Kant's Theses on Existence∗
2008; Taylor & Francis; Volume: 16; Issue: 3 Linguagem: Inglês
10.1080/09608780802200729
ISSN1469-3526
Autores Tópico(s)Philosophy and Historical Thought
ResumoClick to increase image sizeClick to decrease image size Notes ∗This is a developed version of the paper, ‘A Categorical difference in Kant: Reality and Modes of Being’, which I presented at Bilkent University Kant Symposium in April 2005. I would like to thank Stephen Voss, Ilhan Inan and Barry Stocker for their very helpful comments on the earlier drafts of this paper. I would also like to thank the anonymous referee who provided me with valuable feedback in revising the paper. 1Unless otherwise stated, all references to these two works of Kant will be to the following English translations: Immanuel Kant, The One Possible Basis for a Demonstration of the Existence of God=Der einzig mögliche Beweisgrund (hereafter The One Possible Basis), English and German, translated by Gordon Treash (Lincoln and London: University of Nebraska Press, 1979); Immanuel Kant, Critique of Pure Reason (hereafter Critique), translated and edited by P. Guyer and A. W. Wood (Cambridge and New York: Cambridge University Press, 1998). As is standard, references to the Critique are to the pages of the first (A) and second (B) edition. Kant also presents an almost full repetition of his account of the refutation and the thesis in the Critique in his ‘Lectures on the Philosophical Doctrine of Religion’ (1783–1784), edited within Religion and Rational Theology, translated and edited by A. W. Wood and G. Di Giovanni (Cambridge and New York: Cambridge University Press, 2001). 2Kant briefly repeats his views on existence in some of his other post-Critique lectures as well. See, for example, the ‘Ontology’ sections of Metaphysik Mrongovius (1782–1783) and Metaphysik L2 (1790–1791?) in Immanuel Kant, Lectures on Metaphysics, translated and edited by K. Ameriks and S. Naragon (Cambridge and New York: Cambridge University Press, 1997). Hereafter, references to Religion and Rational Theology and Lectures on Metaphysics will be, as provided by the Cambridge edition, to the Academy edition, Kant's gesammelte Schriften, Vols 28 and 29 (Berlin: Walter de Gruyter & Co., 1900–). In Lectures on Metaphysics too Kant tends to make his entry into the issue of existence with reference to the existence of God: ‘This concept, although it is simple, is still quite difficult, because we apply it to concepts which are sublime beyond all experience and example. E.g., to the concept of God’ (28: 554). 3At the end of his refutation in the Critique, Kant explicitly states that what he demonstrated to be ‘only so much trouble and labor lost’ is the Cartesian version of the ontological argument. There he also mentions Leibniz's name as another representative of the ontological argument that failed to prove a priori the possibility of God (Critique, A602/B630). We can conclude that although Kant has in mind Descartes's version as the general axis of the argument to assault, he also considers Leibniz's claim to complete the former with a modal modification. On the other hand, nowhere in his whole corpus of works does Kant mention Anselm's original argument. Some writers claim that he knew nothing of the latter. See, for instance, Chapter 10 of Charles Hartshorne, Anselm's Discovery: A Re-Examination of the Ontological Proof for God's Existence (La Salle: Open Court, 1965). For Descartes's argument, Chapter 5 of René Descartes, Meditations on First Philosophy in Selected Philosophical Writings, translated by John Cottingham, Robert Stoothoff, Dugald Murdoch (Cambridge: Cambridge University Press, 1988); for an extensive discussion of Leibniz's enduring efforts to develop a modal version of the ontological argument, see also Chapters 4 and 8 of Robert Merrihew Adams, Leibniz: Determinist, Theist, Idealist (hereafter Leibniz: DTI) (New York and Oxford: Oxford University Press, 1998). For Anselm's version, see Anselm, Proslogion in The Many-faced Argument, edited by J. H. Hick and A. C. McGill (New York: Macmillan, 1967) 4–6. 4In The One Possible Basis, and also with minor differences in some of his other precritical works such as Nova Dilucidatio (1755) and Inaugural Dissertation (1770), Kant offers an alternative ontological proof that does not rely on the containment of existence as a predicate in the concept of God and its being a real predicate. For the English translations of Nova Dilucidatio and Inaugural Dissertation, see Immanuel Kant, Theoretical Philosophy, 1755–1770. Briefly, Kant argues that for there to be any possibility, its content or what is thought in it must be previously given by something that actually exists, either in it as a determination of it or through it as a consequence (see The One Possible Basis, 83). Therefore, there must be an absolutely necessary being that grounds not only the existence but even the thought or possibility of things in general. In his categorical