The Transcendental Deduction and Skepticism

Stephen P. Engstrom

Journal of the History of Philosophy, Volume 32, Number 3, July 1994, pp. 359-380 (Article) Published by The Johns Hopkins University Press DOI: 10.1353/hph.1994.0069

For additional information about this article

http://muse.jhu.edu/journals/hph/summary/v032/32.3engstrom.html

Access provided by University of Illinois @ Urbana-Champaign (4 Oct 2013 18:42 GMT)

The Transcendental Deduction and Skepticism

STEPHEN ENGSTROM

IT HAS OFTEN BEEN ASSUMED t h a t K a n t ' s aim in t h e T r a n s c e n d e n t a l D e d u c t i o n includes a r e f u t a t i o n o f skepticism.' B u t this a s s u m p t i o n has led i n t e r p r e t e r s to c o n c l u d e n o t o n l y t h a t the D e d u c t i o n is at least in large p a r t a failure, b u t also t h a t K a n t is c o n f u s e d a b o u t h o w his objectives s h o u l d be a c h i e v e d a n d even a b o u t w h a t t h e y are. S u c h a c o n c l u s i o n m a y well raise d o u b t s a b o u t t h e initial a s s u m p t i o n a n d so invite a closer e x a m i n a t i o n o f w h a t K a n t h i m s e l f says c o n c e r n i n g the D e d u c t i o n ' s r e l a t i o n to skepticism. T h e p r e s e n t essay u n d e r takes s u c h a n e x a m i n a t i o n in a n a t t e m p t to d e t e r m i n e w h a t sort o f skepticism K a n t has in view a n d h o w he r e s p o n d s to it. C o n s i d e r a t i o n o f K a n t ' s r e m a r k s will s u g g e s t t h a t t h e D e d u c t i o n d o e s n o t aim at a r e f u t a t i o n o f skepticism, ~ a n d will t h u s o p e n t h e w a y to a n a l t e r n a t i v e c o n c e p t i o n o f its task, briefly e n t e r t a i n e d in the final section. T w o sorts o f skepticism will be c o n s i d e r e d . T h e first is the view K a n t According to P. F. Strawson, for example, "A major part of the role of the Deduction will be to establish.that experience necessarily involves knowledge of objects, in the weighty sense" (The Bounds of Sense [London: Methuen, 1966], 88). And according to Barry Stroud, "the transcendental deduction (along with the Refutation of Idealism) is supposed.., to give a complete answer to the sceptic about the existence of things outside us" ("Transcendental Arguments," Journal of Philosophy 65 [1968]: 242 ). More recently, Paul Guyer has suggested that Kant is ambivalent about whether to answer the skeptic or the empiricist (Kant and the Claims of Knowledge [Cambridge: Cambridge University Press, 1987]; see also "The Failure of the B-Deduction," The Southern Journal of Philosophy 25, Supplement [ 1986]: 67-84). Similar conclusions are reached in Karl Ameriks, "Kant's Transcendental Deduction as a Regressive Argument," Kant-Studien 69 (1978): 273-87; Jonathan Lear, "The Disappearing 'We'," Proceedings of the Aristotelian Society, Supplementary Volume 58 0984): ~x9-42; and Henry Allison, "Reflections on the B-Deduction," The Southern Journal of Philosophy ~5, Supplement (x986): 1-15. (For an intermediate position, see Edwin McCann, "Skepticism and Kant's B Deduction," History of Philosophy Quarterly ~ [1985]: 71-72.) My approach is distinctive primarily in its focus on what Kant himself says concerning the Deduction's relation to skepticism. To keep the discussion manageable, I confine my attention to the Deduction as presented in the second edition, which is where Kant's remarks about skepticism occur.

[359]

360

JOURNAL

OF THE

HISTORY

OF PHILOSOPHY

3 2 : 3 JULY

1994

addresses in the Refutation o f Idealism: "the problematic [material] idealism o f Descartes," "which declares the existence o f objects in space outside u s . . . to be d o u b t f u l a n d indemonstrable" (B274), a r g u i n g that "the only i m m e d i a t e e x p e r i e n c e is i n n e r experience, a n d f r o m it we only infer o u t e r things, and, m o r e o v e r - - a s in all cases w h e r e we are i n f e r r i n g f r o m given effects to determinate c a u s e s - - o n l y in a n u n t r u s t w o r t h y m a n n e r , f o r the cause o f the r e p r e s e n t a tions that we ascribe, p e r h a p s falsely, to o u t e r things, m a y lie in ourselves" (B276).s T h e second is the skepticism K a n t associates with the e m p i r i c i s m o f H u m e . T h i s skepticism, which will be described in g r e a t e r detail below, denies that any o f o u r k n o w l e d g e has its source in reason, m a i n t a i n i n g instead that "what is r e g a r d e d as reason is a universal illusion o f o u r faculty o f k n o w l e d g e " (B128). F o r convenience, we m a y call the f o r m e r "Cartesian.skepticism" a n d the latter " H u m e a n skepticism" (though, as will e m e r g e , these labels can be s o m e w h a t misleading).
1. CARTESIAN SKEPTICISM

1. K a n t m e n t i o n s skepticism only twice in connection with the Deduction, once n e a r the b e g i n n i n g in w , a n d once at the conclusion in w . T h e skepticism m e n t i o n e d in w is H u m e a n , so if the claim that the D e d u c t i o n a t t e m p t s to r e f u t e the Cartesian skeptic is to be s u p p o r t e d by what K a n t himself says a b o u t skepticism, it will have to be based o n the r e m a r k s in w B e f o r e t u r n i n g to those r e m a r k s , we should briefly consider their context, Kant's criticism o f w h a t he calls the " m i d d l e course." 3Listed below are the symbols and abbreviations used in references to Kant's works and the translations on which I have based my own. Page references to the Cr/t/que use the numbering of the first (A) and second (B) editions. Other page references use the numbering of the appropriate volume of Kants gesammdte Schr~wn, herausgegeben van der Deutschen (formerly KCmiglich Preuflischen) Akademie der Wissenschaftan, 99 vols. (Berlin: Walter de Gruyter [and predecessors], 19de- ); citations of Kant's minor works are given, without abbreviated tide, by the volume and page numbers from this edition. AIB Critiqueof Pure Reason, trans. Norman Kemp Smith (London: Macmillan and Company, 19~9). KpV Critiqueof Proztical Reason, trans. Lewis White Beck (Indianapolis: Bobbs-Merrill Company, 1956). KU CritiqueofJudgraem. L Logic, trans. Robert S. Hartman and Wolfgang Schwarz (Indianapolis: Bobbs-Merrill Company, 1974). MAN Metaphysical Foundations of Natural Science, trans. James W. Ellington (Indianapolis: Bobbs-Merrill Company, 197o). P Prolegomena to Any Future Metaphysit.s, trans. Paul Carus (revised by James W. Ellington) (Indianapolis: Hackett Pubfishing Company, t977). Page references to Hume's Treatise of Human Nature use the numbering of the L. A. SelbyBigge edition (Oxford: Oxford University Press, 1888).

