Title of the Paper [16 point font] © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. Speech Act Pluralism in Argumentative Polylogues MARCIN LEWIŃSKI ArgLab, Nova Institute of Philosophy Nova University of Lisbon Campus de Campolide, 1099-032 Lisbon Portugal m.lewinski@fcsh.unl.pt Abstract: I challenge two key as- sumptions of speech act theory, as applied to argumentation: illocution- ary monism, grounded in the idea each utterance has only one (primary) illocutionary force, and the dyadic reduction, which models interaction as a dyadic affair between only two agents (speaker-hearer, proponent- opponent). I show how major contri- butions to speech act inspired study of argumentation adhere to these as- sumptions even as illocutionary pluralism in argumentative pol- ylogues is a significant empirical fact in need of theoretical attention. I demonstrate this with two examples where arguers interacting with multi- ple persons convey plural, argumenta- tively relevant illocutionary forces. Understanding illocutionary pluralism in argumentative polylogues also affords a better account of fallacious and manipulative discourse. Résumé: Je conteste deux hypothèses clés de la théorie des actes de langage appliquée à l'argumentation : l'une du monisme illocutoire, fondée sur l'idée que chaque énoncé n'a qu'une seule force illocutoire (primaire) ; et la réduction dyadique, qui modélise l'interaction comme une affaire dyadique entre seulement deux agents (locuteur-auditeur, promoteur- opposant). Les contributions majeures à l'étude de l'argumentation inspirée par les actes de langage adhèrent à ces hypothèses. Ainsi, la théorie de l'argumentation devrait embrasser le pluralisme illocutoire dans les pol- ylogues argumentatifs. Je démontre le fonctionnement de tels échanges argumentatifs complexes avec deux exemples où des personnes qui argumentent entre elles interagissent avec plusieurs autres personnes et en même temps transmettent plusieurs forces illocutoires et pertinentes dans leur argumentation. Le pluralisme illocutoire dans les polylogues argu- mentatifs permet également de mieux rendre compte des discours fallacieux et manipulateurs. Keywords: argumentative polylogues, fallacies, illocutionary force, speech act theory 422 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. 1. Introduction Speech act theory and argumentation theory provide a powerful account of what human interaction is about. This is especially so when we consider them jointly as mutually reinforcing, comple- mentary theories. Interaction is not a mere exchange of infor- mation; it is, rather, a joint human activity in which people per- form certain intentional acts to achieve their individual and collec- tive goals.1 These intentional acts are predominantly performed in speech: they are, that is, speech acts. And for these speech acts to be performed in the pursuit of individual and collective goals, there needs to exist some presumption of rationality behind them.2 Argumentation is there to protect this presumption when things go wrong. And because things go wrong pretty much all the time— joint activities cannot quite succeed without quality control, and, whenever needed, the resolution of doubt and the management of disagreement—argumentation is a ubiquitous feature of our inter- actions. It is a powerful backup generator constantly connected to our interactional grid. While this generator is fuelled by reasons— there is no argument-making without reason-giving—its energy mix includes many other powerful ingredients: questions, chal- lenges, rebuttals, concessions, retractions, clarifications, and other metalinguistic interventions. Curiously, these ingredients them- selves can be, and have been, analyzed as speech acts. In this way, 1 I consider this to be an impeccably Gricean insight (see also Tomasello 2008). Grice writes: “The conversational maxims, however, and the conversational implicatures connected with them, are specially connected (I hope) with the particular purposes that talk (and so, talk exchange) is adapted to serve and is primarily employed to serve. I have stated my maxims as if this purpose were a maximal- ly effective exchange of information; this specification is, of course, too narrow, and the scheme needs to be generalized to allow for such general purposes as influencing or directing the actions of others.” (Grice 1975, p. 47). Yet, Gricean “as-ifness” tends to be obliterated and the exclusive focus on information exchange remains a persistent analytic practice. For an otherwise excellent formalized model of human interaction, but limited to its information- exchange function, see Roberts (2012). 2 In the 20th century analytic philosophy defended, among others, by Davidson (1973), Dennet (1971), and Grice (1975). For a discussion of this presumption, see Lewiński (2012). Speech Act Pluralism in Argumentative Polylogues 423 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. speech acts and argumentation inescapably rely on one another’s resources. Two qualities of speech act theory and argumentation theory stand out as the key sources of their success. First, our interaction- al contributions, utterances, are treated as functional units: each utterance is aimed to perform an identifiable act that is defined by its function, that is, by the conventionally recognizable job it does in the social world. This function—called illocutionary force in Austin’s pioneering How to do things with words (1962)—is intended, communicated, taken up (or not), and acted upon (or ignored). And much of the speech acts’ powerful attraction in capturing our interactional business relies on the assumption that what is exchanged are monofunctional units. Speech acts are conceptually powerful and parsimonious as basic units of commu- nication precisely because each speech act in a given conversa- tional situation has one and only one (primary) function: when I’m promising, I’m not asking a question; when I’m voting, I’m not complaining about the weather—much in the same way as when I’m cooking a soup, I’m not baking a cake in one and the same string of physical acts (words pronounced, movements of my hands). This assumption can simply be called illocutionary mon- ism (Johnson 2019). Second, the basic model of interaction in- volves two and only two agents: the speaker and the hearer, who in an argumentative situation can instead be called the proponent and the opponent or the protagonist and the antagonist. Interaction, at its very basis, consists in an exchange of monofunctional speech acts between these two agents (Searle 1992). This latter assump- tion can be called the dyadic reduction (see Lewiński 2019). My goal in this paper is to challenge these two assumptions of speech act theory as applied to argumentation.3 While they might be the very conditions of success of the theory, allowing it to work from a conceptually parsimonious model that explains various complex forms of communication as its derivatives, they are simp- ly not correct in many ordinary situations. Instead of treating argumentative discussions in terms of monistic illocutionary acts exchanged in dyadic dialogues, argumentation theory would be 3 For a general speech act theoretic discussion, see Lewiński (2021a). 424 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. better served embracing illocutionary pluralism in argumentative polylogues. This, anyway, is the burden of proof I undertake here. And the way I aim to discharge this burden is as follows. First, in Section 2.1, I set up an incredibly easy target for my critique: models of speech act based argumentative dialogues as used by computer scientists. Of course, the theoretical underpinnings, goals, and applications of these models have little to do with speech act theory as such; yet, focussing just on the speech act aspect allows for the efficient capture of the simplifications that mar other speech act approaches to argumentation. Pragma- dialectics, discussed in Section 2.2., is one such approach that for all its sophistication cannot quite overcome the conceptual limita- tions identified in Section 2.1. In Section 3, I present Levinson’s challenge to monofunctional speech act models of dialogue, grounded in the assumption I call vertical illocutionary pluralism. Aided by this, in Section 4, I present a positive case for horizontal illocutionary pluralism in argumentative polylogues. The basic idea is very simple: when argumentatively interacting with more than one person at the same time (that’s the polylogue compo- nent), we often convey more than one argumentatively relevant illocutionary force (that’s the illocutionary pluralism component). I demonstrate the workings of such complex argumentative ex- changes on two ostensibly simple examples discussed, for some- what different purposes, in literature on conversation and speech acts. I end in Section 5 by presenting a major advantage of illocu- tionary pluralism in argumentative polylogues: it affords a better account of fallacious and otherwise manipulative discourse. 