INFORMAL LOGIC XIII.2, Spring 1991 Argument. Appreciation! Argument-Criticism: The "Aesthetics" of Informal Logic JOEL RUDINOW Sonoma State University Abstract: What rational foundation underlies argument-critical judgements? What are the can- ons of argument criticism and how are they to be "justified"? This paper explores an analogy between art- and argument-criticism and argues that the analogy promises not only to illuminate the nature of argument criticism and capture the cen- tral goals of instruction in informal logic, but also to resolve fundamental problems at the founda- tions of normative theory of argument concerning the "justification" of standards of reasoning. "In dealing with arguments we are critics in much the same way that a film critic is a critic." Ralph Johnson The Problem Some arguments are better than others. Only relativists dispute this. (And the minute they begin to argue for their posi- tion, they've lost the dispute.) But how do we tell that one argument is better than another? And if someone should disagree, how do we establish that one argument is better than another? For this we need a "the- ory of argument eval uation", what some call a "normative theory of argument". Current interest and impetus toward such a normative theory of argument is to be found largely within the recent movement in informal logic. Insofar as argument eval- uation is central to informal logic theory and to the goals of instruction in informal logic, as is widely agreed, a normative the- ory of argument becomes essential. It is worth noting at the beginning that the informal logic movement is motivated in large part by dissatisfaction with formal logic as an approach to understanding and evaluating "real world" arguments, the sorts of argument people actually use and encounter in everyday discourse, Although formal logic can be absolutely decisive with regard to the question of deductive validity, it is by and large reduced to irrelevant impo- tence in the hurly-burly give and take of everyday discourse. Nevertheless, in its search for a new nor- mative theory of argument, informal logic has been unable to fully escape the influ- ence of fonnallogic. The infl uence of fonnal logic on nonnative theory of argument can be seen manifesting itself in at least two ways: 1. The general umbrella concept of argument evaluation current within infor- mal logic, namely "cogency", continues to be articulated almost without exception in the literature of informal logic in terms of deductive validity. Blair and Johnson put it this way: "For too many, the ideal of 'sound- ness' remains the paradigm of cogency, and informal logic is simply applied formal logic without symbols."1 2. The form in which the prevailing accounts of cogency are presented in the lit- erature of informal logic is again almost without exception that of general defini- tions designed to serve as premises in deductive inferences to the effect that a given argument either is or is not cogent. Thus, for example, cogency may be defined as follows: A cogent argument is one whose conclusion is validly drawn from true or justified premises. The student encountering such a definition in an informal logic text will predictably 90 Joel Rudinow understand its intended application to a given argument as follows: This argument is (is not) one whose conclu- sion is validly drawn from true or justified premises. Therefore This argument is (is not) cogent. The influence of formal logic on norma- tive theory of argument is entirely under- standable. Deductive validity is as powerful, as broadly applicable, and as the- oretically satisfying as any normative concept of reasoning ever discovered or devised. Its algorithmic precision and decisiveness make it quite appropriate as at least one par- adigm of argumentative virtue. If deductive validity is not a plus in an argument, it is hard to imagine what would count as a plus. But, as Toulmin, Perelman, Scriven, and generations of instructors of formal logic have been saying, instruction in formal logic is at least inefficient and much of it is irrelevant and even misleading as a prepara- tion for real-world argument analysis and evaluation, all for the simple reason that for- mal logic is not the logic of real-world argu- mentation. Not all arguments are deductive; most real-world arguments are not (purely) deductive; and some arguments which are not deductively valid are better than some arguments which are. Normative theory of argument must take account of this. It must account for inductive arguments and con- ductive arguments and for extended argu- ments which are partly deductive and partly non-deductive and for arguments whose excellence (or lack thereof) hinges upon their humor, irony, timing or rhythm, or upon their narrative realism, or upon the aptness, creativity, and suggestiveness of their central metaphors, or what have you. The "Boot-Strap Problem II Some have worried that normative the- ory of argument is baffled by a "boot-strap problem". The argument