denial of the possibility of a theoretical proof of the existence of God, Kant tacitly dismisses his own precritical proof together with all the traditional ones. I think the shift in Kant's understanding of modality in the critical period justifies this categorical denial to some extent (see also n18); but as his own proof's line of inference is radically different from the other traditional proofs that he claims to refute and his objections to the latter do not in the least apply to the former, it is still curious that he never designs a separate refutation for his own proof. For an extensive discussion of what might be a justification for Kant's rejection of his own proof, see Mark Fisher and Eric Watkins, ‘Kant on the Material Ground of Possibility: From The Only Possible Argument to the Critique of Pure Reason’,The Review of Metaphysics, 52 (1998) No. 2: 369–95. 5See Charles Hartshorne, Anselm's Discovery: A Re-Examination of the Ontological Proof for God's Existence and again his ‘Is the Denial of Existence Ever Contradictory’, The Journal of Philosophy, 63 (17 February 1966) No. 4: 85–93; Norman Malcolm, ‘Anselm's Ontological Arguments’, The Philosophical Review, 69 (January 1960) No. 1: 41–62; Alvin Plantinga, ‘Kant's Objection to the Ontological Argument’, The Journal of Philosophy, 63 (October 1996) No. 19: 537–46; Lenn E. Goodman, God of Abraham (New York: Oxford University Press, 1996). 6See Critique, A593/B621. 7The refutation of the ontological proof in the Critique is the fourth section of a chapter called ‘The Ideal of Pure Reason’, which is one of the most important parts of the whole Critique and a very difficult one to interpret. In the previous three sections, Kant explains how the idea of God as an ens realissimum, an individual being that contains the sum total of all possible predicates of things, is generated by pure reason as a necessary consequence of one of its natural procedures, and is then speculatively matched with the idea of an absolutely necessary being. The difficulty of interpretation arises from the dubiousness of the mentioned part, which allows different ways of reading, namely as a critique of a natural tendency of pure reason itself, as a critique of traditional speculative theologies in general and as a critique of a particular, i.e. Leibniz's, notion of God. I believe that all three readings can be accepted simultaneously. For reasons of economy, I would like only to point out here that the most significant accomplishment of this chapter, which also includes the refutations of the cosmological and what Kant calls the physico-theological proofs, is reformulating God as a transcendental presupposition necessary for the function of the faculty of understanding; and yet being a mere ideal of pure reason, God is, strictly speaking, not an object of possible experience and His existence cannot be proved whatever theoretical means is used. This is a major step in the whole critical project to secularize the concept of God and to relocate it in the practical domain as a regulative principle, away from the need for any ontological argument for His existence. 8My emphases, Critique, A594/B622. 9 Critique, A595/B623. 10In Kant's language, the broader sense of the term ‘objective reality’ (objekive Realität) refers to the applicability of a concept to an object of intuition in general, either empirical or pure. This sense of the term captures Kant's precritical notion of the real element of possibility which denotes, without an empirical modal commitment, the data or material that is represented through a logically possible concept (see also n18). However, since Kant's intention in using the term here, as will be seen in the next section, is to underline the distinction between the ‘logical possibility of concepts’ that can be tested through mere concepts and the ‘real possibility of things’ that can be tested only with an appeal to the ‘principles of possible experience’, my impression is that in the present context ‘objective reality’ has a somewhat more empirical emphasis and refers to a concept's applicability to possible objects of empirical experience. Heidegger and Hanna prefer to use even stronger language in their definitions; the former identifies objective reality simply with actuality or existence, the latter takes it to be the ‘reference or applicability’ of a representation to ‘actual, real, or existing objects’. See Heidegger, Basic Problems of Phenomenology (hereafter Basic Problems), translated by Albert Hofstadter (Bloomington: Indiana University Press, 1982) 34; and Robert Hanna, Kant and the Foundations of Analytic Philosophy (Oxford: Clarendon Press, 2001) 84. Nevertheless, I believe that the sense of objective reality I suggest above serves better to keep the spirit of the distinction between ‘possibility’ and ‘actuality’ of objects to which Kant refers in the Postulates. 11See Norman Malcolm, ‘Anselm's Ontological Arguments’, 58. For Kant's claim which forms the basis of Malcolm's objection, see Critique A595/B623. 