THE TRANSCENDENTAL DEDUCTION AND SKEPTICISM

36X

The middle course purports to be an alternative between what Kant initially presents as the only two ways of conceiving a necessary agreement between concepts and experience: either "experience makes these concepts possible," or "these concepts make experience possible" (B 166). In the case of the categories, Kant insists on the latter alternative on the grounds that it is the only way in which there can be a necessary agreement when the concepts in question are a priori. Thus, as he indicated in w14, the Deduction is to accomplish its primary aim o f showing the legitimacy of our use of the categories in experience by proceeding on the principle that the categories make experience possible. But in w 4 it is also said that the categories make experience possible by making possible its objects, so in proceeding on this principle Kant is following the "Copernican" way of thinking: "objects must conform to our cognition" (Bxvi). The categories can be in necessary agreement with experience and its objects only by making those objects, and hence experience itself, possible. T h e middle course attempts to conceive the necessary agreement by supposing, not that experience depends on the categories, but that both depend on something else, namely, the Creator: "A middle course may be proposed between the two above mentioned, namely, that the categories are neither self-thought first principles a priori of our knowledge nor derived from experience, but subjective dispositions of thought, implanted in us from the first moment of our existence, and so ordered by our Creator that their employment is in complete harmony with the laws of nature in accordance with which experience proceeds--a kind ofpreformation-system of pure reason" (B 167). This idea of a harmony between implanted dispositions and the laws of nature recalls Leibniz's doctrine of preestablished harmony and also Descartes's reliance on the metaphysical claim that God exists and is no deceiver to support his conclusion that there are outer objects corresponding to our ideas of bodies.4 Since metaphysical claims of this sort have been called into question in the Critique, we might be led to conjecture that Kant's objection to the middle course is that the Cartesian conclusion cannot be supported by such questionable claims. The skeptic is introduced when Kant describes the consequences of the middle course as "exactly what the skeptic wishes most of all" (B 168). Now if we read this remark in light of the conjecture just mentioned, itwiU be easy to interpret it as confirming that conjecture, as indicating that Kant has in mind 41n fact, Kant seems to be thinking of Crusius: see P 319n.; cf. also Kant's letter of 2x February 177~ to Marcus Herz (Kant: Philosophical Correspondence z759-99, trans. Arnulf Zweig [Chicago: Universityof Chicago Press, x967],7~-73). But Kant describesthe positionoccupiedby Crusius as employingthe idea of a preestablishedharmony (PhilosophicalCorrespondence,73; MAN 476n.), and he says that for Crusius the Creator cannot deceive(P 3~9n.), so it might be thought that he also has in mind Leibniz and Descartes.

369

JOURNAL

OF T H E H I S T O R Y OF P H I L O S O P H Y

39:3

JULY

~994

a Cartesian skeptic who doubts whether there are outer objects corresponding to our ideas of bodies. Kant himself seems to express this doubt in the Prolegomena when he says, in criticizing the middle course, "we can never know with security what the spirit of truth or the father o f lies may have instilled into us" (P 319n.). So the point of Kant's remark would seem to be that the middle course, unlike the Deduction, fails to refute the outer-world skeptic. We are thus led to the thought that the Deduction's primary aim of legitimating our use of the categories is to be achieved by refuting the skeptic. 2. How mightsuch a legitimation proceed? Since the categories are said to be "concepts of an object in general" (B 198), we might suppose that Kant intends to achieve his aim by showing that the skeptic is committed to employing such concepts of an object and thereby also committed to the outer "objective experience" purportedly called into question. I f we ask what the initial skeptical position is from which these commitments are to be derived, we find a ready answer in wx6, where Kant takes the cogito as a starting point for the Deduction and speaks portentously of the "many consequences" that follow from the "original combination" involved in this Cartesian self-consciousness (B132-33). It is thus tempting to suppose that the Deduction pursues the following line of argument: Cartesian self-consciousness, which the skeptic admits, presupposes the employment of the categories, but this employment constitutes the "objective experience" the skeptic purports to doubt. This line of argument is in tension, however, with the fact that in w 4 and again in w Kant makes clear that the Deduction's central idea is the Copernican proposition that the categories make experience possible. For since Kant is explicit in w that he is concerned with pure, not empirical, apperception, the antiskeptical argument should focus on the possibility of such self-consciousness, not the possibility of experience. This problem cannot be removed by taking the Copernican proposition to be an awkward formulation of the stronger claim that through the categories' use "objective experience" is guaranteed. Kant says the categories secure no more than the possibility of experience and its objects, for he takes h u m a n discursive understanding and divine intuiting understanding to differ precisely in that only the latter can secure the existence as well as the possibility of its object (A92/ BI~' 5, BI38-39, B145). T o avoid the difficulty, it might be proposed that the experience said to be made possible by the categories is the Cartesian "inner experience" on which Kant bases his Refutation of Idealism (B275); perhaps such experience is the starting point in the Deduction as well.5 (Or, to accommodate the fact that wx6 sThat the Deduction followsroughly this line of argument has frequentlybeen suggested. See, for example,Strawson,TheBounds of Sense, 87-88~The twolinesof argumentjust mentioned

THE TRANSCENDENTAL DEDUCTION AND SKEPTICISM

363

b e g i n s with s e l f - c o n s c i o u s n e s s , it m i g h t b e p r o p o s e d t h a t t h e D e d u c t i o n a r gues that pure apperception presupposes inner experience, and that inner e x p e r i e n c e in t u r n p r e s u p p o s e s t h e c a t e g o r i e s a n d " o b j e c t i v e e x p e r i e n c e . " ) But when Kant speaks of experience, he means "objective experience." E x p e r i e n c e , as h e r e p e a t e d l y d e f i n e s it, is " e m p i r i c a l k n o w l e d g e , " k n o w l e d g e o f objects. 6 I n d e e d , h e m a k e s e x p l i c i t in b o t h w14 a n d w (cf. also Bxvii) t h a t his reason f o r m a i n t a i n i n g his C o p e r n i c a n p r o p o s i t i o n t h a t t h e c a t e g o r i e s m a k e e x p e r i e n c e p o s s i b l e is t h a t , since t h e c a t e g o r i e s a r e a p r i o r i c o n c e p t s , t h i s is t h e o n l y w a y in w h i c h t h e i r n e c e s s a r y a g r e e m e n t w i t h e x p e r i e n c e a n d its o b j e c t s is p o s s i b l e . I f t h e r e is t h u s f o r K a n t initially a g e n u i n e q u e s t i o n c o n c e r n i n g t h e p o s s i b i l i t y o f this n e c e s s a r y a g r e e m e n t - - t h e v e r y q u e s t i o n , in fact, t h a t t h e m i d d l e c o u r s e is a d d r e s s i n g in its a p p c a l to a c r e a t o r w h o h a r m o n i z e s o u r correspond to tactics "IIA" and "liB" in Guyer's taxonomy (Guyer claims in effect that Kant is ambivalent about whether to start with pure apperception or inner experience); see C/aims of Knowledge, 77-86. 6See B t47, B 165-66, B218, B~34, B277, Bxvii-xviii, and At ~4. Guyer cites two passages to support his claim that Kant sometimes speaks of experience in a "subjective sense" (C/a/ms of Knowledge, 80. In the first, Kant says, "Now in experience perceptions do indeed [zwar] come together only contingently, so that no necessity in their connection comes to light, or can come to light, from the perceptions themselves" (B219). Guyer says this passage "equates experience with the mere occurrence of perceptual states, rather than with any judgment that perceptions represent an object." But Kant says only that perceptions come together contingendy in experience, not that their occurrence/s experience; so experience, though not equated with judgment, is not said to exclude it either. Moreover, the fact that Kant uses the word zwar in this sentence and then begins the next by saying "But since IDa abet] experience is a knowledge of objects through perceptions..." (my emphasis) suggests that in both sentences he is speaking of experience in the same sense, the sense given in the definition at the beginning of the paragraph: "Experience is an empirical knowledge," which as such involves, in addition to judgment, "a synthesis of perceptions" (B2 t8; cf. P 3oo, 304). The same perceptions in the same experience can come together contingendy yet also be necessarily connected, for these two relations arise from different syntheses: the former, Kant explains, from the synthesis of apprehension, the latter from a dynamical synthesis (cf. B16o, B2otn., and especially P 3o5n.). Thus we have, not two conceptions of experience , but one conception with two levels of analysis. In the second passage, Kant says, "There can be no doubt that all our knowledge begins with experience; for how else should our faculty of knowledge be awakened into exercise if not by objects which, affecting our senses, partly of themselves produce representations and partly bring the activity of our understanding in motion to compare these representations, to connect or to separate them, and so to work up the raw material of sensible impressions into that knowledge of objects which is called experience?" (BI). Claiming (with Lewis White Beck) that this passage contains an equivocation, Guyer says that it makes sense only if the first occurrence of'experience" is taken to mean "merely the raw materials, the subjective states which are the data for all judgment and which may or may not be determined to represent objects." This would be indisputable had Kant said that all our knowledge/s preceded by experience, but as he only says it beg'inswith experience, the passage makes good sense if 'experience' in both its occurrences is taken to mean empirical knowledge. Kant is claiming, with the empiricists, that we have no knowledge before experience, and he is supporting this claim by pointing out that our faculty of knowledge can be awakened into activity only by objects affecting our senses, and that the knowledge of objects that arises in this way is what we call experience.