goes as follows: An adequate normative theory of argument is presupposed in any attempt to argue for or against any such theory. More deeply and specifically, if normative theory of argu- ment must account for the assessment of such a rich variety of argumentative dimen- sions as was just indicated above, it is likely to depend heavily upon arguments as exam- ples. Examples in tum are likely at times to be more or less controversial. Arguments that some normative theorists regard as superior are likely to be regarded by other theorists as inferior. In that event, argu- ments in support of controversial assess- ments of such examples would be called for. But unless these arguments are capable of being assessed without generating similar controversy, normative theory of argument is a non-starter. 2 This is an interesting worry, to which I'll return. However, it needn't baffle the search for an adequate normative theory of argument. I am prepared to assume for the time being that there is an adequate norma- tive theory of argument. We may as well assume this as assume otherwise, particu- larly since an adequate normative theory of argument is what we're searching for. I would go further and assume not only that there is an adequate normative theory of argument, but that we "know", in some intu- itive and pre-theoretical sense of "know", what an adequate normative theory of argument has to say. This assumption is reasonable because we can recognize good arguments and distinguish them from bad ones, and because we can do so with a degree of consensus which can't otherwise be accounted for (even if occasionally our assessments are controversial), and because we can teach people how to do so, and so on. 3 The Proposal: Analogy #1 In the discussion of a paper delivered by Ralph Johnson at the 1989 Sonoma State University Conference on Critical Thinking and Educational Reform, he and I had occa- sion to briefly explore an analogy between art criticism and informal logic.4 In the Socratic spirit of following arguments wherever they may lead, I propose to explore the analogy somewhat further in what follows. Johnson's suggestive employment of the analogy was this: Given the wide range of considerations that are appropriate in real-world argument evaluation, a very good basis for characterizing what we do when we evaluate arguments, and what we want to encourage in students of argument evaluation, is by analogy to what a film critic does in evaluating a film. In evaluat- ing a film a competent critic attends to the multiplicity of the film's salient dimensions (plot, pacing, performance, character devel- opment, dialogue, cinematography, special effects. soundtrack, the place of the film in the director's body of work, and so on) and finally relates these considerations to each other in an overall comprehensive assess- ment of the film. Simllarly with arguments, a comprehensive assessment will ulti- mately depend, not only upon the strength of the connection between its premises and conclusion and the truth or acceptability of the premises, but upon integrating a broader multiplicity of considerations. Then too, just as some films rely more heavily upon dialogue than upon special effects, whereas with others the reverse is true-so that the competent film critic must know where to place the weight of emphasis in a compre- hensive assessment-so it is also with argu- ments. Moreover, just as the aim of film criticism is not to arrive at a snap judgement, a quick and final "thumbs-up" or "thumbs- down", but rather to achieve a full critical appreciation of a complex work, so also with argument. The practice of the mature argument critic is not limited to, nor can it effectively be reduced to, simple algorith- mic judgements of an argument's worth, but consists in sustained in-depth analytic and interpretive commentary which attends to subtlety and nuance and so on. The mature argument critic, in short, is capable of "argument appreciation" in the sense of the Argument Appreciation/Criticism 91 term familiar in connection with the arts. The analogy proposed here can of course be generalized beyond film to literature, music, and so on. The fundamental analogy is to art-critical judgement and discourse. Since analogies are inexact and more than one of them may be illuminating in a given area, they needn't be evaluated in competition with each other. Therefore, in considering the merits and liabilities of this analogy, let us take it that it is not intended to replace or supersede, but rather to supple- ment and enrich other models, such as the jurisprudential modeL The Cognitivity Question Despite whatever promise the analogy may hold, there are those who will regard it as prima facie unwelcome, who will react with dismay at the mere suggestion of an analogy between argument assessment and art criticism, an area of discourse, after all, in which all judgement is but a "matter of taste". The worry here, in the tradition of empiricism, is that the proposed analogy would lead normative theory of argument into relativism or subjectivism. To cast this worry in the vocabulary of twentieth- century analytic epistemology, if argument criticism is like art criticism then it won't be possible to defend it as a "cognitive" area of discourse. 