2. Simple speech act exchanges in argumentative discussions 2.1 The simplest model The simplest model combines the assumption of illocutionary monism and the dyadic reduction via theoretical and empirical simplification. Such is the case with the computational models of “multi-agent” argumentative dialogues as exemplarily developed by Prakken (2000, 2009), McBurney and Parsons (2009), and Reed and Budzyńska (2011). Arguably, these models are con- Speech Act Pluralism in Argumentative Polylogues 425 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. structed not to capture the complexity of natural argumentative interactions, but rather to provide a simple computational basis for artificial agents to engage in interactive forms of reasoning and deliberation. As such, these models successfully overcome some of the entrenched assumptions of formal reasoning—especially monologicity and monotonicity of deductive logic—and perhaps it is ingenuous to require them to take up further burden of proof that would endanger their computational feasibility. Still, recent claims regarding the capacity of such models to precisely grasp the com- plexity of natural argumentative dialogues in context (e.g., Yaskorska-Shah 2021) warrant an objection that in terms of de- scriptive adequacy they might be punching above their weight. I focus here exclusively on these computational models that take speech acts as central conceptual tools to understand argu- mentation.4 Under what to purists is confusingly called “locution rules,” these models stipulate sets of argumentatively relevant illocutions: claim (or assert), argue (or justify), question, chal- lenge, concede, or retract previous speech acts. These illocutions are unambiguously monofunctional: no issues of interpretation, 4 Nonetheless, similar objections apply to the broader computational/argument mining programme (see, however, Musi and Aakhus (2018, 2019) for computa- tional studies that challenge many of the usual limitations, notably the reliance on controlled monological or dyadic examples). Take for example the important contributions of Peldszus and Stede (2013, 2016). Empirically speaking, they use “monologue microtexts” as a basic corpus for mining argumentation struc- tures from natural discourse. These texts are produced in response to yes/no questions such as “Should shopping malls be open on Sundays?” Matching this empirical set-up is a dyadic theoretical framework, where argumentation amounts to a dovetailed, even if expectedly competitive, production of two parties: the proponent who supports her central claim with arguments (and can further defend it with counterattacks) and the challenger who tries to rebut or undercut the proponent’s argumentation. By contrast, in the work on conversa- tional organization of, for example, Roberts (2012), the standard question under discussion (QUD) is an open wh-question that does not partition the logical space into neat dichotomies as yes/no questions inevitably do, but rather into sets of plural and contrastive potential answers to the QUD. As I argued earlier (Lewiński 2017, 2019), in argumentative discourse each such answer can and indeed should be assessed via some form of pro-et-contra considerations, but this dyadic dialectic plays merely an auxiliary, instrumental role in a complex net of contrastive or comparative judgments (e.g., X is better than Y and Z with respect to c, but worse than Y and equal to Z with respect to d). 426 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. hearer-relative uptake, or indirectness are considered (let alone a full-blown illocutionary pluralism advocated below). While the sets of relevant illocutions can be extended by adding further refined categories, for example, by distinguishing between pure questioning, assertive questioning, and rhetorical questioning (Yaskorska-Shah 2021), the basic assumption that each of these speech acts does one argumentative job is left intact. Similar simplifications protect the dyadic assumption: the idea that argumentative exchanges happen between two and only two parties. Prakken does so explicitly by a theoretical fiat: apart from the proponent of a topic t and its opponent, “[t]he remaining par- ticipants, if any, are the third parties with respect to t, assumed to be neutral towards t” (2009, p. 286; see Lewiński and Aakhus 2014). Others do it by a fitting selection of the empirical material used; for example, a radio program with numerous presenters, panellists, witnesses, and audiences is analyzed in terms of a clear- ly defined pro/con debate between two camps (Yaskorska-Shah 2021, p. 5).5 Whatever the source of the reduction, the overall practice is to analyze schematic dyadic exchanges between Paul and Olga or Bob and Wilma, as in Reed and Budzyńska (2011): Bob: You know what? Harry was in Dundee. Wilma: How do you know? Bob: I saw him. Whereby Bob claims or asserts “Harry was in Dundee,” Wilma questions it, and Bob subsequently justifies or agues for it. Taken together, the assumptions of illocutionary monism and the dyadic reduction are at the very core of such “dialogue games.” And they need to be if feasible computational models are to be constructed. However, to a natural human speaker and inter- preter they will remain just that, some contrived “games.” 2.2 A complex model The examples in the previous section reflect the way in which many formal or quasi-formal models of argumentative dialogue 5 For a telling contrast to such a reductive approach, see in-depth discourse analysis of radio programs in Bruxelles and Kerbrat-Orecchioni (2004). Speech Act Pluralism in Argumentative Polylogues 427 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. resort to the concept of speech acts largely for taxonomic purposes (see also Hitchcock 2007; Walton and Krabbe 1995). Broad clas- ses of speech acts—notably Searle’s (1975a) representa- tives/assertives, directives, commissives, expressives, and declara- tions—and more specific types of speech acts provide a good heuristic to capture the set of legal moves in a dialogue game, that is, to efficiently instantiate the “locution rules” of a dialogue. Further, they are also instrumental in defining “protocol rules,” which stipulate how arguments of one party can be questioned by the other party, how critical questions should be responded to with justifications or retractions, etc. Profound reflection into the na- ture of speech acts, their functioning in communication and argu- mentation and, indeed, in the social world as such, is beyond the scope of their interest. As is well known, pragma-dialectical theory is pragmatic pre- cisely because it offers a much deeper integration of the speech act theory with a dialectical approach to argumentation (van Eemeren and Grootendorst 1984, 2004). When it was developed in the late 1970s and early 1980s, it was capable of capturing many of the newly developed insights of the era and reworking them for the purpose of studying argumentation.6 As a result, the pragma- dialectical approach to speech acts amounts to much more than a schematic application of the taxonomic resources of speech act theory (for an excellent summary, see Snoeck Henkemans 2014). One of its crucial innovations is to treat argumentation as an “illo- cutionary act complex,” which is complex in three important respects (van Eemeren and Grootendorst 1984, pp. 32-33): 1) Arguing typically consists of more than one sentence. While classic speech act theory (Austin 1962; Searle 1969) routinely assigned illocutionary forces to one-sentence utterances (“I name this ship the Queen Elizabeth,” “Shoot her!”), argu- 6 Van Eemeren and Grootendorst’s (1984) careful elaboration of the felicity conditions for the speech act of argumentation, discussion of the relations between the illocutionary and perlocutionary aspects of speech acts and between their conventional and intentional features, as well as the distinction between the “recognition” and “correctness” conditions for speech acts, dialogue very well with the contemporary work of Bach and Harnish (1979), Grice (1975), and Searle (1975a, 1975b, Searle and Vanderveken 1985). 428 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. ments characteristically require more linguistic work, some- times even an entire treatise. 2) Further, “the sentences uttered in an argumentation in fact have two illocutionary forces simultaneously” (van Eemeren and Grootendorst 1984, p. 32, italics in the original); when we ar- gue, we also perform assertions, make claims, issue state- ments, or offer hypotheses, all of which are themselves illocu- tionary acts belonging to the class of assertives. 