364

JOURNAL

OF T H E H I S T O R Y

OF P H I L O S O P H Y

32:3 JULY 199 4

concepts with n a t u r e - - t h e n the experience mentioned in the Copernican proposition must be "objective experience," for in the case of inner experience the question of agreement with outer objects does not even arise. Hence the antiskeptical interpretation--whether it starts from selfconsciousness or inner experience--makes it seem that when in w 4 Kant identifies his Copernican proposition as the "principle according to which the whole enquiry must be directed" (A94/B 126), and when in w he sums up the Deduction by calling it the exhibition (Darstellung) of the categories "as principles of the possibility of experience" (B168), he is confused about his own argument. For to exhibit the categories as principles on which the experience of objects depends simply ignores the skeptic's doubt as to whether such experience is something we do indeed have. 3. Here it might be countered that the Copernican rhetoric in w14 and w is not part of the Deduction proper, but belongs instead to a penumbra of commentary that does not provide a reliable guide for negotiating the obscurities of the Deduction's argument. Accordingly, it might be proposed that we are warranted in attributing confusion to that commentary should this be the cost of locating in the Deduction an antiskeptical argument free of the Copernican way of thinking and the confusion that its presence in such an argument would entail. This proposal would require us to ignore not only Kant's identification of the Copernican proposition as the Deduction's guiding principle in w 4, but also his assertion that precisely the paragraph in w 4 that expounds that proposition is "sufficient by itself" for the essential deduction of the categories (Axvii).7 T h e chief difficulty with this proposal, however, is that the Deduction's argument does not address the skeptic's doubt, s After introducing the concept of combination in wi 5, Kant claims in w16 that self-consciousness-the awareness of oneself as the identical subject of all one's representations-presupposes the possibility of uniting all those representations in a single universal consciousness and hence that those representations are subject to the condition (yet unspecified) under which alone such unity of consciousness is possible. Kant then declares in w 7 that an object is "that in the concept of which the manifold of a given intuition is unitecf' and adds that this unification of representations "requires unity of consciousness in the synthesis of them" Kant says the paragraph is sufficient for the Deduction's "objective"side, which, unlike the "subjective" s!de, is "essential" to his purposes because it is intended to "expound and render intelligible the objectivevalidity"of the categories (Axvi-xvii). These remarks occur only in the first edition, but the paragraph in w is retained in the second. s In what followsI summarize only the argument of the first part of the Deduction (w167 15-2o); it is this argument, with its focus on the cogito,that inspires the antiskeptical reading.

THE

TRANSCENDENTAL

DEDUCTION

AND

SKEPTICISM

365

(B137); accordingly, unity of consciousness is "an objective condition of all knowledge," one "under which every intuition must stand in order to become an object for me" (B138). Thus unity of consciousness has a Copernican status: it makes possible the object of knowledge, and hence knowledge itself. Kant goes on to argue in w19 that the unity of a judgment consists in this objective unity of consciousness, and concludes in w that since the categories are just the forms of j u d g m e n t insofar as the manifold of sensible intuition is determined in respect of them and thereby united in a single intuition, "all sensible intuitions stand under the categories, as conditions under which alone their manifold can come together in one consciousness" (B143). In claiming the categories to be requisite for the unity of intuition in w Kant is in effect identifying them with the concept mentioned in w17, the concept in which an intuition has the unity whereby it can be an object for knowledge. Hence the categories are just "concepts of an object in general" (B128; cf. B146 ), concepts that bring a manifold of intuition to the unity of consciousness, the condition "under which every intuition must stand in order to become an object for me," and thus the Copernican status attributed to the unity of consciousness in w is extended to the categories in w they make their objects possible. Faced with this argument, the skeptic might grant that self-consciousness presupposes the possibility of bringing the manifold of intuition to the unity of consciousness under which intuitions must stand in order to become objects for us, and grant further that the categories serve to bring the manifold to this unity, but still doubt whether outer objects exist. In particular, the skeptic can still doubt whether we do indeed have an outer sense by which we experience such objects, or whether the manifold of spatial intuition we attribute to outer sense is merely the invention of our imagination: "the c a u s e . . , may lie in ourselves" (B276). These doubts, which Kant knows the skeptic will raise, and which he addresses in the Refutation of Idealism (B275-77; cf. Bxxxix-xlin.), are not confronted in the Deduction. Kant argues that self-consciousness presupposes that all intuitions are subject to certain conditions (the unity of consciousness and the categories) to which they must conform for there to be objects of knowledge at all and hence for us to have any knowledge of objects in space, but he does not argue that self-consciousness presupposes that we do indeed have such knowledge. 4. Having encountered these difficulties, let us look more closely at Kant's discussion of the middle course. Kant's description of the consequences of the middle course as "exactly what the skeptic wishes most of all" was initially interpreted as linking the middle course to the Cartesian skeptic, and this linkage was taken to imply a concern to refute such a skeptic in the Deduction.

366

JOURNAL

OF T H E H I S T O R Y OF P H I L O S O P H Y

3 2 : 3 JULY ~994

But this interpretation does not fit what Kant says in his "decisive" objection to the middle course: I f the middle course were correct, the necessityof the categories, which belongs essentially to their concept, would then be lacking. For the concept of cause, for instance, which declares the necessity of an outcome under a presupposed condition, would be false if it rested only on an arbitrary subjective necessity, implanted in us, of combining certain empirical representations according to such a rule of [causal] relation. I would not then be able to say that the effect is combined with the cause in the object (that is, necessarily), but only that I am so constituted that I cannot think this representation otherwise than as thus connected. This is exactly what the sceptic wishes most of all. For then all our insight, through the supposed objective validity of our judgments, is nothing but sheer illusion. . . . (B 168) In w14 H u m e ' s skepticism is described as arising in the same way (B 1a7-28; cf. MAN 476n.): Because it did not occur to H u m e that the pure concepts of u n d e r s t a n d i n g might be grounds o f the possibility o f experience, h e was unable to c o m p r e h e n d how they could have their origin in the understanding, and consequently "he was compelled [durch Not gedrungen] to derive them f r o m experience (namely, f r o m a subjective necessity which arises t h r o u g h repeated association in e x p e r i e n c e , a n d which eventually comes mistakenly to be r e g a r d e d as objective, that is, from custom)" (B 1UT). Like H u m e a n empiricism, the middle course traces the concept of cause and the other pure concepts o f u n d e r s t a n d i n g to subjective dispositions and hence implies that all o u r "insight," that is, knowledge through reason (L 65), is illusory. This is Kant's decisive objection. Since for Kant a j u d g m e n t is objectively v a l i d - - t h a t is, true (cf. A 1 2 5 ) - just in case it is necessarily universally valid (necessarily valid for everyone) (P ~'98), he concludes his objection by recasting it in terms o f the latter idea: "nor would there be wanting people who would not admit this subjective necessity (which'must be felt); at the very least one cannot dispute with anyone regarding that which d e p e n d s merely on the m o d e in which one's own subject is organized" (B 168). T h e dispute whose possibility is at stake here is not simply over whether a given event has some specific cause. It concerns something into which we p r e s u m e to have "insight": when we specify a cause for an event, we p r e s u m e to know t h r o u g h reason that the event does indeed have a cause. In o r d e r to dispute with people who deny that a given event has a cause, I must be in a position to require that they must, if their j u d g m e n t is to be objectively valid, agree that the event resulted from some c a u s e - - t h a t is, my j u d g m e n t must be, or p u r p o r t to be, necessarily universally valid. But if I were to suppose that my j u d g m e n t depends on an implanted disposition, then in connecting my representation o f the event with the representation o f a cause in general, I would only be entitled to say that I am so constituted that I cannot think this representation otherwise than as thus connected; I would not be

