1Ted Cohen, ‘Three Problems in Kant’s Aesthetics’, British Journal of Aesthetics ,vol. 42 (2002), pp. 1–12.2Immanuel Kant, Critique of the Power of Judgment, tr. by Paul Guyer and EricMatthews (Cambridge: Cambridge U.P., 2000), § 1, p. 89, Akademie edition (‘Ak.’)5:203. References to Kant not otherwise designated are to this text.3Kant, § 36, p. 169, Ak. 5:288; § 37, p. 169, Ak. 5:289; First Introduction, sec. VIII,ibid., p. 26, Ak. 20:224; cf. sec. VII, p. 77, Ak. 5:191. Three remarks: (i) Kant alsoallows for aesthetic judgements based on displeasure (Unlust), but for the sake ofsimplicity I shall consider only those based on pleasure. (ii) In § 1 Kant defines anaesthetic judgement as a one ‘whose determining ground cannot be other than
subjective’ (p. 89, Ak. 5:203), while in the cited passage in the First Introduction, hedefines it as a judgement ‘whose predicate can never be cognition (concept of anobject)’. He apparently considers these two characterizations to be equivalent. (iii)To appear in British Journal of Aesthetics , vol. 43 (January 2003), pp. 65–74.KANT’S BEAUTIFUL ROSES:A RESPONSE TO COHEN’S ‘SECOND PROBLEM’Miles RindAbstract: According to Kant, the singular judgement ‘This rose is beautiful’ is, ormay be, aesthetic, while the general judgement ‘Roses in general are beautiful’ isnot. What, then, is the logical relation between the two judgements? I argue thatthere is none, and that one cannot allow there to be any if one agrees with Kant thatthe judgement ‘This rose is beautiful’ cannot be made on the basis of testimony. Theappearance of a logical relation between the two judgements can, however, beexplained in terms of what one does in making a judgement of taste. Finally, Idescribe an analogy between Kant’s treatment of judgements of taste and J. L.Austin’s treatment of explicit performative utterances, which I attribute to a deeperaffinity between their respective projects.In a recent publication in this journal,1 Ted Cohen presents three problemsin Kant’s aesthetic theory, one of which, the second of the three, is particu-larly troubling, as it casts doubt on one of Kant’s central claims. The pertinentclaim is the one made in the title of the opening section of the Critique ofAesthetic Judgement, that ‘the judgement of taste is aesthetic’.2 This means,among other things, that the predicate of a judgement like ‘This rose isbeautiful’, when that judgement is properly issued, is not a concept but afeeling of pleasure3—the word ‘predicate’ here signifying not the verbal



KANT’S BEAUTIFUL ROSES2Whatever exactly Kant may mean by saying that the predicate of an aestheticjudgement is a feeling, it is clear that he means that it is not a concept, and that isenough for grasping the problem at hand.4To my mind the best case for denying Kant’s commitment to the thesis of thenon-conceptual character of judgements of taste has been made by Karl Ameriks.See his ‘Kant and the Objectivity of Taste’, British Journal of Aesthetics , vol. 23 (1983),pp. 3–17; and ‘New Views on Kant’s Judgment of Taste’, in Herman Parret, ed.,
Kants Ästhetik/Kant’s Aesthetics  (Berlin and New York: Walter de Gruyter, 1998), pp.predicate of the sentence by which the judgement is expressed but thepredicative component of the corresponding mental act. On the other hand,Kant allows that a judgement like ‘Roses in general are beautiful’ may bederived from a collection of judgements on the beauty of individual roses. Ajudgement of this second kind, however, is not ‘aesthetic’ but, in Kant’sterms, ‘logical’, meaning, again among other things, that it has a concept fora predicate. It follows that the general judgement, contrary to verbalappearances, cannot have the same predicate as the singular one. But thatimplication is at odds with Kant’s concession that the general judgement canbe derived from a collection of singular judgements of beauty. Consequently,either judgements of beauty, whether singular or plural, always have aconcept for a predicate, or else general judgements of beauty cannot beinferred from singular ones; and neither option is compatible with Kant’sdeclared views. At the end of his discussion, Cohen takes up an analogy I onceadvanced between Kantian judgements of taste and Austinian explicitperformative utterances, but confesses himself unable to solve the problemby means of it. I do not think that the analogy will solve the problem, but Ibelieve that Kant can be gotten out of the present