabilities and model evidence are represented internally, by the system itself and on the level of conscious processing. mpe is a highly relevant target for research, because a better understanding of epistemicity may help us to study this aspect in isolation. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 29 simini, 2019, p. 7). the possibility of a fundamental, non-egoic form of reflexivity built into consciousness itself has been discussed for centuries by buddhist philosophers (finnigan, 2018; williams, 1998). following brentano’s classical idea of a single state simultaneously directed at an intentional object and itself (brentano, 1874/1973, p. 153, 2012, p. 25), many fine-grained conceptual models have been developed in more recent analytical philosophy of mind (montague, 2016, chapter 3, 2017; peters, 2013), and the issue is currently attracting a lot of attention (ganeri, 2017; josipovic, 2019; kriegel, 2019; strawson, 2017; thompson, 2011). please note how i am here only interested in the phenomenal character of intrinsic reflexivity itself, in the reported experience of an explicit, introspectively available, nonegoic, non-conceptual, and exclusively self-directed awareness. this character is described in the following case study: case study #14. it’s absolutely alive to itself, fully conscious, and it’s conscious of itself as no-thing, colorless, featureless, totally empty and totally full. the space is conscious of itself at the near end, but it has no personal characteristics. (harding, 2000, p. 20)21 21what makes epistemicity even more interesting – and potentially relates it to transparency/opacity – is the additional phenomenological fact that, while low-complexity mpe states are often described as states of “pure knowing”, apparently very similar configurations are sometimes also later reported as conscious episodes of “pure being”. what is the relationship between “pure knowing” and “pure being”, if once again we strip away all epistemological and metaphysical interpretations of the relevant reports, staying with just the phenomenal character itself? there obviously exists a globalized, non-agentive phenomenology of epistemicity as well as a phenomenological counterpart of pure onticity. the first type of phenomenal character (i.e., globalized epistemicity without an object, often described as “empty cognizance”) can already be found in the first quotation presented above, case study #11. let us look at two examples reporting the experience of a globalized and non-dual form of “empty onticity”, of mere existence as such, the first one coming from william james, the “father” of american psychology himself: case study #15. during the syncope, there is absolute psychic annihilation, the absence of all consciousness; then at the beginning of coming to, one has at a certain moment a vague, limitless, infinite feeling – a sense of existence in general without the least trace of a distinction between the me and the not-me. (james, 1890/1950, p. 273) here is another report of the process of regaining consciousness following a severe accident, as an illustration of what i mean by the “phenomenology of onticity”, again without any form of subject/object dichotomy: case study #16. there was something, and the something was not the nothing. the nearest label for the something might possibly be “awareness,” but that could be misleading, since any awareness i’d ever had before the accident was my awareness, my awareness of one thing or another. […] in contrast, this something, if it be called awareness, had no i as its subject and no content as its object. it just was. (sullivan, 1995, p. 53) this is puzzling and interesting at the same time. there seem to be borderline phenomenologies connecting the region of “pure awareness” with that of “pure being” in phenomenal state-space, linking the experience of knowing as such with the experience of being as such. one logical possimetzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 30 2.2.3 transparency/opacity finally, transparency/opacity shows why mpe is such an interesting limit case for theories of consciousness. please recall that a full satisfaction of low complexity demanded a complete absence of intentional content for mpe, and that “transparency” was defined on the representational level of description (metzinger, 2003b). on this level of analysis, transparency is a property of phenomenal representations, and is brought about by the fact that only their intentional or content properties are available for introspection, not the properties instantiated by the dynamics of earlier processing stages (i.e., of what often is called the “vehicle” in traditional philosophical terminology). from this, it follows that for the special case of a putatively contentless state like mpe, if it really were transparent, there would be nothing introspective attention could ever discover – there would simply be no intentional properties that could ever be available for introspection. if, however, mpe were phenomenally opaque (if it were sometimes experienced as a representation), then we would be forced to conclude that pc1 (wakefulness) and pc3 (self-luminosity) actually refer to an introspective experience of certain properties of the physical carrier itself. for example, such non-intentional “vehicle properties” could simply be earlier processing stages in the brain, certain dynamic aspects of the very construction process that creates the level of phenomenal representation in the first place. as we actually have numerous reports of mpe-like states across the centuries and across cultures, it is now plausible to draw a second conclusion: at least some of these reports may actually be reports about an entirely non-intentional process, not merely a mode-neutral one (josipovic, 2019). now imagine that you would make a previously transparent episode of mpe phenomenally opaque by just the right kind of “looking attentively enough” while at the same time not trying to agentively “fix” attention on it, exactly in the way moore described. a third conclusion is that if the fundamentality constraint is satisfied and mpe really underlies, envelops, or permeates all other forms of conscious content, then – if it were made opaque – we would expect a global phenomenal effect. in richly superimposed states involving perception of body and environment everything should then begin to be experienced as made out of pure consciousness, as merely representational, as a virtual model lacking an inherent, mind-independent existence of its own – not as a world, but as the content of an epistemic space which is now explicitly co-represented (cf. case study #12). bility is that “epistemicity” as well as “onticity” could refer to almost the same kind of global state. this would be again be characterized by the absence of discursive thought, perceptual content, spatiotemporal self-location, agency, etc. the only difference would be that in the first category this global state is still experienced as representational (i.e., as a contentless form of awarenessof ) while in the second case it is not. perhaps the first type of state is reflexive (and therefore experienced as self-representational), while the second one is not. in other words, the relevant difference would then be the difference between a phenomenally opaque and a phenomenally transparent representation of the inner process creating tonic alertness. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 31 in low-complexity situations this would lead to the phenomenology of “empty cognizance” discussed above. if, however, there were rich superimposed content while the internal model of our epistemic space were transparent, then this content would simply appear as “real” and directly given, as the world we live in. looking at mpe in this way begins to yield a theory of consciousness: “appearance” results from a transparent representation of meta-confidence. phenomenally experienced “realness” is a transparent representation of certainty or high bayesian probability – the globalized but currently invisible knowledge that we know. in low-complexity situations the prediction is an experience of “pure being” (see footnote 21). in richly superimposed states, a world would appear. wakefulness is the phenomenally opaque counterpart to the transparent, conscious experience of “realness”. in full-absorption episodes there would be neither introspection nor reportability, just a first-order model of the non-intentional, physical carrier as such. in absorbed but weakly superimposed states already permitting introspective access the phenomenological prediction is that the phenomenal quality of epistemicity (pc5) would be co-instantiated with wakefulness (pc1) and self-luminosity (pc3) – an experience of subjective confidence related to bare wakefulness and the functionally autonomous “mode-neutral brightness” discussed above. 3 the difference between phasic and tonic alertness 3.1 terminology first: alertness vs. wakefulness the empirical literature is often unclear and ambiguous regarding the use of notions like “arousal”, “vigilance”, “sustained attention”, “wakefulness”, and “alertness”. however, there are two semantic elements present in a large majority of scientific treatments: first, epistemic capacity, and second, absence or nonspecificity of representational content. for example, oken, salinsky, and elas speak of “activation states of cerebral cortex that impact the ability to process information where the activation itself contains no specific information” (oken, salinsky, & elsas, 2006, p. 1885). it is exactly the representation of such states which is relevant for mpe, but viewed from a philosophical perspective, the existing terminology to describe such states is highly inconsistent. as oken and colleagues write: unfortunately, no terms are ideal to describe these states of cortical activation since most terms are in broad use with varied associations and there are not [sic] perfect physiological markers. the term vigilance, in particular, has unfortunately been used in different ways by different groups of scientists. […] the field has been hindered by inconsistent or poorly defined terminology. researchers should be particularly careful about the usage metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 32 of the term vigilance, understanding there are factors contributing to vigilance other than just arousal on the sleep–wake dimension. avoidance of the term vigilance because of its varied definitions would be most helpful. the use of the term tonic alertness may be preferable since it does not already have a confusing history. (oken et al., 2006, pp. 1885, 1991) for the purposes of this paper, let us clearly distinguish between the physical, the functional, and the phenomenological levels of description. i will use the following definitions: “arousal”: • a graded physical property of the human brain. → a purely physical, but variable boundary condition determining the depth of cortical information processing available to the organism as a whole. → a vital parameter that must be predictively controlled (e.g., to generate the sleep–wake cycle). “tonic alertness”: • a graded functional property determining the capacity for sustained attention. → a causal function resulting from the successful control of arousal over longer periods of time, in the absence of an external cue. → a functional property causally enabling cognitive capacities like orientation, executive control, attention, and epistemic agency on the mental level. “wakefulness”: • a graded phenomenal property which is sometimes introspectively accessible. → a representation of tonic alertness. → the major component of the phenomenal character of mpe (pc1). → the primary dimension of phenomenal state space. what is the relationship between these three properties? in humans, the functional property of tonic alertness is realized by certain physical properties of the brain, and we do not yet know what exactly these properties are. conceivably, it could be realized by very different physical properties in machines that create an integrated internal model of their own epistemic space as such. for neurotypical humans, it is highly plausible to assume that a critical level of cortical arousal is the most central metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 33 necessary condition. whether the right level of arousal as such is sufficient for tonic alertness is currently unknown, but unlikely. the phenomenal experience of wakefulness is a representation or, plausibly, a predictive bayesian model of the functional property of tonic alertness.22 both the phenomenal property of wakefulness and the functional property of tonic alertness cannot exist without the brain. but tonic alertness, if it is not represented or internally modelled, can exist without the conscious experience of wakefulness (as in some cases of unresponsive wakefulness syndrome; cf. laureys et al., 2010). put differently, an organism can be tonically alert without knowing that it is alert: consciousness is knowing that one is alert. an organism can embody a rich space of epistemic capacities without having an internal model of this fact. what do we know about the functional property connecting the physical and the phenomenological levels? tonic alertness is a global functional property, in part physically realized by the level of cortical arousal, and it is hypothesized to be one of the core functions of the cingulo-opercular (co) network (sadaghiani & d’esposito, 2015). it is one component of what we call the capacity for “attention” in folk-psychological contexts, and its function for the organism consists in “achieving and maintaining a state of high sensitivity to incoming stimuli” (posner, 2008, p. 193). our neuroscientific understanding of the co network is gradually improving, and it already provides us with important pointers that are consistent with our phenomenological analysis: tonic alertness is an a) sustained and b) internally initiated “preparedness to process and respond” which c) implies a capacity for attentional agency, of being able to “co-engage” phasic alertness and selective attention directed at a specific task (sadaghiani & d’esposito, 2015, p. 2763). just like the quality of mindfulness cultivated by contemplative traditions, it can be described as a global and general mechanism “of keeping cognitive faculties available for current processing demands and holding unwanted activity at bay” (sadaghiani & d’esposito, 2015, p. 2764). it is also empirically plausible to assume that tonic alertness is a central causally enabling factor in the realization of two other functional properties which are distinct but closely related: orientation and executive control. here, it is important to note that all three functions – alertness as such, orientation, and executive control in terms of attentional agency – are epistemic capacities which, at a given point in time, a biological organism may have or not have. they are the foundation for what blanke and metzinger called a “strong first-person perspective” (blanke & metzinger, 2009, p. 7). please also note how the paradigmatic cases of “pure” consciousness discussed above are precisely characterized by an absence of spatiotemporal self-location 22it is important to understand that the representational level of description opens up the possibility for misrepresentation, so some instances of mpe could be misrepresentations of tonic alertness. plausibly, the phenomenology of wakefulness can exist in low-arousal states (as in lucid dreamless sleep, metzinger, 2019, sec. 5). this raises many interesting questions, for example whether lucid dreamless sleep would have to count as a misrepresentation of tonic alertness, if you will: as a hallucination or a dream of phenomenal wakefulness. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 34 (i.e., during full-absorption episodes, the subject is not orientated towards time or space) and absence of attentional or cognitive agency (i.e., mental actions like controlling the focus of attention or one’s thought processes). the current proposal says that phenomenality (i.e., a minimal form of conscious experience) emerges precisely whenever the first functional property – tonic alertness, the causal enabler of orientation and executive control – is probabilistically represented. in other words: minimal phenomenal experience is a predictive model of a specific set of epistemic capacities and the representational space opened up by them, for example, “high stimulus-readiness”, “orientation readiness”, plus the mere “capacity for mental self-control” on the level of attention and cognition. however, it is a temporally shallow model as it accumulates evidence only on a very short timescale and does not entertain complex state transitions or sequences over time (friston, 2018; friston, rosch, parr, price, & bowman, 2017, p. 486). this point is in full keeping with the low-complexity constraint and the well-known fact that meditators try precisely to end any sort of inner narrative, by continuously bringing their attention back to the fine details of the present moment, thereby collapsing the deep temporal structure of their model of reality. at the same time it casts doubt on attempts to frame consciousness itself (instead of self -consciousness) as directly related to temporal thickness (friston, 2018; please also see the contribution by george deane, 2020, in this special issue). interestingly, if viewed as a running bayesian model, mpe has a certain degree of model evidence. if this objective property is introspectively available, then this will generate epistemicity, the subjectively experienced confidence that the capacities mentioned above will actually lead to the creation of internal states carrying epistemic value in the future. more simply, we could say that mpe is an expectation of knowledge. if this expectation is transparent, it necessarily creates appearance: the certainty that something knowable exists. 3.2 the phenomenology of phasic vs. tonic alertness we can now formulate the empirical hypothesis that what satisfies wakefulness and epistemicity on the phenomenological level is the existence of a highprobability bayesian representation of tonic alertness. however, on the functional level of description, it is important to first distinguish tonic alertness itself, which is long-lasting and independent of external cues, from short episodes of heightened but externally triggered arousal. the functional difference between these two different kinds of alertness is philosophically relevant, because it generates two fundamentally distinct kinds of phenomenal character. a paradigmatic example is the rapid change in attention due to a brief event, like an unexpected sound in your environment. the phenomenology of phasic alertness is the experience of being forced to take a new attentional perspective, of re-orientating the “arrow of intentionality” (metzinger, 2006, 2017). it is a brief episode in which the first-person perspective as consciously experienced is shifted metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 35 to a new epistemic target, in a bottom-up, saliency-driven way. scientifically, phasic alertness is typically studied with infrequent, unpredictable warning signals preceding the presentation of a stimulus, shortening reaction time. whereas intrinsic, tonic alertness is the basis for operations such as orientating and selective attention, these operations themselves have a distinct phenomenal character of their own: the dynamic process of relocating attentional resources or suddenly shifting to a new target for executive control, typically following an experience of unexpected salience detection. what makes an experience salient depends on many different factors: perceptual features like contrast, clarity, unexpected size or intensity clearly play a causal role, but the overall emotional, motivational, or cognitive context may be equally important in triggering the phenomenology of phasic alertness. in the words of weinbach and henik: alertness is commonly divided into two different modes of function: tonic and phasic alertness. tonic alertness, which is also known as “intrinsic alertness,” designates the internal control of wakefulness or arousal in the absence of an external cue in a top-down manner (sturm et al., 1999; sturm & willmes, 2001). phasic alertness, on the other hand, is a short-lived effect of achieving high levels of alertness following a salient external event (posner, 1978, 2008; sturm & willmes, 2001). (weinbach & henik, 2012, p. 1530) the phenomenology of tonic alertness is the non-dual conscious experience of intrinsic and sustained wakefulness. it fluctuates only on the order of minutes to hours.23 it is entirely non-conceptual, not the result of an agentive, high-level inference, or a judgement about the current behavioural state (whether one is asleep or awake). as such, the experience of bare wakefulness has no grain or internal structure, and can therefore be described as a globalized form of “ultrasmoothness” (see metzinger, 2003a, sec. 3.2.10, and the references given in footnote 16). 4 conclusion: mpe as the phenomenology of tonic alertness this discussion leaves us with a very first, semantically enriched 6d-concept of “minimal phenomenal experience”, which can now, for example, be used as a foundation for systematic psychometric studies. mpe is an inner representation of 23how does this relate to the phenomenology of phasic alertness? on a conceptual level, we can interpret tonic alertness as the epistemic capacity to create and sustain sensitivity to incoming perceptual stimuli. this capacity, in turn, is causally related to the epistemic capacity for orientation: the creation of an integrated and egocentric representational space plus the epistemic capacity for executive control of attention. pc1, the phenomenology of wakefulness, is the conscious representation of the first epistemic capacity. perspectivalness, the phenomenology of being an epistemic agent or a “knowing self”, is the conscious representation of the second and third capacities. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 36 tonic alertness, but it comes in variable degrees of constraint satisfaction. it can be coupled with introspective access or not, it can be reflexive or not, and it may be characterized by a variable degree of transparency. obviously, it can also co-exist with many other forms of conscious content, but because it normally functions as the transparent medium or model of an abstract space in which other contents unfold over time, it is mostly unnoticed. what make it so easy to overlook is the unique combination of transparency with a particularly subtle and abstract kind of phenomenal character, leading to an integration of experiential immediacy with a high degree of counterfactual invariance. in a prototypical form, it would be characterized by a mode-neutral form of bare and functionally autonomous “selfluminous wakefulness” and the accompanying epistemic aspect of subjective confidence, experienced as a quality of “empty cognizance.” the phenomenological prototype would be characterized by a complete absence of time-representation and any form of sensorimotor or high-level cognitive content. further, there would be an absence of low-level embodiment in the sense of spatiotemporal self-location, interoception, and affective background, as well as of higher levels of selfhood like attentional control and cognitive agency. mpe is non-egoic self-modelling: as such it is atemporal, selfless, and not tied to an individual first-person perspective.24 it instantiates a unique but highly abstract kind of phenomenal character which can be described on different levels of analysis – for example, as a maximally abstract form of embodiment or bodily self-awareness, as a dynamic representation of tonic alertness, or as an internal model of an epistemic space. as a representation of a pre-given epistemic space or global medium in which the embodied acquisition of knowledge unfolds over time it has the unique feature of combining the absence of subject/object structure with phenomenal transparency. however, in addition to the specific phenomenological profile already described in sections 2 above and by the resulting 6d-model, it also has certain global aspects that can be introspectively discerned – if there is a prior to initially guide introspective attention, if, as g. e. moore put it in 1903, “we know that there is something to look for”: • openness: the phenomenology of tonic alertness is the experience of openness: the organism is now sensitive to incoming stimuli, it can perceive the environment and internal states of its own body, it is open to the world and it represents this very fact. in this sense, mpe is a non-egoic, non-conceptual form of self-knowledge. • simplicity: all the negative phenomenological characteristics and the low-complexity constraint introduced in section 2.1 are satisfied. this 24if all of the above is correct, mpe is a counter-example to the “projective geometry” model of consciousness formulated by williford, bennequin, friston, & rudrauf (2018). because of its fundamentally aperspectival nature, it is a phenomenal state not covered by this theory. if one additionally assumes that mpe actually is a dynamic, bayesian model of an internal bodily process, then it can be described as bodily self-representation from a god’s-eye point of view. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 37 is the main reason why the phenomenology of tonic alertness is a prime candidate for minimal phenomenal experience. • epistemic capacity: third, the phenomenology of tonic alertness is the experience of possessing a set of epistemic capacities. for example, the organism now has the capacity to orientate itself in time and space, or to generate an epistemic first-person perspective by controlling attention (wiese, 2019). it becomes conscious by non-conceptually knowing about this capacity, but – as a careful investigation into the phenomenology of meditation shows – it does not necessarily exert it. as research on the effortless experience of “pure awareness” in meditation demonstrates, the phenomenology of bare wakefulness is independent of mental agency, time-representation and of self-location in a spatial frame of reference, because it often appears without it.25 phenomenologically, the relevant space of possibilities is indivisible, it has no centre and no boundaries, and it is best described as an epistemic space. • expectation of knowledge: fourth, there is a specific quality of subjective confidence, an ongoing expectation of epistemic states. computationally, wakefulness can be described as a statistical hypothesis, a representation of the probability that veridical perception will actually occur (metzinger, 2019).26 25on the functional level of analysis, tonic alertness and the orientation reflex can be dissociated: tonic alertness is what causally enables spatiotemporal self-location and orientation, but it is not identical with it. the conscious experience of wakefulness is a representation of precisely this epistemic capacity, the mere capacity to know where you are, at what point in time you exist, and how you are currently positioned in the order of things. mpe, the phenomenal experience of wakefulness, is a way in which an organism can non-conceptually know about the existence of this mere capacity, and this is also the reason why the relevant phenomenal character is extremely abstract and subtle. epistemologically, therefore, wakefulness is an abstract, non-egoic form of self-knowledge. however, this knowledge appears under a specific inner mode of presentation, by using an internal representational format which does not yet involve subject/object structure, time-representation, or spatial embodiment. the ensuing phenomenology can therefore be described as selfless, timeless, and non-spatial, as a model of an epistemic space, a space in which knowledge states can occur. 