THE TRANSCENDENTAL DEDUCTION AND SKEPTICISM

367

entitled to require others' a g r e e m e n t , and hence I could not dispute their denial.9 Kant's decisive objection to the m i d d l e course shows that (as we could only expect) he will reject a way o f conceiving how the categories can be in necessary a g r e e m e n t with e x p e r i e n c e if it implies that o u r insight is illusory. But his objection does not show that hc thinks he must demonstrate, by r e f u t i n g the skeptic m e n t i o n e d in w that insight is possible, or that o u t e r objects exist; a n d hence it does not show that the r e f u t a t i o n o f the skeptic belongs to the aim o f the Deduction. 5. Yet what o f the fact, briefly noticed above, that in the Prolegomena Kant says, in criticizing the middle course, "we can never know with security what the spirit o f t r u t h o r the father o f lies may have instilled into us"? Doesn't this suggest that Kant is c o n c e r n e d with Cartesian skepticism in the Deduction? T h e r e m a r k is i n d e e d reminiscent o f Descartes, and it also indicates that, in addition to his decisive objection, K a n t has a f u r t h e r objection to the middle course. But closer examination shows that Kant does not have Cartesian skepticism in mind. Kant objects to the m i d d l e course's way o f accounting for a priori principles o n the g r o u n d s that "since deceptive principles often i n t r u d e thems e l v e s . . , the use o f such a principle [i.e., the middle course] seems very problematic in the absence o f secure criteria to distinguish the genuine origin f r o m the spurious, for we can never know with security what the spirit o f t r u t h o r the f a t h e r o f lies may have instilled into us" (P 319n.).'~ Since the middle course leaves us with no criterion for d e t e r m i n i n g which o f the ostensibly a priori principles p u r p o r t i n g to be in necessary a g r e e m e n t with the laws o f n a t u r e have been instilled by the spirit o f truth a n d which by the father o f lies, it invites a d o u b t about that necessary a g r e e m e n t itself, the very thing it was i n t r o d u c e d to explain and guarantee. This d o u b t is recognizably a kind o f skepticism, yet since it does not e x t e n d to all j u d g m e n t s about o u t e r objects, but only to those p u r p o r t i n g to be a priori, it differs f r o m Cartesian skepticism. T o see this d i f f e r e n c e clearly, we must bear in mind that the middle course is c o n c e r n e d only with concepts and principles that are a priori. In the empirical case, where the necessary a g r e e m e n t is conceivable only on the supposition that o u r concepts and j u d g m e n t s are m a d e possible by experience, it would 9Kant's objection that the middle course implies there can be no dispute over whether an event has a cause does not mean that for him this question is open to dispute. Insofar as the proposition that every event has a cause is an a priori principle known through reason, it does not address a question on which different subjects are free to hold different opinions: it is beyond dispute, whereas the middle course implies that it falls short of dispute. ~oThis objection that the middle course provides no secure criteria can also be found in w "there is no end to be seen of the extent to which one might push the assumption of predetermined dispositions to future judgments" (B167; cf. also A783-84/B811-12 and the criticism of Crusius at L ~l ).

368

J O U R N A L OF T H E H I S T O R Y OF P H I L O S O P H Y

32:3

JULY 1994

make no sense to appeal to implanted dispositions. ~ Consequently, the criterion Kant claims we would lose by adopting the middle course bears only on concepts and principles that purport to be a priori. (For such concepts and principles the only suitable criterion is that provided by the Copernican way of thinking--specifically, by the idea of a condition of the possibility of experience.) Since the empirical criterion (agreement with experience) was never thought to be appropriate for a priori concepts and principles, it would not be undermined by the adoption of the middle course. So the doubt that Kant's further objection identifies the middle course as engendering is not global Cartesian doubt. Here again, as in the case of the doubt described in the decisive objection, the skepticism is of a kind that Kant associates with empiricism (as will become clearer below in section 2). The Cartesian skeptic doubts whether there is an outer world for us to know; the empiricist skeptic associated with the middle course doubts only whether we know that the world is as the categories represent it (namely, as necessarily a mathematically describable world o f substances in law-governed causal interaction).' ' And even if the doubt in question were Cartesian, Kant's rejection of the middle course for engendering it would not show that the Deduction aims to refute it. Like his decisive objection, Kant's further objection indicates that he thinks we should reject a conception of the necessary agreement between the categories and experience if it commits us to skepticism; but this does not imply that he thinks we are not entitled to any such conception unless we can show, conversely and through refutation, that skepticism commits us to it. Neither of Kant's objections implies that the doubt it imputes to the middle course needs to be refuted, and neither conceives of that doubt as Cartesian, so neither provides grounds for supposing that the Deduction aims to refute the Cartesian skeptic. And since this supposition leads to the serious interpretive difficulties noted earlier, we have good reason to reject it. 6. Kant does, of course, aim to refute the Cartesian skeptic in the Refuta-

~l All that is needed is a general principle of nature's suitability for cognition by our understanding. See KU, Intro., w " W e r e the skepticism associated with the middle course to incorporate the idea that empirical concepts and judgments depend on the categories, it could be developed to the point of effectively coinciding with Cartesian skepticism. Yet it should be noted not only that Kant does not describe the doubt as thus developed, but in addition that it would be inappropriate for him to do so, since such development would rely on precisely the Copernican recognition (that the categories make empirical cognition possible) with which the middle course is being contrasted, and hence would be not so much a working out of the consequences of the middle course as a (partial) return to Kant's own alternative position. The skepticism resulting from the middle course is supposed to be a consequence of the failure to adopt the Copernican way of thinking.

THE

TRANSCENDENTAL

DEDUCTION

AND SKEPTICISM

369

tion o f Idcalism.~S But it is noteworthy that there he speaks of Cartesian idealism rather t h a n skepticism. For although he does describe the Cartesian idealist as "skeptical" (A377) , thus implying that this idealist resembles the skeptic, a brief review of what he says about skepticism itself, its sources, and the knowledge claims it questions will suggest that, contrary to the anachronistic expectations we might have today, he does not think o f the skeptic as raising the idealist doubt. Kant opposes skepticism not to realism, but to its traditional foil, dogmatism, the presumption that reason can gain a priori knowledge entirely from its own principles without first engaging in self-critique (see Bxxxv, B7). In its failure to engage in such criticism, dogmatism falls into conflict with itself and thereby leads to skepticism, which, countering dogmatism's presumptive claim to the status of a pure rational sdcnce, opposes its assertions with "equally specious assertions" (B22-23; of. Aix, P 351) a n d thus "undermines the foundations of all knowledge, in order that, if possible, none of its reliability and security remain" (A424/B451; cf. 8:226). Since it first arises from the conflicts within dogmatic metaphysics, skepticism may initially concern only reason's claims to know beyond the limits of experience; but as the principles grounding such claims are gradually recognized to be also at work in reason's empirical application, in the absence of a critique to determine the domain of their legitimate e m p l o y m e n t a f u r t h e r stage of skepticism results, challenging the use o f reason even in experience (P 351). H u m e ' s skepticism is at this second stage. Believing himself to have discovered that "what is regarded as reason is a universal illusion o f o u r faculty o f knowledge" (B 128), H u m e doubted teason's claims (in particular with regard to the concept of cause) in all matters of fact and existence. But skepticism can advance still further, for once it recognizes the propositions o f mathematics to be synthetic (a recognition that H u m e failed to achieve), it will eventually extend even to this science (KpV 52)24 T h e three widening stages of doubt thus challenge the use of reason with r e g a r d to, first, the existence o f things beyond the limits of experience; second, the existence of things in general, even within experience; and third, ,3 But his remarks in the Refutation (and elsewhere; see Bxxxixn.) suggest that he does not think this skeptic has already been refuted in the Deduction. In the Refutation two species of (material) idealism are distinguished: Berkeley's "dogmatic idealism"and Cartesian "problematic idealism." With regard to the former, Kant announces that it has already been disposed of (in the Transcendental Aesthetic), and thus excuses himself from arguing against it in the Refutation. But he does neither of these things in the case of problematic idealism. Why these differences, if he thinks he has already refuted the Cartesian idealist in the Deduction? ,4 In this instance, however, what Kant is prepared to claim about skepticism in general he is not prepared to claim about Hume. Because of his high regard for Hume, Kant is not ready to think that had Hume recognized that mathematical propositions are synthetic he would have taken a further skeptical step. See note ~8 below.