3) Finally, for an illocution to have the force of argumentation, it has to be inferentially connected to another illocution, vari- ously called a conclusion, expressed opinion, point of view, standpoint, thesis, or claim. The very point of making an ar- gument is to provide reasoned support to (or refutation of) this other illocution. Crucial to the discussion here is point 2. Van Eemeren and Grootendorst clearly surpass the limits of simple illocutionary monism in the form of a one-sentence-to-one-illocutionary-force correlation. They do so by drawing a distinction between “elemen- tary illocutions at sentence level” and “compound illocutions [i.e., illocutionary act complexes] at a higher textual level” (1984, pp. 34-35).7 To understand the felicity conditions for the speech act of argumentation, that is, the general conditions under which argu- mentation can justify the expressed opinion to the hearer’s satis- faction, one needs to go beyond individual, sentence-based illocu- tions and consider the higher textual level. This, in turn, is best achieved when argumentation is conceptualized in the context of a complete model of argumentative discussion. Accordingly, prag- ma-dialectics offers an ideal model of a critical discussion, which theorizes all the argumentatively relevant speech acts that the speaker-qua-protagonist can trade with the hearer-qua-antagonist. Argumentation is in fact only one such speech act, central at the argumentation stage of a critical discussion. But other speech acts accompany argumentation: in the first place, the protagonist’s standpoint needs to be expressed by her and then doubted by the 7 As directly discussed by van Eemeren and Grootendorst (1984, pp. 32-33), this is different from another recognized way for speech acts to have a dual illocu- tionary force, namely, via indirect speech acts (Searle 1975b). See also Berme- jo-Luque (2011, p. 60) and Section 3 below. Speech Act Pluralism in Argumentative Polylogues 429 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. antagonist at the confrontation stage. Further, the antagonist needs to challenge the protagonist to defend the standpoint, and both need to agree on premises and discussion rules at the opening stage. At the argumentation stage, while the protagonist advances argumentation, the antagonist can ask critical questions and re- quest further arguments. Finally, at the conclusion stage, the pro- tagonist can uphold or retract her standpoint (van Eemeren and Grootendorst 1984, pp. 95-118; 2004, pp. 62-68). This distribution of speech acts across the stages of a critical discussion resembles the protocols of dialogue games mentioned above, especially when pragma-dialecticians specify them with the help of dialectical profiles (van Eemeren, Houtlosser, and Snoeck Henkemans, 2008). A similar yet original approach has been developed by Berme- jo-Luque (2011). She too considers argumentation to be a complex speech act layered over other basic illocutions. These basic illocu- tions—typically assertive speech acts, but also, possibly, promises, requests, or pieces of advice—constitute “first order speech-acts” by means of which “second order speech acts” of argumentation are performed (Bermejo-Luque 2011, p. 60). Further, Bermejo- Luque theorizes that argumentation is inescapably built out of two interrelated “second order” illocutions, namely, the speech-act of adducing reasons and the speech-act of concluding. These second order illocutionary forces are thus performed simultaneously with first order forces and are nested together with them in the complex speech act of argumentation.8 Such theories thus offer a nuanced approach to illocutions that acknowledges their complexity in the context of argumentative discussions. In particular, in their unique ways they recognize the duality of illocutionary forces performed in argument-making. This duality is, however, a feature of the analytic apparatus of the theory, a feature particularly conspicuous in pragma-dialectics. Once “normatively reconstructed” for the purposes of argument analysis and evaluation, each speech act has one and only one unique function in the broader critical discussion. As a result, all 8 Bermejo-Luque (2011) further elaborates her speech act theory of argumenta- tion by resorting to Bach and Harnish’s (1979) speech-act schema (SAS). 430 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. “dialectically-relevant moves” are, again, clearly determined mono-functional units. Moreover, the context of the discussion is exclusively dyadic: “The pragma-dialectical argumentation theory assumes that, in principle, argumentative language use is always part of an ex- change of views between two parties that do not hold the same opinion, even when the exchange of views takes place by way of a monologue” (van Eemeren and Grootendorst 2004, p. 59, italics added).9 This form of the dyadic reduction is characteristic of any dialectical approach to argumentation and is both theoretically and historically traceable to the fact that the disputing dyad (question- er-answerer, proponent-opponent) is in fact a dialectical analogue of the basic truth values: truth and falsity (Lewiński 2019; see also Dutilh Novaes 2020). While theoretically understandable for the normative purpose of dialogically verifying the validity of reason- ing departing from true premisses, this reductive dialectical ap- proach cannot be claiming in parallel that it is capturing the mech- anisms of actual argumentative discussions, even with a rather complex application of speech act theory. It is a category mistake to confuse the normative dyadic order of truth verification with the descriptive, (possibly) polyadic order of actual argumentative dynamics. In short, despite the nuances of pragma-dialectics and Bermejo- Luque’s linguistic-pragmatic approach, the assumptions of illocu- tionary monism and the dyadic reduction are left largely intact. 3. Levinson’s challenge The basic form of illocutionary pluralism recognized early in the development of speech act theory is the one involved in indirect speech acts whereby a locutionarily direct but illocutionarily secondary speech act is a means to perform a locutionarily indirect 9 For a detailed analysis of various pragma-dialectical strategies for bringing discourse that involves more than two parties into this dyadic mold, see Lew- iński and Aakhus 2014. One such strategy is distinguishing between an overtly addressed but secondary audience—a mere messenger of sorts (e.g., a journalist interviewing a politician)—and an unaddressed but still primary audience that constitutes the actual antagonist of the speaker (e.g., swing voters the inter- viewed politician aims to persuade; van Eemeren 2010, p. 109). Speech Act Pluralism in Argumentative Polylogues 431 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. but illocutionarily primary speech act; via familiar conventions, a question such as “Can you pass me the salt?” functions primarily as a request to pass the salt (Searle 1975b). Since it links illocu- tionary acts via an instrumental by-means-of relation, this form of pluralism can be called vertical. While classic examples involve only two levels, longer chains of instrumental relations among speech acts are possible too. This has been a continuous topic of interest for Levinson. Al- ready in his early empirical work on speech acts, Levinson claimed that “utterance units often seem to involve more than one speech act in a number of different ways,” with indirect speech acts providing “simple examples” of this (Levinson 1981, p. 476). But this point generalizes based on the idea that “the multiplicity of simultaneous functions is really an assignment of more than one intention to the utterance” (Levinson 1981, p. 477). Further, inten- tions are often finely layered, thus generating “a sequence predict- ed by a hierarchical structure of goals” (Levinson, 1981, p. 486). One curious fact about indirect speech acts, such as “Can you pass me the salt?,” is that they can receive both literal uptake as a ques- tion (“Actually, I can’t, someone just took it away”) and indirect uptake as a request (“Here you are!”). However, under proper circumstances, responses such as, You’re not going to drink more tequila, are you!? It’s pretty salty, actually, have you tried yet? would be perfectly in order. Hearers can thus also respond to the speakers’ perlocutionary objects of given speech acts (that is, perlocutionary effects intended by the speakers, see Austin 1962, p. 117)—or, indeed, to broader action sequences they may (rightly or wrongly) ascribe to speakers. To account for such complexities, in his recent work Levinson develops the concept of conversational projects: The notion of project we need for action ascription is not ‘themat- ic thread’ but ‘plan of action’—that is, a course of action that at least one participant is pursuing, which may at first be opaque to others then retrospectively discernible […] and then prospectively projectable. (Levinson 2013, p. 122) 432 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. The crucial point is that in conversation, speakers orient to each other’s projects: Clearly, in conversation, projects are interactionally negotiated, jointly launched, diverted or aborted. Actions then are in the ser- vice of projects, and projects are themselves actions to accom- plish. That is why there is no simple answer to what action this turn is doing: it is doing something local, which governs its re- sponse types, but also part of something more global, which, as soon as it is recognizable, also plays a role in fashioning responses (as in the ‘go ahead’ or ‘blocking’ responses to pre-s). In short, there is a hierarchy of actions within a project. (Levinson 2013, pp. 126-127) To demonstrate the working of conversational projects, Levin- son analyzes the following exchange between a teenage daughter, Virginia, and her mom: 12 Vir: But - you know, you have to have enough mo:ney?, I think 13 ten dollars ’ ud be good. 14 (0.4) 15 Mom: .hhh Ten dollahs a week? 16 Vir: Mm hm. 17 Mom: Just to throw away? 18 (0.5) 19 Vir: Not to throw away, to spe:nd. 20 (.) 