difficulty by other means.In this comment I wish to do four things: (i) to restate Cohen’s problemso as to make clear its importance and its difficulty; (ii) to offer my solution;(iii) to deal with a likely objection; and (iv) to enlarge a bit upon thecomparison of Kant with Austin.I. THE PROBLEMIt is important to appreciate that the problem that Cohen identifies does notarise from any casual or incidental statements that Kant makes, but from oneof the defining ideas of his aesthetic theory. There is, naturally, a certainamount of learned controversy over what exactly Kant means by saying thatthe judgement of taste is ‘aesthetic’, and over how deeply he is committed tothe seemingly extravagant claim that the predicate of a pure judgement oftaste is not a concept.4 But the conviction underlying that claim should be



MILES RIND 3431–47. For the contrary case, see Hannah Ginsborg, ‘Kant on the Subjectivity ofTaste’, ibid., pp. 448–65.5The basis of my attribution of this view to Kant is contained in §§ 32–33, pp.162–66, Ak. 5:281–66. For a fuller account, see Ginsborg, ‘Kant on the Subjectivityof Taste’.6For Kant’s account of the reflective use of the cognitive powers see Kant, sec.VII, pp. 77–78, Ak. 5:189–90; also §§ 9, 21, and 35. This operation is the subject ofCohen’s ‘first problem’.7Kant, § 8, p. 100, Ak. 5:215.8Pluhar’s translation (Indianapolis: Hackett, 1987), which Cohen uses (p. 4),takes Kant’s phrase ‘Vergleichung vieler einzelnen’ to mean a comparison of ‘manysingular roses’. It seems to me evident from the construction of the Germansentence, however, that Kant means ‘many singular judgements’, as in the Guyer–clear enough. Technicalities aside, part of what Kant means by that claim, oris trying to preserve by it, is surely the sense that declaring a thing beautifulis essentially tied to finding it beautiful, in the sense of actually being struckby its beauty, and that this is not the case with declaring a thing to be of acertain colour or origin or moral character or what have you. A description,or the testimony of others, may persuade me that a certain thing is beautiful,but I cannot legitimately express that persuasion by saying ‘X is beautiful’.Rather, I must say something like ‘By all accounts, X is beautiful’; or ‘X mustbe beautiful’; or ‘X is said to be beautiful’.5 I cannot make an epistemicallyunqualified declaration that the thing is beautiful until I have experienced—in Kant’s terms, ‘intuited’—the object for myself and thereby found pleasurein the reflective exercise of my cognitive faculties.6 The question whether thisjustifies Kant’s claim that the predicate of a judgement of taste is not aconcept is one with which I shall deal later in this paper (section III).Given that judgements of taste are essentially tied to intuition in theway just described, it follows that they must be made on objects one at atime; or as Kant says, ‘In regard to logical quantity all judgements of taste aresingular judgements.’7 The point may be supported by considering whatmight seem a counterexample, a judgement to such effect as: ‘The flowers inthat vase are beautiful’. It may seem that this is a non-singular judgement oftaste. But consider: either the subject term refers to the flowers as a singlecollective object of intuition, or else the judgement as a whole means ‘Eachflower in that vase is beautiful’. In the first case the judgement, thoughgrammatically plural, is logically singular, thus confirming Kant’s claim. Inthe second case the judgement is not one of taste at all but a judgement made,presumably, by inference from a series of judgements of taste on theindividual flowers; and this again confirms Kant’s thesis. Kant makes thispoint in terms of the judgement ‘Roses in general are beautiful’, which hesupposes to be made by a ‘comparison of many singular ones’.8 If the