37 ~

JOURNAL OF THE

HISTORY

OF PHILOSOPHY

3 2 : 3 JULY 1 9 9 4

all theoretical sciences, including mathematicg, so that ultimately all o f reason's "insight" c o m e s to be r e g a r d e d as "nothing b u t s h e e r illusion" (B 168).~5 N o w h e r e in this sequence d o e s the d o u b t concern the existence o f o u t e r objects. Except f o r the special case o f reason's e m p l o y m e n t b e y o n d the ]imits o f experience, the d o u b t focuses not on w h e t h e r there is a subject m a t t e r f o r reason's scientific use, but only on w h e t h e r reason has insight into its subject matter. So the fact t h a t K a n t speaks o f " t h e skeptic" in w provides no evidence that he has in m i n d Cartesian skepticism, and thus o u r earlier conclusion that the D e d u c t i o n does not aim to r e f u t e the Cartesian skeptic is c o n f i r m e d .
2. HUMEAN SKEPTICISM AND HUMEAN EMPIRICISM

1. It is m u c h m o r e plausible to s u p p o s e that the skepticism K a n t i m p u t e s to the middle course is k i n d r e d to the skepticism he associates with H u m e ' s empiricism. It should not be i n f e r r e d f r o m this, however, that the D e d u c t i o n aims to r e f u t e the H u m e a n skeptic. F o r such a n aim would not be in k e e p i n g with Kant's u n d e r s t a n d i n g o f the c h a r a c t e r and the origin o f H u m e ' s skepticism. T h e word ' r e f u t a t i o n ' can call u p the image o f o p p o n e n t s who, with a n adversarial t u r n o f m i n d , seek to d e f e a t one a n o t h e r in debate. T h e a d v e r s a r ial attitude is p e r h a p s detectable in the skeptic m e n t i o n e d in w w h o "wishes most o f all" f o r the c o n s e q u e n c e s o f the middle course in virtue o f b e i n g p r e d i s p o s e d to r e g a r d all o u r insight as "nothing b u t s h e e r illusion." But t h o u g h K a n t occasionally attributes such an attitude to the skepdc, h e rarely if ever does so to H u m e in particular, '8 n o r does he view H u m e as his o p p o n e n t . F o r Kant, skepticism results w h e n reason fails to see a way o f achieving its necessary aims: it is "a way o f thinking in which reason does such violence to itself that it could n e v e r h a v e arisen save f r o m c o m p l e t e d e s p a i r [Verzweiflung] o f e v e r achieving satisfaction with r e g a r d to its most i m p o r t a n t p u r p o s e s " (P 271).'7 H u m e ' s skepticism is n o exception. As we have seen, Kant says H u m e was compelled to derive the p u r e concepts o f u n d e r s t a n d i n g f r o m e x p e r i e n c e ( f r o m a subjective necessity, custom) because he failed to hit u p o n the C o p e r n i can solution to the p r o b l e m o f h o w those concepts can be in necessary agree,5Throughout these stages, skepddsm, itself a cultivated exercise of reason Ca principle of technical and scientific ignorance" [A494/B451]), challenges the use of reason in the sciences. Whether skepticism might undermine even the prescientific employment of common reason, Kant leaves to the reader to judge (KpV5~), though he also says there is no danger, since "the sound understanding will always maintain its rights" with regard to the principles of experience (P 351; see also Bxxxiv). 'r See P ~73 for a disputable instance. '~ We could say that skepticism will amount to misology if the violence reason does to itself involves self-hatred, which could result if reason, failing to see a way of achieving its necessary aims, turns on itself in frnstration (cf. A75fi/B784). In this way we might account for the perverse wish of the skeptic mentioned in w

THE TRANSCENDENTAL

DEDUCTION

AND S K E P T I C I S M

371

ment with experience (B127)--the solution that makes comprehensible how such ostensibly a priori concepts can have their origin in the understanding. But since H u m e could find no alternative to the empirical derivation, skepticism was unavoidable: "Hume's empiricism with regard to principles inevitably leads to skepticism" (KpV 5e). Thus, Kant says, H u m e believed himself to have discovered that "what is regarded as reason is a universal illusion of our faculty of knowledge," and so "gave himself over entirely to skepticism"(B x28). Hume's skepticism thus arose from his "complete despair" of comprehending how the pure concepts could have their origin in the understanding. Hence Kant sees Hume, not as a skeptical adversary to be refuted, but as a philosopher who was compelled to adopt a skeptical position for understandable reasons.X8 2. The language of refutation can also be used in nonadversarial contexts where the purpose in argument is not to defend a thesis, but simply to determine the truth. Any argument might be called a refutation if it demonstrates the truth of a claim that someone happens to deny. In order for there to be such a demonstration, however, the initial presumption must of course favor the position to be refuted at least to the extent that the argument against it cannot assume it to be in error. But as will become clear in what follows, Kant does not regard Hume's skepticism as thus favored in the initial presumption. So even when the language of refutation is not intended to imply any adversarial relation, it would be inappropriate to describe the Deduction's aim as including a refutation of the H u m e a n skeptic. As we have seen, Kant says H u m e was "compelled" to derive the pure concepts of understanding from experience. But Kant also regards H u m e as having been "overly hasty and incorrect" in concluding that the concept of cause arises from this source (P ~58). If H u m e was "overly hasty," then the empirical derivation must not have been his only available option, even if he was in a sense compelled to adopt it. Kant apparently supposes that, although the failure to comprehend how the pure concepts could have their origin in the understanding will compel one to derive them from experience/f one is committed to providing an account of their origin (even if it has skeptical consequences), that failure does not compel one absolutely, for one can always
,s Kant goes so far as to say that had Hume recognized that his problem concerning the concept of cause was merely one component of a general problem concerning synthetic a priori knowledge, a problem that arises even in the case of mathematics, he would have recognized the necessity of providing a general solution, and since he was much too insightful to base mathematics upon experience, he would have been stimulated (awakened from his empiricist slumber, as it were) to seek a solution along the very lines Kant develops in the Cr/t/que (P 272-73; see also BI 9 2o). (I shall not consider to what extent Kant's understanding of Hume is accurate, Since my purpose is m determine Kant's objectives in the Deduction, I am confining my attention to Kant's own understanding of his reladon to Hume.)