12Malcolm suggests that Kant's real view cannot be that ‘necessity is properly predicated only of propositions (judgments) not of things’. As for his ground of objection, he refers to Kant's discussion of ‘The Postulates of Empirical Thought in General’, where Kant establishes the criterion of necessary existence. See Malcolm's n33, in ‘Anselm's Ontological Arguments’. I will try to show how inaccurate Malcolm's understanding of the postulate of necessity is in the final section. 13James Van Cleve, Problems from Kant, (New York and Oxford: Oxford University Press, 1999) 192. 15Ibid., A597/B625. 14For Kant, the ens realissimum notion of God and the nominal definition of God as that whose non-being is impossible are generated by two distinct procedures of pure reason, the former being the consequence of pure reason's drive for the complete or thoroughgoing determination of things, the latter being the consequence of the drive to reach the unconditioned in the series of conditions. Kant humourously narrates the story of the speculative match that pure reason makes between these two originally distinct notions: ‘First it convinces itself of the existence of some necessary being. In this it recognizes an unconditioned existence. Now it seeks for the concept of something independent of all conditions, and finds it … in that which contains all reality’ (ibid., A587/B615). However, although he thinks that the notion of an absolutely necessary being is fundamentally confused, he holds the view that the notion of a most real being, as will be seen, is logically possible. 16 Critique, A596/B624. 17Kant, in the present context, does not explicitly make a distinction between positive and negative predicates. He rather seems to take ‘predicates’ as positive attributes that make up the real content of things: ‘Logical negation … is never properly attached to a concept, but rather only to its relation to another concept in a judgment, and therefore it is far from sufficient to designate a concept in regard to its content’ (ibid., A574/B602). In the case of the idea of ens realissimum, as it contains all reality, it lacks no real content and thus must have only ‘positive’ predicates. That the ens realissimum has the sum total only of positive predicates can be further confirmed by resorting to Leibniz, whose notion of God, I believe, Kant considers here. However, the question about the ens realissimum is whether containing only positive predicates entails freedom from internal contradiction and thus ensures logical possibility, because there are positive predicates that are not logically opposite to each other, and yet cannot be contained in the same subject. For Kant's solution to this problem that preoccupied Leibniz for a long time, see n43. For the moment, as the logical criterion of possibility is only a formal one and abstracts from all content, it is still not wrong to conclude that the concept of a being that contains only positive predicates is logically possible. 18This distinction is a good example of how Kant amended his precritical understanding of modality in the Critique. In The One Possible Basis, Kant makes an apparently similar distinction in the concept of possibility, between what he calls the formal or logical element and the real element. The formal element of possibility, in conformity with the logical possibility of concepts, refers to the Leibnizian notion of possibility that is constituted by freedom from contradiction alone. On the other hand, the real element of possibility, which Kant borrowed from Baumgarten, refers to the givenness of the material or data of possibility through which what is to be possible is thought or represented. However, this wide field of possibility which applies to all that is representable, given in whatever way, is narrowed in the Critique, by imposing not only conceptual but also sensible conditions, i.e. space and time, on the givenness and consequently on the real possibility of things. (As possibility itself is one of the modal categories, this further, critical constraint on the notion of possibility goes along with the requirement that categories must be applicable to sensible intuition to provide real cognition of things.) As discussed in the final section of this essay, what is possible is thus identified with a possible object of experience. This is how the precritical notion of the real element of possibility is developed into the notion of real possibility in Kant's critical programme of modality. For the precritical distinction, see The One Possible Basis, 67. For Kant's terminological connection with Baumgarten, see also Heidegger, Basic Problems, 34–5, and the translator's introduction to The One Possible Basis, 19–21; and for Kant's specific indebtedness to Baumgarten for the form of the ontological argument that Kant claims to refute, see Charles Hartshorne, Anselm's Discovery: A Re-Examination of the Ontological Proof for God's Existence, Part Two, Chapter 10. 