5 I think the point needs to be made at this juncture that the worry over what we may call the "cognitivity question" does not by itself weigh against the proposed analogy. Indeed, so far the analogy between argu- ment criticism and art criticism seems to hold up remarkably well. Consider: Some films are better than others. Only aesthetic relativists dispute this. But how do we tell that one film is better than another? And if someone should disagree, how do we estab- lish that one film is better than another? For this we need an answer to the cognitivity question with regard to film criticism, or more generally art criticism. This is pre- cisely the position we are in with regard to 92 Joel Rudinow normative theory of argument. Moreover, the worry, as expressed so far, simply assumes that art criticism is non-cognitive. There is of course a long- standing though inconclusive tradition of argument, traceable through Hume perhaps even as far as Plato's Ion, to the effect that art criticism is non-cognitive. But despite this, art critics and art lovers continue to engage in discussions which they regard as perfectly meaningful and open to familiar forms of challenge, verification, and refuta- tion. They continue to engage in disputes which they regard as genuine. They offer and evaluate arguments, they weigh evi- dence and in general continue to behave as though their discourse is perfectly cogni- tive. None of this is by itself decisive. But there is, accordingly, also a long-standing tradition of argument in support of the cog- nitivity of art criticism. For present pur- poses the fact that art critics and theorists of art criticism have been wrestling with the cognitivity question for some time does suggest at least the reasonability of explor- ing the proposed analogy. An Answer to The Cognitivity Question: Analogy #2 The question of cognitivity in a given area of discourse is generally understood as turning on the availability of decision pro- cedures, or criteria, for settling disagree- ment. This is not the place to canvass the extensive literature on the issue of the cog- nitivity of art criticism. However, one par- ticular line of argument for the cognitivity of art criticism will be worth considering here for the light it may shed on the logic of argument assessment: namely the argument based analogically on the cognitivity of color attribution. Color attribution is an interesting point of departure because it is generally taken as a paradigm of cognitiv- ity. It is generally assumed, by partisans on all sides of the cognitivity debate in aesthet- ics, ethics, and elsewhere, that there are rec- ognized and accepted procedures for settling disagreements about the colors of things. We needn't here concern ourselves with the issue of the metaphysical status of colors. Colors have, since Locke, generally been taken as exemplary of the so-called "secondary qualities" of an object. Suffice to say that, whatever view was taken about the difference between primary and second- ary qualities and the metaphysical status of secondary qualities, there has never been a serious challenge, outside of the framework of a radical scepticism or relativism, to the idea that there are ways to tell the color of an object. So, how do we tell the color of an object? What are the decision procedures, the crite- ria, for settling disagreements about color? And how might this help us ground the cog- nitivity of art- or argument-criticism? Colors are perceptual phenomena with a physical basis in the behavior of light. Thus the notions most frequently appealed to in treatments of color cognitivity are the notions of empirical measurement and intersubjectivity. The general outlines of an account of color cognitivity would be sketched accordingly in one of two direc- tions: in terms of measuring the light reflec- tion and absorption characteristics of things or by appeal to intersubjectivity in compar- ing one's color-perceptual experience with that of other observers. The question arises: Which of these two directions is more fundamental? It is com- monly assumed (no doubt largely because it is so natural to assume, for perceptual phe- nomena with foundations in physics, such as color) that straightforward empirical appeals to the behavior of light are more fundamental, final, and decisive as means of determining the colors of things than the intersubjective appeal to comparative per- ceptual experience, and thus are theoreti- cally prior as a basis for the cognitivity of color attribution. However, there are rea- sons to think the reverse is true. First, there is no good reason to think color attribution was anything other than a paradigm of cog- nitivity before the physical regularities of the behavior of light were