KANT’S BEAUTIFUL ROSES4Matthews translation. An interesting incidental question is: what exactly does Kantmean by ‘roses in general’? For the sake of simplicity I have proceeded as if hemeant ‘every rose’; but it is possible that he means something more nuanced, suchas ‘any rose that is not defective qua rose (e.g., wilted, torn, blighted, deformed,. . .)’. 9For the distinction between general (generale) and universal (universale)judgements, see the Kant–Jäsche Logic, § 21, note 2. In the Critique of Pure Reason(A 71/B 96), the distinction is between gemeingültige and allgemeine judgements, alsocustomarily translated as ‘general’ and ‘universal’. The distinction, or someequivalent one, is, I believe, standard in scholastic logic, and even in present-daylogic, quantifiers—signs of generali ty—may be either universal or particular.inference seems too great a leap, Kant could equally well have used anexample in which the generalization is across a finite number of objects of acertain kind, such as the flowers in a particular vase.Indeed, Kant’s point can be made without even mentioning a judge-ment of universal logical quantity. What matters is only that the judgementis general; and general judgements may be particular as well as universal.9Thus, if the inference from a series of singular judgements on the beauty ofindividual roses to the universal judgement ‘Roses in general are beautiful’lacks credibility, take instead the unimpeachable inference from ‘This rose isbeautiful’ to ‘Some rose is beautiful’. Kant’s point holds good: the lattercannot be a judgement of taste, because it does not express the speaker’sfinding some object beautiful. It is made, not by an exercise of taste, but by aninference from someone’s exercise of taste (presumably one’s own).I take it to be evident, then, that, given Kant’s view that a genuinejudgement of taste can only be made by the reflective exercise of one’scognitive powers upon an object of one’s own intuition, one must accept theimplication that such judgements can only be singular and not general (anda fortiori not universal). So far, there is no problem. But then we must recall Kant’s claim that in logically general judge-ments of beauty, the predicate corresponding to the word ‘beautiful’ is aconcept, while in the singular judgement of taste it is not. It follows that thetwo kinds of judgement do not share a common predicate; from which inturn it follows that the general judgement, contrary to verbal appearances,cannot be a generalization of the singular one. And that seems to fly in theface of manifest fact. At the very least, it makes it difficult to explain how ageneral judgement of beauty can be inferred from a singular one, or from anycollection of singular ones. Hence Cohen’s observation: ‘It seems incredible,and it is more than a little frustrating, that the logic of this inference is sodifficult to formulate’ (p. 6).



MILES RIND 510Kant, § 8, p. 100, Ak. 5:215.11That is, the definitions of the beautiful as the object of a satisfaction ‘withoutany interest’ (p. 96, Ak. 5:211); as ‘that which pleases universally without a concept’(p. 104, Ak. 5:219); as ‘the form of purposiveness of an object, insofar as it isperceived in it without representation of an end’ (p. 120, Ak. 5:236); and as ‘thatwhich is cognized without a concept as the object of a necessary satisfaction’ (p. 124,Ak. 5:240).12This of course raises the issue: what is the relation between the act of judgingand the feeling of pleasure in a judgement of taste? Kant poses this question in § 9,and gives the surprising answer that the judging precedes the pleasure (p. 102, Ak.5:216–17). Commentators are divided between those who hold that there are twoacts of judging, one that gives rise to pleasure and another that requires everyoneto share that pleasure, and those who hold that there is just one act of judging,which of itself somehow constitutes the pleasure whose universal sharing itrequires. The ‘two-acts’ view is defended by Paul Guyer in Kant and the Claims of
Taste, 2nd ed. (Cambridge: Cambridge U.P., 1997), pp. 133–41, the ‘one-act’ view byII. THE SOLUTIONIf there is, among the claims just presented, one to which Kant is not clearlycommitted, it is surely the claim that a judgement like ‘Every rose isbeautiful’ is the universal generalization of ‘This rose is beautiful’. In fact,Kant never even makes that claim: he merely says that the general judgement‘arises from the comparison of many singular ones’ and ‘is no longerpronounced merely as an aesthetic judgement, but as an aestheticallygrounded logical judgement.’10 The difficulty is that, if the one judgement isnot the universal generalization of the other, it is unclear what the logicalrelation between the two judgements is, or how the one may be inferred fromthe other.It seems to me that Cohen, in a couple of separate remarks, actuallysuggests the way out of this difficulty, though he does not take the right pathhimself. First, he observes that, although Kant says or implies that thepredicate of a general judgement of beauty