372

JOURNAL

OF T H E H I S T O R Y

OF P H I L O S O P H Y

3 2 : 3 JULY ~994

leave the problem open to f u r t h e r consideration. Kant thus appears to regard H u m e ' s adoption o f the empirical derivation as contingent upon an overly eager desire to account for the origin of the pure concepts, a desire that led him to attempt a solution to the problem he confronted prior to grasping it in its full generality (cf. P 260, 272-73; B 19-20). Determined to account for the p u r e concepts, yet failing to c o m p r e h e n d how they could have their origin in the u n d e r s t a n d i n g a n d hence unable to comprehend how synthetic a priori knowledge is possible, H u m e was compelled to derive them from experience, despite the fact that this derivation had the skeptical consequence that synthetic a priori knowledge is impossible. But what justifies Kant in regarding Hume's conclusion that the pure concepts derive f r o m experience as "overly hasty and incorrect"? T h e following passage suggests the answer: We are actually [wirklich] in possession of synthetic knowledge a priori, as is shown by the principles of understanding, which anticipate experience. Now if someone cannot at all comprehend the possibility of these principles, be may indeed at first doubt whether they actually dwell in us a priori; but he cannot yet proclaim this [inability to comprehend their possibility] to be an impossibility of these principles through mere powers of the understanding, nor proclaim all steps that reason takes under their guidance to be null. He can say only this: if we had insight into their origin and authenticity, we would be able to determine the scope and limits of our reason; but until we do, all assertions of the limits of reason are blindly ventured. In this manner a thoroughgoing doubt of all dogmatic philosophy, which goes its way without a critique of reason itself, is perfectly well grounded; but reason cannot on these grounds be altogether denied such forward steps, if they are prepared and secured through a better laying of the ground. (A762-63/B79o-9 l) Kant's t h o u g h t here is that the inability to c o m p r e h e n d the possibility o f something does not add u p to knowledge of its impossibility. Although a failure to c o m p r e h e n d how synthetic a priori knowledge is possible can raise a legitimate doubt as to whether we actually have such knowledge (and thereby pose a challenge to dogmatic claims to the contrary), it does not justify an inference to the skeptical conclusion that such knowledge is impossible. Since the empirical derivation entails this skeptical conclusion, a d o p d n g it as a result o f failing to c o m p r e h e n d the possibility of synthetic a priori knowledge is t a n t a m o u n t to making this unjustified inference. So although there is a sense in which H u m e was understandably compelled to derive the p u r e concepts from experience, this derivadon is not one to which he was strictly speaking entitled. Since his derivation was not based on an adequate investigation o f the problem o f synthetic a priori knowledge, his skeptical conclusions were themselves, in a sense, dogmatic a n d subject to doubt (cf. A767-68/B795-96). 3. B u t K a n t ' s criticism of H u m e a n skepticism is not confined to the sugges-

T H E T R A N S C E N D E N T A L D E D U C T I O N AND S K E P T I C I S M .

373

tion that Hume's conclusions are insufficiently supported. For Kant, the empirical derivation provided by Locke and H u m e is manifestly inadequate as an account of the origin of the categories. Before he even begins the Deduction, Kant tells us that the empiricists' derivation is already refuted: "But this erap/rical derivation, in which both philosophers agree, cannot be reconciled with the actuality [Wirklichkeit] of the scientific a priori knowledge that we possess, namely, pure mathematics and general natural science, and it is therefore refuted by this fact [durch das Faktum widerlegt]" (BI27-28; cf. also B 4 - 5, B2o). (The impossibility of achieving the reconciliation in question here is of course just what H u m e has admitted in giving himself over to skepticism.) That we actually possess (scientific) synthedc a priori knowledge is something Kant has presumed from the start. It is pointed out in w and w of the Introduction (B3ff., B 14ff.), and it is taken for granted in the formulation of the central problem of the Cr/t/que in w (B19ff.). Since as we have just seen the empirical derivation is unsupported, and since it conflicts with the actuality of the a priori sciences, it is sufficiently refuted by this fact.'9 In the second Critique Kant goes even further, suggesting that insofar as the empirical derivation implies the skeptical denial of the possibility of synthetic a priori knowledge, it is refuted by itself. After briefly describing his critical project of laying a secure g r o u n d for a systematic scientific philosophy, both theoretical and practical, he says (in a remark that again reflects the presumption that synthetic a priori knowledge is actually in our possession):
,9 Kant's appeal to the fact of the a priori sciences fits with the Prolegomena'sanalytic method, which assumes that .the a priori sciences of pure mathematics and p u r e (or general) natural science are actual and seeks to ascertain the grounds of their possibility; but it may appear to be at odds with the synthetic method of the Crit~, which does not assume the actualit~ of such sciences, but confines itself to the concept of pure reason as a faculty of knowledge and after determining the elements of this faculty develops from them the system of synthetic a priori knowledge whose actuality the analytic method took as its starting point (see P 263, 274, 279). However, there is no real conflict here, for different questions are at stake. Because it leads to skepticism, the empirical derivation is in effect a negadve answer to the question whether synthetic a priori knowledge is possible. But in both the Prolegomena and the Cr/t/que the fundamental question is how such knowledge is possible (P w B 19): Thus both methods, analydc and synthetic, take for granted that synthetic a priori knowledge is possible. They can do so because the possibility of this knowledge is shown by the actuality of the a priori sciences. The difference between the two methods lies in their explanations of how such knowledge is possible. Whereas the Prolegomena relies again on the actuality of the a priori sciences, this fact is not appealed to in the synthetic explanation furnished by the Cr/t/que. In that explanation, Kant first analyzes our power of cognition to isolate the elements (such as the categories) in virtue of which experience and synthetic knowledge/n general are possible and t h e n employs the Copernican way of thinking, according to which these elements or conditions make possible the objects of experience, to make comprehensible how synthetic a pr/or/knowledge is possible (see A154-58/B 193-97). Thus in the Cr/t/que Kant does not argue fr0ra the assumption that synthetic a priori knowledge is actual, but neither does he argue against the empiricist's claim that it is impossible.

374

. J O U R N A L OF T H E H I S T O R Y OF P H I L O S O P H Y 3 2 : 3 .JULY 1 9 9 4

N o t h i n g worse could befall these labors, t h a n that s o m e o n e should m a k e the u n e x pected discovery that t h e r e n e i t h e r is n o r can be any a priori k n o w l e d g e at all. But h e r e t h e r e is n o d a n g e r . It w o u l d be the s a m e as i f s o m e o n e w a n t e d to p r o v e t h r o u g h r e a s o n that t h e r e is n o reason. F o r we only say that we know s o m e t h i n g t h r o u g h reason w h e n we are conscious that we c o u l d have k n o w n it e v e n if it h a d not c o m e b e f o r e us in e x p e r i e n c e ; thus rational k n o w l e d g e a n d a p r i o r i k n o w l e d g e are o n e a n d the same. T o want to squeeze necessity from an empirical proposition (expum/ce aquam) and with this [sc. necessity] to secure t r u e u n i v e r s a l i t y . . , f o r a j u d g m e n t is a direct contradiction.