discovered or even imagined. Nor is there any good rea- son to suppose that the cognitivity of color attribution would be undercut by the dis- covery of a range of color phenomena which defied these regularities. Suppose, for example, that a species of flower were discovered whose blossom absorbed and reflected exactly the same wavelengths of light as a standard garden-variety red rose but which nevertheless appeared to every normal observer under all viewing condi- tions to be bright blue. Of course it is not obvious what color such a flower would be. One might insist (however dogmatically) that such a flower is red, though it doesn't "look" it. However, I would maintain that it is at least as plausible and serviceable to regard such a flower as blue despite its extraordinary light absorption and reflection characteristics, or, in other words, that inter- subjectivity is theoretically prior as a basis for the cognitivity of color attribution. Moreover, for our present analogical purposes the notion of intersubjectivity holds more promise than its rival. As cate- gories, "artworks" and "arguments" only partly overlap. Arguments and artworks are compositions. But though an object might "embody" an argument, arguments, unlike artworks, are never objects. They lie essen- tially outside the realm of the physical. There's nothing to them but words, mean- ings, propositions, ideas, and various rela- tionships among these. Thus one cannot coherently look forward to developing direct empirical measurements for assess- ing argument strength. There is nothing one could do with arguments that would be analogous to measuring the light reflection and absorption characteristics of an object. Whatever the colors analogy may have to offer for our purposes must therefore lie with the theoretically deeper notion of intersubjectivity-a notion on which the idea and practice of scientific measurement, in its essential repeatability, are based-and the way it functions in the area of color att ribution. 6 A car speeds past at dusk. One observer Argument Appreciation/Criticism 93 says, "Did you see that black convertible?" Another responds, "It was midnight blue." Normally the first step in settling such dis- agreements is to arrange better viewing conditions. At a minimum you want more than a fleeting glimpse, preferably in full sunlight. Optimal viewing conditions are those which maximize the observer's opportunity to exercise perceptual discrimi- natory capacity and minimize the likeli- hood of error. But suppose the disagreement persists under optimal viewing conditions. This raises the question of authority or expertise. Are some observers more author- itative than others, and if so on what basis does this authority rest? On my view, color attribution is one of those areas whose cognitivity is tied to a ref- erence group of observers. Disagreement over the colors of things is settled ultimately by appeal to the experiences (under the appropriate perceptual conditions) of mem- bers of the qualified color reference group, whose consensual judgements are regarded as authoritative. There are two main criteria for reference group membership: statistical normality and "sensitivity", or discrimina- tory capacity. The first criterion, statistical normality, which picks out those whose color perceptions constitute the agreed majority or plurality among observers, has its roots in the very notion of intersubjectiv- ity, which reflects the value of consensus in the domain of epistemology. The second criterion, sensitivity, which picks out those capable of the most and the finest (most sub- tle) discriminations, is a reasonable response to the inevitable difficulty in achieving complete consensus, and seeks in tum its own foundation, which in the case of color attribution it finds in standard empiri- cal discrimination tests to determine color blindness. It is generally agreed that the observers to whose experiences the cognitivity of color attribution is tied are both statistically normal and capable of more color discrimi- nations than members of competing groups, in other words, that color sensitivity is the norm. But this is a contingent matter. We 94 Joel Rudinow can easily imagine color sensitivity's being or becoming statistically rare. This a point with surprisingly interesting philosophical implications. First, it shows the common- place view that color disagreement is to be settled simply by appeal to the experiences of normal observers to be superficial and theoretically inadequate. The question of who would decide the colors of things if color sensitivity were rare remains philo- sophically interesting. Second, and most interesting for present analogical purposes, is the criteriological notion that disagreements are referred to observers identified by their sensitivity or discriminatory capacity. This presents what is for some a very welcome model on which to account for the cognitivity of art criti- cism, a model which supplies a quite inter- esting and subtle answer to the