21 Mom: ((shrilly)) On [WHAT? That ’ s what I been tryin ’ a fi nd = 22 Pr?: [eh hih hih 23 Mom: = out. besides McDo:nalds?, Mom’s “Ten dollahs a week?” (line 15) is ostensibly locally a clarification question (per week or per month?) to a proposal for more money; but also, querying the amount produces an oppor- tunity for daughter Virginia to justify the amount, where, if those justifications prove inadequate, grounds are thereby provided for rejecting the proposal. Virginia’s project, asking for more pocket money (earlier more clothes), is countered by Mom’s project of holding the status quo. Other-initiated repair, an information re- Speech Act Pluralism in Argumentative Polylogues 433 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. quest, a challenge to produce reasons, a pre-accusation and thus likely refusal to grant the request, are all visible in the one turn. (Levinson 2013, pp. 126-127) In discussing Levinson’s nuanced approach to action ascription in conversational activities (see also Levinson 1979, 1981), Sbisà identifies his speech act pluralism with “the plural potentialities of sequential positioning” (2013, p. 239) for speech acts in conversa- tion. We thus have a certain possible sequence of acts more or less rationally and recognizably linked in an overarching project, a sequence we can possibly project from any speech act utterance in conversation. This is a flexible and sophisticated approach to vertical plurality far exceeding the limits of Searle’s conventional approach, both in the way ascriptions are made and in their com- plexity (2 vs. n-levels). This goes quite some way to undermining the possible counterargument of classical speech act theory that illocutionary pluralism is not so important because it can be, in the end, explained away by the tools developed in the theory’s salad days: chiefly, the concept of indirect illocutionary forces. Howev- er, it is still a vertical model; it is based on a projection of various illocutionary forces (and, further, perlocutionary objects) linked in a “by way of” or “by means of” manner to the literally uttered act through some kind of sequential hierarchy. For someone like Levinson, genuinely attentive to the interac- tive dynamics of actual conversations, some form of speech act pluralism is a natural feature of our communication. Its dismissal is indeed one of the “essential inadequacies of the speech act models of dialogue” (Levinson 1981) that cannot be easily over- come if an empirically adequate description of communicative reality is one of its avowed goals. Instead of developing dialogue models in terms of fixed sets of rules applying over a fixed set of uniform speech act types, an approach sensitive to a hierarchically organized, context-dependent, and possibly strategic, goal- structure of each speaker’s conversational projects should be adopted (Levinson 1979, 1981, 2013). This has not escaped the attention of argumentation scholars. Indeed, Jackson and Jacobs’s work on conversational argument implements a very similar, flexible, and dynamic approach (Jack- 434 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. son and Jacobs, 1980; Jacobs and Jackson, 1981, 1989; for insight- ful discussion, see Jacobs 1989 and Snoeck Henkemans 2014). In their view, a simple rule-governed level of analysis that under- stands argumentative speech acts in terms of structural units of conversation, and even a more sophisticated approach of the early speech act theory that views them as conventionally recognized, functional entities, should give way to the analysis focused on what they call a rational level of discourse organization, oriented precisely to speakers’ complex plans organized via practical rea- soning in a given context of activity (Jacobs and Jackson 1989; Jacobs 1989; Jackson 1992). On such an account, the discursive manifestations and the felicity conditions of the speech act of arguing vary across activities and arguers’ objectives. Next to paradigmatic assertoric arguments advanced by the protagonist (as analyzed by van Eemeren and Grootendorst), one can distinguish hypothetical arguments, devil’s advocacy, arguments constructed collaboratively by the protagonist and the antagonist, third-party arguments, and even arguments made solely for the purpose of demonstrating a speaker’s beliefs without any genuine attempt at rationally persuading the doubters and objectors. Overall, “[i]nstead of an isolable and homogeneous speech act, one finds a family of act types that vary in function and pragmatic logic de- pending upon the context of their use and the form of their expres- sion” (Jacobs 1989, p. 350). Considered from this perspective, the simple approach of com- puter scientists based on a schematically rule-governed, taxonomic understanding of speech act theory grounded in the assumption of illocutionary monism is a far cry from the logic of actual conversa- tion. In addition, while most of Jackson and Jacobs’s examples are still prototypically dyadic—most often, they involve a couple arguing in the private comfort of their car, their kitchen, or over the phone—their speech act approach extends to genuine multi- party events, such as mediation sessions, group arguments among friends, or public speeches on campus. This means that the dyadic reduction is not explicitly imposed by theoretical fiats or oppor- tune data selection—but neither are the theoretical consequences of examining polylogical argumentation consistently drawn. Speech Act Pluralism in Argumentative Polylogues 435 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. 4. Plural speech acts in a polylogue In my constructive response to such complexities of actual argu- mentation, I start from challenging the dyadic reduction. In princi- ple, illocutionary pluralism and polylogue are independent of one another: a speaker can intentionally and conventionally perform a plural illocution even in a dyadic context. Elsewhere, I described dilemmatic deliberations and strategic (illocutionary) ambiguity as two instances of this (Lewiński 2021a). While such complex dyad- ic strategies exist, a prototypical case of illocutionary pluralism is in the context of a polylogue—and, vice versa, a typical polylogue is very likely to generate illocutionarily plural speech acts. Polylogue, in its simplest form, is a conversation that involves more than two people (Lewiński 2014). However, typical argu- mentative polylogues are complex activities in which multiple players discuss their distinct, and often incompatible, positions across a variety of places (Aakhus and Lewiński 2017). Online discussions, political debates among liberals, socialists, and con- servatives, or public controversies over climate change and energy production are all cases of polylogues. Polylogue is thus both a natural form of complex communication ordinary speakers are competent in from very young age (see Tomasello 2008) and an open framework from which various aspects of complex argumen- tation can be consistently studied. One key characteristic of polylogues from a speech act perspec- tive is that each speaker engages in a speech activity with more than one hearer (see also Clark and Carlson 1982). This seemingly innocuous fact has the potential to turn speech act theory on its head. To start with, the whole communicative transaction—the speaker’s complex intention (that involves at least the locutionary intention to mean something, the illocutionary intention to perform a socially recognizable conventional act, the perlocutionary inten- tion to influence the hearers, and the communicative intention to have these intentions recognized by the hearers) and the hearers’ uptake—cannot be described anymore as a dyadic affair. A speak- er might