19 Critique, A597/B625. This surely is not a contradiction in the strict, logical sense of the term. What Kant means, I think, is that while through the mere concept of a thing what could only be tested is the concept's logical possibility which would not even suffice to prove the real possibility of the thing itself, introducing existence as a predicate into the concept of a thing and thereby making an existential commitment for the thing in question would be falling into a fundamental confusion concerning the use of modal concepts. 20See Anselm, Proslogion (Chs 2–4) in The Many-faced Argument, 4–6. 21‘[I] am not free to think of God without existence (that is, a supremely perfect being without a supreme perfection) …' (Descartes, Meditations on First Philosophy in Selected Philosophical Writings, 107). 22Heidegger emphasizes the radical character of Kant's objection: ‘Kant's thesis … does not assert merely that existence cannot belong to the concept of the most perfect being … It goes further. It says, fundamentally, that something like existence does not belong to the determinateness of a concept at all’ (Heidegger, Basic Problems, 32). 23The problem of the possibility of an existence producing machine is first pointed out by Gaunilo, the immediate critic of Anselm's argument. He argues that if the minor premise of Anselm's argument that ‘that which exists is more excellent than that which stands in relation to my understanding only’ is accepted, then the mere idea of the most excellent ‘lost island’, by virtue of its perfection, would suffice to prove that it necessarily exists. See Gaunilo and Anselm, ‘Criticism and Reply’ in The Many-faced Argument, 22–3. There are various modern objections to Gaunilo's argument on the ground that the idea of a greatest possible island is neither analogous to the concept of God, nor a consistent idea for there is no end to the greatness of an island. See, for instance, Alvin Plantinga, God, Freedom and Evil (New York: Harper & Row, 1977) 91. However, as the purest form of the syllogism in the minor premise of the ontological argument exhibits, the problem here is the very introduction of existence into the mere concept of something, not the way in which or, as Kant would say, the disguise under which this introduction is conducted. Thus, even if he may be said to fail to give a convincing example, Gaunilo's anticipation of Kant's objection is still impressive. 24 Critique, A597/B625. 25Ibid., A7/B11. 26Ibid., A598/B626. 27Although it has never been as popular as his thesis that existence is not a real predicate, Kant's claim to the categorical syntheticity of all existential propositions has received serious criticisms from the literature on the refutation. One specific criticism that has been frequently repeated by not only the proponents but also the opponents of the ontological argument is that the latter thesis is not compatible with the former one. For only three of the typical examples of this sort of allegation, see Jerome Schaffer, ‘Existence, Predication, and the Ontological Argument’, Mind, New Series, 71 (1962) No. 283: 309: Kant defines [real predicate] as something ‘which is added to the concept of the subject and enlarges it’. This is a most unfortunate definition for Kant to use, however, since it leads to contradiction with another important doctrine of his, that existential propositions are synthetic. Nicholas Everitt, The Non-Existence Of God (London and NewYork: Routledge, 2004) 52: This appears to commit him to saying that ‘exists’ does enlarge the subject concept, and hence that ‘exists’ passes both tests for being a real predicate… Indeed, it seems that Kant's own words commit him to denying the thesis as well as asserting it; and Lenn E. Goodman, God of Abraham, 67: ‘His claim that all existential propositions are synthetic is inconsistent with his thesis that “exists” is not a real predicate’. In the second part, I will discuss how this way of identifying the syntheticity of propositions with the predicate's enlarging of the subject or its being a real predicate fails to capture the true senses of Kant's notions of syntheticity and existence. 28 Critique, A598/B626. Norman Kemp-Smith translates the original term ‘Bestimmung’ as ‘determining predicate’; compare the same quotation in Immanuel Kant, Immanuel Kant's Critique of Pure Reason, translated by Norman Kemp-Smith (New York: Palgrave, 1929). The idea of determination being a real predicate that adds to the subject concept comes from pure reason's principle of thoroughgoing or complete determination of things: ‘Every thing … as to its possibility … stands under the principle … according to which, among all possible predicates of things, in so far as they are compared with their opposites, one must always apply to it; (Critique, A572/B600). The principle is in fact a manifestation of what is required to know things completely with their determinate contents. To know something completely, one has to know the sum total of all possible predicates (the idea which brings about the concept of ens realissimum, of the individual being that is determined by this idea alone) and determine the thing through them, either affirmatively or negatively (see ibid., A573/B60). Real or determining predicates are therefore ones that contribute to this ideal process of thoroughgoing determination. 