cognitivity question. On this model there emerges an authoritative reference group of observers distinguished by their aesthetic sensitivity, their ability to make subtle aesthetic dis- criminations, manifest primarily in their discourse about works of art. Where a sub- stantial consensus regarding the interpreta- tion or assessment of some object of critical attention exists among the members of this authoritative reference group, then the mat- ter is settled. Thus, as an area of discourse, art criticism is a cognitive one, meaning that . there are relevant and applicable criteria available for settling disagreements within the area of discourse. But not every dis- agreement which arises within the area of discourse is in fact settled by appeal to these criteria. Where substantial disagreement regarding the interpretation or assessment of some object of critical attention persists among the members of the reference group, then the matter is not settled. For example, most anyone with any claim to aesthetic sensitivity will agree that Chaplin's work in film is greater in depth and substance than the work of Arnold Schwarzenegger. But is Fellini better than Bufiuel? There are many points about which persistent disagreement among sensitive and informed art critics is to be expected. Yet that does not mean that there is no cognitive substance to the debate over such points, much less that the entire area of discourse lacks criteria for the rational resolution of disagreement. That substantial room is left for dis- agreement among the members of the refer- ence group without destroying its authority as a reference group is among the most promising features of this analogical line of inquiry, both for art criticism and for nor- mative theory of argument. In addition to the answer it provides to the cognitivity question, it shows a way out of the "boot- strap problem" mentioned above. Argu- ment critics may persist in disagreement as to the relative merits of many an argument without thereby undermining the cognitiv- ity of argument assessment, so long as there remains substantial consensus among them regarding at least some significant core of "paradigm examples" of argument strength and weakness. Moreover the fact that this account or model of cognitivity derives ultimately from a perceptual domain, which is one of the accepted paradigms of cognitivity, and in which sensitivity is empirically demon- strable, itself constitutes a considerable hedge against scepticism and relativism. Limits of the Analogies On the other hand, this last point calls attention to the chief liability inherent in arguments from analogy: overlooking rele- vant differences. In an argument based on multiple analogies, this liability is multi- plied. So we may expect to find a few. Since colors are perceptual phenomena, color attribution is not normally mediated by inferences. Thus, on my account of color attribution, the consensual experience of qualified observers is but one theoretical layer beneath the truth about the colors of things. Art- and argument-criticism are both matters of judgement, as opposed to perception, so in these areas there's an inter- mediate layer of inferential practice and principle to consider. Secondly, as was just mentioned, color sensitivity is empirically demonstrable, where aesthetic sensitivity, manifested pri- marily in discourse expressing judgements about artworks, is not. Similarly, sensitivity to the nuances of reasoning, manifested pri- marily in inferential practice, is not. This means that claims to sensitivity in these areas are problematic in a way in which claims to color sensitivity are not. The prob- lem in the problematic cases is what to sub- stitute for the straightforward empirical foundation available for claims to authorita- tive levels of color sensitivity. The problem is thus essentially one of justifying claims to sensitivity, and hence, on this account, to legitimacy as an authority. Scepticism aside for the moment, let us assume that there are indeed art -critical authorities and authorities in the area of argument criticism, just as there are authori- ties in the medical, ecological, political, his- torical, military, and many other fields. How in fact do such authorities become established? The authorities in such fields typically constitute what we might call a "community of discourse". They recognize each other. They address and respond to each other on matters that fall within their purview. They make judgements in each other's presence. They review, and in gen- eral respect, each other's judgement. To be an authority is to be recognized as one by other recognized authorities. A large and central part of what it takes to gain such rec- ognition has to do with demonstrable mas- tery of the relevant sort of inferential practice or judgement. Such mastery is demonstrated typically by making mean- ingful contributions to the discourse of the community. This in tum typically requires being conversant with, and generally in line with, the conventions of the discourse, including especially the paradigm exam- ples and precedents and the principles, can- ons, and criteria which reinforce and stabilize each other and thereby inform the discourse. 