skillfully distribute her intentions differently to various hearers. Via one unique utterance, such as “Next time you do this, papa will start reading boring argumentation books as your bed- time stories!,” a mother might be uncompromisingly scolding her 436 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. beloved child, while inviting her partner to live up to the child- bearing consensus they reached. Further, there is not just a simple “common ground” between the dyad of the speaker and the hearer, but rather various “shared grounds” of the subsets of participants, next to the one common to all of them. A child might not, and perhaps even should not, be privy to the discussions and agree- ments her parents have. As a result, the whole scheme of mono- functional social activities understood as isolated games in which monofunctional moves directed at one unique other player are performed pretty much collapses. Take a simple case of a (pre-COVID face-to-face) academic conference: apart from, let us hope, some knowledge sharing and testing, it is a social activity rife with image management, job seeking, friends making, flirting, etc. A competent social agent knows she is playing different games at the same time. The result of it is that, as aptly put by McGowan, “a single act can also sim- ultaneously be a move in two (or even more) different norm- governed activities” (2019, p. 96). Importantly, however, for McGowan these acts would be most naturally “parallel acts,” akin to Bach and Harnish’s “collateral acts” (1979, pp. 96-107)10 but defined by some rules of the game rather than a speaker’s inten- tion. In an academic question and answer session, I can, for in- stance, advise a colleague to look into a certain body of literature (illocutionary act) and, simultaneously and often unbeknownst to me, via the same words pronounced make someone win a bet (“Didn’t I tell you he’s gonna quote Foucault again?!”), or make the chair intervene with “we agreed questions be kept short!” (parallel acts). The picture we have here is one of a plurality of acts realized by speech and composed of a single illocutionary act accompanied by multiple (possible) parallel acts. 10 Bach and Harnish (1979) distinguish between proper speech acts, character- ized by the communicative presumption, namely, the presumption that when a speaker says something, she is saying it with “some recognizable illocutionary intent” (1979, p. 7) and collateral acts, “conversational acts performed in con- junction with or in lieu of illocutionary acts” (1979, p. 97). Collateral acts can be overt (e.g., joking or storytelling) or covert (e.g., innuendos or deliberate ambiguities); the latter “succeed (the intention with which they are performed is fulfilled) only if their intent is not recognized, or at least not recognized as intended to be recognized” (Bach and Harnish 1979, p. 101). Speech Act Pluralism in Argumentative Polylogues 437 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. But the argument I want to make here is precisely that the plu- rality can consist in a plurality of fully-fledged speech acts with overt illocutionary forces. To illustrate this, let me move directly to the example. Example 1 Imagine good old pre-COVID-19 days where a philosopher, Mr. X, comments in person on a conference paper of another philosopher, Mr. Y. At a conference dinner, having a one-on-one cigarette with the author, Mr. Y, X says: 1.1 This was the best paper of the Argumentation and Speech Acts 2021 Conference I’ve seen. Assuming that “this” refers to Y’s paper, of all things imagina- ble, this is a compliment, and a felicitous one: it expresses praise over the hearer’s characteristics, actions, products, or possessions that are considered undeniably praiseworthy, especially if these are rare or unique (see Aakhus and Aldrich 2002). Sacks (1992) calls such compliments “safe” and “strong”; they make the only ad- dressee, Y, feel good and risk hardly anything for X. We do not feel anything much about them, this is just a part of the social academic business, full of strategic niceties, etc. Now, for contrast, imagine a philosopher, Mr. X, commenting on a conference paper of Mr. Y. During his commentary in a conference room with quite a few other scholars present, X says: 1.2 This was the best paper of the Argumentation and Speech Acts 2021 Conference I’ve seen. So it happens that X is a member of a panel evaluating papers for a prize. He’s had quite a few discussions about it with two other panel members present in the room. One of them, A, had all along claimed Y’s paper is weak beyond discussion. The other, B, staunchly defended it. X was hesitant (“Let’s wait and see the last version…”). Beyond this complication, X’s colleague C who yesterday was a little bit too smart—arrogant, even—in discussing 438 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. his paper is also in the room. X knows the trade of academic di- plomacy and has the following illocutionary intentions issued with an overt communicative intention, namely, an intention for these illocutionary intentions to be recognized as intended11: a) Compliment Y. b) Externalize a difference of opinion with A. c) Concede an argument (and perhaps also agree on the con- clusion) of B. d) Rank C’s work as inferior. Finally, he might also, whether deliberately or inadvertently: e) Denigrate, offend, or insult C, as well as other audience members whose natural reaction could well be: “How about my paper you attended earlier today? You think it was dumb, huh?” Because of possibilities d) and e), Sacks (1992) called such compliments in a multi-party conversation “unsafe” given that being comparative ascriptions of personal qualities or achieve- ments, while complimenting one person, they might also insult or offend another. The question is: Which illocutionary act has Mr. X performed? To start with, option e) would, on the most standard analysis, be considered a perlocutionary effect rather than an illocutionary 11 See Grice (1957, 1975), Strawson (1964), Bach and Harnish (1979). Else- where (Lewiński 2021a) I called such intentional performance of plural illocu- tionary forces “illocutionary pluralism proper” in contradistinction to “illocu- tionary relativism,” which instead relies on a plural ascription of various illocu- tionary forces by different hearers, possibly in a way entirely unintended by the speaker (cf. Sbisà 2013; Johnson 2019). This is an important qualification. Especially in polylogical argumentation over social media, each claim and argument can be reported, reused, and relayed beyond a speaker’s control and even knowledge. In this process, quite expectedly, a plurality of different illocutionary forces can be attributed to the speaker (e.g., Donald Trump: “recommended that bleach be used to treat COVID-19” vs. “inquired whether bleach could be used to treat COVID-19” vs. “joked that bleach be used to treat COVID-19”). While fascinating, these phenomena are not directly addressed here. Speech Act Pluralism in Argumentative Polylogues 439 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. act—either a perlocutionary object intended by X, or a perlocu- tionary sequel, unintended but nevertheless caused by X’s illocu- tion (see Austin 1962, pp. 117-118). Let us then focus on a-d here. As a first approximation of an answer, let me now organize this via a somewhat pedantic speech act analysis: Level 1: By way of an assertion (assertive): “This was the best pa- per of the Argumentation and Speech Acts 2021 Conference I’ve seen.” Level 2: X is complimenting Y (another assertive),12 by way of which he is: Level 3: 1) Externalizing a difference of opinion with A. (On the pragma-dialectical reading of “mixed” argumentative disputes by van Eemeren and Grootendorst (1984, 2004), this would be both an assertive of the confrontation stage, that of expressing X’s standpoint, and a commissive of the same stage, namely, non-acceptance of A’s standpoint.) 2) Conceding an argument (and perhaps also agreeing on the conclusion) of B. (Again, for pragma-dialecticians, this would be a commissive of the argumentation or the concluding stage: acceptance of ar- gumentation or of a standpoint, respectively.) 3) Ranking C’s work as inferior. (A curious speech act; in Searle’s taxonomy, likely a mix of assertive/representative and declarative speech acts, as in the subclass of “representative declarations” [1975a, pp. 360-361]. For Austin, this would be a clear case of a “verdictive” speech act [1962, pp. 152-154] of a rather informal kind. In argumen- tation, given the comparative “better than” relation inherent in ranking, it can be an act of stating a key premise of a complex 12 According to Searle and Vanderveken, similarly to complaining and boasting, complimenting can be “either an assertive or an expressive” illocutionary act (1985, pp. 213-215). The exclusive disjunction here (“either…or…”) indicates that there are two mutually exclusive versions of these acts, depending largely on the propositional content: one assertive (as in example 1.2), another expres- sive (“I was totally amazed by your talk!”). Thus, Searle and Vanderveken’s account preserves the assumption of illocutionary monism. 