26to give an example, it is part of the phenomenology of waking up from dreamless deep sleep that in the very beginning there is a sense of confidence that perceptual states will very soon occur, that not only is one now “open to the world” and knows about one’s epistemic capacity, but that one can and will very soon know where one is, what time it is, and so on. one will also know who one is: phenomenologically, the gradual transition from unconsciousness to the wake state is characterized by another subtle and intuitive the wake state is characterized by another subtle and intuitive presentiment, something one might perhaps term a phenomenal “foreshadowing of selfhood”. this is an expectation not merely of knowledge, but of egoic self-knowledge. it involves an anticipation of mental agency and the capacity for global self-control (wiese, 2019), the subjective but as yet non-egoic confidence that a full-blown first-person perspective involving executive control on the mental level plus an extended autobiographical self-model will very soon appear. therefore, the phenomenal quality of subjective confidence not only relates to knowing metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 38 • spontaneity: finally, there is a fifth global aspect characterizing the phenomenology corresponding to the functionally autonomous process of tonic alertness as bare, apparently self-generating and apparently ever-fresh wakeful awareness, and it is related to the absence of time-representation – to the “temporal thinness” mentioned above (friston et al., 2017). mpe has a specific quality of spontaneity, as it is ahistorical and contains not even an implicit representation of its own causal history. it is the “natural state” – a spontaneous presence of epistemic capacity. taken together, these global aspects give us a first more general phenomenological description of the relevant target state: it is a mostly transparent experience of openness to the world, in combination with an abstract, non-conceptual representation of mere epistemic capacity, plus a spontaneously occurring, domaingeneral sense of confidence, of the likelihood of knowledge – but as yet without object. the target state is experienced as timeless, unconstructed, and functionally autonomous. mpe is aperspectival and it does not instantiate mps, but it definitely has representational content and a correspondingly unique, phenomenal character sui generis. however, the question of whether and in what sense it can count as “fundamental”, and whether it is the only truly minimal state of consciousness, has not been answered. if the above is correct, phenomenality is a transparent, temporally thin, and global representation of subjective confidence, relative to the abstract possibility of knowledge. this contribution is the first in a series of publications aiming at advancing our understanding of “minimal phenomenal experience” (mpe). the primary goal of this paper has been to lay the first foundations for a minimal-model explanation of consciousness (cf. section 1.2). one aim is to develop a new entry point for consciousness research by taking a closer look at a specific phenomenological subset, namely “pure consciousness” experiences in meditation (section 2). the first result is that such experiences clearly exist. a second result is that there are actually fewer paradigmatic reports of such experiences in the literature than previously assumed. third, while such descriptions have strong common phenomenological denominators, they are always “multidimensional” in the sense that they express multiple features and not one single, unequivocal form of phenomenal character.27 however, the phenomenological prototype to which all such reports about the environment, but it can also extend to the likelihood of egoic self-knowledge, to the capacity of “predicting oneself into existence” (friston, 2010; hohwy, 2016). 27fourth (and not discussed in this paper, but see metzinger, 2019), a first analysis of the relevant literature clearly shows that the attempt to develop a finer-grained and more empirically grounded phenomenology of “pure consciousness” episodes is confronted by a number of serious methodological obstacles. they are the problem of embodied theory contamination (cf. section 4.2 in metzinger, 2019), the problem of a positive characterization of the relevant type of phenomenal character (cf. section 4.3 in metzinger, 2019), and the problem of performative self-contradiction (cf. section 4.5 in metzinger, 2019). these problems seem to have affected and hindered scientific research on meditation in the past. as one early reviewer stated: “none of the studies reviewed metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 39 refer is clearly associated with a fundamental and non-egoic internal representation of tonic alertness (section 3). its primary dimensions are wakefulness and epistemicity. possibly, it can conceptually be described as a model of an unpartitioned epistemic space. however, much remains to be clarified. alongside more detailed philosophical analysis, mpe needs investigating using psychometric, computational, and neuroscientific methods. acknowledgments i would like to thank two particularly careful and constructive anonymous reviewers of the selfless minds project, as well as raphaël millière, adriana alcaraz sánchez, thomas andrillon, olivia carter, monima chadha, tom clark, cyril costines, nils donselaar, martin dresler, kristina eichel, sascha fink, tiziano furlanetti, vittorio gallese, jakob hohwy, amol kelkar, carsten korth, rafi malach, marcello massimini, lucy mayne, ulrich ott, arnaud poublan, danaja rutar, katharine shapcott, ronald sladky, heleen slagter, giulio tononi, nao tsuchiya, wanja wiese, jennifer windt, mateusz wozniak and audiences at the 2018 carnap lectures in bochum, the interdisciplinary college in günne (ik 2018), the predictive processing group at the donders institute in nijmegen, the association for the scientific study of consciousness in kraków (assc 22), the ernst strüngmann institute in frankfurt am main, the trends in mindfulness conference at the university of torino, the max planck institute for human development in berlin, the melbourne consciousness group at monash university, all participants in the mpe workshop in london, ontario (assc 23), and the audience at the selfless minds workshop at the frankfurt institute for advanced studies for comments and constructive criticism. i am greatly indebted to dr emily troscianko for editorial support. references albahari, m. (2009). witness-consciousness: its definition, appearance and reality. journal of consciousness studies, 16(1), 62–84. retrieved from https://www.ingentaconnect.com/content/imp/jcs/2009/00000016/00000001/art00003 albahari, m. (2020). beyond cosmopsychism and the great i am: how the world might be grounded in universal advaitic consciousness. in w. seager (ed.), the routledge handbook of panpsychism. new york: routledge. alexander, c. (1988). a conceptual and phenomenological analysis of pure consciousness during sleep. lucidity letter, 7(2). retrieved from https://journals.macewan.ca/lucidity/article/view/818 ataria, y., dor-ziderman, y., & berkovich-ohana, a. (2015). how does it feel to lack a sense of boundaries? a case study of a long-term mindfulness meditator. consciousness and cognition, 37, 133–147. https://doi.org/10.1016/j.concog.2015.09.002 bachmann, j. k. (2014). accounting for pure consciousness: an examination of the ability of the representationalist approach to phenomenal consciousness to account for pure consciousness experiences. edmonton: university of alberta. batterman, r. w. (2002). asymptotics and the role of minimal models. the british journal for the philosophy of science, 53(1), 21–38. https://doi.org/10.1093/bjps/53.1.21 batterman, r. w., & rice, c. c. (2014). minimal model explanations. philosophy of science, 81(3), 349–376. https://doi.org/10.1086/676677 bayne, t., hohwy, j., & owen, a. m. (2016). are there levels of consciousness? trends in cognitive sciences, 20(6), 405–413. https://doi.org/10.1016/j.tics.2016.03.009 bayne, t., seth, a. k., & massimini, m. (2019). are there islands of awareness? trends in neurosciences, 43(1), 6–16. https://doi.org/10.1016/j.tins.2019.11.003 […] met the five criteria: adequate reliability and validity, a comprehensive phenomenological assessment, and control for introspective sensitization and demand characteristics” (pekala, 1991, p. 78). metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://www.ingentaconnect.com/content/imp/jcs/2009/00000016/00000001/art00003 https://journals.macewan.ca/lucidity/article/view/818 https://doi.org/10.1016/j.concog.2015.09.002 https://doi.org/10.1093/bjps/53.1.21 https://doi.org/10.1086/676677 https://doi.org/10.1016/j.tics.2016.03.009 https://doi.org/10.1016/j.tins.2019.11.003 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 40 blanke, o., & metzinger, t. (2009). full-body illusions and minimal phenomenal selfhood. trends in cognitive sciences, 13(1), 7–13. https://doi.org/10.1016/j.tics.2008.10.003 brentano, f. (1973). psychology from an empirical standpoint (a. c. rancurello, d. terrell, & l. mcalister, trans.). london & new york: routledge. (original work published 1874.) brentano, f. (2012). descriptive psychology. london & new york: routledge. buswell, r. e. (1991). tracing back the radiance: chinul’s korean way of zen. honolulu: university of hawaii press. chatterjee, t. (1982). the concept of sāksin. journal of indian philosophy, 10(4), 339–356. retrieved from https://www.jstor.org/stable/23445372 deane, g. (2020). dissolving the self: active inference, psychedelics, and ego-dissolution. philosophy and the mind sciences, 1(i), 2. https://doi.org/10.33735/phimisci.2020.i.39 dentico, d., ferrarelli, f., riedner, b. a., smith, r., zennig, c., lutz, a., et al. (2016). short meditation trainings enhance non-rem sleep low-frequency oscillations. plos one, 11(2), e0148961. https://doi.org/10.1371/journal.pone.0148961 dunne, j. d., thompson, e., & schooler, j. (2019). mindful meta-awareness: sustained and non-propositional. current opinion in psychology, 28, 307–311. https://doi.org/10.1016/j.copsyc.2019.07.003 fasching, w. (2008). consciousness, self-consciousness, and meditation. phenomenology and the cognitive sciences, 7(4), 463–483. https://doi.org/10.1007/s11097-008-9090-6 fasching, w. (2011). “i am of the nature of seeing”: phenomenological reflections on the indian notion of witness-consciousness. in m. siderits, e. thompson, & d. zahavi (eds.), self, no self? oxford: oxford university press. ferrarelli, f., smith, r., dentico, d., riedner, b. a., zennig, c., benca, r. m., et al. (2013). experienced mindfulness meditators exhibit higher parietal-occipital eeg gamma activity during nrem sleep. plos one, 8(8), e73417. https://doi.org/10.1371/journal.pone.0073417 fink, s. b. (2020). look who’s talking! varieties of ego-dissolution without paradox. philosophy and the mind sciences, 1(i), 3. https://doi.org/10.33735/phimisci.2020.i.40 finnigan, b. (2017). buddhist idealism. in t. goldschmidt & k. l. pearce (eds.), idealism (pp. 178–199). oxford: oxford university press. finnigan, b. (2018). is consciousness reflexively self-aware? a buddhist analysis. ratio, 31(4), 389–401. https://doi.org/10.1111/rati.12200 flavell, j. h. (1979). metacognition and cognitive monitoring: a new area of cognitive–developmental inquiry. american psychologist, 34(10), 906–911. https://doi.org/10.1037/0003-066x.34.10.906 forman, r. k. c. (1986). pure consciousness events and mysticism. sophia, 25(1), 49–58. https://doi.org/10.1007/bf02789849 forman, r. k. c. (1998). what does mysticism have to teach us about consciousness? journal of consciousness studies, 5(2), 185–201. retrieved from https://www.ingentaconnect.com/content/imp/jcs/1998/00000005/00000002/art00005 fort, a. o. (1980). the concept of suṣupta in advaita vedānta. annals of the bhandarkar oriental research institute, 61(1/4), 221–228. retrieved from www.jstor.org/stable/41691868 fort, a. o. (1984). the concept of sākṣin in advaita vedānta. journal of indian philosophy, 12(3), 277–290. retrieved from https://www.jstor.org/stable/23444147 fox, k. c. r., & christoff, k. (eds.). (2018). the oxford handbook of spontaneous thought: mind-wandering, creativity, and dreaming. oxford: oxford university press. friston, k. (2010). the free-energy principle: a unified brain theory? nature reviews neuroscience, 11(2), 127–138. https://doi.org/10.1038/nrn2787 friston, k. (2018). am i self-conscious? (or does self-organization entail self-consciousness?) frontiers in psychology, 9, 579. https://doi.org/10.3389/fpsyg.2018.00579 friston, k., rosch, r., parr, t., price, c., & bowman, h. (2017). deep temporal models and active inference. neuroscience and biobehavioral reviews, 77, 486–501. https://doi.org/10.1016/j.neubiorev.2017.04.009 ganeri, j. (2017). mental time travel and attention. australasian philosophical review, 1(4), 353–373. https://doi.org/10.1080/24740500.2017.1429794 giacino, j. t., ashwal, s., childs, n., cranford, r., jennett, b., katz, d. i., et al. (2002). the minimally conscious state. definition and diagnostic criteria. neurology, 58(3), 349–353. https://doi.org/10.1212/wnl.58.3.349 giustina, a., & kriegel, u. (2017). fact-introspection, thing-introspection, and inner awareness. review of philosophy and psychology, 8(1), 143–164. https://doi.org/10.1007/s13164-016-0304-5 gupta, b. (1998). the disinterested witness: a fragment of advaita vedānta phenomenology. evanston: northwestern university press. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.tics.2008.10.003 https://www.jstor.org/stable/23445372 https://doi.org/10.33735/phimisci.2020.i.39 https://doi.org/10.1371/journal.pone.0148961 https://doi.org/10.1016/j.copsyc.2019.07.003 https://doi.org/10.1007/s11097-008-9090-6 https://doi.org/10.1371/journal.pone.0073417 https://doi.org/10.33735/phimisci.2020.i.40 https://doi.org/10.1111/rati.12200 https://doi.org/10.1037/0003-066x.34.10.906 https://doi.org/10.1007/bf02789849 https://www.ingentaconnect.com/content/imp/jcs/1998/00000005/00000002/art00005 www.jstor.org/stable/41691868 https://www.jstor.org/stable/23444147 https://doi.org/10.1038/nrn2787 https://doi.org/10.3389/fpsyg.2018.00579 https://doi.org/10.1016/j.neubiorev.2017.04.009 https://doi.org/10.1080/24740500.2017.1429794 https://doi.org/10.1212/wnl.58.3.349 https://doi.org/10.1007/s13164-016-0304-5 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 41 harding, d. e. (2000). face to no-face: rediscovering our original nature. agoura hills, ca: inner directions publishing. hasenkamp, w. (2018). catching the wandering mind: meditation as a window into spontaneous thought. in k. c. r. fox & k. christoff (eds.), the oxford handbook of spontaneous thought (pp. 539–551). oxford: oxford university press. hasenkamp, w., wilson-mendenhall, c. d., duncan, e., & barsalou, l. w. (2012). mind wandering and attention during focused meditation: a fine-grained temporal analysis of fluctuating cognitive states. neuroimage, 59(1), 750–760. https://doi.org/10.1016/j.neuroimage.2011.07.008 hellie, b. (2007). that which makes the sensation of blue a mental fact: moore on phenomenal relationism. european journal of philosophy, 15(3), 334–366. https://doi.org/10.1111/j.1468-0378.2007.00274.x higgins, d. (2012). an introduction to the tibetan dzogchen (great perfection) philosophy of mind. religion compass, 6(10), 441–450. https://doi.org/10.1111/rec3.12004 higgins, d. (2013). the philosophical foundations of classical rdzogs chen in tibet: investigating the distinction between dualistic mind (sems) and primordial knowing (ye shes). wien: arbeitskreis für tibetische und buddhistische studien universität wien. hohwy, j. (2016). the self–evidencing brain. noûs, 50(2), 259–285. https://doi.org/10.1111/nous.12062 james, w. (1950). the principles of psychology. new york: dover publications. (original work published 1890.) josipovic, z. (2019). nondual awareness: consciousness-as-such as non-representational reflexivity. progress in brain research, 244, 273–298. https://doi.org/10.1016/bs.pbr.2018.10.021 kawagoe, t., onoda, k., & yamaguchi, s. (2019). the neural correlates of “mind blanking”: when the mind goes away. human brain mapping, 40, 4934–4940. https://doi.org/10.1002/hbm.24748 kelly, l. (2019). the way of effortless mindfulness. boulder, co: sounds true. kriegel, u. (2009). subjective consciousness: a self-representational theory. oxford: oxford university press. kriegel, u. (2012). précis of subjective consciousness: a self-representational theory. philosophical studies, 159(3), 443–445. https://doi.org/10.1007/s11098-011-9761-y kriegel, u. (2019). dignāga’s argument for the awareness principle: an analytic refinement. philosophy east and west, (69), 144–156. https://doi.org/10.1353/pew.2019.0003 kriegel, u. (2020). the three circles of consciousness. in m. guillot & m. garcia-carpintero (eds.), the sense of mineness. oxford: oxford university press. kunsang, e. p., schmidt, m., & tweed, m. (eds.). (2012). perfect clarity: a tibetan buddhist anthology of mahamudra and dzogchen. berkeley, ca: north atlantic books. laureys, s., celesia, g. g., cohadon, f., lavrijsen, j., león-carrión, j., sannita, w. g., et al. (2010). unresponsive wakefulness syndrome: a new name for the vegetative state or apallic syndrome. bmc medicine, 8, 68. https://doi.org/10.1186/1741-7015-8-68 laureys, s., owen, a. m., & schiff, n. d. (2004). brain function in coma, vegetative state, and related disorders. the lancet neurology, 3(9), 537–546. https://doi.org/10.1016/s1474-4422(04)00852-x limanowski, j., & friston, k. (2020). attenuating oneself: an active inference perspective on “selfless” experiences. philosophy and the mind sciences, 1(i), 6. https://doi.org/10.33735/phimisci.2020.i.35 lingpa, k. (2014). the tibetan book of the dead. morrisville, nc: lulu press, inc. mackenzie, m. (2008). self-awareness without a self: buddhism and the reflexivity of awareness. asian philosophy, 18(3), 245–266. https://doi.org/10.1080/09552360802440025 mañjuśrīmitra, lipman, k., simmons, b., & norbu, n. (1987). primordial experience: an introduction to rdzogs-chen meditation. boston, ma: shambhala. margolis, e., & laurence, s. (2014). concepts. in edward n. zalta (ed.), the stanford encyclopedia of philosophy. retrieved from https://plato.stanford.edu/archives/sum2019/entries/concepts/ maruthai, n., nagendra, r. p., sasidharan, a., srikumar, s., datta, k., uchida, s., & kutty, b. m. (2016). senior vipassana meditation practitioners exhibit distinct rem sleep organization from that of novice meditators and healthy controls. international review of psychiatry, 28(3), 279–287. https://doi.org/10.3109/09540261.2016.1159949 mason, l. i., alexander, c. n., travis, f. t., marsh, g., orme-johnson, d. w., gackenbach, j., et al. (1997). electrophysiological correlates of higher states of consciousness during sleep in long-term practitioners of the transcendental meditation program. sleep, 20(2), 102–110. https://doi.org/10.1093/sleep/20.2.102 mason, l. i., & orme-johnson, d. w. (2010). transcendental consciousness wakes up in dreaming and deep sleep. international journal of dream research, 3, 28–32. https://doi.org/10.11588/ijodr.2010.1.595 metcalfe, j., & shimamura, a. p. (eds.). (1996). metacognition: knowing about knowing (1st mit press paperback ed.). cambridge, ma: mit press. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.neuroimage.2011.07.008 https://doi.org/10.1111/j.1468-0378.2007.00274.x https://doi.org/10.1111/rec3.12004 https://doi.org/10.1111/nous.12062 https://doi.org/10.1016/bs.pbr.2018.10.021 https://doi.org/10.1002/hbm.24748 https://doi.org/10.1007/s11098-011-9761-y https://doi.org/10.1353/pew.2019.0003 https://doi.org/10.1186/1741-7015-8-68 https://doi.org/10.1016/s1474-4422(04)00852-x https://doi.org/10.33735/phimisci.2020.i.35 https://doi.org/10.1080/09552360802440025 https://plato.stanford.edu/archives/sum2019/entries/concepts/ https://doi.org/10.3109/09540261.2016.1159949 https://doi.org/10.1093/sleep/20.2.102 https://doi.org/10.11588/ijodr.2010.1.595 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 42 metzinger, t. (ed.). (2000). neural correlates of consciousness: empirical and conceptual questions. cambridge, ma: mit press. metzinger, t. (2003a). being no one: the self-model theory of subjectivity. cambridge, ma: mit press. metzinger, t. (2003b). phenomenal transparency and cognitive self-reference. phenomenology and the cognitive sciences, 2(4), 353–393. https://doi.org/10.1023/b:phen.0000007366.42918.eb metzinger, t. (2006). conscious volition and mental representation: toward a more fine-grained analysis. in n. sebanz & w. prinz (eds.), disorders of volition (pp. 19–48). cambridge, ma: mit press. metzinger, t. (2013). the myth of cognitive agency: subpersonal thinking as a cyclically recurring loss of mental autonomy. frontiers in psychology, 4, 931. https://doi.org/10.3389/fpsyg.2013.00931 metzinger, t. (2014). how does the brain encode epistemic reliability? perceptual presence, phenomenal transparency, and counterfactual richness. cognitive neuroscience, 5(2), 122–124. https://doi.org/10.1080/17588928.2014.905519 metzinger, t. (2017). the problem of mental action. in t. k. metzinger & w. wiese (eds.), philosophy and predictive processing. retrieved from https://predictive-mind.net/papers/the-problem-of-mental-action metzinger, t. (2018). why is mind wandering interesting for philosophers? in k. c. r. fox & k. christoff (eds.), the oxford handbook of spontaneous thought (pp. 97–111). oxford: oxford university press. metzinger, t. (2019). minimal phenomenal experience: the aras-model theory: steps toward a minimal model of conscious experience as such. https://doi.org/10.31231/osf.io/5wyg7 metzinger, t., & windt, j. m. (2015). what does it mean to have an open mind? in t. k. metzinger & j. m. windt (eds.), open mind. frankfurt am main: mind group. meyniel, f., sigman, m., & mainen, z. f. (2015). confidence as bayesian probability: from neural origins to behavior. neuron, 88(1), 78–92. https://doi.org/10.1016/j.neuron.2015.09.039 millière, r. (2020). the varieties of selflessness. philosophy and the mind sciences, 1(i), 8. https://doi.org/10.33735/phimisci.2020.i.48 montague, m. (2016). the given: experience and its content (first edition). oxford: oxford university press. montague, m. (2017). what kind of awareness is awareness of awareness? grazer philosophische studien, 94(3), 359–380. https://doi.org/10.1163/18756735-09403004 moore, g. e. (1903). the refutation of idealism. mind, 12(48), 433–453. retrieved from https://www.jstor.org/stable/2248251 namgyal, d. t. (2001). clarifying the natural state: a principal guidance manual for mahamudra. hong kong: rangjung yeshe publications. namgyal, d. t. (2006). mahamudra the moonlight: quintessence of mind and meditation. somerville, ma: wisdom publications. namgyal, d. t. (2019). moonbeams of mahāmudrā. (with “dispelling the darkness of ignorance” by wangchung dorje, the ninth karmapa); (e. m. callahan, trans.). boulder, co: snow lion publications. oken, b. s., salinsky, m. c., & elsas, s. m. (2006). vigilance, alertness, or sustained attention: physiological basis and measurement. clinical neurophysiology : official journal of the international federation of clinical neurophysiology, 117(9), 1885–1901. https://doi.org/10.1016/j.clinph.2006.01.017 pagnoni, g. (2019). the contemplative exercise through the lenses of predictive processing: a promising approach. progress in brain research, 244, 299–322. https://doi.org/10.1016/bs.pbr.2018.10.022 pekala, r. j. (1991). the phenomenology of meditation. in m. a. west (ed.), the psychology of meditation (pp. 59–80). https://doi.org/10.1093/acprof:oso/9780198521945.003.0004 peters, f. (2013). theories of consciousness as reflexivity. the philosophical forum, 44(4), 341–372. https://doi.org/10.1111/phil.12018 picard, f. (2013). state of belief, subjective certainty and bliss as a product of cortical dysfunction. cortex; a journal devoted to the study of the nervous system and behavior, 49(9), 2494–2500. https://doi.org/10.1016/j.cortex.2013.01.006 piccinini, g., & craver, c. (2011). integrating psychology and neuroscience: functional analyses as mechanism sketches. synthese, 183(3), 283–311. https://doi.org/10.1007/s11229-011-9898-4 posner, m. i. (1978). chronometric explorations of mind. hillsdale, nj: erlbaum. posner, m. i. (2008). measuring alertness. annals of the new york academy of sciences, 1129, 193–199. https://doi.org/10.1196/annals.1417.011 ramm, b. j. (2019). pure awareness experience. inquiry, 16(1), 1–23. https://doi.org/10.1080/0020174x.2019.1592704 rgyal-ba-g’yung-drung, b. (2017). the pith instructions for the stages of the practice sessions of the a khrid system of bon rdzogs chen (great completion) meditation (second edition; g. s. gurung & d. p. brown, trans.). occidental, ca: bright alliance. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1023/b:phen.0000007366.42918.eb https://doi.org/10.3389/fpsyg.2013.00931 https://doi.org/10.1080/17588928.2014.905519 https://predictive-mind.net/papers/the-problem-of-mental-action https://doi.org/10.31231/osf.io/5wyg7 https://doi.org/10.1016/j.neuron.2015.09.039 https://doi.org/10.33735/phimisci.2020.i.48 https://doi.org/10.1163/18756735-09403004 https://www.jstor.org/stable/2248251 https://doi.org/10.1016/j.clinph.2006.01.017 https://doi.org/10.1016/bs.pbr.2018.10.022 https://doi.org/10.1093/acprof:oso/9780198521945.003.0004 https://doi.org/10.1111/phil.12018 https://doi.org/10.1016/j.cortex.2013.01.006 https://doi.org/10.1007/s11229-011-9898-4 https://doi.org/10.1196/annals.1417.011 https://doi.org/10.1080/0020174x.2019.1592704 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org minimal phenomenal experience 43 rice, c., rohwer, y., & ariew, a. (2019). explanatory schema and the process of model building. synthese, 196(11), 4735–4757. https://doi.org/10.1007/s11229-018-1686-y rinpoche, k. t., & namgyal, l. t. (2011). the ninth karmapa’s ocean of definitive meaning. boulder, co: snow lion publications. rinpoche, t. u. (1995). rainbow painting: a collection of miscellaneous aspects of development and completion. hong kong: rangjung yeshe publications. rinpoche, t. u. (2013a). as it is, volume i. new york: rangjung yeshe publications. rinpoche, t. u. (2013b). as it is, volume ii (2nd ed.). new york: rangjung yeshe publications. russell, i. f. (2013). the ability of bispectral index to detect intra–operative wakefulness during total intravenous anaesthesia compared with the isolated forearm technique. anaesthesia, 68(5), 502–511. https://doi.org/10.1111/anae.12177 sadaghiani, s., & d’esposito, m. (2015). functional characterization of the cingulo-opercular network in the maintenance of tonic alertness. cerebral cortex, 25(9), 2763–2773. https://doi.org/10.1093/cercor/bhu072 sanders, r. d., tononi, g., laureys, s., & sleigh, j. w. (2012). unresponsiveness ≠ unconsciousness. anesthesiology: the journal of the american society of anesthesiologists, 116(4), 946–959. https://doi.org/10.1097%2faln.0b013e318249d0a7 seli, p., kane, m. j., metzinger, t., smallwood, j., schacter, d. l., maillet, d., et al. (2018). the family-resemblances framework for mind-wandering remains well clad. trends in cognitive sciences, 22(11), 959–961. https://doi.org/10.1016/j.tics.2018.07.007 seli, p., kane, m. j., smallwood, j., schacter, d. l., maillet, d., schooler, j. w., & smilek, d. (2018). mind-wandering as a natural kind: a family-resemblances view. trends in cognitive sciences, 22(6), 479–490. https://doi.org/10.1016/j.tics.2018.03.010 severeide, c. j. (1990). physiological and phenomenological aspects of transcendental meditation. scientific research on maharishi’s transcendental meditation and tm-sidhiprogram, maharishi european research university publication, (3121), l556–84. shear, j. (2007). eastern methods for investigating mind and consciousness. in s. schneider & m. velmans (eds.), the blackwell companion to consciousness (pp. 697–710). malden, ma: john wiley & sons. siegel, s. (2007). how can we discover the contents of experience? the southern journal of philosophy, 45(s1), 127–142. https://doi.org/10.1111/j.2041-6962.2007.tb00118.x stace, w. t. (1960). mysticism and philosophy. london: macmillan. strawson, g. (2017). self-intimation. in g. strawson (ed.), the subject of experience (pp. 136–164). https://doi.org/10.1093/acprof:oso/9780198777885.003.0008 sturm, w., de simone, a., krause, b., specht, k., hesselmann, v., radermacher, i., et al. (1999). functional anatomy of intrinsic alertness: evidence for a fronto-parietal-thalamic-brainstem network in the right hemisphere. neuropsychologia, 37(7), 797–805. https://doi.org/10.1016/s0028-3932(98)00141-9 sturm, w., & willmes, k. (2001). on the functional neuroanatomy of intrinsic and phasic alertness. neuroimage, 14(1 pt 2), s76–84. https://doi.org/10.1006/nimg.2001.0839 sullivan, p. r. (1995). contentless consciousness and information-processing theories of mind. philosophy, psychiatry, & psychology, 2(1), 51–59. retrieved from https://muse.jhu.edu/article/245074 thompson, e. (2011). self-no-self? memory and reflexive awareness. in m. siderits, e. thompson, & d. zahavi (eds.), self, no self? oup oxford. thompson, e. (2015a). dreamless sleep, the embodied mind, and consciousness. in t. k. metzinger & j. m. windt (eds.), open mind. https://doi.org/10.15502/9783958570351 thompson, e. (2015b). waking, dreaming, being: self and consciousness in neuroscience, meditation, and philosophy. new york; chichester: columbia university press. travis, f., & pearson, c. (2000). pure consciousness: distinct phenomenological and physiological correlates of “consciousness itself”. international journal of neuroscience, 100(1-4), 77–89. https://doi.org/10.3109/00207450008999678 wainwright, w. j. (1981). mysticism: a study of its nature, cognitive value and moral implications. madison, wi: university of wisconsin press. ward, a. f., & wegner, d. m. (2013). mind-blanking: when the mind goes away. frontiers in psychology, 4, 650. https://doi.org/10.3389/fpsyg.2013.00650 weinbach, n., & henik, a. (2012). the relationship between alertness and executive control. journal of experimental psychology. human perception and performance, 38(6), 1530–1540. https://doi.org/10.1037/a0027875 weisberg, m. (2012). simulation and similarity: using models to understand the world. oxford: oxford university press. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1007/s11229-018-1686-y https://doi.org/10.1111/anae.12177 https://doi.org/10.1093/cercor/bhu072 https://doi.org/10.1097%2faln.0b013e318249d0a7 https://doi.org/10.1016/j.tics.2018.07.007 https://doi.org/10.1016/j.tics.2018.03.010 https://doi.org/10.1111/j.2041-6962.2007.tb00118.x https://doi.org/10.1093/acprof:oso/9780198777885.003.0008 https://doi.org/10.1016/s0028-3932(98)00141-9 https://doi.org/10.1006/nimg.2001.0839 https://muse.jhu.edu/article/245074 https://doi.org/10.15502/9783958570351 https://doi.org/10.3109/00207450008999678 https://doi.org/10.3389/fpsyg.2013.00650 https://doi.org/10.1037/a0027875 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org thomas metzinger 44 wiese, w. (2018). experienced wholeness. integrating insights from gestalt theory, cognitive neuroscience, and predictive processing. cambridge, ma: mit press. wiese, w. (2019). explaining the enduring intuition of substantiality: the phenomenal self as an abstract ‘salience object’. journal of consciousness studies, 26(3-4), 64–87. https://doi.org/https://www.ingentaconnect.com/content/imp/jcs/2019/00000026/f0020003/art00004 wiese, w., & metzinger, t. k. (2017). vanilla pp for philosophers: a primer on predictive processing. in t. k. metzinger & w. wiese (eds.), philosophy and predictive processing. https://doi.org/10.15502/9783958573024 williams, p. (1998). the reflexive nature of awareness: a tibetan madhyamaka defence. richmond, uk: curzon. williford, k., bennequin, d., friston, k., & rudrauf, d. (2018). the projective consciousness model and phenomenal selfhood. frontiers in psychology, 9, 732. https://doi.org/10.3389/fpsyg.2018.02571 windt, j. m. (2015a). dreaming: a conceptual framework for philosophy of mind and empirical research. cambridge, ma: mit press. windt, j. m. (2015b). just in time—dreamless sleep experience as pure subjective temporality. in t. k. metzinger & j. m. windt (eds.), open mind. https://doi.org/10.15502/9783958571174 windt, j. m., nielsen, t., & thompson, e. (2016). does consciousness disappear in dreamless sleep? trends in cognitive sciences, 20(12), 871–882. https://doi.org/10.1016/j.tics.2016.09.006 winter, u., levan, p., borghardt, t. l., akin, b., wittmann, m., leyens, y., & schmidt, s. (2020). content-free awareness: eeg-fcmri correlates of consciousness as such in an expert meditator. frontiers in psychology, 10, 3064. https://doi.org/10.3389/fpsyg.2019.03064 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. metzinger, t. (2020). minimal phenomenal experience: meditation, tonic alertness, and the phenomenology of “pure” consciousness. philosophy and the mind sciences, 1(i), 7. https://doi.org/10.33735/phimisci.2020.i.46 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/https://www.ingentaconnect.com/content/imp/jcs/2019/00000026/f0020003/art00004 https://doi.org/10.15502/9783958573024 https://doi.org/10.3389/fpsyg.2018.02571 https://doi.org/10.15502/9783958571174 https://doi.org/10.1016/j.tics.2016.09.006 https://doi.org/10.3389/fpsyg.2019.03064 https://doi.org/10.33735/phimisci.2020.i.46 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org introduction from mps to mpe scientific explanation by minimal models sketching the epistemic goal of the mpe approach mpe on the phenomenological level of analysis pure consciousness in contemplative phenomenology consciousness as such in western philosophy of mind introspective availability epistemicity transparency/opacity the difference between phasic and tonic alertness terminology first: alertness vs. wakefulness the phenomenology of phasic vs. tonic alertness conclusion: mpe as the phenomenology of tonic alertness vicarious attention, degrees of enhancement, and the contents of consciousness. vicarious attention, degrees of enhancement, and the contents of consciousness azenet lopeza (lopez.azenet@lrz.uni-muenchen.de) abstract how are attention and consciousness related? can we learn what the contents of someone’s consciousness are if we know the targets of their attention? what can we learn about the contents of consciousness if we know the targets of attention? although introspection might suggest that attention and consciousness are intimately connected, a good body of recent findings in cognitive psychology and cognitive neuroscience brings compelling reasons to believe that they are two separate and independent processes. this paper attempts to bring attention and consciousness back together to make the study of attentional distributions an essential ingredient for the study of the contents of consciousness. my proposal has two main components. first, i introduce a framework for systematizing the relations between attention in its different forms and consciousness in its different forms. although philosophers and cognitive scientists have repeatedly highlighted the importance of such systematization, most details are still to be worked out. here i take an initial stab at this project based on the notion of degrees of informational enhancement. second, i introduce the notion of vicarious attention to account for a kind of additional processing benefit that comes for free when attention is allocated to a target. i then propose that this kind of processing must also be considered when mapping attention targets into contents of consciousness. keywords attention ∙ attentional blink ∙ consciousness ∙ motion-induced blindness ∙ troxler fading 1 the explanatory project what is the relation between attention and consciousness? can we learn what the contents of someone’s consciousness are if we know the targets of their attention? what can we learn about the contents of consciousness if we know the targets of attention? the project of providing answers to these questions (in particular, an affirmative answer to the second one and a substantive answer to the third) is the explanatory project, that is, the project of using attention to contribute to our understanding of consciousness.1 aludwig maximilian university of munich, germany 1prominent theorists endorsing a project in this spirit include prinz (2012) and graziano (2013; 2019). lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://philosophymindscience.org https://orcid.org/0000-0002-6647-5709 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 2 the explanatory project is motivated by both pre-theoretical and theoretical considerations. pre-theoretically, the things that fill in our conscious experiences in our everyday lives seem to follow closely the things we attend to. an experience of pain feels intensified when you focus your attention on the aching body part, and it feels lessened when you direct your attention away from it. attending to the feeling of the floor against your feet seems sufficient for bringing this feeling to your conscious experience. it might also be necessary: before you directed your attention to your feet, the feeling of the floor was perhaps a very faint component of your overall conscious experience, if it was a part of it at all.2 these observations suggest that if we want to know the contents of someone’s conscious experiences, all we need to do is look at the things they are attending to.3 there are also theoretical considerations that lead to treating attention as a guide to the contents of consciousness. one prominent example is the attended intermediate representations theory of consciousness, articulated by the philosopher jesse prinz (2012). according to prinz, empirical research reveals that the neuronal correlates and the functional roles of attention are intertwined with those of consciousness so that attention is both necessary and sufficient for consciousness. on this view, a conscious experience of a pain arises when and only when a representation of that pain is modulated by attention.4 another prominent example is the global workspace theory (baars, 2002, 2005; dehaene & naccache, 2001). on this theory, for you to have a conscious experience of a pain is for information of that pain to become globally available. this means that information must be encoded by central cognitive mechanisms integrating inputs from different processing channels, including channels that process different kinds of sensory and proprioceptive information. attention to the relevant information is necessary and sufficient for it to become thus available. if theories like global workspace and attended intermediate representations are on the right track, attention turns out to be the key to understand consciousness. however, a wealth of recent findings in cognitive psychology and cognitive neuroscience strongly suggests that attention and consciousness are, in fact, two 2eric schwitzgebel (2007) conducted a retrospective sampling study to assess whether participants had this form of peripheral and subtle experiences when they were not attending, especially in the visual and tactile modalities. he gave participants a beeper that would randomly go off and instructed them to note the experiences they were having immediately before hearing the beep. results were mixed: they did not conclusively favor a thin view (according to which we never have these experiences) nor a rich view (according to which these experiences are pervasive). instead, the data pointed to a moderate view: we likely experience some things outside the focus of attention. the proposal i develop in this paper also takes this moderate stance. i am thankful to jennifer windt for bringing this study to my consideration. 3for interesting surveys of commonsense intuitions about the relations between attention and consciousness, see mole (2008b), de brigard (2010). 4this is a simplification of prinz’s proposal. on his view, the representation is of a particular kind, namely, a representation at the intermediate level in the hierarchy of perceptual processing. two controversial tenets of prinz’s account that i cannot discuss here are that all consciousness is located at this intermediate level, and that all consciousness is perception-based. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 3 separate and independent processes. this might further suggest that attention does not really stand in a privileged relation to consciousness. if so, then the study of attention cannot really offer us a unique window into consciousness. studying attention would then be no more relevant for understanding consciousness than studying other mental processes. this paper attempts to bring attention and consciousness back together so that the study of attention remains an essential ingredient for the study of the contents of consciousness. in a nutshell, i will argue that the picture of attention and consciousness currently given by the empirical sciences does still support the existence of tight and essential connections between their underlying processes. i propose that if attention is understood as an informational enhancer (a characterization strongly suggested by the way attention is conceptualized and operationalized in the empirical literature), it is possible to systematically map different forms of attention into different forms of consciousness by looking at the degree to which the relevant information is enhanced. here is the plan for what follows. sections 2 and 3 are mostly stage-setting. in section 2, i clarify the notions of attention and consciousness and pin down the kinds this paper addresses more directly, namely, selective attention and phenomenal consciousness. i also review some empirical findings that have been taken to support a tight connection between attention and consciousness of these kinds. in section 3, i review some of the most prominent empirical findings evincing a dissociation between attention and consciousness and point at the specific ways how these findings could jeopardize the explanatory project. together, sections 2 and 3 give a general sense of our current empirical picture of the relations between attention and consciousness. section 4 is a bridging one, where i introduce the main lines of my positive proposal, explain how it fits into our empirical picture, and identify potential difficulties. the most substantive discussion comes in sections 5 and 6. in section 5, i discuss the core of the proposal, namely, the enhancement framework. i elaborate on some of its central notions, compare it with other akin approaches, and defend it from initial empirical objections. i also identify a crucial assumption for defusing the discussed counterexamples: the claim that when attention is allocated to a target, the elicited processing benefits often extend to other things beyond the selected target. in section 6, i introduce the term “vicarious attention” to refer to this kind of processing, and i offer some initial conceptual and empirical reasons for believing that attention does behave in this way. 2 selective attention as a window into phenomenal consciousness “attention” and “consciousness” are terms that pick out a variety of related yet distinguishable kinds of processing. this paper focuses on phenomenal consciousness, the kind of consciousness constituted by the subjective experience in which lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 4 the world appears to us in a certain way, and which is characterized by its what-itis-like qualities (block, 1995; nagel, 1974).5 in this sense, for you to be conscious of a pain is for there to be something it is like for you to feel that pain, to experience a specific qualitative character. notice that if we are concerned with this kind of consciousness, then a theory like global workspace might not turn out to be relevant (but see raffone & pantani, 2010). global workspace theory rather seems to concern consciousness of another familiar kind, namely, access consciousness. this kind of consciousness is defined by its functional role in the processing of information. it is characterized by the subject’s cognitive access to the experienced contents, which enables them to report these contents (block, 1995).6 phenomenal consciousness must also be distinguished from consciousness of a third important kind, namely, reflective consciousness. this kind of consciousness involves internal monitoring and self-awareness, that is, awareness of one’s conscious states (carruthers, 2000; lycan, 1996; rosenthal, 1997). at least conceptually, neither reflective consciousness nor access consciousness entails phenomenal consciousness, and phenomenal consciousness entails neither. regarding attention, here i adopt a characterization that is fairly standard amongst empirical researchers. i take attention to be a process involving the selection of a subset of the available sensory information for preferential processing, a mental focus in a particular aspect of information, and the prioritization of some parts of the information that is available at any given moment (carrasco, 2011; desimone & duncan, 1995; posner, 1980; cohen et al., 2012; maier & tsuchiya, 2021; naghavi & nyberg, 2005; nani et al., 2019; see also pitts et al., 2018). nevertheless, we must note that this characterization only pins down a specific kind of attention, namely selective attention. there are other kinds of attentional processing that are not selective in any straightforward sense, for instance, distributed or diffuse attention. in addition, more general and global processes like those related to alertness and arousal are also sometimes counted as members of the attentional family (petersen & posner, 2012; posner & petersen, 1990). importantly, we must note that selective attention can also be of different types. it can be voluntary and controlled, guided by the subject’s goals and task demands, or it can be involuntary and automatic, driven by stimulus salience and environmental conditions. in what follows, i refer to these varieties as endogenous and 5most of the empirical literature discussed in this paper explicitly states the purpose to address this kind of consciousness nani et al. (2019). 6access consciousness is sometimes defined in terms of attention. for example, daniel stoljar (2019) proposes that for content to be conscious in this sense, it must be poised for use in the rational control of thought and action, and it must be thus poised because it has been attended. although this view entails the dependence of access consciousness on attention, it does not need to entail the same for phenomenal consciousness. the moral is that attention might be necessary for some forms of consciousness but not for others. more precisely: some forms of attention might be necessary for some forms of consciousness, as we will see below. i thank an anonymous referee for prompting me to make this clarification. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 5 exogenous attention, respectively. finally, selective attention can also be spatial, feature-based, or object-based. each of these varieties gives rise to different profiles of processing advantages. note that this characterization is neutral as to whether the selection and prioritization of information are accessible at the personal level and reportable by the subject. this characterization is thus compatible with sub-personal selection and prioritization that is not accessible and reportable. in this way, attention is characterized as a process that, in principle, can occur with or without subjective awareness, and consequently with or without phenomenal consciousness.7 what is, then, the relation between selective attention and phenomenal consciousness? can we learn about the things that constitute the contents of an organism’s subjective experience of the world if we know which subset of information is selected and prioritized for preferential processing within their cognitive economy? what can we learn about the contents of an organism’s subjective experience of the world if we know which subset of information is selected and prioritized for preferential processing? our answers to the second and third questions depend to a good extent on our answer to the first question. furthermore, one might think that our ability to provide interesting answers to the second and third questions hinges on a tight connection between selective attention and phenomenal consciousness. a natural line of thought is: if attention and consciousness are very tightly connected so that they always co-occur, then attention can be used as a window into consciousness because the contents of conscious experience will always be identical to the targets of attention. one might also think that if attention can be used as a window into consciousness at all, then attention and consciousness must be very tightly connected so that they always co-occur. indeed, a good body of empirical findings lends initial support for the existence of such an intimate connection. on the behavioral front, studies on inattentional blindness strongly suggest that we fail to see stimuli we do not attend to (mack & rock, 1998; simons & chabris, 1999). in recent years, these results have been extended to properties that were initially thought to be immune to the effect, including the gist of a scene (cohen et al., 2011; li et al., 2002; mack & clarke, 2012), statistical scene properties (jackson-nielsen et al., 2017)8 and even highly meaningful stimuli, like faces or one’s own name (devue et al., 2009). a variant of inattentional blindness is change blindness: major and meaningful changes in a scene can go unnoticed when attention is drawn away from them (rensink et al., 1997; tse, 2004; wolfe et al., 2006). it has also been observed that when a stimulus 7this is an important point for the present discussion, for some might define attention as conscious selection and prioritization of information. this view seems implied by william james’ (1890) celebrated characterization of attention as a concentration and focalization of consciousness, and more recently, by declan smithies’ (2011). i thank an anonymous reviewer from this journal for prompting me to make this clarification 8contra bronfman et al. (2014); block (2014). lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 6 captures attention, a second stimulus rapidly following it often goes unseen; this phenomenon is known as the attentional blink (chun & potter, 1995; raymond et al., 1992). attention also has various measurable effects on conscious perception. for example, attending to a stimulus increases its perceived contrast, brightness, size, distance, speed, or direction of motion (carrasco et al., 2004; carrasco & barbot, 2019). relatedly, visual sensitivity decreases when attention is distracted; this effect is known as load-induced blindness (macdonald & lavie, 2008). there is also neuroscientific data suggesting a close link between attention and consciousness. famously, cases of hemispatial neglect reveal that damage to brain areas related to attentional processing is accompanied by impairments in conscious perception (driver, 1998). relevant brain areas are located in the frontal and parietal cortices, suggesting that frontoparietal involvement is required for attentional and conscious processing. in addition, sub-cortical areas like the pulvinar nucleus of the thalamus have been associated with filtering information on the one hand and with the level of consciousness on the other (corbetta & shulman, 2002; nani et al., 2019; zhao et al., 2019). it has also been observed that both attention and consciousness involve common ways of information processing in the brain. prominently, they both have been linked to the amplification of neural signals and oscillatory synchrony. a variety of techniques, including single-unit recording, electroencephalography (eeg), and functional magnetic resonance imaging (fmri), have revealed that attention amplifies signals from targets (corbetta & shulman, 2002; desimone & duncan, 1995; kastner & ungerleider, 2000; luck et al., 2000; raz & buhle, 2006; reynolds & chelazzi, 2004; treue, 2003). these techniques have also revealed that attention actively suppresses signals from distractors (andersen & müller, 2010; chelazzi et al., 1993; hopf et al., 2006; luck et al., 1997; moran & desimone, 1985; reynolds et al., 1999; thut et al., 2006; vanduffel et al., 2000; worden et al., 2000). in turn, activity in sensory regions has been observed to be greater when stimuli are consciously perceived, as opposed to not perceived (dehaene et al., 2001; haynes et al., 2005; hesselmann et al., 2011; polonsky et al., 2000; ress et al., 2000; ress & heeger, 2003; sergent et al., 2011; tong et al., 1998; tse et al., 2005; vuilleumier et al., 2001). finally, both attention and consciousness have been observed to correlate with synchronic oscillatory activity, especially on the gamma frequency band. synchrony between neurons and between populations of neurons seems to be an efficient way how brain regions encoding different bits of information can communicate with each other in order to create the unified percepts that constitute our conscious experiences of things in the world (engel & singer, 2001; singer & gray, 1995). synchronic gamma activity in the visual cortex differentiates seen from unseen stimuli wyart & tallon-baudry (2009). this kind of correlation has been observed in a variety of paradigms, including masking (melloni et al., 2007), bistable perception (doesburg et al., 2005), and flash suppression in monkeys (wilke et al., 2006), as well as in modalities other than vision (gross et al., 2007). in turn, synchrony has been considered a way lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 7 the selective filtering associated with attention is implemented, insofar as it generates windows of increased or decreased excitability (buzsáki & draguhn, 2004; steriade et al., 1993; womelsdorf & fries, 2007). a neural population that is not synchronized to another will most likely fail to transmit its signal within the timeframes for optimal processing (fries, 2005). attention has also been observed to enhance gamma-band oscillations with a variety of experimental paradigms in monkeys and humans (bauer et al., 2006; engell & mccarthy, 2010; fries et al., 2001; gruber et al., 1999; tallon-baudry et al., 1997; tallon-baudry et al., 2005; vidal et al., 2006). findings like these suggest that (1) there is an intimate relation between attention and consciousness, which can be either a one-way or a two-way dependence, and which perhaps ultimately boils down to identity, (2) yes, knowing which subset of information is selected and prioritized for further processing is sufficient for knowing what the contents of someone’s consciousness are, and (3) by knowing what information is selected and prioritized we know exactly what the contents of consciousness are. alas, the road from attention to consciousness is not that straightforward. on the one hand, almost all these findings have been contested, and there are ongoing debates about what the relevant data proves. one salient example concerns inattentional blindness and its variant change blindness: it has been proposed that these experiments could involve failures of memory rather than failures of conscious perception, so that unattended stimuli are seen but quickly forgotten (wu, 2014). another prominent example concerns frontoparietal involvement: these brain areas could be required for reporting conscious experience and not for conscious experience itself (nani et al., 2019; tallon-baudry, 2012). accordingly, hemispatial neglect can be reinterpreted as a case of inaccessible or non-reportable conscious experience.9 finally, oscillatory synchrony could be a general information code in the brain, usable by any cognitive function including but not limited to attentional and conscious processing (tallon-baudry, 2012). on the other hand, there is also an overwhelming amount of empirical and theoretical support for the idea that selective attention and phenomenal consciousness are rather independent of each other (haladjian & montemayor, 2015; maier & tsuchiya, 2021; montemayor & haladjian, 2015; tallon-baudry, 2012; tallonbaudry et al., 2018; van boxtel et al., 2010b). i will now provide an overview of this evidence. 