7 It should surprise no reader of this jour- nal, any more than it would surprise readers Argument Appreciation/Criticism 95 of The Journal of Aesthetics and Art Criti- cism, that one of the central themes in the ongoing conversation among recognized authorities in normative theory of argu- ment, and art criticism as well, is the debate over the paradigms, principles, precedents, canons, criteria, and so on which inform inferential practice in these areas. Norma- tive theory of argument can boast greater stability than art criticism in this regard, both in the area of paradigm examples and standards and rules. Art criticism, con- cerned as it is with a domain in which origi- nality and creativity are among the preeminent values, has had to remain rela- tively flexible. Historical examples, which serve as relatively stable points of reference and comparison in the critical discourse concerning the various artistic genre, never- theless remain open to ongoing challenge and reassessment, as do the canons of criti- cism. By contrast, normative theory of argument has evolved a substantial and rela- tively stable core of exemplary inferences and inference types as well as norms and standards exemplified by them to support argument assessments, though here too room remains for reasoned discussion. Both areas, however, are vulnerable to sceptical challenge directed against the form of "justification" sketched here. Is it a genuine form of justification? Not only do we find the form of circularity where partic- ular judgements and inferences are justified by appeal to general principles and stand- ards which are justified by appeal to partic- ular judgements and inferences. There is in addition the circularity of a body of authori- ties self-selected through mutual recogni- tion. The question becomes: Is the circularity involved in this form of justifica- tion "virtuous" or "vicious"?~ Does the Primary Analogy "Justify" Elitism? Let me conclude by exploring one line of argument to the effect that such a line of justification would be viciolls. The idea of 96 Joel Rudinow referring disagreement to observers distin- guished by their sensitivity makes it possi- ble to hold that some small but specially endowed or cultivated minority enjoys an authoritative position in a given area of dis- course. In the area of color attribution this has not posed a problem because, as was mentioned earlier, it is generally held that color sensitivity is normal. Nor has it caused much concern in the area of art criticism, though for another reason. In the art world there are many who are not particularly bothered by the idea that the cognitivity of aesthetic claims is grounded in the judge- ment of a few anointed scholars and critics. Aesthetes, or many of them, apparently like to think of themselves as an elite.9 Nor does this bother the rest of the world very much. After all, the aesthetes continue to disagree among themselves about the matters that interest them the most, so why not write the whole business off as a pretentious matter of taste? But the idea that the cognitivity of argument assessment should be grounded in the judgements of an elite priesthood of informal logic, whose members reserve the right to persistently disagree among them- selves in their judgements, even about the very rules of their "practice", and whose membership conditions seem to outsiders to amount to no more than mutual recognition, would pose a big problem for normative theory of argument. How, on this analogical account, will normative theory of argument protect itself from the charge of elitism? The answer, it seems to me, is this: Rea- soning is an essentially public domain and must be regarded as fully open to any mem- ber of the community of rational agents (i.e. persons). This does not mean that people are innately endowed with sensitivity to the nuances of argument or that all persons are or will ever be equally good at argument assessment. It might tum out that at some time and place (here and now, maybe) only some small and dwindling minority of the people is any good at dealing with argu- ments. Like aesthetic sensitivity, sensitivity to the nuances of argument must be devel- oped and is subject to lifelong cultivation and refinement. What the essential public- ity of reasoning does entail is that the princi- ples and standards of argument assessment, normative theory of argument, be intelligible to any rational being. In lay terminology, argument assessment must be teachable. And of course it is, as for that matter is aesthetic sensitivity. On this point, that argument assessment is teachable, there is overwhelming consensus among those with sensitivity to the nuances of argument. I'm not aware of anyone associated with the informal logic movement, for example, who takes the position