440 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. practical argument [Lewiński 2017] supporting a conclusion: “Y [and not C, etc.] should win the best essay prize”). Now, the key point here is the following: Mr. X performs a simple one-sentence locution in a well-defined social activity of a confer- ence commentary. But this unique locution is a polylogical contri- bution: it intentionally acts on interactions with a number of other participants in the event and it does so in a way that is convention- ally recognized. As such, it performs a number of parallel illocu- tionary acts. Since this includes multiple argumentative illocution- ary acts at the same level (level 3 as described above), I call this form of pluralism horizontal illocutionary pluralism in contrast to vertical pluralism constituted by hierarchical relations among illocutions (level 1-3 here; see Section 3). For pragma- dialecticians, that would likely mean Mr. X is having a number of dyadic critical discussions with different antagonists at the same time. Think of a chess master playing simultaneously against multiple opponents on different boards: what happens on one board or in one discussion stays in that discussion. But this would precisely be the form of dyadic reduction that renders a description (and evaluation!) of actual argumentative activities utterly inade- quate. We have one board here—one communicative event—and so what happens in one discussion obviously “leaks” to another; for example, X is now in a discursive coalition with B, thus consti- tuting a collective antagonist against A’s standpoint. They now have a majority and thus an upper hand in their collective delibera- tion with A. Socially speaking, X might be encouraging the hum- ble Y and publicly defending her praiseworthy work against con- stant denigrations of boastful C, etc., etc. A lot of illocutionary weight can be carried by a single utter- ance. Example 2 This example has been discussed in McGowan’s (2018, 2019) recent work on exercitive speech acts. Her original analysis had a different purpose, namely, that of demonstrating how conversa- tional contributions can enact new, activity-specific norms, thus Speech Act Pluralism in Argumentative Polylogues 441 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. altering the activity’s normative status. For example, after Jane’s question, “Is Ralph eligible?,” the department members cannot, among other things, completely ignore Ralph as a potential de- partment chair (at the very least, they are expected to justify why he is not eligible, should this be the case). On McGowan’s well- known analysis, such acts thus constitute conversational exerci- tives: speech acts that, following Austin (1962), are exercitives “since they enact facts about what is permissible in some realm,” and, moreover, “they are conversational exercitives since the realm in question is a particular conversation” (McGowan 2019, p. 34). Imagine a department meeting of the type described by Lew- iński and Aakhus (2014). Peter is a white male full professor and chair of the philosophy department. Jane is a recently hired young assistant professor, and while she is white, she is the only woman in the department. Ralph is a long-time permanent instructor and the only person of colour in the department. PETER: We really need to figure this out. I am stepping down as chair at the end of this year and someone really needs to step up. Nobody wants to do it but someone has to. [Awkward silence.] JANE: Well, I know that I am not eligible to be chair. I just got here and I don’t have tenure but is Ralph eligible? [Long awkward silence.] JANE: [with her hands in the air] I am just trying to be clear on the eligibility conditions so we can be sure that we are consid- ering all of our options. I am not sure that this would be of in- terest [turns to Ralph sitting alongside wall] to you, Ralph, but if it does and we could make it happen, it would be good for everyone—especially given that no one else seems to want to do it! [Many slightly reluctant nods in the room] PETER: Ahh, well we’ll come back to this. (McGowan 2018, p. 192; 2019, pp. 186-187) Along with McGowan, let us focus here on Jane’s question, “Is Ralph eligible?” On the most straightforward speech act interpre- tation, this is a real, informative question, as is also ostensibly confirmed in Jane’s follow-up (“I am just trying to be clear on the 442 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. eligibility conditions”). But here it comes as a response to Peter’s act of soliciting proposals, a speech act typical of deliberative argumentation (Corredor 2020; Lewiński 2021b). This solicitation of proposals has also a very specific nature here: Peter’s opening can be glossed as “any volunteers?” because the person to be proposed needs to be a current department member, that is, belong to the group present in that very meeting. At this juncture, Jane seems to be reasoning along the lines of a disjunctive syllogism: the next chair “has to” be chosen from among us, Peter will not continue, “nobody [else] wants to do it,”13 and “I [Jane] am not eligible to be chair.” So, we might just get down to Ralph. This interpretation makes it seem as though Jane’s reasoning is tinged with identifying an eliminative, last-resort option, some- thing McGowan is actually out to challenge with this example. However, given the contextual background, Jane at least presents Ralph as an option to consider in the first place. So, over and above the innocent question (and in spite of her disingenuous disclaimer, “I am just trying to be clear on the eligibility condi- tions”), Jane is performing a speech act that elsewhere I called a conversational nomination: a speaker picks out one of the mem- bers of the interacting group as having certain attributes relevant to the contextual task at hand (Lewiński 2021a). Admittedly, in this context, Jane is doing it via what conversational analysts call a pre-sequence or what speech act theorists can simply call an indi- rect speech act—she is asking about one of the basic felicity con- ditions (namely, a preparatory condition) for a successful proposal; in this institutional setting, one cannot propose an ineligible candi- date. Still, similarly to many other recognizable indirect speech acts, doing so often just counts as performing the primary speech act, which is, here, that of issuing a proposal.14 Further in the 13 Notice a quantifier domain restriction (Stanley and Szabó 2000): in Peter’s opening remark, “nobody” is likely restricted to “nobody you would routinely consider to be our department chair; likely, a white, male, full professor.” 14 Compare: “Oh, I hurt my arm, someone else has to drive.” Asking in this context, “Does Ralph have a driving license?” just counts as a request for Ralph to drive (if indeed he does). Admittedly, it would be even stronger if the nega- tive variants of these questions were asked: “Isn’t Ralph eligible?,” “Doesn’t Ralph have a driving license?” as these are moving more in the direction of rhetorical questions with an obvious answer. Speech Act Pluralism in Argumentative Polylogues 443 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. exchange, she discharges the initial burden of proof for her pro- posal (see Kauffeld 1998) by judiciously counterbalancing the negative, eliminative reason (“no one else seems to want to do it”) with a positive reason (“it would be good for everyone”). Given the context of practical argumentation within a delibera- tive activity (Corredor 2020; Lewiński 2021b), Jane’s indirect proposal is thus the main speech act performed by way of her asking a question that constitutes a conversational nomination. But, also, given that Ralph is present in that meeting, and given Jane’s follow up where she directly addresses him as “you, Ralph,” “Is Ralph eligible?” also functions as an act of daring, encouraging, or challenging Ralph to run for the office, all of which seem to be straightforward illocutionary acts. In terms of pragma-dialectical argumentation theory, Jane’s question can also be attributed a number of simultaneously valid illocutionary forces. It can be seen as expressing a standpoint, and, as McGowan recognizes, a rather surprising or even provocative one (confrontation stage). It can also be seen as a challenge to the implicitly standing, conventionally expected standpoint that it should be a white, male, full professor. What Jane is conveying can be read as: “Why don’t you instead consider someone out of the box for once.” She might also be seen as adding a common starting point to their deliberative discussion (opening stage): “long-time, Afro-American instructors are eligible to be depart- ment chairs too.” Yet another option is counter-argumentation at the argumentation stage: “Contrary to what you might think, Ralph is formally eligible too (so why not consider him).” As in previous examples, these different illocutionary forces are not merely dif- ferent plausible interpretations of Jane’s speech act; they might all be intended by her as conventionally recognizable acts performed to different department members present in the room. Or she might be engaging in strategic illocutionary ambiguity (see Lewiński 2021a) that lets her engage in plausible deniability if pressed, which is something she already pre-empts by saying, “I am just trying to be clear.” All of these are cases of illocutionary pluralism in an argumen- tative polylogue. 