(KpV 1~)

Kant's point here appears to be the following: To deny that (synthetic) a priori knowledge is possible is to assert that all our knowledge is necessarily derived from experience; but since this (synthetic) assertion involves necessity, it can never be justified by experience and hence cannot be grounded unless it has an a priori basis in reason; yet the possibility of such a basis is precisely what the assertion denies. The "contradiction" that this argument points out lies, not within the proposition that synthetic a priori knowledge is impossible, but between this proposition and what is implied in supposing it to be knowable. The argument is thus meant to show, not that synthetic a priori knowledge is possible (its possibility is shown by its actuality), but rather that it is impossible to know that synthetic a priori knowledge is impossible. Although the argument does not mention Humean empiri{:ism and skepticism by name, it is fair to say in light o f the ensuing discussion (KpV I :t-14) that Kant takes it to show that the H u m e a n skeptical denial of the possibility of synthetic a priori knowledge directly refutes itself, and that insofar as the empirical derivation implies this denial, it too is self-refuting. ~~ The empirical derivation and the skepticism it implies are thus already, prior to any appeal to the considerations entertained in the Deduction, refuted both on a priori grounds and also by appeal to the actuality of the a priori sciences. Evidently, then, there is no initial presumption in favor of the empirical derivation and the skepticism to which it gives rise. The issue is therefore not whether synthetic a priori knowledge is possible, but how (B 19),
*oThe attempt to squeeze water from pumice can be seen in the opening pages of Hume's

Treatise, in which the empirical derivation is introduced as an empirical hypothesis and then

employed as if it were an a priori principle. Hume first treats his claim that for every simple idea there is a resembling simple impression as an inductively supported empirical proposition, and he challenges anyone who might deny this daim to produce a counterexample (Treatise, 3-4). But later, when considering the category of substance, he proceeds as though this proposition had already been established as an exceptionless rule. In supposing that the idea of substance cannot be "distinct from that of a collection of particular qualities" unless it "be deriv'd from the impressions of sensation or reflexion" (Treatise, 15-16), he excludes in advance the possibility that it might be a simple idea for which no resembling impression can be found and so just the sort of counterexample he had previously challenged the unconvinced reader to supply.

T H E T R A N S C E N D E N T A L D E D U C T I O N AND S K E P T I C I S M

375

and accordingly the Deduction aims neither to refute the skeptic nor to refute the empiricist. 4- Yet even though they have already been refuted, empiricism and its attendant skepticism have not thereby been satisfactorily removed. For their refutation does not eliminate the difficulty that leaves those seeking an account of the origin o f the pure concepts feeling compelled to adopt the empirical derivation. As long as H u m e is unable to comprehend how the pure concepts could have their origin in the understanding, he will be left with the sense of having no way of accounting for them except by deriving them from experience, despite the implication that synthetic a priori knowledge is impossible. ~ Hence what is needed is "a better laying of the ground," a way of understanding the categories that will afford us "insight" into the "origin and authenticity" of the principles of pure understanding and thereby enable us to see how the synthetic a priori propositions of the Sciences are possible. This is what Kant attempts to provide in the Deduction. Thus, in the second Critique, speaking o f the category of cause in particular, he explains that his deduction of it as an a priori concept consisted in "show[ing] its possibility from pure understanding without empirical sources" (KpV 53). He continues: "So, after doing away [Wegschaffung] with the empiricism of the origin of this concept, I was able to utterly remove [aus dem Grunde heben] its inevitable consequence, skepticism . . . . Thus I utterly removed the total doubt of everything into which theoretical reason professes to have insight" (KpV 53-54). So the Deduction removes skepticism and does so by indirect, even doubly indirect means: skepticism is removed by doing away with its cause, empiricism, and empiricism is done away with by showing the categories' possibility from pure understanding without empirical sources and thereby removing what made empiricism seem inevitable to Hume--despite its manifest inadequacy as an account of the origin of the categories--namely, the mistaken assumption that this possibility cannot be known. For Kant, the empiricist's skepticism is a sign of the inadequacy of the empirical derivation. The proper response, given this diagnosis, is to show
'~ From the inability to comprehend how the pure concepts could have their origin in the understanding there can also arise, even without adoption of the empirical derivation, a doubt concerning the actuality of synthetic a priori knowledge. For this inability is the basis of the failure to comprehend how the principles of synthetic a priori knowledge are possible, and that failure, as we have seen, may raise a doubt concerning whether they "actually dwell in us a priori" (A763/ B79o ). This doubt is not sufficient, however, to eradicate the presumption that we actually possess such principles, and hence it is not sufficient to undermine the refutation of the empirical derivation that this presumption provides. For as we have noted, "the sound understanding will always maintain its rights" with regard to the principles of experience (P 351). Without eradicating the presumption, the doubt may persist alongside it until comprehension is achieved of how the pure concepts can originate in the understanding.

376

JOURNAL OF THE HISTORY OF PHILOSOPHY 3~:3 JULY 1994

how it is possible for the categories to have their origin in the understanding. Skepticism is not the position of an adversary, but the symptom of an infirmity. It is not to be vanquished, but rather prevented; and when it arises the cure must work on the underlying cause. 3. THE DEDUCTION'S DIALECTICAL TASK

1. Although the Deduction does not aim to refute skepticism, it does, of course, address a doubt concerning the categories insofar as it aims to show how it is possible for them to have their origin in the understanding. This doubt is raised in w13 in the form of a question of our right to employ them. The fact that the doubt is raised as a question of right suggests that Kant's approach in the Deduction (and in the Cr/t/que as a whole) is dialectical in the sense that it seeks to reconcile apparently conflicting claims of reason. For on the one hand, asking whether we have the right to employ the categories presumes that we have all along, even if only implicidy, taken ourselves to have this right. Otherwise the question would be idle. And on the other hand, the question also involves the presumption that there is some reason, an apparently conflicting claim, on which its challenge is based. Thus in raising the question of right Kant indicates that the problem of the Deduction consists in an apparent conflict between claims recognized by reason. ~* But if the claims are recognized by reason, they are not arbitrary or contingent, and hence we cannot remove the conflict by simply relinquishing one of them. In such a case the problem is dialectical; the aim must be to reconcile the claims by showing them to be in systematic agreement. The question o f right is based on the Metaphysical Deduction's a priori doctrine that the origin of the categories lies in the understanding. This doctrine appears to conflict with our presumption that we have a right to apply the categories--indeed, to apply them a priori--in synthetic judgments to objects that are external to the understanding in the sense that they are not simply consequences of the understanding's activity, but are given to the understanding only to the extent that they affect the mind. Insofar as the Metaphysical Deduction bases its doctrine on a systematic derivation of the categories from the understanding conceived as fundamentally a faculty of judgment, and insofar as the presumption that we have a right to apply the categories originates in "the sound understanding," which "will always maintain its rights" (P 351), neither of these apparently conflicting claims can be simply relinquished. *"For further discussionof Kant's use of the question of right, see Dieter Henrich, "Kant's Notion of a Deductionand the MethodologicalBackgroundof the First Crit/que,"in Kant's Transcendental Deductions: The Three "Critiques" and the "Opus postumum," ed. Eckart F6rster (Stanford: Stanford UniversityPress, 1989), ~9-46-