is a concept, he does not say whatthis concept is (p. 5). What then could it be? Presumably it is the concept ofthe beautiful. But what is the content of that concept? A number of differentanswers would be compatible with the various things that Kant says aboutthe content or purport of the judgement of taste (such as his four ‘definitionsof the beautiful’).11 I propose to adopt the formula ‘capable of being judgedwith pleasure in mere reflection’ as an analysis of the concept of thebeautiful. ‘Mere reflection’ is one of Kant’s descriptions of the operation ofthe cognitive faculties that gives rise to the distinctive pleasure of taste;‘judging’ is the aspect of this operation whereby the sharing of this pleasureis required of all who judge of the object.12



KANT’S BEAUTIFUL ROSES6Hannah Ginsborg in ‘On the Key to Kant’s Critique of Taste’, Pacific Philosophical
Quarterly, vol. 72 (1991), pp. 290–313. In my view, only the ‘one-act’ interpretationis compatible with Kant’s thesis that the judgement of taste is essentially tied to thejudging person’s own intuition of the object of the j udgement; but there is not spacefor me to justify this claim here.13The question, please note, is how Kant can maintain this; in section III, I shalladdress the question why he maintains this, or rather why he must do so.14See Critique of Pure Reason, A 70/B 95. To be sure, Kant would say that thisis only a ‘general-logical’ characterization of the act of judgement: there would alsobe a ‘transcendental-logical’ characterization in terms of the synthesis of a manifoldof empirical intuition in accordance with the transcendental unity of apperceptionand so forth. The first purports to describe how concepts are related to one anotherin the judgement, the second, how intuitions are ‘brought to concepts’ in the firstplace.This much granted, the question arises: how can Kant maintain that theword ‘beautiful’ expresses that concept, or any concept, in a so-calledaesthetically grounded logical judgement, but not in a judgement of taste?13The answer to this can again be derived from a remark of Cohen’s. Defendersof Kant, Cohen observes, often insist that for Kant a judgement is not astatement or a proposition but an act of the mind (p. 7). So what is the act inquestion? In the case of ‘Every rose is beautiful’, it would be the act ofcombining the concept of a rose with the concept of beauty using what Kantcalls the logical functions of the understanding, specifically those of thecategorical, universal, affirmative, assertoric form of judgement.14 In the caseof ‘This rose is beautiful’, the act would consist in judging the object withpleasure in mere reflection. From this it follows that, in making the judgement of taste, one does notassert that the rose in question is capable of being judged with pleasure inmere reflection; rather, one simply does so judge it. That one does so judgethe rose implies that it is capable of being so judged. That is why, once I havejudged that this rose is beautiful, I am in a position to make the logicaljudgement ‘Some rose is beautiful’, or, once I have made the judgement ‘Thisrose is beautiful’ of a certain number of roses, or perhaps of every rose in acertain vase, I am in a position to make the judgement ‘Roses in general arebeautiful’, or ‘Every rose in that vase is beautiful’. I infer the logicaljudgement from the judgements of taste that I have made, in the sense thatI infer it from having made those judgements. I do not infer it from them astheir logical consequence, as I might infer ‘Some man is mortal’ from ‘Theman Socrates is mortal’.If Cohen does not see the possibility of this solution, or does not acceptit as a solution, that may be because he assumes that the items from whichan inference is made must be propositions or statements. Thus, even when



MILES RIND 715This is what I take Cohen to be claiming in the second paragraph on p. 7,though I find the exact purport of this passage difficult to make out.16‘By hypothesis’ because there is nothing about the words ‘This rose isbeautiful’ that guarantees that they are the expression of a genuine judgement oftaste: someone might utter those words, and mean them, but on the basis of thejudgements of others, say.17I am aware that this implies that a judgement of taste can have nocontradictory. However, the seeming logical opposition betw een ‘X is beautiful’,uttered as a judgement of taste, and ‘X is not beautiful’ may be accounted for alongthe lines of the explanation already given: ‘X is beautiful’ betokens the performanceof a mental act whose possibility is denied by ‘X is not beautiful’. It will be notedthat this presumes that ‘X is not beautiful’ is a logical judgement, not a judgementof taste. Such I believe to be the implication of Kant’s declared views. I discuss therelated issue of whether ‘X is ugly’ can be a Kantian pure judgement of taste (andargue that it cannot) in ‘Can Kant’s Deduction of Judgments of Taste Be Saved?’,