444 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. 5. Conclusion: Speech act pluralism and fallacies The analytic and evaluative advantages of speech act pluralism for argumentation theory are yet to be fully understood. It is a com- plex issue in itself since, in a proper sense, speech act pluralism extends across three axes, namely, locutionary pluralism, dealing mostly with classic issues of semantic underdetermination and contextuality of meaning (Cappelen and Lepore 2005); illocution- ary pluralism discussed here (see also Sbisà 2013; Johnson 2019; Lewiński 2021a); and perlocutionary pluralism investigating cognitive and behavioural effects of language use on hearers and especially manipulations of language generating underhand perlo- cutionary effects. This last issue would properly belong to an area of cognitive linguistics, and cognitive pragmatics in particular (see de Saussure 2018; Schumann, Zufferey, and Oswald 2019, 2021; Lombardi Valauri et al. 2020; Müller 2020 for good recent exam- ples). Only having fully accounted for these three forms of plural- ism would we be able to satisfactorily realize Austin’s original postulate to fully elucidate “[t]he total speech act in the total speech situation” (Austin 1962, p. 147). Here, I merely sketched an account of illocutionary pluralism in argumentative discussions. But even with this limited account, some advances in argumentation theory can be made. One peren- nial paradox in the theory is how to account for the fact that falla- cies are argumentative moves that are, at the same time, irrational and convincing (Hamblin 1970). This paradox is evident in the context of two important considerations. First, as empirical evi- dence consistently indicates, ordinary arguers are quite skilful at detecting and rejecting fallacious argumentation (van Eemeren, Garssen, and Meuffels 2009; Schumann, Zufferey, and Oswald 2019, 2021). Second, should we think of fallacies exclusively in the context of dyadic discussions, would not the victim of a fallacy be the first to point out the abuse and to insist on its correction (Casey 2020)? A convincing fallacy sounds like an oxymoron. One possible interpretation of this is that fallacies, and especial- ly fallacies such as the straw man fallacy, resort to fine, manipula- tive, linguistic devices in order to appear reasonable, while they are not (Lewiński and Oswald 2013; Müller 2020; Schumann, Zufferey, and Oswald 2019, 2021). There exists, in speech act Speech Act Pluralism in Argumentative Polylogues 445 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. theory terms, some deceitful relation between the locutionary form and the perlocutionary effects of fallacious speech acts. But anoth- er interpretation, recently entertained in the field (Aikin and Talis- se 2019; Casey 2020; de Saussure 2018), is that fallacies succeed in being treacherously convincing while being irrational because the context of argumentative discussion is not dyadic but, as I would call it, polyadic. Ad hominem, ad baculum, and straw man fallacies succeed not because they aim to gain recognition from the attacked opponent—the first to detect the abuse and express indignation—but rather from the larger ring of other participants and onlookers. They let the abuser present himself as a smarter, wittier, and altogether more potent reasoner than his victim who, in turn, is either silenced, thus letting the abuser’s argument pre- vail before others, or burdened with the tedious task of rectifying the abuse. The challenge to the dyadic reduction, and hence the embracing of a polylogue, is thus the condition of possibility for this promising account. But how about illocutionary pluralism? It is very likely, but not strictly necessary. As just mentioned, some perfidious effects might trade on the complex relations between the locutionary and perlocutionary aspects of fallacious speech acts, this including purely perlocutionary effects on the audience. Yet, an illocutionary variant is certainly plausible. For instance, when in a public context, X is responding to Y’s argumentation with an ad hominem argument; she is performing a (fallacious) speech act of counter-argumentation vis-à-vis Y: “How can you be that naïve to even suspect Biden might have won?! Don’t you have eyes to see all the vote-rigging?” Vis-à-vis the larger audience, however, this illocutionary act simultaneously constitutes an act of ranking Y as not being cognitively competent enough to engage in critical argumentative exchanges. This interpretation, while sur- prisingly in line with the pragma-dialectical take on ad hominem (it violates the attacked arguer’s freedom to engage in a critical discussion; see van Eemeren and Grootendorst 1984, 2004), does require illocutionary pluralism. In the present scenario, the abuser X does not aim to rationally convince the victim Y of Y’s own cognitive deficiencies, nor even to do it in some covert, under- hand, perlocutionary way. Instead, the abuser tries to convince others of it and does so in an overt, intentionally and conventional- 446 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. ly recognized, and thus illocutionary fashion: “I hereby want you to recognize she’s really not up to it.” All this requires further careful investigation.15 One thing, however, is undeniable: the long-time relation between argumenta- tion theory and speech act theory is a treasure trove that continues to deliver. Acknowledgements This work has been supported by the Portuguese Foundation for Science and Technology (FCT) and CHIST-ERA Programme via the project CHIST-ERA/0002/2019. References Aakhus, M. and A. Aldrich. 2002. Crafting communication activity: Understanding felicity in “I wish I…” compliments. Research on Language and Social Interaction 35(4): 395-425. Aakhus, M. and M. Lewiński. 2017. Advancing polylogical analysis of large-scale argumentation: Disagreement management in the fracking controversy. Argumentation 31(1): 179–207. Aikin, S. and R. Talisse. 2019. Why we argue: A guide to disagreement in an age of unreason. 2nd ed. New York: Routledge. Austin, J. L. 1962. How to do things with words. Oxford: Clarendon Press. Bach, K. and R. Harnish. 1979. Linguistic communication and speech acts. Cambridge, MA: MIT Press. Bermejo-Luque, L. 2011. Giving reasons: A linguistic-pragmatic ap- proach to argumentation theory. Dordrecht: Springer. Bruxelles, S. and C. Kerbrat-Orecchioni. 2004. Coalitions in polylogues. Journal of Pragmatics 36(1): 75-113. Cappelen, H. and E. Lepore. 2005. Insensitive semantics. A defense of semantic minimalism and speech act pluralism. Oxford: Blackwell. Casey, J. P. 2020. The persuasive force of the ad baculum. In Proceed- ings OSSA 12: Evidence, Persuasion & Diversity, Windsor, ON: OS- SA Conference Archive. 15 Elsewhere (Lewiński 2021a), I show how illocutionary pluralism in the context of polylogue works behind a number of other forms of manipulative discourse discussed in the recent speech act literature, for example, back-door speech acts (Langton 2018) and dog-whistles (Saul 2018). Speech Act Pluralism in Argumentative Polylogues 447 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. Clark, H. H. and T. B. Carlson. 1982. Hearers and speech acts. Lan- guage 58(2): 332-373. Corredor, C. 2020. Deliberative speech acts: An interactional approach. Language & Communication 71: 136-148. Davidson, D. 1973. Radical interpretation. Dialectica 27(3-4): 313-328. Dennett, D. C. 1971. Intentional systems. The Journal of Philosophy 68(4): 87-106. Dutilh Novaes, C. 2020. The dialogical roots of deduction. Historical, cognitive, and philosophical perspectives on reasoning. Oxford: Ox- ford University Press. Eemeren, F. H. van. 2010. Strategic maneuvering in argumentative discourse: Extending the pragma-dialectical theory of argumenta- tion. Amsterdam: John Benjamins. Eemeren, F. H. van, B. Garssen and H. L. M. Meuffels. 2009. Fallacies and judgments of reasonableness: Empirical research concerning the pragma-dialectical discussion rules. Dordrecht: Springer. Eemeren, F. H. van and R. Grootendorst. 