31Henry E. Allison, Kant's Transcendental Idealism: An Interpretation and Defense (New Haven and London: Yale University Press, 1983) 71; my emphasis. In some of his critical works Kant prefers to call real predicates synthetic predicates, and thereby applies the analytic/synthetic distinction to predicates themselves in a way that allows Allison's interpretation. See, for example, Lectures on Metaphysics, 28: 552: ‘We call determinations not analytic predicates [praedicata analytica] but rather synthetic predicates [praedicata synthetica]'. However, as will be discussed in the rest of this essay, given the exceptional status of modal propositions, the syntheticity of propositions does not warrant qualifying their predicates as real, determining or synthetic predicates. Kant puts this clearly: ‘… we have introduced the categories: possibility, actuality, and necessity, and then we deemed that they are not at all determinations of a thing, or synthetic predicates’ (ibid., 29: 822). 29Sebastian Gardner, Kant and the Critique of Pure Reason (London: Routledge, 1999) 239. 30See Heidegger, Basic Problems, 33–4. 32 Critique, A599/B627. 33Compare the mentioned quotation in Immanuel Kant, Kritik der reinen Vernunft (Werkausgabe Band IV), (Frankfurt am Main: Shurkamp, 1996): ‘Sein ist offenbar kein reales Prädikat …’. 34 The One Possible Basis, 57; my emphasis. Compare ibid., 56: ‘Das Dasein ist gar kein Prädikat oder Determination von irgend einem Dinge’. This shortcut formulation of Kant might have been thought as the textual source of the widely exemplified misconception of his thesis in the literature, i.e. that it is not a predicate at all; but then in the same text, Kant goes on to explain the sense in which existence may be used as a predicate. 35See n28. 36Heidegger grounds his interpretation of Kant's thesis on the importance of this sense of ‘reality’, which, he claims, is adopted by Kant from Scholastic terminology: When Kant talks about the omnitudo realitatis, the totality of all realities, he means not the whole of all beings actually extant but, just the reverse, the whole of all possible thing-determinations, the whole of all thing-contents or real-contents, essences, possible things. Accordingly, realitas is synonymous with Leibniz's term possibilitas, possibility. Realities are the what-contents of possible things in general without regard to whether or not they are actual, or ‘real’ in our modern sense; (Heidegger, Basic Problems, 34) For a similar articulation of the distinction between reality and existence in Kant's language, see also Wolfgang Schwarz, ‘Kant's Categories of Reality and Existence’, Philosophy and Phenomenological Research, 48 (December 1987) No. 2: 343–6. As Heidegger finds it synonymous with the Leibnizian notion of possibility, I think this sense of reality is better captured by Kant's precritical notion of ‘real element of possibility’ that I alluded to in n18. Just like the real element of possibility, it designates a neutral notion of possibility. It is free of any modal reference in the critical sense; that is, it does not involve a reference to the agreement of the object with the conditions of experience in general, but it refers solely to the data or content that is represented by the concept. Traces of this precritical neutral notion of possibility are still present in the Critique, and at times it becomes a really taxing business to differentiate the sense in which Kant uses the term ‘possibility’. 37Richard Campbell, in his ‘Real Predicates and “Exists”‘, suggests a misleading reading of Kant's discourse of real or determining predicates. Campbell's idea is that ‘although the notion of a determining predicate is to be understood relative to a given judgment’, the notion of a real predicate is not. He obviously assumes a distinction between the meanings of the two terms, which, I believe, Kant uses interchangeably to refer to the same notion. Without giving a satisfactory account of the distinction he has in mind, he bases his own, revised definition of a real predicate upon this alleged distinction: ‘… a real predicate is one which is apt to serve as a determining predicate’, or again in the footnote to the former definition, ‘one which could add a determination in some judgment’ (96). I have the impression that Campbell takes ‘real predicate’ as a broader notion to mean a concept that has a capacity to function as a determining predicate, whether or not it actually does so. That is probably why he thinks the notion of a real predicate, in contradistinction with