Archiv für Geschichte der Philosophie, vol. 84 (2002), pp. 20–42, at pp. 27–29.he takes up the idea that a Kantian judgement is an act of the mind, he seemsto take this to imply that a judgement of taste should be represented in aninference by a statement to the effect that a certain person judges a certainobject with a certain sort of pleasure.15 But it need not, and, if I am right,should not be so represented. On the account that I have proposed, I do notdraw my inference from any premises—any statements or propositions—atall, but from my own prior acts of judging. To be sure, it is not clear how one should characterize this operationbeyond what I have already said, namely that, having made certainjudgements of taste, I am in a position to make (indeed am committed to) acertain general judgement. I call the act an inference because it is therecognition, in a judgement, of the implication of another judgement that onehas made. One may, if so minded, restrict the term ‘inference’ to the drawingof logical consequences from propositions, so long as one recognizes that inthe present case, the making of one sort of judgement has consequences forwhat other judgements one can legitimately make.What, then—to return to our troubling question—is the logical relationbetween the judgement of taste ‘This rose is beautiful’ and a generaljudgement like ‘Every rose is beautiful’ or ‘Some rose is beautiful’, on Kant’stheory? I believe it best to say that, properly speaking, there simply is nological relation between those judgements. The first judgement being byhypothesis aesthetic,16 it cannot enter into any logical relation, properly socalled, with another judgement at all.17 Making that judgement, however,does have implications for what other judgements one can legitimately make.To sum up: When I make a judgement of taste, I perform an act ofreflection whereby I both derive a certain pleasure and require that pleasure



KANT’S BEAUTIFUL ROSES8of everyone with respect to the object of the judgement. By doing that, I putmyself in a position to affirm the proposition that the object is capable ofbeing the object of such an act, along with whatever may be a logicalconsequence of that proposition. That is why, having made a favourablejudgement of taste on each rose in a certain vase, I may make the generaljudgement that every rose in the vase is beautiful (or that some rose isbeautiful). Both judgements are brought to expression with the word‘beautiful’, but only in the second case does that word correspond to aconcept in the act of judgement itself. My logical judgement is thus, as Kantsays, aesthetically grounded, but is not, despite verbal appearances, a logicalconsequence of the preceding judgement or judgements of taste. Judgementsof taste have no logical consequences properly so called.III. AN OBJECTION ADDRESSEDI am aware that the interpretation of Kant offered here may strike somereaders as an oversubtle attempt to avoid the obvious, the obvious herebeing, supposedly, that the predicate of a judgement of taste, like thepredicate of any other judgement, surely is a concept. Why not simplyembrace this fact and avoid the need for so much laborious finesse?The reply to this is that one cannot embrace that putative fact withoutrelinquishing the observation that I put forward earlier as the main supportof Kant’s thesis that ‘the judgement of taste is aesthetic’. This was theobservation that, unless I have actually found a thing beautiful, in the senseof being struck by its beauty, I am in no position to affirm without qualifica-tion that it is so. I may say, on the basis of testimony or description, that thething must be beautiful, that it is said to be beautiful, that it is supposed tobe beautiful, or other things of the sort; but I cannot say outright that it isbeautiful. This could not be the case if in judgements of the form ‘X isbeautiful’, when properly made, the word ‘beautiful’ expressed a concept; forif it did so, then testimony or description could in principle yield sufficientevidence to justify a judgement of beauty apart from any exercise of taste onthe part of the judging person. The objection, however, may be pressed further. My reply, it may besaid, presumes that Kant’s non-conceptuality thesis is the only possibleexplanation of why neither testimony nor description can justify anunqualified singular judgement of beauty. But an alternative explanationappears to be available, namely that an inference from testimony ordescription is subject to empirical uncertainty. To infer from a descriptionthat a certain thing is beautiful (so the explanation would run), I mustcommand some laws or reliable universal statements correlating observablefeatures of things with beauty; and unfortunately no one has yet established



MILES RIND 918To deal with the other part of the proffered explanation, the part concerningjudgements from description, would require a longer discussion. To show that onepart of the explanation fails is enough to show that the explanation as a whole fails.19One may be tempted to think that this is true even in Kant’s theory, for onthat theory, judgements of taste are not made on the basis of anything that can becalled ‘evidence’ at all. But then it is incompetent to say that such judgements exceedthe evidence.any such statements. In order to infer from the judgements of others that athing is beautiful, I must be assured of their competence as judges and of thepropriety of their exercise of taste with respect to the particular object inquestion; and there is much room for error on both counts. Thus, it seems, wecan account for our initial observation without accepting Kant’s non-conceptuality claim: the reason why I cannot rely on testimony or descriptionin order to affirm without epistemic qualification that a thing is beautiful, wemay say, is not that the term ‘beautiful’ is sometimes non-conceptual, butsimply that those sources of evidence are not sufficiently reliable.But this explanation will not work. Consider only the supposeduncertainty of inference from testimony.18 If such uncertainty were responsi-ble for the need to add epistemic qualifications to judgements of beautymade on the basis of testimony, then it would require us just as much to addsuch qualifications to judgements of beauty that we make from our ownexercise of taste; for I have just as much reason to doubt my own capacitiesand the propriety of my exercise of them as I have to doubt those of others.In that case, there would be no such thing as a legitimate judgement of taste,or a legitimate epistemically unqualified singular judgement of beauty: tocall a thing beautiful, without qualification, would always be to make anassertion in excess of the evidence.19 Thus the explanation on offer entails therejection of the fact that was to be explained, namely that judgements of tastecan only (which also means that they can) be made on the basis of a certainacquaintance with their objects. Those who would reject that claim may havetheir reasons for doing so, but they can only reject Kant’s account ofjudgements of taste wholesale. In sum, one cannot modify Kant’s aesthetictheory by rejecting his thesis of the non-conceptual character of judgementsof taste, but must accept or reject theory and thesis together.IV. KANT AN D AUSTINFinally, I want to return to the analogy by which, as Cohen reports, I oncecompared Kantian judgements of taste with Austinian explicit performativeutterances. The analogy was intended to have the following purport. It will



KANT’S BEAUTIFUL ROSES10 20A caution: that all utterances are performative means only that to saysomething is always to do something (beyond just saying something). It does notmean that every utterance is an explicit performative, like ‘I promise’, ‘I accept’, etc.See J. L. Austin, ‘Performative Utterances’, in his Philosophical Papers, 3rd edn(Oxford: Oxford U.P., 1979), pp. 244, 249.be agreed (I hope) that the word ‘promise’ has the same sense in ‘I promiseto be there’ (when uttered by someone—say me—to make a promise) as in‘M. R. promised to be there’; but, according to Austin, in the second case itdescribes or reports someone’s act of promising while in the first case it doesnot. Thus a word may have both a descriptive and a non-descriptive functionwithout therefore having two different senses. Note that the second sentenceis of a