1984. Speech acts in argumen- tative discussions. Dordrecht: Floris. Eemeren, F. H. van and R. Grootendorst. 2004. A systematic theory of argumentation: The pragma-dialectical approach. Cambridge: Cam- bridge University Press. Eemeren, F. H. van, P. Houtlosser and A. F. Snoeck Henkemans. 2008. Dialectical profiles and indicators of argumentative moves. Journal of Pragmatics 40(3): 475-493. Fogal, D., D. W. Harris and M. Moss (eds.). 2018. New work on speech acts. Oxford: Oxford University Press. Grice, H. P. 1957. Meaning. The Philosophical Review 66(3): 377-388. Grice, H. P. 1975. Logic and conversation. In Syntax and semantics, Vol. 3: Speech acts, eds. P. Cole and J. L. Morgan, 41-58. New York: Ac- ademic Press. Hamblin, C. L. 1970. Fallacies. London: Methuen. Hitchcock, D. 2007. Informal logic and the concept of argument. In Philosophy of logic, ed. D. Jacquette, Volume 5 of Handbook of the philosophy of science, eds. D. M. Gabbay, P. Thaggard and J. Woods, 101–129. Amsterdam: Elsevier. Jackson, S. 1992. “Virtual standpoints” and the pragmatics of conversa- tional argument. In Argumentation illuminated, eds. F. H. van Eeme- ren, R. Grootendorst, J. A. Blair and C. A. Willard, 260-269. Amster- dam: SicSat. Jackson, S. and S. Jacobs. 1980. Structure of conversational argument: Pragmatic bases for the enthymeme. Quarterly Journal of Speech 66(3): 251-265. 448 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. Jacobs, S. 1989. Speech acts and arguments. Argumentation 3(4): 345- 365. Jacobs, S. and S. Jackson. 1981. Argument as a natural category: The routine grounds for arguing in conversation. Western Journal of Speech Communication 45(2): 118-132. Jacobs, S. and S. Jackson. 1989. Building a model of conversational argument. In Rethinking communication. Vol 2: Paradigm exemplars, eds. B. Dervin, L. Grossberg, B. J. O’Keefe and E. Wartella, 153- 171. Newbury Park: Sage. Johnson, C. R. 2019. Investigating illocutionary monism. Synthese 196(3): 1151-1165. Kauffeld, F. J. 1998. Presumptions and the distribution of argumentative burdens in acts of proposing and accusing. Argumentation 12(2): 245-266. Langton, R. 2018. Blocking as counter-speech. In New work on speech acts, eds. D. Fogal, D.W. Harris and M. Moss, 144-164. Oxford: Ox- ford University Press. Levinson, S. C. 1979. Activity types and language. Linguistics 17(5-6): 365-400. Levinson, S. C. 1981. The essential inadequacies of speech act models of dialogue. In Possibilities and limitations of pragmatics, eds. H. Par- ret, M. Sbisà, J. Verschueren, 473-492. Amsterdam: John Benjamins. Levinson, S. C. 2013. Action formation and ascription. In Handbook of conversation analysis, eds. T. Stivers and J. Sidnell, 103-130. Malden, MA: Wiley-Blackwell. Lewiński, M. 2012. The paradox of charity. Informal Logic 32(4): 403- 439. Lewiński, M. 2014. Argumentative polylogues: Beyond dialectical understanding of fallacies. Studies in Logic, Grammar and Rhetoric 36(1): 193-218. Lewiński, M. 2017. Practical argumentation as reasoned advocacy. Informal Logic 37(2): 85-113. Lewiński, M. 2019. Argumentative discussion: The rationality of what? TOPOI: An International Review of Philosophy 38(4): 645-658. Lewiński, M. 2021a. Illocutionary pluralism. Synthese. Lewiński, M. 2021b. Conclusions of practical argument: A speech act analysis. Organon F 28(2): 420-457. Lewiński, M. and M. Aakhus. 2014. Argumentative polylogues in a dialectical framework: A methodological inquiry. Argumentation 28(2): 161-185. https://doi.org/10.1007/s11229-021-03087-7 Speech Act Pluralism in Argumentative Polylogues 449 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. Lewiński, M. and S. Oswald. 2013. When and how do we deal with straw men? A normative and cognitive pragmatic account. Journal of Pragmatics 59: 164-177. Lombardi Valauri, E., L. Baranzini, D. Cimmino, F. Cominetti, C. Coppola and G. Mannaioli. 2020. Implicit argumentation and persua- sion: A measuring model. Journal of Argumentation in Context 9(1): 95-123. McBurney, P. and S. Parsons. 2009. Dialogue games for agent argumen- tation. In Argumentation in artificial intelligence, eds. I. Rahwan, G. R. Simari, 261–280. Dordrecht: Springer. McGowan, M. K. 2018. Responding to harmful speech: More speech, counter speech, and the complexity of language use. In Voicing dis- sent: The ethics and epistemology of making disagreement public, ed. C. R. Johnson, 182-199. New York: Routledge. McGowan, M. K. 2019. Just words: On speech and hidden harm. Ox- ford: Oxford University Press. Müller, M-L. 2020. Non-propositional meanings and commitment at- tribution: More arguments in favor of a cognitive approach. Journal of Argumentation in Context 9(1): 148-166. Musi, E. and M. Aakhus. 2018. Discovering argumentative patterns in energy polylogues: A macroscope for argument mining. Argumenta- tion 32(3): 397-430. Musi, E. and M. Aakhus. 2019. Framing fracking: Semantic frames as meta-argumentative indicators for knowledge-driven argument min- ing of controversies. Journal of Argumentation in Context 8(1): 112- 135. Peldszus, A. and M. Stede. 2013. From argument diagrams to argumen- tation mining in texts: A survey. International Journal of Cognitive Informatics and Natural Intelligence 7(1): 1-31. Peldszus, A. and M. Stede. 2016. Rhetorical structure and argumentation structure in monologue text. In Proceedings of the 3rd workshop on argument mining, ACL 2016, 103–112. Association for Computation- al Linguistics. Prakken, H. 2000. On dialogue systems with speech acts, arguments, and counterarguments. In Logics in artificial intelligence, eds. M. Ojeda- Aciego, I. P. de Guzmán, G. Brewka, L. Moniz Pereira, 224-238. Berlin: Springer. Prakken, H. 2009. Models of persuasion dialogue. In Argumentation in artificial intelligence, eds. I. Rahwan, G. R. Simari, 281–300). Dor- drecht: Springer. Reed, C. and K. Budzyńska. 2011. How dialogues create arguments. In Proceedings of the Seventh Conference of the International Society 450 Lewiński © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. for the Study of Argumentation (ISSA), eds. F. H. van Eemeren, B. Garssen, D. Godden and G. Mitchell, 1633-1645. Amsterdam: SicSat. Roberts, C. 2012. Information structure in discourse: Towards an inte- grated formal theory of pragmatics. Semantics & Pragmatics 5, Arti- cle 6: 1-69. Sacks, H. 1992. Lectures on conversation. Vol. I & II. Oxford: Black- well. Saul, J. 2018. Dogwhistles, political manipulation, and philosophy of language. In New work on speech acts, eds. D. Fogal, D. W. Harris, M. Moss, 360-383). Oxford: Oxford University Press. Saussure, L. de. 2018. The straw man fallacy as a prestige-gaining device. In Argumentation and language: Linguistic, cognitive and discursive explorations, eds. S. Oswald, T. Herman and J. Jacquin, 171-190. Cham: Springer. Sbisà, M. 2013. Some remarks about speech act pluralism. In Perspec- tives on pragmatics and philosophy, eds. A. Capone, F. Lo Piparo and M. Carapezza, 227-244. Cham: Springer. Schumann, J., S. Zufferey and S. Oswald. 2019. What makes a straw man acceptable? Three experiments assessing linguistic factors. Journal of Pragmatics 141: 1-15. Schumann, J., S. Zufferey, and S. Oswald. 2021. The linguistic formula- tion of fallacies matters: The case of causal connectives. Argumenta- tion 35(3): 361-388. Searle, J. R. 1969. Speech acts: An essay in the philosophy of language. Cambridge: Cambridge University Press. Searle, J. R. 1975a. A taxonomy of illocutionary acts. In Language, mind, and knowledge, vol. 7, ed. K. Günderson, 344-369. Minneap- olis: University of Minnesota Press. Searle, J. R. 1975b. Indirect speech acts. In Syntax and semantics, 3: Speech acts, ed. P. C. J. L. Morgan, 59-82. New York: Academic Press. Searle, J. R. 1992. Conversation. In (On) Searle on conversation, eds. J. R. Searle et al., 7-29). Amsterdam: John Benjamins. Searle, J. R. and D. Vanderveken. 1985. Foundations of illocutionary logic. Cambridge: Cambridge University Press. Snoeck Henkemans, A. F. 2014. Speech act theory and the study of argumentation. Studies in Logic, Grammar and Rhetoric 36(1): 41- 58. Stanley, J. and Z. G. Szabó. 2000. On quantifier domain restriction. Mind & Language 15(2-3): 219-261. Strawson, P. F. 1964. Intention and convention in speech acts. The Philosophical Review 73(4) 439-460. Speech Act Pluralism in Argumentative Polylogues 451 © Marcin Lewiński. Informal Logic, Vol. 41, No. 3 (2021), pp. 421–451. Tomasello, M. 2008. Origins of human communication. Cambridge, MA: MIT Press. Walton, D. N. and E. C. W. Krabbe. 1995. Commitment in dialogue: Basic concepts of interpersonal reasoning. Albany, NY: State Uni- versity of New York Press. Yaskorska-Shah, O. 2021. Managing the complexity of dialogues in context: A data-driven discovery method for dialectical reply struc- tures. Argumentation. https://doi.org/10.1007/s10503-020-09543-x