9in support of this interpretation, split-brain cases suggest that there might be separate streams of consciousness within the same brain. usually, only the stream connected to the brain’s language center can be verbally reported on, but other measures can be designed to tap into the other stream (for instance, reporting by drawing). see volz & gazzaniga (2017). lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 8 3 a dissociation on three fronts current grounds in favor of a dissociation between phenomenal consciousness and selective attention span three prominent fronts: behavioral, neuroscientific, and evolutionary. behavioral studies suggesting a dissociation between the contents of phenomenal consciousness and selective attention have been further supported by neuroscientific studies relying on diverse techniques. fmri, eeg, magnetoencephalography (meg), and single-unit recording evince different neural correlates for selective attention and conscious perception (maier & tsuchiya, 2021; van boxtel et al., 2010b). moreover, as pointed out above, findings previously taken to reveal common neural correlates for attention and consciousness have recently been re-assessed in the light of further empirical findings, methodological refinements, and more finely-grained conceptual distinctions (tallon-baudry, 2012; tallon-baudry et al., 2018). finally, it has been argued that attention and consciousness must diverge because anatomical and functional considerations suggest that they have evolved separately: while attention seems to be an early adaptation already present in very simple organisms, consciousness appears much later in evolutionary history, and only in organisms with a certain level of cognitive sophistication (haladjian & montemayor, 2015; montemayor & haladjian, 2015; see also graziano, 2013). in what follows, i present a somewhat more detailed yet quick survey of some relevant research. i sort it out into four groups: attention in the absence of consciousness, consciousness in the absence of attention, independence of both kinds of processing, and opposing functions and effects of the two of them. this survey admittedly fails to be comprehensive and to make full justice to all the arguments involved. still, it can lend us a good sense of the empirical grounds for considering attention and consciousness as independent processes. 3.1 attention without consciousness behavioral studies using a masked priming paradigm have shown that subjects can attend to a feature of a stimulus (for instance, color) without having seen either the stimulus or the feature (ansorge et al., 2010; kentridge et al., 2008; kiefer & martens, 2010; scharlau, 2002; scharlau & ansorge, 2003; scharlau & neumann, 2003). furthermore, a combination of this technique with a paradigm for testing object-based attention (egly et al., 1994; kahneman et al., 1992) has been used to show that subjects can attend to objects that they cannot see norman et al. (2015). notably, the cues used to direct attention in these studies were fully visible. however, it has been shown that attention can be manipulated with invisible cues as well (fuchs et al., 2013; kim & blake, 2005). moreover, even “high-level” properties such as being male or female can capture attention despite being invisible (jiang et al., 2006). at the intersection of cognitive psychology and the neurosciences, it has been observed that patients with blindsight, who have impaired or severely degraded lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 9 visual experiences due to damage in their primary visual cortex, are still able to effectively use visual attentional cues to detect targets in their blind field (kentridge et al., 1999; kentridge et al., 2008; kentridge, 2011; schurger et al., 2008). also, studies probing the tilt-aftereffect for seen and unseen adaptor stimuli have shown that both can be modulated by either spatial (bahrami et al., 2008) or feature-based attention (kanai et al., 2006), with stronger effects elicited for attended adaptors. furthermore, fmri has revealed increased attention-related processing in the parts of the visual cortex that encode unseen stimuli on dual-task (bahrami et al., 2007) and binocular suppression paradigms (watanabe et al., 2011). also, priming effects on event-related potentials can be enhanced by attention even when the prime is not consciously seen (kiefer & brendel, 2006). so attention can operate in the absence of awareness of the attended targets. this tells us that knowing that someone is attending to something still leaves it open whether they have a conscious experience of that thing. thus, for the explanatory project to work, we need to address the question: when is selective attention accompanied by phenomenal consciousness? is there a systematic way to pin this down? are there common features that group together all instances of phenomenally unconscious selective attention, on the one hand, and all instances of phenomenally conscious selective attention, on the other?10 my proposal in section 5 attempts to answer these questions. 3.2 consciousness without attention on the behavioral front, studies of iconic memory with the partial report paradigm suggest that it is possible to experience many more elements from a visual scene than those picked up for report by selective attention (block, 2007, 2008; landman et al., 2003; sperling, 1960). in addition, studies with a dual-task paradigm suggest that it is possible to have conscious experiences of elements in a visual scene outside the focus of attention, including natural scenes and animals (li et al., 2002) and faces (reddy et al., 2004, 2006).11 finally, studies of visuospatial crowding suggest that it is possible to have conscious experiences of things one cannot attend to because the grain of attentional selection seems to be coarser than the grain of visual spatial resolution (block, 2012; intriligator & cavanagh, 2001; pelli, 2008).12 10as noted below (section 5), a second important factor is stimulus strength (braun, 2007; dehaene et al., 2006; pitts et al., 2018). 11see also matthews et al. (2018). 12to be sure, there is controversy about what the results of partial report, dual-task, and crowding studies show. besides the methodological challenges mentioned below, alternative accounts of the findings are available, which do not require phenomenal consciousness of the relevant stimuli. for example, henry taylor and bilge sayim (2018) argue that participants in the crowding experiments infer rather than see the identity of the middle items, based on the information they pick up from the flankers in conjunction with an assumption of homogeneity. along similar lines, an anonymous reviewer for this journal has suggested that the predictive nature of perception could explain this performance. predictive processing theorists propose that conscious percepts lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 10 these behavioral results have been further supported by neuroscientific research. for instance, for the dual-task paradigm, fmri reveals an increased response in the brain’s fusiform face area, relative to baseline, for pictures of faces presented in the periphery (reddy et al., 2007). this activity was independent of whether participants were fully attending to the faces or not, thus suggesting that perception of faces could occur with or without attention.13 however, the findings in this group face a critical methodological challenge: to date, there are no straightforward methods for testing consciousness that do not do it by means of attention. paradigms designed for assessing consciousness typically rely on a task that requires that stimulus to be consciously perceived. however, participants will likely allocate at least some attention to any task-relevant stimulus (cohen et al., 2012). furthermore, reporting whether a stimulus was seen or not requires participants to attend to the stimulus. in other words, even if the stimulus was consciously seen before participants selected it for report, report-based methods cannot tap into this consciousness without thereby “contaminating” it with attention, so that what ends up being tested is no longer an unattended conscious content (wu, 2018). to be sure, this is also a problem for the studies that purportedly evince the dependence of consciousness on attention, such as inattentional blindness and change blindness studies: by removing attention from the stimuli of interest, the means for assessing whether that stimulus is consciously perceived are removed as well (stazicker, 2011; wu, 2017). to overcome these challenges, behavioral researchers have proposed new methods such as no-report paradigms (tsuchiya et al., 2015) and paradigms that test the visibility or invisibility of task-irrelevant stimuli (pitts et al., 2018). it has also been proposed that behavioral methods are not likely to be up to the task, and that the methodological disentanglement of attention and consciousness will probably have to be done with neuroscientific tools, such as optogenetics (van boxtel et al., 2010b). i find the latter approach highly congenial. the positive proposal offered in section 5 goes along these lines. result from the brain’s best predictions about the cause of an incoming sensory signal (clark, 2013). thus, in the face of a noisy signal from the middle stimulus, the brain might take its best guess about the stimulus’s identity. if this is the case, one might wonder whether participants genuinely have a conscious experience of the middle item rather than an illusory one. 13note, however, that there is conflict in the interpretations of the crucial finding in the dual-task paradigm, namely, that participants can use information from stimuli outside their focus of attention to complete a task. the studies described in this section treat these peripheral stimuli as consciously seen but unattended. however, as we saw in section 3.1, these stimuli are sometimes treated as attended but unseen (bahrami et al., 2007). my proposal (in section 5) is to treat these stimuli as both attended and consciously seen, albeit to lessened degrees than the stimuli in the central task. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 11 3.3 opposite functions and effects in addition to the studies suggesting that consciousness might neither require nor be required for attention, a dissociation between attention and consciousness has also been proposed on the basis of general function. as a selective process, attention is often conceptualized as a mechanism for analyzing and filtering information (van boxtel et al., 2010b). in contrast, it is not clear what the function of consciousness is or whether it even has a function (block, 1995; haladjian & montemayor, 2015). nonetheless, when functionally characterized at all, consciousness is regarded as a more integrative and synthetic process (for instance, see tononi et al., 2016). other research shows that even in cases when attention and consciousness do subserve the same task-specific function, they do so in different ways. for example, although manipulations of attention and consciousness both strengthen priming effects, attending to the prime does so by speeding up reaction times to congruent targets, while increasing the prime’s visibility does so by slowing down reaction times to incongruent targets (van den bussche et al., 2010). more strikingly, some studies have suggested that attention and consciousness sometimes have opposing effects, so that attention sometimes seems to impair consciousness. it has been observed that attention can diminish the intensity of perceptual experience (de weerd et al., 2006). in a motion-induced blindness paradigm, where a salient stimulus is rendered invisible by a surrounding cluster of moving dots, purposefully attending to the stimulus makes it more likely to disappear (geng et al., 2007; schölvinck & rees, 2009). the effect is greater for more salient stimuli (bonneh et al., 2001). similarly, attention can accelerate the troxler fading effect: while holding fixation, some peripheral stimuli disappear faster when attention is directed to them (lou, 1999).14 finally, distracting attention from the target stimuli in the attentional blink paradigm makes it easier to see the second target, which is missed when attention is fully engaged on the detection task (olivers & nieuwenhuis, 2005). these findings tell us that attention and consciousness are not identical at the level of brain mechanisms or computational functions. furthermore, they tell against the idea that attentional processing is in the business of improving conscious perception, and that the former might be required for the latter: the studies suggest that sometimes one can only have a conscious experience of something if one is not attending to it. this opens up an cluster of interesting questions of relevance for the explanatory project, and especially for the proposal offered in section 5, which is based on the idea that both attention and consciousness are correlated with informational enhancement. i will come back to these studies in that section. 14relatedly, it was previously argued that attention and consciousness had opposite effects on afterimage duration: more attention to a stimulus appeared to decrease afterimage strength and duration while increasing stimulus visibility produced a corresponding increase on these dimensions (koch & tsuchiya, 2007; van boxtel et al., 2010a). however, recent evidence shows that attention and consciousness both increase afterimage duration (travis et al., 2017). lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 12 3.4 double dissociation and independence finally, there is a good body of neuroscientific research supporting a double dissociation between attention and consciousness. a double dissociation means that consciousness can occur without attention, and attention can occur without consciousness. thus, the studies discussed in this section go beyond those discussed in sections 3.1 and 3.2, which only support a single dissociation. notably, a double dissociation between attention and consciousness does not entail that they always occur without the other, which is clearly false. nonetheless, it does entail that attention and consciousness are largely independent processes. for starters, attention and consciousness are associated with different underlying cortical activation patterns (chica et al., 2016; nani et al., 2019; webb et al., 2016) and with different temporal structures (davidson et al., 2018; fiebelkorn & kastner, 2020). the two of them also exhibit different meg and eeg profiles. meg revealed a correlation between stimulus visibility and mid-gamma activity for a sustained period (250-500ms) in the occipital cortex. in contrast, attention correlated with high-gamma activity in the parietal cortex at a slightly later point (350-500ms) (wyart & tallon-baudry, 2008).15 in turn, eeg revealed greater responses for consciously perceived stimuli than for non-perceived stimuli, regardless of whether the stimuli were attended or not (chica et al., 2010; forschack et al., 2017). furthermore, the eeg components of attention have been observed for both visible and invisible cues (travis et al., 2019). it has also been argued, on neuroscientific grounds, that there is an intrinsic aspect of phenomenal consciousness for which attention does not and cannot account, namely, the first-person perspective (tallon-baudry et al., 2018).16 while attention can provide the contents of a conscious state, it cannot provide the egocentric reference frame necessary for subjectively experiencing these contents.17 this element must be provided by a separate process: the internal monitoring of bodily states, specifically, the monitoring of visceral input from the heart and the gut.18 a double dissociation has also been advocated from a more theoretical standpoint, based on considerations about the separate evolution of attention and consciousness (haladjian & montemayor, 2015; montemayor & haladjian, 2015). montemayor and haladjian argue that although there is some overlap between the two processes, represented in the kind of processing that they call conscious attention, there are also many primitive forms of selective attention that do not re15mid-gamma: 54-64hz; high-gamma: 76-90hz. 16i am thankful to a reviewer from this journal for bringing this research to my attention. 17despite the similarity of this characterization to what i called reflective consciousness above, the authors do not intend to single out this specific kind of consciousness. on their account, the first-person perspective is an essential component of phenomenal consciousness. 18evidence favoring this view includes observations that heartbeat evoked response correlates with the self-relevance of the thoughts that a subject is currently entertaining, associated activity in brain areas independently correlated with the self, and studies of the experience of body ownership. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 13 quire consciousness of the phenomenal kind. these include spatial attention and feature-based attention, and the slightly higher-level object-based attention, all of which aid animals in spatial navigation and detection of relevant environmental stimuli (haladjian & montemayor, 2015).19 furthermore, the authors also point out that most conceptual theories of consciousness entail that there are forms of consciousness that do not require selective attention. for example, the distinction between phenomenal and access consciousness (block, 1995) entails such a dissociation. if phenomenal consciousness is sometimes not accompanied by access consciousness, then phenomenal consciousness may sometimes not be accompanied by attention (montemayor & haladjian, 2015, p. 88).20 in this way, the authors’ proposal goes beyond supporting a distinction between phenomenally conscious and phenomenally unconscious attention (a distinction the authors endorse): it also makes room for inattentive phenomenal consciousness. to be sure, montemayor and haladjian acknowledge that some forms of selective attention are probably necessary for phenomenal consciousness. one example is cross-modal attention, which evolves out of the more primitive attentional forms like spatial, feature-based, or object-based attention. this kind of attention enables the kind of informational integration with an egocentric perspective that is essential for phenomenal consciousness. however, even if all forms of phenomenal consciousness involve selective cross-modal attention, there might still be some forms of phenomenal consciousness that do not require selective attention of other kinds. in this sense, one can say that phenomenal consciousness is independent of selective attention of these specific sorts. summing up, there is reason to think that selective attention and phenomenal consciousness must come apart along many dimensions, despite partial overlap and occasional co-occurrence. this is a stronger claim than the one supported by the discussion in section 3.3, which only states that attention and consciousness do come apart. now it is time to consider the implications for the explanatory project, that is, the project of understanding of consciousness through the study of attention. as we saw, this project comprises three central questions, namely: (1) what is the relation between selective attention and phenomenal consciousness? (2) can we learn what the contents of someone’s phenomenal consciousness are if we know the targets of their attention? and finally (3) what can we learn about the contents of consciousness if we know the targets of attention? 19in their (2015) paper, the authors note that not all animals with these functions have conscious awareness. however, it is not clear whether this means that these animals lack phenomenal consciousness. the latter claim might be somewhat hasty, given the current methodological challenges for assessing the presence of this kind of consciousness. the former claim, on the contrary, seems plausible, if by* awareness* they mean something like what i above called reflective consciousness. 20interestingly, montemayor and haladjian point out that even opponents of the phenomenal/access consciousness distinction, such as daniel dennett (2005), seem to be committed to dissociating selective attention and phenomenal consciousness (see the discussion in montemayor & haladjian, 2015, pp. 88–89). lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 14 3.5 taking stock this quick overview suggests that (1) attention and consciousness are neither identical nor dependent on each other; in fact, each of them is a separate process that can occur without the other and can even hinder the other and (2) no, knowing which subset of information is selected and prioritized for further processing is not sufficient for knowing what the contents of someone’s consciousness are. selective attention to a target, be it endogenous or exogenous, and be it spatial, featurebased, or object-based, sometimes is and sometimes is not accompanied by a conscious experience of the attended thing. moreover, there may be no principled way of mapping the former onto the latter, and it might well be that the two of them are separable as a matter of necessity. hence, studying selective attention may not be of particular value for learning when information of things in the world becomes the content of a conscious experience. another consequence is that theories like attended intermediate representations theory and global workspace theory are mistaken, at least insofar as they are regarded as general theories of consciousness encompassing phenomenal consciousness.21 but what about (3)? even if attention and consciousness are highly independent, can we not learn something about the contents of someone’s conscious experience by knowing what things they are selectively attending to? in fact, there is reason to think that we can. 4 tying attention and consciousness back together as mentioned above, even a double dissociation is compatible with a partial overlap of selective attention and phenomenal consciousness. indeed, the current evidence allows some forms of attention to be necessary for phenomenal consciousness, for instance, cross-modal attention (haladjian & montemayor, 2015; montemayor & haladjian, 2015).22 but if at least some forms of selective attention are necessary for phenomenal consciousness, and if there is a partial overlap between both kinds of processing, then it is possible to use selective attention as a guide into 21as noted above, some theorists have defended a version of global workspace theory that is explicitly construed as a theory of phenomenal consciousness (raffone & pantani, 2010). it is also worth noting that attended intermediate representations is not intended to be about selective attention. prinz notes that his definition of attention can be extended to diffuse attention and processes like vigilance, alertness, and arousal (prinz, 2012, p. 91). however, prinz’s characterization of attention (as a mechanism for making representations available for working memory encoding) does rely on the idea that attended information is information selected and prioritized for preferential processing. 