that argument assessment can't be taught, though of course there's a great deal of disagreement about how to go about it. On this controversial question of peda- gogy, which is quite beyond the scope of the present paper, let me nevertheless close with one last reference to the analogy we've been considering, for the analogy may offer some guidance, even if it leaves many of the large and deep questions unanswered. Like the cultivation of aesthetic sensitivity, the teaching of argument assessment depends not upon the codification and transmission of a set of algorithmic decision procedures that cover the field. Like aesthetic sensitiv- ity, sensitivity to the nuances of argument is likely to be best cultivated and refined by exposure to the object of appreciation! criticism, in this case arguments. 10 Notes 1. Anthony Blair and Ralph H. Johnson, "The Current State of Infonnal Logic and Critical Thinking", Informal Logic 9 (1987), p. 147. 2 lbe argument as expressed here may be under- stood as an instance of the larger problem of jus- tifying standards of reasoning generally. For example, the question arises how one justifies a general rule of deductive inference. In address- ing this question, Nelson Goodman argues that "principles of deductive inference are justified by their conformity with accepted deductive practice. Their validity depends on accordance with the particular deductive inferences we actually make and sanction. If a rule yields unacceptable inferences, we drop it as invalid. Justification of general rules thus derives from judgements rejecting or accepting particular inferences." This occasions the puzzle of how to escape the "flagrant circularity" of justifying particular deductive inferences by appeal to general inference rules which are justified by appeal to particular deductive inferences. See Nelson Goodman, Pact, Fiction, and Porecast, 3rd ed. (Indianapolis: Bobbs-Merrill, 1973), pp.63-64. J In this I stop short of the charmingly simple solution Goodman proposes for the puzzle of the circular justification of general rules of inference, namely that the circle is a "virtuous" one of rules and examples alike being justified by "being brought into agreement with each other." This solution, elegant and powerful though it is, has embarrassing consequences for normative theory of argument. See Stephen P. Stich and Richard E. Nisbett, "Justification and the Psychology of Human Reasoning", Philos- ophy of Science 47 (1980), pp. 188-202. Stich and Nisbell offer an amendment to Goodman which results in an account of the justification of standards of reasoning similar to the one pro- posed here. I am indebted to one of this jour- nal's anonymous reviewers for the reference. 4 Ralph H. Johnson, "New Wine in Old Wine- skins", presented at the 9th International Con- ference on Critical Thinking and Educational Reform at Sonoma State University in August 1989. The opening epigraph is taken from this talk. 5 In this and what follows I use terminology inherited from empiricism through logical posi- tivism because this tradition makes the most stringent of criteriological demands. If one can articulate criteria that would satisfy a "positiv- ist", one can be pretty sure that one's criteria will at least be adequately stringent. Argument Appreciation/Criticism 97 6 For a fuller discussion of what follows see my "Colours, Cognitivity and Aesthetics", The British Journal of Aesthetics 17 (Autumn 1977), pp. 320-44. And cf. Stich and Nisbett, pp. 198-202. 7 This is essentially the amended Goodmanian account offered by Stich and Nisbett, ibid. For a fuller discussion of the notion of communities of discourse and paradigm examples in aesthet- ics, see my "Duchamp's Mischief', Critical Inquiry 7 (Summer 1981), p. 749. Recall Goodman's proposed solution to Hume's riddle of induction. His first move is to invoke a paradigm of accepted inferential prac- tice, namely deduction, as a model for the justi- fication of inferential practice generally, and then goes on to argue that deduction is justified in a circular way which is nevertheless, on his view, virtuous. Stich and Nisbett's objections to Goodman's proposed solution amount to the claim that the circle is too tight, and their amendment amounts to the proposal that it be expanded by the inclusion of authoritative judgement. In effect, they are prepared to endorse circular justifications which are accepted by authorities. Thus they are prepared to accept Goodman's account of the justifica- tion of deduction as well as his assessment of its virtue. 9 For a critique of aesthetic elitism, see my "Duchamp's Mischief', pp. 753-59. 10 An earlier version of this paper was presented at the 10th International Conference on Critical Thinking and Educational Reform at Sonoma State University in August 1990. I am indebted to Ralph Johnson and this journal's referees for helpful criticisms of an earlier draft. JOEL RUDINOW DEPARTMENTOP PIIlLOSOPHY SONOMA STATE UNIVERSITY 1801 EASTCOTATl AVENUE ROHNERT PARK. CA 94928 o