kind that can function in logical inference, while the first is not: onedraws inferences not from the sentence ‘I promise to be there’, but fromsomeone’s uttering it. These inferences nevertheless reflect genuineimplications of the utterance. With this linguistic precedent in view, it shouldbe easier to accept that the word ‘beautiful’ can have the function ofexpressing a concept in ‘Some rose is beautiful’ but not in ‘This rose isbeautiful’, without thereby changing its sense; and also to accept that onedraws inferences, not from the sentence ‘This rose is beautiful’, but fromsomeone’s uttering or thinking it. In proposing this analogy, I did notsuppose that the workings of Austin’s theory of explicit performativeutterances would serve to explain what is going on in the judgement of taste,nor am I inclined to follow Cohen’s proposal that the word ‘beautiful’, asused in a judgement of taste, ‘[makes] explicit what act is being performedby the judge’ (p. 8). The point was merely to make it easier to accept thepeculiar character that Kant’s account of judgements of taste requires us toattribute to the word ‘beautiful’.There is, however, a further significance to the comparison with Austin.One of the things that Austin brought to the attention of Anglophonephilosophers was that human utterance is subject to conditions andimplications quite distinct from those customarily called logical, but fully asirremissible, and in that sense as rigorous, as logical ones. These are theconditions and implications, not so much of the sentences that we utter, asof our acts of uttering them. Austin’s aim in first setting up and thenundercutting the term ‘performative utterance’, as I understand him, is to getus first to recognize the peculiar character of such conditions and implica-tions, and then to recognize their pervasiveness: all intelligible utterancesturn out to be performative.20Something similar, I believe, occurs in Kant’s third Critique. To be sure,for Kant the primary object of examination is not the utterance but thejudgement, which he seems habitually to think of, in most un-Austinian



MILES RIND 1121See §§ 7–8, pp. 97–99, Ak. 5:212–14. For an eloquent account of the affinitybetween Kant’s account of judgements of taste and the claims of ordinary-languagephilosophy, see Stanley Cavell, ‘Aesthetic Problems of Modern Philosophy’, in his
Must We Mean What We Say? (Cambridge: Cambridge U.P., 1976), pp. 86–96. Seealso Stanley Bates and Ted Cohen, ‘More on What We Say’, Metaphilosophy, vol. 3(1972), pp. 1–24, at pp. 22–24.22It may be pointed out that Kant does not analogously hold that alljudgements have a ‘taste’ aspect. Indeed not, but, as has been repeatedly observed,his account of judgements of taste has at least an appearance of implying that thepleasure of taste must accompany all cognitive judgements. I discuss this matter in‘Can Kant’s Deduction of Judgments of Taste Be Saved?’23I thank Ted Cohen and Lauren Tillinghast for their comments on an earlierversion of this paper.fashion, as a kind of private mental performance. Yet it is striking how muchof Kant’s characterization of aesthetic judgements is in terms of how they areexpressed, and in terms of the peculiar force and burden of using such aword as ‘beautiful’, as against some other, such as ‘agreeable’.21 It is morestriking still that a critique of the power of judgement should find its primaryobject in a judgement marked by its non-conceptual, or in Kant’s terms(which in this instance turn out to be less eccentric than they may haveseemed at first) its non-‘logical’ character. Kant maintains that the judgementof taste cannot be explained in terms of what is asserted in it, but only interms of what one does in making it, namely to engage in a peculiarly ‘free’,reflective operation of the cognitive faculties. That operation is supposed toreveal the nature of our very capacity to make judgements, cognitive orother, and thus to be no less fundamental than the forms and functions oflogic. The analogue here is Austin’s use of the explicit performative utteranceto reveal the performative aspect of utterance in general.22 I make no claimsfor the plausibility of Kant’s account of the mental operation supposedlyunderlying judgements of taste. I merely draw attention to the boldness ofhis undertaking, and to its affinity with Austin’s.23Miles RindDepartment of Philosophy (MS 055)Brandeis UniversityWaltham, MA 02140USAEmail: rind@brandeis.edu