22note that one of the virtues of haladjian and montemayor’s arguments for this claim is that, being based on considerations about the evolution of different brain areas and functions, they can bypass the methodological challenges for the behavioral and neuroscientific assessment of consciousness in the absence of attention (discussed in section 3.2). lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 15 phenomenal consciousness, provided that the conditions under which these two co-occur are carefully delimited. borrowing montemayor and haladjian’s terms, this project amounts to tracing the boundaries of conscious attention. these boundaries shall at least separate conscious attention from unconscious attention. an additional virtue would be that these boundaries also delimit conscious attention from inattentive consciousness, provided that the latter exists. i propose that the current research on selective attention and phenomenal consciousness suggests that (1) they are not identical but are bound to co-occur in some specific cases, (2) no, knowing which subset of information is selected and prioritized for further processing is not sufficient for knowing what the contents of consciousness are because we also need to know whether the process of selection and prioritization is of the right kind, and (3) knowing which information is selected and prioritized and what kind of selection and prioritization is being used does tell us which things constitute the contents of conscious experience. in the remainder of the paper, i will introduce a framework that will help us define the specific cases when attention and consciousness are bound to co-occur, thus tracing the boundaries of conscious attention. the following discussion is mainly focused on question (1), that is, in clarifying the relations between attention and consciousness. my proposal can be summarized in four steps. first, it is a plausible consequence of the characterization of attention as a process of selection and prioritization that the selected information is enhanced within the relevant cognitive system. second, information can be more or less enhanced: enhancement comes in degrees. third, it might be possible to pin down a specific degree of enhancement such that, when information is enhanced to that degree, it becomes the content of a conscious experience. if we learn what that degree is, and if we can pin down the forms of attention that enhance information to that degree, we can rely on attention as a window into consciousness. finally, it might also be possible to pin down a specific degree of enhancement such that reaching that degree is necessary for information to enter the contents of conscious experience. degrees of attention have been previously proposed as a promising framework for systematizing the relations between attention and consciousness. for example, on the empirical front, michael pitts and colleagues (2018) propose that different “amounts” of attention might give rise to different forms of attention, which might enable us to identify the forms of attention required for consciousness. they conceptualize these amounts of attention as points in a continuum that goes from zero to full attention and where a threshold for “minimal attention” can be identified. on the conceptual front, daniel stoljar (2019) construes attention as a graded kind of processing and proposes that there is an amount of attention such that, when that amount of attention is allocated to a target, the target becomes conscious.23 23to be sure, stoljar uses this framework as a way of clarifying when content becomes access conscious. in his view, increasing the degree of attention marks the transition between content being merely disposed for rational use and being poised for such use. the latter defines access lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 16 initial empirical motivation for endorsing this kind of framework comes from the observations that attention and consciousness correlate with boosting effects on neuronal responses (see the studies on signal amplification cited at the end of section 2). it also comes from studies suggesting that attention has effects similar to increasing some perceptual dimensions, such as brightness or contrast (carrasco & barbot, 2019; fazekas & nanay, 2021; also see the studies on perceptual adaptation cited in section 3.1). furthermore, the framework can accommodate the further finding that the response increases associated with consciousness are significantly greater than those associated with attention (kouider et al., 2016), a finding that is sometimes thought to evince a dissociation between attention and consciousness (maier & tsuchiya, 2021). to be sure, adopting this framework requires an explanation of the findings suggesting that attention and consciousness have opposite effects (from section 3.3). it also requires a response to an argument presented by catherine tallonbaudry (2012), which disputes the idea that signal amplification is a common mechanism for attention and consciousness. both things can be done. however, this will require fleshing out the proposal some more. 5 degrees of enhancement24 i will now offer a fuller elaboration of the enhancement framework. i start by clarifying the notions of informational enhancement and thresholds of enhancement, how these notions relate to similar proposals, and how they can be the basis for systematizing the relations between attention and consciousness (section 5.1). i then defend this framework from the objections identified above (sections 5.2 and 5.3) and from an additional objection (section 5.4). finally, i conclude this part by highlighting a crucial assumption for defusing the discussed counterexamples: sometimes, attention is allocated vicariously. this idea will be the focus of the final section of this paper (section 6). 5.1 degrees of enhancement, amounts of attention and measures of integration i said that the conception of attention as a process for selecting and prioritizing information entails that attended information is enhanced. however, i have not yet explained what i mean by this notion.25 as i understand it, enhancement is a higher-order property that information has when some first-order properties are consciousness. i thank an anonymous reviewer from this journal for bringing this proposal to my consideration. 24i am thankful to two anonymous reviewers for their valuable suggestions, from which the discussion in this section has greatly benefitted. 25i offer further discussion on the notion of informational enhancement elsewhere (lopez, 2020). lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 17 present, in virtue of which the organism can make better use of that information. some of those first-order properties include perceptual dimensions like detail, determinacy, precision, or vividness. they also include computational properties, such as facilitation and suppression effects at the neuronal level. indeed, suppression of responses to non-target or distractor information is sometimes necessary to optimize target information processing for the system’s use.26 one might think that enhancement is positively correlated with neuronal activity so that more neuronal activity corresponds to a higher degree of enhancement. this picture is suggested by many of the studies reviewed above, such as those evincing that attention amplifies stimulus signals (see section 2) and those evincing increased neuronal activity for task-relevant as opposed to task-irrelevant stimuli (section 3.2).27 however, there are some reasons not to be too hasty in endorsing this correlation. notably, there are cases where attending to a stimulus might result in a decreased neuronal response, such as feature-based attention (reynolds & heeger, 2009; thiele & bellgrove, 2018). for example, a neuron hardwired to “prefer” upward motion might fire below baseline when presented with a stimulus moving downwards, and attention is directed to that stimulus (martinez-trujillo & treue, 2004). characterizing enhancement as a higher-order property, rather than identifying it with a first-order property like neuronal spiking, enables us to bypass these difficulties. firing below its baseline for its less preferred stimulus can be a way of conveying enhanced information of this stimulus’ presence to neighboring neurons, thus biasing the competition more in favor of this stimulus. enhancement is essentially a comparative notion. for a piece of information to be enhanced, there must be a condition with respect to which the enhancement occurs. this condition might be the previous processing of the same piece of information or the processing of a different piece of information at the same time. it can also include both. for example, when you direct your attention to the pain in your leg, pain information is enhanced with respect to how that same information was processed a moment before and with respect to how other pieces of information are being processed at the same time (say, it becomes more enhanced than the information of the article that you are reading). i have also proposed that two thresholds of informational enhancement might be distinguished: one is the threshold that information must reach in order to be consciously experienced at all, while the other one is the threshold that, when 26because attention can inhibit processing in this way, one might think that boosting information past a threshold (as i suggest below) cannot be the only function of attention. although the present account is compatible with a plurality of first-order functions, such as response boosting and inhibition, the proposal is that all of these first-level functions subserve the higher-order function of enhancing information. even when attention elicits suppression effects, this is to aid in raising the degree of enhancement of some piece of information to a higher threshold. i take suppression effects to be related to enhancement in the same way as the firing below baseline discussed in the main text. i thank an anonymous reviewer from this journal for prompting me to clarify this point. 27for a defense of the view that attention is a signal amplifier, see fazekas & nanay (2021). lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 18 reached, guarantees that information will be consciously represented. call these t– and t+, respectively. now, given the characterization of enhancement as a comparative magnitude, t– and t+ would be better characterized as relative rather than absolute values. whether they are reached on a single occasion for a piece of information may thus strongly depend on what other information the system is processing at the moment. this can also be influenced by stimulus strength and background conditions (such as context) or global states of the system (such as level of arousal). it is, of course, an empirical question whether our neurocognitive systems implement something like t– and t+, and how exactly these thresholds should be identified. that is, if enhancement is a measurable magnitude, determining which values or ranges of values of this magnitude correspond to t– and t+ can only be determined empirically. still, there is reason to think that this project is on the right track. there are at least two ways of starting to pin down these thresholds. on the one hand, the proposal by pitts and colleagues (2018) mentioned in section 4 relies on the hypothesis that there are different “checkpoints” in information processing, which might separate conscious from unconscious attention (and indeed, also might separate kinds of consciousness, such as phenomenal vs. access consciousness). the authors use this framework to classify a selection of some recent studies exploring the relations between attention and consciousness, on the basis of the kind of attentional strategy that each of them demands, in combination with the kind of stimulus used and its presentation conditions. see a graphic illustration of their proposal in figure 1. in this picture, tcould be pinned down at the intersection of cells 1 and 2, while t+ could be pinned down at the intersection between 5 and 6. the proposal that there are degrees of enhancement that are either necessary or sufficient for phenomenal consciousness and that are given by relative thresholds is also congenial in spirit to a prominent theoretical approach to the study of consciousness, namely, integrated information theory (iit). according to iit, a system becomes conscious when it reaches the maximum integration level of the information processed by its parts (oizumi et al., 2014; tononi et al., 2016). very roughly, information is integrated when it has an irreducible causal profile. this occurs when the information carried by the whole system cannot be reduced to the sum of the information carried by the system’s parts so that information carried by the whole generates causal effects that the sum of the information carried by the parts cannot generate. information of the whole can also be causally affected in ways that need not be reflected in the information of the parts. a measure of the amount of informational integration in the system is symbolized as φ. when φ has a positive value, the system can be regarded as conscious. moreover, higher values of φ predict “more consciousness,” or consciousness of more complex and sophisticated types. thus, a second way to pin down t+ and tcould start by considering the relations of informational enhancement to informational integration. some interestlopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 19 figure 1: an expanded taxonomy of the relations between attention and consciousness. different “checkpoints” in information processing lead up to more sophisticated forms of consciousness. the amount of attention allocated to the relevant stimuli defines some of these transitions and, importantly, the transition from pre-conscious to phenomenally conscious states. republished with permission of the royal society (u.k.) from “the relationship between attention and consciousness: an expanded taxonomy and implications for ‘no-report’ paradigms”; pitts, m. a., lutsyshyna, l. a., & hillyard, s. a.; philosophical transactions of the royal society of london b: biological sciences, 37(1755), copyright 2018; permission conveyed through copyright clearance center, inc. ing interactions might arise. for starters, one might propose that t+ is the same measure as a positive φ since both signal the point where conscious processing appears. along these lines, one might think that degrees of enhancement are just amounts of integration, such that more enhanced information is more integrated information, or vice-versa. note, however, that tcannot be straightforwardly correlated with a degree of integration since iit conceptualizes integration as sufficient for consciousness. in other words, once information is integrated, even if the measure of integration is very low, the system counts as conscious, although its consciousness can be of a very basic kind. iit thus predicts that there is no such a thing as integrated unconscious information.28 in contrast, enhancement is 28against this prediction, haladjian and montemayor (2015) note that sometimes information is integrated outside of consciousness. on the other hand, it has also been argued that sometimes there is consciousness without integration (brogaard et al., 2021). a supplementation of intelopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 20 not like this: information can be enhanced below t-, and thus, there is enhanced unconscious information. one possible way to go would be to characterize tas the minimum degree of enhancement that information must reach to become integrated. enhancement and integration will then come apart in this respect: enhancement is a magnitude that is necessary but not sufficient for phenomenal consciousness, whereas integration is a magnitude that is both necessary and sufficient. alternatively, tand t+ might signal different amounts of φ, such that information reaching tgives rise to consciousness of more rudimentary forms (the ones corresponding to lower values of φ), while information reaching t+ gives rise to more sophisticated forms of consciousness (the ones corresponding to higher values of φ). even if a further specification is still needed, the elements we have in hand already enable us to outline how the enhancement framework conceptualizes some cases of purported unconscious attention (from section 3.1) and inattentive consciousness (from section 3.2) in a way that supports the explanatory project. i mentioned at the end of section 3.1 that it would be nice to identify some features that characterize unconscious attention. i now propose that one such feature is that the relevant information fails to reach the higher threshold of enhancement t+. in many of these cases, there are limiting factors for the amount of information that the organism can extract from the stimulus and the extent to which the system can use this information. some of these factors are masks (kentridge et al., 2008; norman et al., 2013) or damage to cortical areas responsible for extracting specific kinds of stimulus information (kentridge et al., 1999). on the other hand, cases of purported inattentive consciousness can be tentatively characterized as cases where information of the relevant stimuli is enhanced at least up to t-, while leaving it an open question whether the enhancement reaches t+ as well. note that characterizing t+ as sufficient for phenomenal consciousness does not entail that lesser degrees of enhancement are never conscious. however, one might distinguish between conscious information above and below t+ so that, for instance, the former is more elaborated, distinct, or sophisticated than the latter. so if we characterize information of “unattended” stimuli in partial report, dual-task, or crowding as enhanced above tbut not necessarily above t+, we might count these contents as conscious, albeit this consciousness is of a lesser degree than the one for attended stimuli. it is now time to address some potential empirical objections to the present proposal. two of these objections were mentioned towards the end of section 4. the first is an argument from the opposing effects of selective attention and phenomenal consciousness. if attention is an information enhancer, and if consciousness requires a certain degree of enhancement, then why is it that in some cases, attention makes stimuli disappear faster or impairs perception, as in motioninduced blindness and troxler fading, or the attentional blink? (see section 3.3). grated information theory with the enhancement framework could provide a way to accommodate these cases. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 21 the second objection is tallon-baudry’s (2012) observation that signal amplification might not be a common correlate for attention and consciousness, despite initial appearances. i will tackle these worries in reverse order. 5.2 what if attention and consciousness amplify signals in different ways? tallon-baudry disputes the idea that the neural mechanisms used by attention are directly equivalent to those mediating contrast increases in the stimulus. although attending to a stimulus has similar perceptual and behavioral effects than increasing its contrast (as evinced, for instance, in carrasco et al., 2004, among other studies), she points out that these effects might be implemented in different ways at the neural level. to support this point, she cites studies that reveal an earlier peaking of neuronal responses in the monkey primary visual cortex when physical stimulus contrast is increased (gawne et al., 1996). she then compares these results with those observed in meg readings during spatial attention in humans (hopf et al., 2006), which do not reveal such a temporal shift in neuronal responses. furthermore, she mentions other studies which suggest that if spatial attention affects the temporal profile of neuronal responses at all, it must do it only at a later temporal point (after 200ms, which contrasts starkly with the 30-40ms that it takes for the effects of increases in physical contrasts to manifest; see noguchi et al., 2007). finally, a direct comparison of the effects of contrast increases and spatial attention in monkey area v4 showed that although both factors affected the magnitude of single-cell responses, attention did not produce an earlier spiking (lee et al., 2007). based on these findings, tallon-baudry argues that it is premature to identify something like “neural amplification” as a common mechanism for attention and consciousness. however, her proposal seems to be that there are fine-grained differences in how attentional modulations and manipulations of visibility affect neural responses, not that either the former or the latter fail to amplify neural responses. sure, one could individuate neural mechanisms very finely so that different amplification mechanisms underpin attention and consciousness. but in any case, the proposal that consciousness requires reaching a threshold of enhancement does not require that a single mechanism underlies both attentional modulations and modulations in response to physical changes in the stimulus affecting the stimulus’ visibility. indeed, there might be a diversity of mechanisms constituting the first-order properties that underpin the higher-order property of enhancement. 5.3 why does attention sometimes make stimuli disappear from consciousness? let us now consider how the enhancement framework can accommodate some of the effects of attention that seem to go against the idea that attention enhances information of a stimulus to make it more suitable for conscious experience. we lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 22 have three puzzling cases of this sort: motion-induced blindness, troxler fading, and the attentional blink. motion-induced blindness and troxler fading are perceptual effects that seem to occur due to hardwired information processing constraints. in both cases, a stimulus that remains physically present throughout disappears intermittently from perception. motion-induced blindness occurs when a static stimulus becomes momentarily invisible due to the presence of a field of moving distractors (bonneh et al., 2001; schölvinck & rees, 2009). in turn, troxler fading is the gradual disappearance of peripheral stimuli when the gaze remains rigidly fixated on an object for several seconds (troxler, 1804). the puzzling observation is that purposefully directing attention to these stimuli facilitates their disappearance. this is especially problematic for two tenets of the enhancement framework, namely, that attending to a target enhances information of that target to a greater degree and that the higher the degree of enhancement, the most likely it is for a piece of information to enter the contents of a conscious experience. on the face of motion-induced blindness and troxler fading, it looks like one of these tenets has to go: either attention is not an informational enhancer, or else degrees of enhancement do not correlate with consciousness. however, once we consider the cases more closely, we will see that they need not be inconsistent with either claim. in the motion-induced blindness study by marieke schölvinck and geraint rees (2009), participants were instructed to press a key when they noted that one of two yellow dots disappeared. the yellow dots were presented behind a grid of rotating blue crosses. see figure 2 for illustration.29 participants were prompted to focus their attention on either the right or the left dot while maintaining fixation on a center point. once they noted that a dot disappeared, participants must press a different key indicating which of the two dots it was. the experimenters observed that attended dots were significantly more likely to be reported as disappearing.30 29see an animated version of the stimulus at: https://commons.wikimedia.org/wiki/file:motionbl indnessf.gif#/media/file:motionblindnessf.gif 30 interestingly, the second experiment in this study assessed the effect when attention was globally withdrawn from both the yellow dots and its rotating distractors by being engaged in a different task (detecting a stimulus in a rapid series presented at fixation). in this experiment, the motioninduced blindness display was presented in the upper-right quadrant of the visual periphery. three conditions were examined: no attentional load, low attentional load, and high attentional load; these were defined by the difficulty of the central task. this experiment showed that the greater the attentional load, and thus the lesser the amount of attention that could be used for the motion-induced blindness task, the longer the yellow dot remained invisible. this suggests that attentional processing indeed has a role in making stimuli pop into conscious experience. still, these results may need to be contrasted with those of the attentional blink studies discussed below, which evince improved visibility under peripheral task conditions (see footnote 34 for further discussion). lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://commons.wikimedia.org/wiki/file:motionblindnessf.gif#/media/file:motionblindnessf.gif https://commons.wikimedia.org/wiki/file:motionblindnessf.gif#/media/file:motionblindnessf.gif https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 23 figure 2: testing attentional modulations of motion-induced blindness. a standard trial in schölvinck and rees’ study started with an arrow cueing attention to the left or the right. then, two yellow dots and a grid of rotating blue crosses appeared. participants must maintain fixation and press a key when they noted the disappearance of one of the dots. redrawn from journal of vision 9(1), 38, schölvinck, m. l., and rees, g., “attentional influences on the dynamics of motioninduced blindness,” 31-39, copyright 2009, with permission from association for research in vision and ophthalmology. according to the enhancement framework, when attention is directed to the yellow dot, information of this dot is enhanced to a higher degree than information of the unattended dot. the account also predicts that other things being equal, the attended dot should be consciously represented if the unattended one is. schölvinck and rees’s findings falsify this prediction. but importantly, other things are not equal. directing attention to one of the dots not only enhances information of the dot; it also enhances information of the contextual cues that trigger the dot’s disappearance as a result of hardwired processing constraints. furthermore, one cannot neatly disentangle object-based attention from spatial attention. when you attend to the dot, you also attend to the dot’s location. at that location, there is information of the moving grid; this information is also enhanced and computed for the generation of a conscious percept. crucially, information of moving stimuli is prone to take over. for one thing, it is beneficial for an organism to have information of the things that are moving around it, as this information might signal important changes in the environment correlated with food or danger. in contrast, information of a stationary and unchanging stimulus is not as prone to keep drawing on the organism’s processing resources once it has been registered. to sum up: attending to the yellow dot also enhances some contextual information, which is more potentially significant and thus enhanced to a higher degree, trumping competing information in reaching the comparative threshold for conscious perception t+. this interpretation is in line with schölvinck and reese’s account of their observations. they conceive motion-induced blindness as a completion process of the distracters into one homogeneous field, resulting in the target being hidden from lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 24 figure 3: testing attentional modulations of troxler fading. in lou’s study, participants must direct their attention to either the green or the orange disks while maintaining fixation on the center and press a key to report the disappearance of a disk. adaptation (in color) from lou, l., perception 28(4), p. 521, copyright © 1999 by sage publications. reprinted by permission of sage publications. view (graf et al., 2002). completion implies neural competition between target and distracters. directing attention within the motion-induced blindness display “favors” the distracters by enhancing distracter field completion, inducing target disappearance. consider now lianggang lou’s (1999) study of troxler fading. here, participants were instructed to attend to either the green or the orange disks in the periphery of a display while maintaining fixation on the center (see figure 3). participants’ task was to press a key when at least one disk disappeared and keep it pressed until the disk appeared again. they were then asked how many disks of each color had disappeared. lou observed that the attended disks were more likely to disappear first.31 like motion-induced blindness, this effect creates trouble for the enhancement framework because this framework dictates that attending to a disk enhances information of that disk and that this information is more likely to become consciously represented than competing information. but in lou’s study, the opposite happened: attended information was more likely to go unconscious. an important difference between this study and schölvinck and rees’ is the type of attention each relies on. schölvinck and rees’ study relies primarily on object-based attention: attention is directed on the basis of a stimulus with objectlike properties. in contrast, lou’s study relies primarily on feature-based attention: attention is directed on the basis of a property shared by several stimuli in 31interestingly, color was a relevant factor: green attended disks disappeared first about 95% of the time, while orange ones disappeared first only about 68% of the time. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 25 the display, namely color. while object-based attention is usually constrained to enhance information from the specific location defined by the object’s boundaries, feature-based attention is usually more evenly spatially distributed to bring out the relevant feature at any location where it is present. that said, one crucial methodological aspect of lou’s study introduces an interesting interaction between feature-based and object-based attention. participants in his study were encouraged to treat the three attended disks as forming a triangle, as this would facilitate attending to all three of them simultaneously. if participants followed the instruction, their attention was not singling out any particular disk during this task. instead, their attention picked out the object constituted by the three disks sharing the same color. accordingly, the enhanced information was information of this object as a whole, not information of any given single disk. this already lessens the worries for the enhancement framework: since there was no trial where all three disks disappeared, there was no trial where the enhanced object faded entirely from consciousness.32 still, a worrisome aspect of this study is that the unattended disks remained visible throughout. on the face of it, this contradicts the prediction that these disks should be less likely to make it into consciousness. however, here, too, considerations about contextual conditions and hardwired constraints are relevant. information of the three-disk object is being enhanced in a specific context, which involves peripheral presentation and particular ranges of luminance and contrast.33 it is plausible that information of the stimulus is not enhanced independently of this contextual information. contextual information is also enhanced and included in the computations that determine what the conscious percept will be. in addition to these considerations, there is also an alternative way of characterizing informational enhancement, accommodating both schölvinck and rees’s and lou’s findings. on this alternative characterization, a piece of information is enhanced when its capacity to drive neuronal responses is boosted. recall the case of feature-based attention making an “upward motion” neuron fire below its baseline when attention is directed to a stimulus moving downwards. we said that this pattern of firing conveys more information about the presence of a downward moving stimulus. one can also say that when attention is directed to the stimulus moving downwards, the capacity of this stimulus to drive the cell’s response is boosted. the way the boost is expressed depends on how the cell typically responds to stimuli of this kind. for a neuron preferentially tuned for upward motion, information of downward motion presence is enhanced by suppression of firing. analogously, enhancing the information of the yellow dot in the motion-induced blindness paradigm or information of the three-disk triangle in the troxler fading paradigm means potentiating the responses that these stimuli would typically elicit. since these responses involve momentary stimulus disappearance, it is expected that this effect is more likely to occur. 32on average, 1.2-1.4 disks disappeared per trial (depending on eccentricity). 33for troxler fading to occur, the stimuli must be faint. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 26 figure 4: the attentional blink paradigm. participants must report two targets appearing within a rapid succession of distractors, all presented at fixation. when the second target appears within 200-400ms after the first, participants fail to report it. this effect is known as the attentional blink. these lines of explanation can also be extended to the remaining problem case, namely the attentional blink. the attentional blink is the inability to identify the second of two targets presented in rapid succession (chun & potter, 1995; olivers & nieuwenhuis, 2005; raymond et al., 1992). figure 4 depicts a standard paradigm for testing this effect. the received explanation of the attentional blink is that the first target captures a great amount of attentional resources so that there is not a lot left for detecting the second. the effect is more pronounced when the lag between the two targets is shorter, which suggests that the “replenishment” of attentional resources after being used up in the first target takes some time. as noted above (section 2), this effect supports the idea that selective attention is necessary for conscious perception. when participants’ selective attention has been exhausted, they fail to see the second target. in the study by christian olivers and sander nieuwenhuis’s (2005), the targets were two numbers presented within a string of letters. remarkably, the experimenters found that detection of the second target improves when attention is drawn away from the target-detection task, this finding seems at odds with the enhancement framework. the framework predicts that when less attention is available for a target, that target is less likely to make it into consciousness. however, the study by olivers and nieuwenhuis suggests otherwise. once again, on a closer look, things need not be so. note that the stimulus of interest is target 2. what is important is how much attention is available for enhancing this specific piece of information in every case. fully engaging in the detection task enables the first target to use up all attention available so that there is no attention or insufficient attention for processing the second target. contrastingly, performing a simultaneous unrelated task hinders the first target’s salience, thus allowing that the overall balance of attention reaches the second target as well. hence, what we are looking at when comparing a standard attentional blink lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 27 paradigm with an attentional blink as a concurrent task is not a case where allocating less attention to a specific stimulus (i.e., locally) makes that stimulus more likely to enter consciousness. although less attention is allocated to the overall task where the stimulus is embedded (i.e., globally), this setting favors that attention is distributed in a way that more attention is allocated to the stimulus of interest.34 so far, i have explained how the three cases discussed can plausibly involve enhancing information so that the result is a conscious percept of the relevant information. in doing so, i have primarily focused on the facilitation effects that underpin enhancement. however, it is very likely that the suppression effects of attention also play an essential role in explaining these cases.35 attention can inhibit responses to information other than its targets. this can happen synchronically or diachronically. for example, it is possible that in the attentional blink, the information of the second target is actively suppressed in order to encode the information of the first target better. if this were the case, then diverting attention away, as olivers and nieuwenhuis did, would reduce the active suppression effects, thus making the second target more likely to be consciously perceived. similarly, it is possible that the attention-expedited troxler fading and motion-induced blindness are partially due to the interactions between information whose processing is directly facilitated and information that is actively suppressed.36 5.4 attention and consciousness without enhancement? a final challenge for the degrees of enhancement framework is posed by effortless attention (bruya, 2010; montemayor & haladjian, 2015).37 in effortless attention, information seems to be processed without enhancement, at least in this sense: the distribution of attention seems flat, rather than having “peaks” for more at34 as suggested in footnote 30, it might be further illuminating to compare olivers and nieuwenhuis’ results with those obtained in the second experiment in schölvinck & rees (2009), where embeddedness in a peripheral task diminished stimuli visibility. the account given in the main text could also explain this apparent conflict: in schölvinck and rees’ experiment, making the relevant task peripheral plausibly diminished the amount of attention available for the stimulus of interest (compared to the amount available when the task is central). notably, in this experiment, both tasks were in the same modality (visual), whereas in the olivers and nieuwenhuis study, they were not. while the detection task was visual and was indeed presented at the center of the visual field, the concurrent tasks were either “mental” (engaging in free associations on task-irrelevant topics) or auditory (listening to music). 35this possibility was brought to my consideration separately by slawa loev and an anonymous reviewer from this journal. i am thankful to both of them. 36a further example of this interaction between facilitation and suppression is multiple-object tracking (for instance, in pylyshyn, 2004; pylyshyn, 2006; see also alvarez & scholl, 2005). as an anonymous reviewer from this journal has kindly pointed out, active suppression of non-target information plausibly plays a role in the subjects’ ability to isolate their targets in this paradigm. thus, target information is enhanced partly due to the inhibition of non-target information. 37i thank an anonymous reviewer from this journal for bringing this case to my consideration. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 28 tended stimuli.38 if enhancement requires the peaks, that is, if it requires greater attention on some parts of the sensory input than others, then effortless attention would be attention without enhancement. furthermore, if effortless attention is accompanied by phenomenal feelings (as it plausibly is), then there is phenomenal consciousness without enhancement.39 one initial response to this challenge is that the enhancement framework is primarily intended to capture selective attention. effortless attention, on the other hand, might not be of the selective type. however, the case remains of interest for the more overarching project of extending the enhancement framework to account for attention and its relations with phenomenal consciousness more generally, beyond just the selective types of attention. 5.5 wrapping up in this section, i have argued that the enhancement framework is explanatorily useful, empirically well-motivated, and even consistent with evidence that initially seemed to contradict it directly. it also provides us with conceptual guidelines for approaching the daunting body of empirical literature from a systematic point of view. in the final section of this paper, i will elaborate on one assumption that i have made in my treatment of the paradoxical effects of attention. i suggested that in many cases, it is not easy to disentangle different targets of attentional enhancing, for example, an object and the location the object is sitting in (especially if the object is static). i take this to support the idea that attending to some targets often involves enhancing information of things other than these targets alone so that attentional benefits can be observed for things other than the selected targets. i will now offer some more elaboration of this idea and some initial conceptual and empirical support. 6 vicarious attention “vicarious attention” is the term that i will use to refer to attentional allocations that “come for free” when our attention is directed at a given target. for example, suppose that you are focusing your attention on a dark spot on the wall. by the same act, you are also allocating attention to a specific location in space, and the properties present there. alternatively, suppose that you are sitting at a noisy café, and then your attention is suddenly captured by the sound of your name.40 by the same act, your attention is allocated to other things, such as the region where the voice came from or the voice’s pitch. i propose that when the sound of your 38i borrow these suggestive terms from watzl (2017). 39similar considerations may apply to mind wandering (see footnote 49). 40this is the cocktail-party effect (cherry, 1953; moray, 1959; treisman, 1969), which reveals that stimuli of great significance are preferentially processed even with very little attention. i thank an anonymous reviewer for pointing this out to me. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 29 name captures your attention or when you purposefully focus your attention on the dark spot on the wall, your attention also becomes allocated to the locations and properties of these stimuli. in attending to the dark spot on the wall, you also attend to these things, albeit indirectly.41 in this sense, you are attending vicariously to these things. conceptual and empirical considerations support the idea that attention behaves in this way. prominently, robert kentridge (2011) has argued that a distinction between attending to an object and attending to the object’s spatial location is not theoretically useful. suppose that o is an object sitting at location l. kentridge argues that there is no substantial difference in attention-related benefits between allocating attention directly to o and allocating attention to o by allocating attention to l. he argues that absent good grounds for drawing a distinction, we should believe that attending to a spatial location is sufficient for attending to an object at that location. his view seems to be that there is nothing more to attending to a location than enhancing the representations of the things present there. the notion of vicarious attention enables us to accommodate these insights without collapsing spatial and object-based attention. the idea that there is no substantive distinction between the two of them is captured by the proposal that both targets can be picked out with a single act of attending, in which one of them is primarily targeted and the other one benefits from this allocation vicariously. on spatial-based allocations, attended locations are the primary targets, whereas, on object-based allocations, the primary targets are the attended objects. behavioral studies suggest that the function of spatial attention is to enhance goal-specific properties of stimuli falling in the attended space, since attending to a spatial region typically produces enhanced processing of properties of objects falling within that region, rather than of properties of space per se (kentridge, 2011; remington & folk, 2001). furthermore, kentridge argues that this point is also supported by studies showing that attention alters the perceived properties of objects, such as perceived contrast and perceived hue (carrasco et al., 2004). these studies show that spatial attention does more than speeding responses to items presented in a spatial location: it improves the processing of these objects by improving the processing of their properties. this picture is further supported by neuroscientific studies suggesting that attending to an object produces attentional benefits for all its features, regardless of whether these are required for the task at hand. for example, in a study by o’craven, downing and kanwisher (1999), attention was directed to one of two superimposed images (a house or a face), one of which was slowly oscillating. as expected, there was increased activity in the brain areas dedicated to the attended stimulus type (the para-hippocampal place area, or the fusiform face area, respectively). however, activity was also observed in the area of the brain that encodes motion when the attended stimulus was oscillating, even though these oscillations were task-irrelevant. 41one might use the “in virtue of” language to capture these relations: vicarious attention occurs in virtue of attention being allocated to something else. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 30 figure 5: excerpts from datta and deyoe’s “library” of attentional distributions. attending to one section of the visual field brings about attentional benefits for neighboring and non-neighboring regions. yellow and red indicate increased neuronal activity in the areas of medial occipital visual cortex encoding each of the eighteen sections in datta and deyoe’s stimulus grid, when participants were directing their attention to sections 6, 9 and 17, respectively (from left to right). yellow indicates greater benefits. adapted from vision research, may 49 (10), datta, r. and e. a. deyoe, “i know where you are secretly attending! the topography of human visual attention revealed with fmri.” 1037-1044, copyright (2009), with permission from elsevier. many thanks to zara morgue for creating this figure. finally, a study by ritobrato datta and edgar a. deyoe (2009) on the topography of visual attention shows that attending to a specific region in the visual field gives rise to attentional effects beyond that region. indeed, the attentional benefits spread over other regions that were not purposely attended themselves.42 datta and deyoe presented participants with a stimulus grid that divided their visual field into eighteen regions within 28 visual degrees around fixation.43 they found that directing attention to one of these regions brought about increased activity in neurons coding for other regions as well, regardless of whether these other regions were immediately contiguous to the purposely attended regions. figure 5 depicts some of their observations. these findings strongly suggest that at any given time when attention is allocated to a stimulus, the total attentional distribution at that time is not exhausted by attention to that stimulus. more conservatively, datta and deyoe’s results suggest that focused attention to a spatial location does not entail that this is the only visual field region that benefits from attention. further, their findings also suggest that an organism’s allocations of attention are not entirely constrained by voluntary control. attention can spread over things or regions in space other than the intended target of attention.44 42suppression effects for regions other than the attended one were also observed, as discussed below. 43see this grid in figure 4 from datta & deyoe (2009). 44datta and deyoe’s findings also illustrate the suppression effects that focally attending to a section of the visual field has on other sections (depicted in blue in the figure). these effects must be lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 31 i have presented some motivation for the idea that attending to some targets often involves enhancing the information of things other than these targets alone. hence, sometimes we attend to things vicariously, both in everyday life and in controlled lab settings. however, it remains an open question whether and how vicarious allocations of attention can be regimented. for the present discussion, what is important is to note that the notion of vicarious attention introduces a more fine-grained way of thinking about the stimuli that can and should become contents of consciousness. therefore, this proposal qualifies the way we understand cases of unconscious attention (section 3.1) and especially inattentive consciousness (section 3.2). regarding unconscious attention, the notion of vicarious attention supports the idea that information of the relevant targets is enhanced to a lessened degree, insofar as many of these targets benefitted from attention only because participants were attending to something else. thus, blindsight patients (kentridge et al., 1999) succeed in attending to the unseen target only because they attend to a location in their blind field.45 likewise, the color of a masked prime (kentridge et al., 2008) is attended only because the location of the prime is attended, and masked objects (norman et al., 2013) are attended only because a (visible) cue captures attention. on the other hand, if attention can behave in the proposed ways, then cases of purported inattentive consciousness might involve more attention than initially thought. for instance, in partial report paradigms (landman et al., 2003; sperling, 1960), attention might be vicariously allocated to the non-cued stimuli in virtue of being directed at the cued ones. in turn, in crowding paradigms (block, 2012), attention might be vicariously allocated to the middle items in virtue of being directed to the flankers. admittedly, much remains to be clarified about the notion of vicarious attention. for one thing, one might wonder how the distinction between vicarious attention and the kind of attention it piggybacks on (call it primary attention) interacts with other distinctions sometimes made in the attention literature. two prominent examples are the distinction between selective and non-selective attenconsidered when regimenting the way attention is vicariously transferred from one spatial location to another. as an anonymous reviewer from this journal has pointed out, information from the suppressed regions might fail to enter consciousness if the suppression is sufficiently strong. hence, attending to a spatial region will not always bring about enhancement benefits for neighboring regions, or at least not for all neighboring regions. overall, these considerations further highlight the need for integrating suppression effects into our account of attention. conceptualizing enhancement as a higher-order property that results from a combination of facilitation and suppression is a step in this direction. 45this proposal must be distinguished from the one previously advanced by christopher mole (2008a). according to mole, blindsighters merely attend to target locations, not to the targets themselves. thus, the attended thing (a spatial location) is not the same as the non-experienced thing (an object at that location). he then concludes that the case does not support a dissociation between targets of attention and contents of consciousness. although to my lights this rationale is on the right track, the considerations just discussed tell against it. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 32 tion and the distinction between spatially focused and spatially diffuse attention (prettyman, 2014).46 one might think that vicarious attention is never selective or focused, at least in the sense that vicarious targets are by definition different from the targets purportedly attended. however, depending on how focused attention is defined, vicarious attention can sometimes be focused. when i focus my attention on the dark spot on the wall and thereby attend vicariously to its dark shade, this allocation may count as focused, if by focused we understand something like attention constrained to a local stimulus (as opposed to a global one), or attention constrained to a delimited sub-region of the visual field. some non-focal forms of attention that operate alongside focused attention will be vicarious by definition. for instance, consider the suggestion by adrienne prettyman (2014) that focused attention is often accompanied by a halo of spatially distributed or diffuse attention, which comes at no additional processing cost.47 since this allocation “comes for free,” it is vicarious by our count. however, diffuse attention does not need to be always vicarious. one might purposefully broaden the focus of attention: this happens, for instance, in open-monitoring meditation (lutz et al., 2007). relatedly, one might wonder whether vicarious attention is a kind of spatial attention. one reason for thinking this is that vicarious targets often spatially overlap or are spatially contiguous with the targets they piggyback. however, one reason not to characterize vicarious attention as essentially spatial from the start is that it has the profile of feature-based attention in some cases. for example, a spatial cue might vicariously direct attention to a specific color, and information of that color becomes enhanced at other locations as well.48 i have suggested that lack of purposefulness might be a way to define when a target is vicarious. if one is attending to x but not on purpose, then one is attending to x vicariously. nonetheless, this idea might need to be revised in the light of effortless attention. to borrow an example from montemayor and haladjian (2015): if you are deeply engrossed in writing a novel, your attention is engaged in this action but in a way that does not involve subjectively felt effort. although the action of writing the novel is purposeful, one might also say that attention is not purposefully directed to the objects involved (say, the keyboard and the screen) or to the action itself. does this mean that effortless attention is vicarious? my initial answer is that insofar as writing the novel is an intentional action, any attention used to carry it out is not vicarious. however, the way we attend to the objects involved in the action (the keyboard and the screen) might as well be vicarious.49 46see also prettyman (2021). 47prettyman suggests that this halo of diffuse attention might be responsible for participants’ performance in the dual-task paradigm (see section 3.2). thus, although peripheral stimuli are processed without focused attention, they are not processed independently of attention. 48this happens in the (2008) masked priming study by kentridge and colleagues, where participants were cued to attend to the invisible prime’s location, yet attentional benefits for the same color of the prime at other locations in the display were observed. 49 along these lines, one might also wonder how vicarious attention interacts with mind wandering. like effortless attention, mind wandering plausibly involves a “flat” attentional distribution (see lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 33 task conditions are another way of determining when attention is vicarious. generally, a task-irrelevant target will be a vicarious one. an interesting suggestion related to this point is that vicarious attention might pick out information that is not directly relevant for attention’s functional role, for instance, high-level or conceptual information. in other words: when i attend to the apple on my desk, the information needed for the functional role of attention, that is, for enhancement, is seemingly information of color and shape, but maybe not the information that this is an apple. yet, attending to the apple’s low-level features might bring along enhancement of its more high-level features. nonetheless, these roles might be reversed depending on task demands or contextual demands. for example, when your attention gets captured by the sound of your name, the attention you allocate to low-level features of this stimulus like pitch and timber piggybacks on your attention to its high-level semantic features. the considerations offered in the preceding paragraphs are not intended to provide a complete and systematic characterization of vicarious attention. however, they can provide some guidelines for embarking on such a project. importantly, a complete regimentation of vicarious attention might be the basis for a deeper understanding of the relations between the different layers of phenomenally conscious contents.50 section 5.4). however, unlike effortless attention, mind wandering is not associated with any specific activity (for example, writing a novel or dancing); instead, it is often characterized as task-unrelated thought (smallwood & schooler, 2015; but see irving, 2016 for dissent). mindwandering might also be plausibly considered an attentionless state. since it plausibly involves phenomenal consciousness, it qualifies as a potential case of inattentive consciousness that falls within the scope of the present proposal. although i cannot discuss this case at length here, here are some inchoate considerations about some directions this discussion might take. if mind wandering involves vicarious attention rather than being an attentionless state, then it would no longer exemplify inattentive consciousness. one reason to think that mind-wandering could involve vicarious attention is its lack of purposefulness. however, one worry with this proposal is that vicarious attention must piggyback on attention to some other target, and it is not clear what this target might be for mind wandering. on the other hand, vicarious attention might not be needed to defuse this kind of counterexample: one could conceptualize mind wandering as an unregulated and spontaneous succession of targets of attention, where each of these targets is the primary occupant of the organism’s processing resources at a time (the view proposed in christoff et al., 2016 seems to go along these lines). i wish to thank jennifer windt for prompting me to think about this case. 50here is a quick example for illustrating this project. when i look at the white rabbit in the yard, my visual experience has a layer constituted by distal stimulus properties and a layer constituted by what is phenomenally apparent to me. this second layer could remain the same even if the stimulus out there changed (for instance, if it was replaced by a robot rabbit that looks exactly like the real rabbit). vicarious attention could capture the sense in which i am attending to the properties of the stimulus that are not necessarily reflected in my visual experience, such as being a rabbit vs. being a robot rabbit. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 34 7 concluding remarks: back to the explanatory project the discussion in this paper has been motivated by an attempt to answer three questions: (1) what is the relation between phenomenal consciousness and selective attention?, (2) can we know what the contents of phenomenal consciousness are, if we know what are the targets of attention?, and (3) what can we know about the contents of consciousness if we know the targets of attention? i have proposed that (1) the relation between phenomenal consciousness and selective attention is defined by informational enhancement: attention determines the degree of enhancement, and phenomenal consciousness arises when a threshold is reached. i also proposed that (2) to know what the contents of consciousness are, we need to know what information is enhanced and the degree to which that information is enhanced. question (3), in turn, remains open. it is plausible that some things that about the contents of conscious experience will remain unknown even if we know everything there is to know about how attention is distributed. for instance, we might remain ignorant about what it feels like when that information is enhanced. still, attention might provide a useful approximation to the scientific and objective investigation of the contents of phenomenal consciousness. acknowledgments writing this paper was possible thanks to a one-year research grant from the german service for academic exchange (daad). i am most thankful to ophelia deroy for hosting me and advising me during this time, as well as to my co-workers at the cvbe’s research group at the ludwig maximilian university of munich, not only for their interdisciplinary input and the rich and stimulating research environment they create but also for their various support as i adjusted to life in a new country: oriane armand, bahador bahrami, lucas battich, lenka gorman, jurgis karpus, anita keshmirian, slawa loew, louis longin, sofiia rappe, zahra rezazadeh, anna strasser, justin sulik, paul taylor, mark wulff carstensen, and jimena zapata. i am also most thankful to berit brogaard for the many illuminating discussions on the ideas in this paper and to e.j. green, magdalena balcerakjackson, and eli chudnoff for their very appreciated written feedback. many thanks to edgar deyoe, michael pitts, and marieke schölvinck for kindly granting their permission for presenting their figures here. finally, i wish to extend special thanks to two anonymous referees from this journal for their careful comments on previous versions of the manuscript. references alvarez, g. a., & scholl, b. j. (2005). how does attention select and track spatially extended objects? new effects of attentional concentration and amplification. journal of experimental psychology: general, 134(4), 461–476. https://doi.org/10.1037/0096-3445.134.4.461 andersen, s. k., & müller, m. m. (2010). behavioral performance follows the time course of neural facilitation and suppression during cued shifts of feature-selective attention. proceedings of the national academy of sciences, 107(31), 13878–13882. https://doi.org/10.1073/pnas.1002436107 ansorge, u., horstmann, g., & worschech, f. (2010). attentional capture by masked colour singletons. vision research, 50(19), 2015–2027. https://doi.org/https://doi.org/10.1016/j.visres.2010.07.015 lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1037/0096-3445.134.4.461 https://doi.org/10.1073/pnas.1002436107 https://doi.org/10.1016/j.visres.2010.07.015 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 35 baars, b. j. (2002). the conscious access hypothesis: origins and recent evidence. trends in cognitive sciences, 6(1), 47–52. https://doi.org/10.1016/s1364-6613(00)01819-2 baars, b. j. (2005). global workspace theory of consciousness: toward a cognitive neuroscience of human experience. in s. laureys (ed.), the boundaries of consciousness: neurobiology and neuropathology (vol. 150, pp. 45–53). elsevier. https://doi.org/https://doi.org/10.1016/s0079-6123(05)50004-9 bahrami, b., carmel, d., walsh, v., rees, g., & lavie, n. (2008). spatial attention can modulate unconscious orientation processing. perception, 37(10), 1520–1528. https://doi.org/10.1068/p5999 bahrami, b., lavie, n., & rees, g. (2007). attentional load modulates responses of human primary visual cortex to invisible stimuli. current biology, 17(6), 509–513. https://doi.org/10.1016/j.cub.2007.01.070 bauer, m., oostenveld, r., peeters, m., & fries, p. (2006). tactile spatial attention enhances gamma-band activity in somatosensory cortex and reduces low-frequency activity in parieto-occipital areas. the journal of neuroscience, 26(2), 490–501. https://doi.org/10.1523/jneurosci.5228-04.2006 block, n. (1995). on a confusion about a function of consciousness. behavioral and brain sciences, 18(2), 227—247. https://doi.org/10.1017/s0140525x00038188 block, n. (2007). consciousness, accessibility, and the mesh between psychology and neuroscience. behavioral and brain sciences, 30(5-6), 481–499. https://doi.org/10.1017/s0140525x07002786 block, n. (2008). consciousness and cognitive access. proceedings of the aristotelian society, 108(1pt3), 289–317. https://doi.org/10.1111/j.1467-9264.2008.00247.x block, n. (2012). the grain of vision and the grain of attention. thought: a journal of philosophy, 1(3), 170–184. https://doi.org/10.1002/tht3.28 block, n. (2014). rich conscious perception outside focal attention. trends in cognitive sciences, 18(9), 445–447. https://doi.org/10.1016/j.tics.2014.05.007 bonneh, y. s., cooperman, a., & sagi, d. (2001). motion-induced blindness in normal observers. nature, 411(6839), 798–801. https://doi.org/10.1038/35081073 braun, j. (2007). vision: attending the invisible. current biology, 17(6), r202–3. https://doi.org/10.1016/j.cub.2007.01.039 brogaard, b., chomanski, b., & gatzia, d. e. (2021). consciousness and information integration. synthese, 198(3), 763–792. https://doi.org/10.1007/s11229-020-02613-3 bronfman, z. z., brezis, n., jacobson, h., & usher, m. (2014). we see more than we can report: “cost free” color phenomenality outside focal attention. psychological science, 25(7), 1394–1403. https://doi.org/10.1177/0956797614532656 bruya, brian. (2010). effortless attention : a new perspective in the cognitive science of attention and action. mit press. https://public.ebookcentral.proquest.com/choice/publicfullrecord.aspx?p=3339135 buzsáki, g., & draguhn, a. (2004). neuronal oscillations in cortical networks. science, 304(5679), 1926–1929. https://doi.org/10.1126/science.1099745 carrasco, m. (2011). visual attention: the past 25 years. vision research, 51(13), 1484–1525. https://doi.org/10.1016/j.visres .2011.04.012 carrasco, m., & barbot, a. (2019). spatial attention alters visual appearance. current opinion in psychology, 29, 56–64. https://doi.org/10.1016/j.copsyc.2018.10.010 carrasco, m., ling, s., & read, s. (2004). attention alters appearance. nature neuroscience, 7(3), 308–313. https://doi.org/10.1038/nn1194 carruthers, p. (2000). phenomenal consciousness: a naturalistic theory. cambridge university press. https://doi.org/10.101 7/cbo9780511487491 chelazzi, l., miller, e. k., duncan, j., & desimone, r. (1993). a neural basis for visual search in inferior temporal cortex. nature, 363(6427), 345–347. https://doi.org/10.1038/363345a0 cherry, e. c. (1953). some experiments on the recognition of speech, with one and with two ears. journal of the acoustical society of america, 25, 975–979. https://doi.org/10.1121/1.1907229 chica, a. b., bayle, d. j., botta, f., bartolomeo, p., & paz-alonso, p. m. (2016). interactions between phasic alerting and consciousness in the fronto-striatal network. scientific reports, 6, 31868. https://doi.org/10.1038/srep31868 chica, a. b., lasaponara, s., lupiáñez, j., doricchi, f., & bartolomeo, p. (2010). exogenous attention can capture perceptual consciousness: erp and behavioural evidence. neuroimage, 51(3), 1205–1212. https://doi.org/10.1016/j.neuroimage.2010.03.002 christoff, k., irving, z. c., fox, k. c. r., spreng, r. n., & andrews-hanna, j. r. (2016). mind-wandering as spontaneous thought: a dynamic framework. nature reviews neuroscience, 17(11), 718–731. https://doi.org/10.1038/nrn.2016.113 chun, m. m., & potter, m. c. (1995). a two-stage model for multiple target detection in rapid serial visual presentation. journal of experimental psychology: human perception and performance, 21(1), 109–127. https://doi.org/10.1037//00961523.21.1.109 lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/s1364-6613(00)01819-2 https://doi.org/10.1016/s0079-6123(05)50004-9 https://doi.org/10.1068/p5999 https://doi.org/10.1016/j.cub.2007.01.070 https://doi.org/10.1523/jneurosci.5228-04.2006 https://doi.org/10.1017/s0140525x00038188 https://doi.org/10.1017/s0140525x07002786 https://doi.org/10.1111/j.1467-9264.2008.00247.x https://doi.org/10.1002/tht3.28 https://doi.org/10.1016/j.tics.2014.05.007 https://doi.org/10.1038/35081073 https://doi.org/10.1016/j.cub.2007.01.039 https://doi.org/10.1007/s11229-020-02613-3 https://doi.org/10.1177/0956797614532656 https://public.ebookcentral.proquest.com/choice/publicfullrecord.aspx?p=3339135 https://doi.org/10.1126/science.1099745 https://doi.org/10.1016/j.visres.2011.04.012 https://doi.org/10.1016/j.visres.2011.04.012 https://doi.org/10.1016/j.copsyc.2018.10.010 https://doi.org/10.1038/nn1194 https://doi.org/10.1017/cbo9780511487491 https://doi.org/10.1017/cbo9780511487491 https://doi.org/10.1038/363345a0 https://doi.org/10.1121/1.1907229 https://doi.org/10.1038/srep31868 https://doi.org/10.1016/j.neuroimage.2010.03.002 https://doi.org/10.1038/nrn.2016.113 https://doi.org/10.1037//0096-1523.21.1.109 https://doi.org/10.1037//0096-1523.21.1.109 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 36 clark, a. (2013). whatever next? predictive brains, situated agents, and the future of cognitive science. behavioral and brain sciences, 36(3), 181–204. https://doi.org/10.1017/s0140525x12000477 cohen, m. a., alvarez, g. a., & nakayama, k. (2011). natural-scene perception requires attention. psychological science, 22(9), 1165–1172. https://doi.org/10.1177/0956797611419168 cohen, m. a., cavanagh, p., chun, m. m., & nakayama, k. (2012). the attentional requirements of consciousness. trends in cognitive sciences, 16(8), 411–417. https://doi.org/10.1016/j.tics.2012.06.013 corbetta, m., & shulman, g. l. (2002). control of goal-directed and stimulus-driven attention in the brain. nature reviews neuroscience, 3(3), 201–215. https://doi.org/10.1038/nrn755 datta, r., & deyoe, e. (2009). i know where you are secretly attending! the topography of human visual attention revealed with fmri. vision research, 49, 1037–1044. https://doi.org/10.1016/j.visres.2009.01.014 davidson, m. j., alais, d., boxtel, j. j. van, & tsuchiya, n. (2018). attention periodically samples competing stimuli during binocular rivalry. elife, 7. https://doi.org/10.7554/elife.40868 de brigard, f. (2010). consciousness, attention and commonsense. journal of consciousness studies, 17(9-10), 189–201. https://www.ingentaconnect.com/content/imp/jcs/2010/00000017/f0020009/art00006 de weerd, p., smith, e., & greenberg, p. (2006). effects of selective attention on perceptual filling-in. journal of cognitive neuroscience, 18(3), 335–347. https://doi.org/10.1162/089892906775990561 dehaene, s., changeux, j.-p., naccache, l., sackur, j., & sergent, c. (2006). conscious, preconscious, and subliminal processing: a testable taxonomy. trends in cognitive sciences, 10(5), 204–211. https://doi.org/10.1016/j.tics.2006.03.007 dehaene, s., & naccache, l. (2001). towards a cognitive neuroscience of consciousness: basic evidence and a workspace framework. cognition, 79(1-2), 1–37. https://doi.org/10.1016/s0010-0277(00)00123-2 dehaene, s., naccache, l., cohen, l., bihan, d. l., mangin, j. f., poline, j. b., & rivière, d. (2001). cerebral mechanisms of word masking and unconscious repetition priming. nature neuroscience, 4(7), 752–758. https://doi.org/10.1038/89551 dennett, d. c. (2005). sweet dreams: philosophical obstacles to a science of consciousness (vol. 54, pp. 475–476). mit press. desimone, r., & duncan, j. (1995). neural mechanisms of selective visual attention. annual review of neuroscience, 18(1), 193–222. https://doi.org/10.1146/annurev.ne.18.030195.001205 devue, c., laloyaux, c., feyers, d., theeuwes, j., & brédart, s. (2009). do pictures of faces, and which ones, capture attention in the inattentional-blindness paradigm? perception, 38(4), 552–568. https://doi.org/10.1068/p6049 doesburg, s. m., kitajo, k., & ward, l. m. (2005). increased gamma-band synchrony precedes switching of conscious perceptual objects in binocular rivalry. neuroreport: for rapid communication of neuroscience research, 16(11), 1139– 1142. https://doi.org/10.1097/00001756-200508010-00001 driver, j. (1998). the neuropsychology of spatial attention. in attention. (pp. 297–340). psychology press/erlbaum (uk) taylor & francis. egly, r., driver, j., & rafal, r. d. (1994). shifting visual attention between objects and locations: evidence from normal and parietal lesion subjects. journal of experimental psychology: general, 123(2), 161–177. https://doi.org/10.1037//0096-3445.123.2.161 engel, a. k., & singer, w. (2001). temporal binding and the neural correlates of sensory awareness. trends in cognitive sciences, 5(1), 16–25. https://doi.org/10.1016/s1364-6613(00)01568-0 engell, a. d., & mccarthy, g. (2010). selective attention modulates face-specific induced gamma oscillations recorded from ventral occipitotemporal cortex. journal of neuroscience, 30(26), 8780–8786. https://doi.org/10.1523/jneurosci.1575-10.2010 faivre, n., & kouider, s. (2011). multi-feature objects elicit nonconscious priming despite crowding. journal of vision, 11(3). https://doi.org/10.1167/11.3.2 fazekas, p., & nanay, b. (2021). attention is amplification, not selection. british journal for the philosophy of science, 72(1), 299–324. https://doi.org/10.1093/bjps/axy065 fiebelkorn, i. c., & kastner, s. (2020). functional specialization in the attention network. annual review of psychology, 71, 221–249. https://doi.org/10.1146/annurev-psych-010418-103429 forschack, n., nierhaus, t., müller, m. m., & villringer, a. (2017). alpha-band brain oscillations shape the processing of perceptible as well as imperceptible somatosensory stimuli during selective attention. the journal of neuroscience, 37(29), 6983–6994. https://doi.org/10.1523/jneurosci.2582-16.2017 fries, p. (2005). a mechanism for cognitive dynamics: neuronal communication through neuronal coherence. trends in cognitive sciences, 9(10), 474–480. https://doi.org/10.1016/j.tics.2005.08.011 fries, p., reynolds, j. h., rorie, a. e., & desimone, r. (2001). modulation of oscillatory neuronal synchronization by selective visual attention. science, 291(5508), 1560–1563. https://doi.org/10.1126/science.1055465 lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1017/s0140525x12000477 https://doi.org/10.1177/0956797611419168 https://doi.org/10.1016/j.tics.2012.06.013 https://doi.org/10.1038/nrn755 https://doi.org/10.1016/j.visres.2009.01.014 https://doi.org/10.7554/elife.40868 https://www.ingentaconnect.com/content/imp/jcs/2010/00000017/f0020009/art00006 https://doi.org/10.1162/089892906775990561 https://doi.org/10.1016/j.tics.2006.03.007 https://doi.org/10.1016/s0010-0277(00)00123-2 https://doi.org/10.1038/89551 https://doi.org/10.1146/annurev.ne.18.030195.001205 https://doi.org/10.1068/p6049 https://doi.org/10.1097/00001756-200508010-00001 https://doi.org/10.1037//0096-3445.123.2.161 https://doi.org/10.1016/s1364-6613(00)01568-0 https://doi.org/10.1523/jneurosci.1575-10.2010 https://doi.org/10.1167/11.3.2 https://doi.org/10.1093/bjps/axy065 https://doi.org/10.1146/annurev-psych-010418-103429 https://doi.org/10.1523/jneurosci.2582-16.2017 https://doi.org/10.1016/j.tics.2005.08.011 https://doi.org/10.1126/science.1055465 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 37 fuchs, i., theeuwes, j., & ansorge, u. (2013). exogenous attentional capture by subliminal abrupt-onset cues: evidence from contrast-polarity independent cueing effects. journal of experimental psychology: human perception and performance, 39(4), 974–988. https://doi.org/10.1037/a0030419 gawne, t. j., kjaer, t. w., & richmond, b. j. (1996). latency: another potential code for feature binding in striate cortex. journal of neurophysiology, 76(2), 1356–1360. https://doi.org/10.1152/jn.1996.76.2.1356 geng, h., song, q., li, y., xu, s., & zhu, y. (2007). attentional modulation of motion-induced blindness. chinese science bulletin, 52(8), 1063–1070. https://doi.org/10.1007/s11434-007-0178-0 graf, e. w., adams, w. j., & lages, m. (2002). modulating motion-induced blindness with depth ordering and surface completion. vision research, 42(25), 2731–2735. https://doi.org/10.1016/s0042-6989(02)00390-5 graziano, m. s. a. (2013). consciousness and the social brain. oxford university press. graziano, m. s. a. (2019). attributing awareness to others: the attention schema theory and its relationship to behavioural prediction. journal of consciousness studies, 26(3-4), 17–37. https://www.ingentaconnect.com/content/imp/jcs/2019/ 00000026/f0020003/art00002 gross, j., schnitzler, a., timmermann, l., & ploner, m. (2007). gamma oscillations in human primary somatosensory cortex reflect pain perception. plos biology, 5(5), e133. https://doi.org/10.1371/journal.pbio.0050133 gruber, t., müller, m. m., keil, a., & elbert, t. (1999). selective visual-spatial attention alters induced gamma band responses in the human eeg. clinical neurophysiology, 110(12), 2074–2085. https://doi.org/10.1016/s1388-2457(99)00176-5 haladjian, h. h., & montemayor, c. (2015). on the evolution of conscious attention. psychonomic bulletin and review, 22(3), 595–613. https://doi.org/10.3758/s13423-014-0718-y haynes, j.-d., driver, j., & rees, g. (2005). visibility reflects dynamic changes of effective connectivity between v1 and fusiform cortex. neuron, 46(5), 811–821. https://doi.org/10.1016/j.neuron.2005.05.012 hesselmann, g., hebart, m., & malach, r. (2011). differential bold activity associated with subjective and objective reports during ”blindsight” in normal observers. the journal of neuroscience, 31(36), 12936–12944. https://doi.org/10.1523/jn eurosci.1556-11.2011 hopf, j.-m., boehler, c. n., luck, s. j., tsotsos, j. k., heinze, h.-j., & schoenfeld, m. a. (2006). direct neurophysiological evidence for spatial suppression surrounding the focus of attention in vision. proceedings of the national academy of sciences of the united states of america, 103(4), 1053. https://doi.org/10.1073/pnas.0507746103 intriligator, j., & cavanagh, p. (2001). the spatial resolution of visual attention. cognitive psychology, 43(3), 171–216. https://doi.org/10.1006/cogp.2001.0755 irving, z. (2016). mind-wandering is unguided attention: accounting for the “purposeful” wanderer. philosophical studies, 173(2), 547–571. https://doi.org/10.1007/s11098-015-0506-1 jackson-nielsen, m., cohen, m. a., & pitts, m. a. (2017). perception of ensemble statistics requires attention. consciousness and cognition, 48, 149–160. https://doi.org/10.1016/j.concog.2016.11.007 james, w. (1890). the principles of psychology (vol. 1). dover publications. jiang, y., costello, p., fang, f., huang, m., & he, s. (2006). a genderand sexual orientation-dependent spatial attentional effect of invisible images. proceedings of the national academy of sciences, 103(45), 17048. https://doi.org/10.1073/pn as.0605678103 kahneman, d., treisman, a., & gibbs, b. j. (1992). the reviewing of object files: object-specific integration of information. cognitive psychology, 24(2), 175–219. https://doi.org/10.1016/0010-0285(92)90007-o kanai, r., tsuchiya, n., & verstraten, f. a. j. (2006). the scope and limits of top-down attention in unconscious visual processing. current biology, 16(23), 2332–2336. https://doi.org/10.1016/j.cub.2006.10.001 kastner, s., & ungerleider, l. g. (2000). mechanisms of visual attention in the human cortex. annual review of neuroscience, 23, 315–341. https://doi.org/10.1146/annurev.neuro.23.1.315 kentridge, r. w. (2011). attention without awareness : a brief review. in c. mole, d. smithies, & w. wu (eds.), attention: philosophical and psychological essays (pp. 228–246). oxford university press. kentridge, r. w., heywood, c. a., & weiskrantz, l. (1999). attention without awareness in blindsight. proceedings of the royal society b biological sciences, 266(1430), 1805–1811. https://doi.org/10.1098/rspb.1999.0850 kentridge, r. w., nijboer, t. c. w., & heywood, c. a. (2008). attended but unseen: visual attention is not sufficient for visual awareness. neuropsychologia, 46(3), 864–869. https://doi.org/10.1016/j.neuropsychologia.2007.11.036 kiefer, m., & brendel, d. (2006). attentional modulation of unconscious “automatic” processes: evidence from event-related potentials in a masked priming paradigm. journal of cognitive neuroscience, 18(2), 184—198. https://doi.org/10.1162/ 089892906775783688 kiefer, m., & martens, u. (2010). attentional sensitization of unconscious cognition: task sets modulate subsequent masked semantic priming. journal of experimental psychology: general, 139(3), 464–489. https://doi.org/10.1037/a0019561 lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1037/a0030419 https://doi.org/10.1152/jn.1996.76.2.1356 https://doi.org/10.1007/s11434-007-0178-0 https://doi.org/10.1016/s0042-6989(02)00390-5 https://www.ingentaconnect.com/content/imp/jcs/2019/00000026/f0020003/art00002 https://www.ingentaconnect.com/content/imp/jcs/2019/00000026/f0020003/art00002 https://doi.org/10.1371/journal.pbio.0050133 https://doi.org/10.1016/s1388-2457(99)00176-5 https://doi.org/10.3758/s13423-014-0718-y https://doi.org/10.1016/j.neuron.2005.05.012 https://doi.org/10.1523/jneurosci.1556-11.2011 https://doi.org/10.1523/jneurosci.1556-11.2011 https://doi.org/10.1073/pnas.0507746103 https://doi.org/10.1006/cogp.2001.0755 https://doi.org/10.1007/s11098-015-0506-1 https://doi.org/10.1016/j.concog.2016.11.007 https://doi.org/10.1073/pnas.0605678103 https://doi.org/10.1073/pnas.0605678103 https://doi.org/10.1016/0010-0285(92)90007-o https://doi.org/10.1016/j.cub.2006.10.001 https://doi.org/10.1146/annurev.neuro.23.1.315 https://doi.org/10.1098/rspb.1999.0850 https://doi.org/10.1016/j.neuropsychologia.2007.11.036 https://doi.org/10.1162/089892906775783688 https://doi.org/10.1162/089892906775783688 https://doi.org/10.1037/a0019561 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 38 kim, c.-y., & blake, r. (2005). psychophysical magic: rendering the visible ‘invisible.’ trends in cognitive sciences, 9(8), 381–388. https://doi.org/10.1016/j.tics.2005.06.012 koch, c., massimini, m., boly, m., & tononi, g. (2016). neural correlates of consciousness: progress and problems. nature reviews neuroscience, 17(5), 307–321. https://doi.org/10.1038/nrn.2016.22 koch, c., & tsuchiya, n. (2007). attention and consciousness: two distinct brain processes. trends in cognitive sciences, 11(1), 16–22. https://doi.org/10.1016/j.tics.2006.10.012 kouider, s., barbot, a., madsen, k. h., lehericy, s., & summerfield, c. (2016). task relevance differentially shapes ventral visual stream sensitivity to visible and invisible faces. neuroscience of consciousness, 2016(1). https://doi.org/10.1093/ nc/niw021 landman, r., spekreijse, h., & lamme, v. a. f. (2003). large capacity storage of integrated objects before change blindness. vision research, 43(2), 149–164. https://doi.org/10.1016/s0042-6989(02)00402-9 lee, j., williford, t., & maunsell, j. h. r. (2007). spatial attention and the latency of neuronal responses in macaque area v4. the journal of neuroscience, 27(36), 9632–9637. https://doi.org/10.1523/jneurosci.2734-07.2007 li, f. f., vanrullen, r., koch, c., & perona, p. (2002). rapid natural scene categorization in the near absence of attention. proceedings of the national academy of sciences of the united states of america, 99(14), 9596–9601. https://doi.org/10.1 073/pnas.092277599 lopez, a. (2020). information gating and the structure of consciousness [phd thesis]. https://scholarship.miami.edu/discov ery/fulldisplay?context=l&vid=01uoml_inst:researchrepository&search_scope=research&tab=research&docid =alma991031524284802976 lou, l. (1999). selective peripheral fading: evidence for inhibitory sensory effect of attention. perception, 28(4), 519–526. https://doi.org/10.1068/p2816 luck, s. j., girelli, m., mcdermott, m. t., & ford, m. a. (1997). bridging the gap between monkey neurophysiology and human perception: an ambiguity resolution theory of visual selective attention. cognitive psychology, 33(1), 64–87. https://doi.org/10.1006/cogp.1997.0660 luck, s., woodman, g., & vogel, e. (2000). event-related potential studies of attention. trends in cognitive sciences, 4(11), 432–440. https://doi.org/10.1016/s1364-6613(00)01545-x lutz, a., dunne, j. d., & davidson, r. j. (2007). meditation and the neuroscience of consciousness. in p. d. zelazo, m. moscovitch, & e. thompson (eds.), cambridge handbook of consciousness (pp. 19–497). cambridge university press. lycan, w. g. (1996). consciousness and experience (vol. 72). mit press. macdonald, j. s. p., & lavie, n. (2008). load induced blindness. journal of experimental psychology: human perception and performance, 34(5), 1078–1091. https://doi.org/10.1037/0096-1523.34.5.1078 mack, a., & clarke, j. (2012). gist perception requires attention. visual cognition, 20(3), 300–327. https://doi.org/10.1080/ 13506285.2012.666578 mack, a., & rock, i. (1998). inattentional blindness. the mit press. maier, a., & tsuchiya, n. (2021). growing evidence for separate neural mechanisms for attention and consciousness. attention, perception & psychophysics, 83(2), 558–576. https://doi.org/10.3758/s13414-020-02146-4 martinez-trujillo, j. c., & treue, s. (2004). feature-based attention increases the selectivity of population responses in primate visual cortex. current biology, 14(9), 744–751. https://doi.org/10.1016/j.cub.2004.04.028 matthews, j., schröder, p., kaunitz, l., boxtel, j. j. a. van, & tsuchiya, n. (2018). conscious access in the near absence of attention: critical extensions on the dual-task paradigm. philosophical transactions of the royal society london b biological sciences, 373(1755). https://doi.org/10.1098/rstb.2017.0352 melloni, l., molina, c., pena, m., torres, d., singer, w., & rodriguez, e. (2007). synchronization of neural activity across cortical areas correlates with conscious perception. the journal of neuroscience, 27(11), 2858–2865. https://doi.org/10 .1523/jneurosci.4623-06.2007 mole, c. (2008a). attention in the absence of consciousness? trends in cognitive sciences, 12(2), 44; author reply 44–5. https://doi.org/10.1016/j.tics.2007.11.001 mole, c. (2008b). attention and consciousness. journal of consciousness studies, 15(4), 86–104. https://www.ingentaconne ct.com/content/imp/jcs/2008/00000015/00000004/art00006 montemayor, c., & haladjian, h. h. (2015). consciousness, attention, and conscious attention. mit press. moran, j., & desimone, r. (1985). selective attention gates visual processing in the extrastriate cortex. science, 229(4715), 782–784. https://doi.org/10.1126/science.4023713 moray, n. (1959). attention in dichotic listening: affective cues and the influence of instructions. the quarterly journal of experimental psychology, 11, 56–60. https://doi.org/10.1080/17470215908416289 nagel, t. (1974). what is it like to be a bat? philosophical review, 83, 435–450. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.tics.2005.06.012 https://doi.org/10.1038/nrn.2016.22 https://doi.org/10.1016/j.tics.2006.10.012 https://doi.org/10.1093/nc/niw021 https://doi.org/10.1093/nc/niw021 https://doi.org/10.1016/s0042-6989(02)00402-9 https://doi.org/10.1523/jneurosci.2734-07.2007 https://doi.org/10.1073/pnas.092277599 https://doi.org/10.1073/pnas.092277599 https://scholarship.miami.edu/discovery/fulldisplay?context=l&vid=01uoml_inst:researchrepository&search_scope=research&tab=research&docid=alma991031524284802976 https://scholarship.miami.edu/discovery/fulldisplay?context=l&vid=01uoml_inst:researchrepository&search_scope=research&tab=research&docid=alma991031524284802976 https://scholarship.miami.edu/discovery/fulldisplay?context=l&vid=01uoml_inst:researchrepository&search_scope=research&tab=research&docid=alma991031524284802976 https://doi.org/10.1068/p2816 https://doi.org/10.1006/cogp.1997.0660 https://doi.org/10.1016/s1364-6613(00)01545-x https://doi.org/10.1037/0096-1523.34.5.1078 https://doi.org/10.1080/13506285.2012.666578 https://doi.org/10.1080/13506285.2012.666578 https://doi.org/10.3758/s13414-020-02146-4 https://doi.org/10.1016/j.cub.2004.04.028 https://doi.org/10.1098/rstb.2017.0352 https://doi.org/10.1523/jneurosci.4623-06.2007 https://doi.org/10.1523/jneurosci.4623-06.2007 https://doi.org/10.1016/j.tics.2007.11.001 https://www.ingentaconnect.com/content/imp/jcs/2008/00000015/00000004/art00006 https://www.ingentaconnect.com/content/imp/jcs/2008/00000015/00000004/art00006 https://doi.org/10.1126/science.4023713 https://doi.org/10.1080/17470215908416289 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 39 naghavi, h. r., & nyberg, l. (2005). common fronto-parietal activity in attention, memory, and consciousness: shared demands on integration? consciousness and cognition, 14(2), 390–425. https://doi.org/10.1016/j.concog.2004.10.003 nani, a., manuello, j., mancuso, l., liloia, d., costa, t., & cauda, f. (2019). the neural correlates of consciousness and attention: two sister processes of the brain. frontiers in neuroscience, 13, 1169. https://doi.org/10.3389/fnins.2019.01169 noguchi, y., tanabe, h. c., sadato, n., hoshiyama, m., & kakigi, r. (2007). voluntary attention changes the speed of perceptual neural processing. european journal of neuroscience, 25(10), 3163–3172. https://doi.org/10.1111/j.14609568.2007.05552.x norman, l. j., heywood, c. a., & kentridge, r. w. (2013). object-based attention without awareness. psychological science, 24(6), 836–843. https://doi.org/10.1177/0956797612461449 norman, l. j., heywood, c. a., & kentridge, r. w. (2015). exogenous attention to unseen objects? consciousness and cognition, 35, 319—329. https://doi.org/10.1016/j.concog.2015.02.015 o’craven, k. m., downing, p. e., & kanwisher, n. (1999). fmri evidence for objects as the units of attentional selection. nature, 401(6753), 584–587. https://doi.org/10.1038/44134 oizumi, m., albantakis, l., & tononi, g. (2014). from the phenomenology to the mechanisms of consciousness: integrated information theory 3.0. plos computational biology, 10(5), e1003588. https://doi.org/10.1371/journal.pcbi.1003588 olivers, c. n. l., & nieuwenhuis, s. (2005). the beneficial effect of concurrent task-irrelevant mental activity on temporal attention. psychological science, 16(4), 265–269. https://doi.org/10.1111/j.0956-7976.2005.01526.x pelli, d. g. (2008). crowding: a cortical constraint on object recognition. current opinion in neurobiology, 18(4), 445–451. https://doi.org/10.1016/j.conb.2008.09.008 petersen, s. e., & posner, m. i. (2012). the attention system of the human brain: 20 years after. annual review of neuroscience, 35, 73–89. https://doi.org/10.1146/annurev-neuro-062111-150525 pitts, m. a., lutsyshyna, l. a., & hillyard, s. a. (2018). the relationship between attention and consciousness: an expanded taxonomy and implications for ’no-report’ paradigms. philosophical transactions of the royal society london b biological sciences, 373(1755). https://doi.org/10.1098/rstb.2017.0348 polonsky, a., blake, r., braun, j., & heeger, d. j. (2000). neuronal activity in human primary visual cortex correlates with perception during binocular rivalry. nature neuroscience, 3(11), 1153–1159. https://doi.org/10.1038/80676 posner, m. i. (1980). orienting of attention. quarterly journal of experimental psychology, 32(1), 3–25. https://doi.org/10.1 080/00335558008248231 posner, m. i., & petersen, s. e. (1990). the attention system of the human brain. annual review of neuroscience, 13, 25–42. https://doi.org/10.1146/annurev.ne.13.030190.000325 prettyman, a. (2014). attention and conscious perception. [phd thesis]. university of toronto. prettyman, a. (2021). what is diffuse attention? mind and language. https://2a013fac-9cf9-4aa3-b07e-aff848aae421.filesus r.com/ugd/e09b7d_33fc1ad31deb4d408111d0a72a7a7ac0.pdf prinz, j. (2012). the conscious brain: how attention engenders experience. oxford university press usa. pylyshyn, z. (2004). some puzzling findings in multiple object tracking: i. tracking without keeping track of object identities. visual cognition, 11(7), 801–822. https://doi.org/10.1080/13506280344000518 pylyshyn, z. w. (2006). some puzzling findings in multiple object tracking (mot): ii. inhibition of moving nontargets. visual cognition, 14(2), 175–198. https://doi.org/10.1080/13506280544000200 raffone, a., & pantani, m. (2010). a global workspace model for phenomenal and access consciousness. consciousness and cognition: an international journal, 19(2), 580–596. https://doi.org/10.1016/j.concog.2010.03.013 raymond, j. e., shapiro, k. l., & arnell, k. m. (1992). temporary suppression of visual processing in an rsvp task: an attentional blink? journal of experimental psychology: human perception and performance, 18(3), 849–860. https://doi.org/10.1037//0096-1523.18.3.849 raz, a., & buhle, j. (2006). typologies of attentional networks. nature reviews neuroscience, 7(5), 367–379. https://doi.org/ 10.1038/nrn1903 reddy, l., moradi, f., & koch, c. (2007). top-down biases win against focal attention in the fusiform face area. neuroimage, 38(4), 730–739. https://doi.org/10.1016/j.neuroimage.2007.08.006 reddy, l., reddy, l., & koch, c. (2006). face identification in the near-absence of focal attention. vision research, 46(15), 2336–2343. https://doi.org/10.1016/j.visres.2006.01.020 reddy, l., wilken, p., & koch, c. (2004). face-gender discrimination is possible in the near-absence of attention. journal of vision, 4(2), 4–4. https://doi.org/10.1167/4.2.4 remington, r. w., & folk, c. l. (2001). a dissociation between attention and selection. psychological science, 12(6), 511–515. https://doi.org/10.1111/1467-9280.00394 lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1016/j.concog.2004.10.003 https://doi.org/10.3389/fnins.2019.01169 https://doi.org/10.1111/j.1460-9568.2007.05552.x https://doi.org/10.1111/j.1460-9568.2007.05552.x https://doi.org/10.1177/0956797612461449 https://doi.org/10.1016/j.concog.2015.02.015 https://doi.org/10.1038/44134 https://doi.org/10.1371/journal.pcbi.1003588 https://doi.org/10.1111/j.0956-7976.2005.01526.x https://doi.org/10.1016/j.conb.2008.09.008 https://doi.org/10.1146/annurev-neuro-062111-150525 https://doi.org/10.1098/rstb.2017.0348 https://doi.org/10.1038/80676 https://doi.org/10.1080/00335558008248231 https://doi.org/10.1080/00335558008248231 https://doi.org/10.1146/annurev.ne.13.030190.000325 https://2a013fac-9cf9-4aa3-b07e-aff848aae421.filesusr.com/ugd/e09b7d_33fc1ad31deb4d408111d0a72a7a7ac0.pdf https://2a013fac-9cf9-4aa3-b07e-aff848aae421.filesusr.com/ugd/e09b7d_33fc1ad31deb4d408111d0a72a7a7ac0.pdf https://doi.org/10.1080/13506280344000518 https://doi.org/10.1080/13506280544000200 https://doi.org/10.1016/j.concog.2010.03.013 https://doi.org/10.1037//0096-1523.18.3.849 https://doi.org/10.1038/nrn1903 https://doi.org/10.1038/nrn1903 https://doi.org/10.1016/j.neuroimage.2007.08.006 https://doi.org/10.1016/j.visres.2006.01.020 https://doi.org/10.1167/4.2.4 https://doi.org/10.1111/1467-9280.00394 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 40 rensink, r. a., o’regan, j. k., & clark, j. j. (1997). to see or not to see: the need for attention to perceive changes in scenes. psychological science, 8(5), 368–373. https://doi.org/10.1111/j.1467-9280.1997.tb00427.x ress, d., backus, b. t., & heeger, d. j. (2000). activity in primary visual cortex predicts performance in a visual detection task. nature neuroscience, 3(9), 940–945. https://doi.org/10.1038/78856 ress, d., & heeger, d. j. (2003). neuronal correlates of perception in early visual cortex. nature neuroscience, 6(4), 414–420. https://doi.org/10.1038/nn1024 reynolds, j. h., & chelazzi, l. (2004). attentional modulation of visual processing. annual review of neuroscience, 27, 611–647. https://doi.org/10.1146/annurev.neuro.26.041002.131039 reynolds, j. h., chelazzi, l., & desimone, r. (1999). competitive mechanisms subserve attention in macaque areas v2 and v4. the journal of neuroscience, 19(5), 1736–1753. https://doi.org/10.1523/jneurosci.19-05-01736.1999 reynolds, j. h., & heeger, d. j. (2009). the normalization model of attention. neuron, 61(2), 168–185. https://doi.org/10.101 6/j.neuron.2009.01.002 rosenthal, d. m. (1997). a theory of consciousness. in n. block, o. j. flanagan, & g. guzeldere (eds.), the nature of consciousness. mit press. scharlau, i. (2002). leading, but not trailing, primes influence temporal order perception: further evidence for an attentional account of perceptual latency priming. perception & psychophysics, 64(8), 1346–1360. https://doi.org/10.3758/bf03194777 scharlau, i., & ansorge, u. (2003). direct parameter specification of an attention shift: evidence from perceptual latency priming. vision research, 43(12), 1351–1363. https://doi.org/10.1016/s0042-6989(03)00141-x scharlau, i., & neumann, o. (2003). perceptual latency priming by masked and unmasked stimuli: evidence for an attentional interpretation. psychological research, 67(3), 184–196. https://doi.org/10.1007/s00426-002-0116-3 schölvinck, m. l., & rees, g. (2009). attentional influences on the dynamics of motion-induced blindness. journal of vision, 9(1), 38.1–9. https://doi.org/10.1167/9.1.38 schurger, a., cowey, a., cohen, j. d., treisman, a., & tallon-baudry, c. (2008). distinct and independent correlates of attention and awareness in a hemianopic patient. neuropsychologia, 46(8), 2189–2197. https://doi.org/10.1016/j.neur opsychologia.2008.02.020 schurger, a., cowey, a., & tallon-baudry, c. (2006). induced gamma-band oscillations correlate with awareness in hemianopic patient gy. neuropsychologia, 44(10), 1796–1803. https://doi.org/10.1016/j.neuropsychologia.2006.03.015 schwitzgebel, e. (2007). do you have constant tactile experience of your feet in your shoes? or is experience limited to what’s in attention? journal of consciousness studies, 14(3), 5–35. https://www.ingentaconnect.com/content/imp/jcs /2007/00000014/00000003/art00001 sergent, c., ruff, c. c., barbot, a., driver, j., & rees, g. (2011). top-down modulation of human early visual cortex after stimulus offset supports successful postcued report. journal of cognitive neuroscience, 23(8), 1921–1934. https://doi. org/10.1162/jocn.2010.21553 simons, d. j., & chabris, c. f. (1999). gorillas in our midst: sustained inattentional blindness for dynamic events. perception, 28(9), 1059–1074. https://doi.org/10.1068/p281059 singer, w., & gray, c. m. (1995). visual feature integration and the temporal correlation hypothesis. annual review of neuroscience, 18, 555–586. https://doi.org/10.1146/annurev.ne.18.030195.003011 smallwood, j., & schooler, j. w. (2015). the science of mind wandering: empirically navigating the stream of consciousness. annual review of psychology, 66(1), 487–518. https://doi.org/10.1146/annurev-psych-010814-015331 smithies, d. (2011). attention is rational-access consciousness. in c. mole, d. smithies, & w. wu (eds.), attention: philosophical and psychological essays (pp. 247–273). oxford university press. sperling, g. (1960). the information available in brief visual presentations. psychological monographs: general and applied, 74(11), 1–29. https://doi.org/10.1037/h0093759 stazicker, j. (2011). attention, visual consciousness and indeterminacy. mind and language, 26(2), 156–184. https://doi.org/10.1111/j.1468-0017.2011.01414.x steriade, m., nuñez, a., & amzica, f. (1993). a novel slow (< 1 hz) oscillation of neocortical neurons in vivo: depolarizing and hyperpolarizing components. the journal of neuroscience, 13(8), 3252–3265. https://doi.org/10.1523/jneurosci. 13-08-03252.1993 stoljar, d. (2019). in praise of poise. in a. pautz & d. stoljar (eds.), blockheads! essays on ned block’s philosophy of mind and consciousness. tallon-baudry, c. (2012). on the neural mechanisms subserving consciousness and attention. frontiers in psychology, 2, 397–397. https://doi.org/10.3389/fpsyg.2011.00397 tallon-baudry, c., bertrand, o., delpuech, c., & permier, j. (1997). oscillatory gamma-band (30-70 hz) activity induced by a visual search task in humans. the journal of neuroscience, 17(2), 722–734. https://doi.org/10.1523/jneurosci.1702-00722.1997 lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1111/j.1467-9280.1997.tb00427.x https://doi.org/10.1038/78856 https://doi.org/10.1038/nn1024 https://doi.org/10.1146/annurev.neuro.26.041002.131039 https://doi.org/10.1523/jneurosci.19-05-01736.1999 https://doi.org/10.1016/j.neuron.2009.01.002 https://doi.org/10.1016/j.neuron.2009.01.002 https://doi.org/10.3758/bf03194777 https://doi.org/10.1016/s0042-6989(03)00141-x https://doi.org/10.1007/s00426-002-0116-3 https://doi.org/10.1167/9.1.38 https://doi.org/10.1016/j.neuropsychologia.2008.02.020 https://doi.org/10.1016/j.neuropsychologia.2008.02.020 https://doi.org/10.1016/j.neuropsychologia.2006.03.015 https://www.ingentaconnect.com/content/imp/jcs/2007/00000014/00000003/art00001 https://www.ingentaconnect.com/content/imp/jcs/2007/00000014/00000003/art00001 https://doi.org/10.1162/jocn.2010.21553 https://doi.org/10.1162/jocn.2010.21553 https://doi.org/10.1068/p281059 https://doi.org/10.1146/annurev.ne.18.030195.003011 https://doi.org/10.1146/annurev-psych-010814-015331 https://doi.org/10.1037/h0093759 https://doi.org/10.1111/j.1468-0017.2011.01414.x https://doi.org/10.1523/jneurosci.13-08-03252.1993 https://doi.org/10.1523/jneurosci.13-08-03252.1993 https://doi.org/10.3389/fpsyg.2011.00397 https://doi.org/10.1523/jneurosci.17-02-00722.1997 https://doi.org/10.1523/jneurosci.17-02-00722.1997 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org vicarious attention, degrees of enhancement, and the contents of consciousness 41 tallon-baudry, c., bertrand, o., hénaff, m.-a., isnard, j., & fischer, c. (2005). attention modulates gamma-band oscillations differently in the human lateral occipital cortex and fusiform gyrus. cerebral cortex, 15(5), 654–662. https://doi.org/ 10.1093/cercor/bhh167 tallon-baudry, c., campana, f., park, h.-d., & babo-rebelo, m. (2018). the neural monitoring of visceral inputs, rather than attention, accounts for first-person perspective in conscious vision. cortex, 102, 139–149. https://doi.org/10.101 6/j.cortex.2017.05.019 taylor, h., & sayim, b. (2018). crowding, attention and consciousness: in support of the inference hypothesis. mind & language, 33(1), 17–33. https://doi.org/10.1111/mila.12169 thiele, a., & bellgrove, m. a. (2018). neuromodulation of attention. neuron, 97(4), 769–785. https://doi.org/10.1016/j.neur on.2018.01.008 thut, g., nietzel, a., brandt, s. a., & pascual-leone, a. (2006). alpha-band electroencephalographic activity over occipital cortex indexes visuospatial attention bias and predicts visual target detection. the journal of neuroscience, 26(37), 9494–9502. https://doi.org/10.1523/jneurosci.0875-06.2006 tong, f., nakayama, k., vaughan, j. t., & kanwisher, n. (1998). binocular rivalry and visual awareness in human extrastriate cortex. neuron, 21(4), 753–759. https://doi.org/10.1016/s0896-6273(00)80592-9 tononi, g., boly, m., massimini, m., & koch, c. (2016). integrated information theory: from consciousness to its physical substrate. nature reviews neuroscience, 17(7), 450–461. https://doi.org/10.1038/nrn.2016.44 travis, s. l., dux, p. e., & mattingley, j. b. (2017). re-examining the influence of attention and consciousness on visual afterimage duration. journal of experimental psychology: human perception and performance, 43(12), 1944–1949. https://doi.org/10.1037/xhp0000458 travis, s. l., dux, p. e., & mattingley, j. b. (2019). neural correlates of goal-directed enhancement and suppression of visual stimuli in the absence of conscious perception. attention, perception & psychophysics, 81(5), 1346–1364. https://doi.org/10.3758/s13414-018-1615-7 treisman, a. m. (1969). strategies and models of selective attention. psychological review, 76(3), 282–299. https://doi.org/10.1037/h0027242 treue, s. (2003). visual attention: the where, what, how and why of saliency. current opinion in neurobiology, 13(4), 428–432. https://doi.org/10.1016/s0959-4388(03)00105-3 troxler, i. p. v. (1804). über das verschwinden gegebener gegenstände innerhalb unseres gesichtskreises. ophthalmologische bibliothek, 2(2), 1–53. https://www.worldcat.org/title/ophthalmologische-bibliothek/oclc/491712012 tse, p. u. (2004). mapping visual attention with change blindness: new directions for a new method. cognitive science, 28(2), 241–258. https://doi.org/10.1016/j.cogsci.2003.12.002 tse, p. u., martinez-conde, s., schlegel, a. a., & macknik, s. l. (2005). visibility, visual awareness, and visual masking of simple unattended targets are confined to areas in the occipital cortex beyond human v1/v2. proceedings of the national academy of sciences of the united states of america, 102(47), 17178–17183. https://doi.org/10.1073/pnas.0508010102 tsuchiya, n., wilke, m., frässle, s., & lamme, v. a. f. (2015). no-report paradigms: extracting the true neural correlates of consciousness. trends in cognitive sciences, 19(12), 757–770. https://doi.org/10.1016/j.tics.2015.10.002 van boxtel, j. j. a., tsuchiya, n., & koch, c. (2010a). opposing effects of attention and consciousness on afterimages. proceedings of the national academy of sciences of the united states of america, 107(19), 8883–8888. https://doi.org/10 .1073/pnas.0913292107 van boxtel, j., tsuchiya, n., & koch, c. (2010b). consciousness and attention: on sufficiency and necessity. frontiers in psychology, 1, 217. https://doi.org/10.3389/fpsyg.2010.00217 van den bussche, e., hughes, g., humbeeck, n. v., & reynvoet, b. (2010). the relation between consciousness and attention: an empirical study using the priming paradigm. consciousness and cognition, 19(1), 86–97. https://doi.org/10.1016/j. concog.2009.12.019 vanduffel, w., tootell, r. b., & orban, g. a. (2000). attention-dependent suppression of metabolic activity in the early stages of the macaque visual system. cerebral cortex, 10(2), 109–126. https://doi.org/10.1093/cercor/10.2.109 vidal, j. r., chaumon, m., o’regan, j. k., & tallon-baudry, c. (2006). visual grouping and the focusing of attention induce gamma-band oscillations at different frequencies in human magnetoencephalogram signals. journal of cognitive neuroscience, 18(11), 1850–1862. https://doi.org/10.1162/jocn.2006.18.11.1850 volz, l. j., & gazzaniga, m. s. (2017). interaction in isolation: 50 years of insights from split-brain research. brain, 140(7), 2051–2060. https://doi.org/10.1093/brain/awx139 vuilleumier, p., sagiv, n., hazeltine, e., poldrack, r. a., swick, d., rafal, r. d., & gabrieli, j. d. e. (2001). neural fate of seen and unseen faces in visuospatial neglect: a combined event-related functional mri and event-related potential study. proceedings of the national academy of sciences, 98(6), 3495–3500. https://doi.org/10.1073/pnas.051436898 lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1093/cercor/bhh167 https://doi.org/10.1093/cercor/bhh167 https://doi.org/10.1016/j.cortex.2017.05.019 https://doi.org/10.1016/j.cortex.2017.05.019 https://doi.org/10.1111/mila.12169 https://doi.org/10.1016/j.neuron.2018.01.008 https://doi.org/10.1016/j.neuron.2018.01.008 https://doi.org/10.1523/jneurosci.0875-06.2006 https://doi.org/10.1016/s0896-6273(00)80592-9 https://doi.org/10.1038/nrn.2016.44 https://doi.org/10.1037/xhp0000458 https://doi.org/10.3758/s13414-018-1615-7 https://doi.org/10.1037/h0027242 https://doi.org/10.1016/s0959-4388(03)00105-3 https://www.worldcat.org/title/ophthalmologische-bibliothek/oclc/491712012 https://doi.org/10.1016/j.cogsci.2003.12.002 https://doi.org/10.1073/pnas.0508010102 https://doi.org/10.1016/j.tics.2015.10.002 https://doi.org/10.1073/pnas.0913292107 https://doi.org/10.1073/pnas.0913292107 https://doi.org/10.3389/fpsyg.2010.00217 https://doi.org/10.1016/j.concog.2009.12.019 https://doi.org/10.1016/j.concog.2009.12.019 https://doi.org/10.1093/cercor/10.2.109 https://doi.org/10.1162/jocn.2006.18.11.1850 https://doi.org/10.1093/brain/awx139 https://doi.org/10.1073/pnas.051436898 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org azenet lopez 42 watanabe, m., cheng, k., murayama, y., ueno, k., asamizuya, t., tanaka, k., & logothetis, n. (2011). attention but not awareness modulates the bold signal in the human v1 during binocular suppression. science, 334(6057), 829–831. https://doi.org/10.1126/science.1203161 watzl, s. (2017). structuring mind: the nature of attention and how it shapes consciousness. oxford university press. webb, t. w., igelström, k. m., schurger, a., & graziano, m. s. a. (2016). cortical networks involved in visual awareness independent of visual attention. proceedings of the national academy of sciences, 113(48), 13923–13928. https://doi.org/10.1073/pnas.1611505113 wilke, m., logothetis, n. k., & leopold, d. a. (2006). local field potential reflects perceptual suppression in monkey visual cortex. proceedings of the national academy of sciences, 103(46), 17507–17512. https://doi.org/10.1073/pnas.0604673103 wolfe, j. m., reinecke, a., & brawn, p. (2006). why don’t we see changes?: the role of attentional bottlenecks and limited visual memory. visual cognition, 14(4-8), 749–780. https://doi.org/10.1080/13506280500195292 womelsdorf, t., & fries, p. (2007). the role of neuronal synchronization in selective attention. current opinion in neurobiology, 17(2), 154–160. https://doi.org/10.1016/j.conb.2007.02.002 worden, m. s., foxe, j. j., wang, n., & simpson, g. v. (2000). anticipatory biasing of visuospatial attention indexed by retinotopically specific alpha-band electroencephalography increases over occipital cortex. the journal of neuroscience, 20(6), rc63. https://doi.org/10.1523/jneurosci.20-06-j0002.2000 wu, w. (2014). attention. routledge. https://doi.org/10.4324/9781315777603 wu, w. (2017). attention and perception: a necessary connection? in b. nanay (ed.), current controversies in philosophy of perception (pp. 148–162). routledge. wu, w. (2018). the neuroscience of consciousness. in e. n. zalta (ed.), the stanford encyclopedia of philosophy (winter 2018). metaphysics reseach lab, stanford university. https://plato.stanford.edu/archives/win2018/entries/consciousnessneuroscience/ wyart, v., & tallon-baudry, c. (2008). neural dissociation between visual awareness and spatial attention. the journal of neuroscience, 28(10), 2667–2679. https://doi.org/10.1523/jneurosci.4748-07.2008 wyart, v., & tallon-baudry, c. (2009). how ongoing fluctuations in human visual cortex predict perceptual awareness: baseline shift versus decision bias. the journal of neuroscience, 29(27), 8715–8725. https://doi.org/10.1523/jneurosc i.0962-09.2009 zhao, t., zhu, y., tang, h., xie, r., zhu, j., & zhang, j. h. (2019). consciousness: new concepts and neural networks. frontiers in cellular neuroscience, 13, 302. https://doi.org/10.3389/fncel.2019.00302 open access this article is distributed under the terms of the creative commons attribution 4.0 international license (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, as long as you give appropriate credit to the original author(s) and the source, provide a link to the creative commons license, and indicate if changes were made. lopez, a. (2022). vicarious attention, degrees of enhancement, and the contents of consciousness. philosophy and the mind sciences, 3, 1. https://doi.org/10.33735/phimisci.2022.9194 ©the author(s). https://philosophymindscience.org issn: 2699-0369 https://doi.org/10.1126/science.1203161 https://doi.org/10.1073/pnas.1611505113 https://doi.org/10.1073/pnas.0604673103 https://doi.org/10.1080/13506280500195292 https://doi.org/10.1016/j.conb.2007.02.002 https://doi.org/10.1523/jneurosci.20-06-j0002.2000 https://doi.org/10.4324/9781315777603 https://plato.stanford.edu/archives/win2018/entries/consciousness-neuroscience/ https://plato.stanford.edu/archives/win2018/entries/consciousness-neuroscience/ https://doi.org/10.1523/jneurosci.4748-07.2008 https://doi.org/10.1523/jneurosci.0962-09.2009 https://doi.org/10.1523/jneurosci.0962-09.2009 https://doi.org/10.3389/fncel.2019.00302 https://doi.org/10.33735/phimisci.2022.9194 https://creativecommons.org/licenses/by/4.0/ https://philosophymindscience.org the explanatory project selective attention as a window into phenomenal consciousness a dissociation on three fronts attention without consciousness consciousness without attention opposite functions and effects double dissociation and independence taking stock tying attention and consciousness back together degrees of enhancement degrees of enhancement, amounts of attention and measures of integration what if attention and consciousness amplify signals in different ways? why does attention sometimes make stimuli disappear from consciousness? attention and consciousness without enhancement? wrapping up vicarious attention concluding remarks: back to the explanatory project