that of a determining predicate, is not relative to a propositional context. However, besides introducing an unnecessary and unexplained distinction between real and determining predicates, Campbell understands Kant's negative thesis concerning existence as a positive definition of a real predicate, and thus reads the ‘could’ in ‘Being is not … a concept of something that could add to the concept of a thing’ as a capacity that every real predicate, in isolation from any propositional context, must have. I suggest, however, that the ‘could’ in question refers to the context-independent incapacity of ‘being’ or ‘existence’ to add to the concept of a thing as a predicate, and thus classifies existence as a non-real predicate, tout court. On the other hand, a real or determining predicate is one which actually adds (a further determination) to the concept of a thing in a given proposition. What Campbell defines as a concept that could or is apt to add a determination to the concept of a thing, without any mention of a propositional context, may correspond in my account only to what I classified above as a potentially real or determining predicate. See Richard Campbell, ‘Real Predicates and ‘Exists’, Mind, New Series, 83 (January 1974) No. 329: 95–9. 38See Heidegger, Basic Problems, 29–34. 39Heidegger devotes the entire second chapter of his Basic Problems to the history of this assumption. 40This distinction can be traced back to Aristotle and his early medieval commentators such as Aquinas. At many places in his Metaphysics, Aristotle distinguishes the categories as the figures of predication, or more precisely, as the attributes that constitute the essence of an individual thing, from the other ways in which a thing is said to be, i.e. being accidentally, being as truth, being potentially or actually. See especially, Aristotle, Metaphysics in The Complete Works of Aristotle, edited by Jonathan Barnes. (New Jersey: Princeton University Press, 1984) 1026a 33ff. Aquinas, in his On Being and Essence, clarifies that the term ‘essence’ applies only to the first sense of being that is divided into ten categories, and points out that as it is what is signified by the definition of what the thing is, essence is also called by the name ‘whatness;’ see St Thomas Aquinas, Selected Writings, translated and edited by Ralph McInerny (London: Penguin Books, 1998) 31. In his Lectures on Metaphysics, Kant himself repeatedly draws this categorical distinction between essentia and existentia, or what and modes of being and calls modes ‘extraessential’ properties that do not belong to the essence of a thing; see, for example, ibid., 28: 553. 41 Critique, A80, B106. 42‘[T]ranscendental affirmation … is called reality (thinghood)’ (ibid., A574/B602). 43Kant makes a warning about the original function of the ens realissimum in the derivation of the particular possibilities of things: The derivation of all other possibility from this original being … cannot be regarded as a limitation of its highest reality and as a division, as it were of it; for then the original being would be regarded as a mere aggregate of derivative beings … Rather, the highest reality would ground the possibility of all things as a ground and not as a sum total; the manifoldness of the former rests not on the limitation of the original being itself, but on its complete consequences. (ibid., A579/B607) Kant is aware that his previous discourse of ‘possessing/containing all reality’ may lead the reader to a Spinozistic notion of God, which he thinks, as containing all realities within itself, suffers from real internal conflicts. Kant addresses this problem of real conflicts in the Amphiboly, where he directly criticizes Leibniz: the principle that realities (as mere affirmations) never logically oppose each other … signifies nothing at all either in regard to nature nor overall in regard to anything in itself (of this we have no concept). For real opposition always obtains … i.e., where one reality, if combined in one subject with another, cancels out the effect of the latter (ibid., A273/B329) This problem, which Kant has had in mind even in the precritical period, was what required a revision in the Leibnizian–Wolffian notion of possibility, whose only criterion is freedom from logical contradiction; hence the Baumgartenian distinction between the logical and real element of possibility that I explained in n18. However, Kant's ground notion of God, together with the notion of particular possibilities (or realities) as its consequences, avoids such real oppositions. In The One Possible Basis, Kant proposes a similar solution by dividing possibilities into two classes with respect to their relation with God, namely the ones that belong to its own determinations and the ones that belong to its consequences. See, ibid., 83–5. Here, it is also useful to keep in mind the distinction between the idea of the sum total (of all reality) and the ideal of the most rea
Referência(s)