THE

TRANSCENDENTAL

DEDUCTION

AND SKEPTICISM

377

T h e T r a n s c e n d e n t a l Deduction thus faces the dialectical task o f reconciling these a priori claims. Dialectic in the sense just i n t r o d u c e d is distinct f r o m dialectic in Kant's technical sense a n d in particular f r o m his "transcendental dialectic," which consists in the criticism and diagnosis o f reason's illegitimate metaphysical claims to know b e y o n d the limits o f experience (see A 6 o - 6 4 / B 8 5 - 8 8 ) . Yet it is related to transcendental dialectic as genus to species, for while transcendental dialectic is distinctive in that it employs the special means o f exposing an illusion u n d e r l y i n g reason's illegitimate claims, its ultimate aim is to reconcile the claims o f reason and thereby justify or vindicate reason itself. T h u s Kant says at one point that it was the antinomy o f p u r e reason that first awoke him f r o m his dogmatic slumber and d r o v e him to the critique o f reason "in o r d e r to r e m o v e [heben] the scandal o f the seeming [scheinbaren] contradiction o f reason with itself."~s Reason's self-criticism is u n d e r t a k e n for the sake o f selfjustification, which consists in exhibiting the systematic c o h e r e n c e a m o n g reason's constituting principles and i n t e r e s t s . O n Kant's dialectical approach, then, the claims o f reason are initially p r e s u m e d to be true.24 Since d o u b t arises only t h r o u g h a p p a r e n t conflict, only reconciliation will r e m o v e it. a. Kant's idea that philosophy is c o n c e r n e d with the justification o f reason echoes, o r r a t h e r reconceives, the Leibnizian idea o f philosophy as theodicy, and accordingly the Deduction's dialectical task can be elucidated by pointing out an analogy it bears to the traditional task o f theodicy. T h e task o f theodicy is not to r e f u t e o r to convert the infidel, but r a t h e r to justify o r make intelligible the ways o f G o d to those who believe, or are at least initially disposed to believe, yet who are at a loss to u n d e r s t a n d how God's justice, benevolence, wisdom, and p o w e r can be reconciled with the manifest evil in the world. W h a t is desired is not a p r o o f that G o d exists, but r a t h e r a d e m o n s t r a t i o n o f how God's attributes are compatible with the defeat and suffering o f the innocent and the t r i u m p h and prosperity o f the wicked. T h e task o f theodicy is accordingly to provide a vindication o f the divine attributes in the face o f the existence o f evil. Similarly, Kant's task in the Deduction does not require that he address 9s Letter to Christian Garve, 21 September 1798 (PhilosophicalCorrespondence,~52; my translation); see also Bxxxiv. In revealing the illusion underlying reason's illegitimate claims, transcendental dialectic shows that such claims stem not from any contradiction within reason itself, but merely from a misunderstanding, to which we are naturally susceptible, of reason's proper employment. The illegitimacy is shown to consist in the misapplication of principles that are in themselves free of contradiction and whose proper employment (as regulative principles) is therefore fullyjustified. (Cf. A64~-43/B67o-7 l, A669/B697.) ,4 Hence the fact, noted earlier, that Kant takes pure mathematics and general natural science for granted as actual (as Fakta) is a further sign that his approach is dialectical.

378

J O U R N A L OF T H E H I S T O R Y OF P H I L O S O P H Y 3 2 : 3 JULY x 9 9 4

someone initially disposed to radically doubt that there are any pure concepts of understanding. That the categories are such concepts is presumed; the task is to show how it is possible for these pure concepts--concepts we have been employing all along--to relate a priori to objects (A85/BlI7). Just as a theodicy addresses those who are disposed to believe in God and attempts to show how God's attributes are compatible with mundane evil, s o the Deduction addresses those who regard the categories as a priori concepts and attempts to show how the categories' origin in the understanding is compatible with their reladng a priori to objects. Thus in the Prolegomena when he explains how he has solved Hume's problem (and in so doing indicates how the pure concepts can relate to objects), Kant says that his soludon "rescues" or "vindicates" (retteOfor the pure concepts their a priori origin (P 313; see also KpV 54). It might even be said that the Transcendental Deduction is meant to "vindicate" the Metaphysical Deduction, for an explanation of how concepts that purport to relate a priori to objects can in fact do so would remove the doubt to which the table o f categories gives rise and thereby remove the assumption that made empiricism seem inevitable to Hume, namely, that such an explanation is impossible. 3. How is the task to be accomplished? A detailed answer to this question lies well beyond the scope of this paper, but the outline of an answer is provided by the Copernican proposition that, as was noted above, Kant introduces in w14 as the principle he will follow in showing how the categories can relate a priori to objects--the principle that the categories make experience and its objects possible. T h e doubt as to the possibility of the categories' reladng a priori to objects is to be removed by showing that given this principle, the proposition that the categories originate in the understanding and the proposition that the objects to which they purport to relate a priori are given to the understanding from without, although seemingly in conflict, are in fact in necessary agreement. This principle can yield the needed reconciliation if, as Kant argues in w 4 and w there is no other way o f conceiving how concepts originating in the understanding can relate a priori to objects, and if, when this relation is conceived according to the principle, the two apparently conflicting propositions are in systematic agreement. And it is clear, at least in outline, how this systematic agreement is supposed to be shown. For if, as the Metaphysical Deduction asserts, the categories originate in the understanding, as conditions of thought, then given the Copernican principle that such conditions make experience and its objects possible, the categories will be in agreement with those objects and not merely contingently. The categories' a priori relation to objects, which initially appeared to conflict with their origin in the understanding, will turn out to be a consequence of it. 4. Interpreting Kant's task as dialectical may appear to make it too easy.

THE

TRANSCENDENTAL

DEDUCTION

AND SKEPTICISM

379

Kant does not rule out the possibility of subjects whose sensibility has forms other than space and time, and it may be asked whether, despite the categories' purported universal validity, there can be subjects whose understanding has different pure concepts, or even no pure concepts at all. On the interpretation outlined above, the Deduction will fail to exclude the possibility of such subjects, just as a theodicy fails to exclude the possibility that God does not exist. But if the Deduction does not establish the categories' universal validity, it will not exclude the possibility that they are merely "the mode in which one's own subject is organized," which is "what the skeptic wishes most of all" (B ~68). T h o u g h the Deduction will not lead to skepticism, it will not exclude it either. If the interpretation outlined above is correct, however, the possibility of subjects that think but without the categories has already been excluded by the Metaphysical Deduction's argument that the categories have their origin in the understanding. The middle course, in conceiving of the categories as implanted dispositions, takes their origin to lie outside the understanding; by failing to recognize that they belong to the understanding's constitution, it invites the thought that there might be subjects in whom other dispositions or perhaps no dispositions at all are implanted. In the Metaphysical Deduction, on the other hand, Kant first identifies judgment as the fundamental act of discursive understanding (A69/B94) and then undertakes a systematic elaboration of the forms of judgment, from which he derives the table of categories. This derivation enables him to claim of the understanding that "it is because it contains these concepts that it is a pure understanding; for through them alone can it understand anything in the manifold of intuition, that is, think an object for intuition" (A8o/B lo6). In thus arguing that the categories are conditions of understanding, the Metaphysical Deduction, if successful, establishes that they constitute the understanding and hence excludes the possibility of a subject that thinks but without the categories. This conclusion is then vindicated by the Transcendental Deduction. Yet even if the argument of the Metaphysical Deduction is granted, a doubt may linger. The forms o f space and time constitute o u r sensibility, but not sensibility in general, so why suppose that the Metaphysical Deduction establishes anything more than that the categories constitute o u r understanding, rather than understanding in general? Simply to deny the possibility of subjects that think but without the categories would be to lapse into dogmatism. Kant does deny this possibility, but after acknowledging that such a denial will initially seem hard to accept, he gives his reason: "we must assign to things, necessarily and a priori, all the properties that constitute the conditions under which alone we think them. Now I cannot have the least representation of a thinking being through any outer experience, but only through self-

380

JOURNAL

OF T H E H I S T O R Y

OF P H I L O S O P H Y

3 2 : 3 JULY 199 4

consciousness. Objects of this kind are, therefore, nothing more than the transference o f this consciousness of mine to other things, which in this way alone can be represented as thinking beings" (A346-47/B4o4-4o5; cf. A 3 5 3 54). T h e possibility o f subjects whose forms o f sensibility are other t h a n space a n d time cannot be r u l e d o u t , because space and time are neither implicated in o u r conception o f sensibility in general (as receptivity), nor conditions o f self-consciousness, and hence not conditions u n d e r which alone we think a thinking being. But the categories are precisely such conditions (if, as the Metaphysical Deduction argues, they are conditions o f thought). So the "transference" of self-consciousness whereby I represent another thinking being is the transference o f the categories themselves, and hence they are valid for everything that thinks. Needless to say, the preceding remarks characterize the Deduction's task only in the most general terms. T h e y are only meant to open the way to an interpretation that does not impose on the Deduction the task of refuting a skeptic, but recognizes instead that for Kant, although skepticism should not be the conclusion of philosophy, neither should it be the premise. 2~

Universityof Pittsburgh
,51 am grateful to James Conant, Hannah Ginsborg, Frederick Neuhouser, Charles Parsons, John Rawls, Manley Thompson, and two anonymous readers for their detailed comments and suggestions. I also thank the participants of a conference at Rice Universityon first principles in philosophy, at which an earlier version of this paper was presented.