Appendix IV

How To Be a Theist without Proving that God Exists

 

 

1. Kant’s Critique of the Traditional Theoretical Arguments

          As mentioned in IV.1, Kant groups all theoretical arguments for the existence of God under three headings: the ‘ontological’, ‘cosmological’, and ‘physicotheological’ types.[1] Given his preference for fourfold divisions, we might expect Kant to have exposed not three Ideals of reason, but four, especially because such a 2LAR would accord well with the fourfold divisions of the paralogisms and anti­nomies. A fourfold division of ideals would bring the Dialectic of Kt1 into perfect conformity with the structure of the categories, making it a com­plete architectonic system of the twelve basic illusions into which the speculative metaphysician can fall when reasoning about questions concerning ultimate reality. Despite his apparently threefold division, Kant does not disappoint our architectonic ex­pec­ta­tions. For the ‘Ideal of Pure Reason’ [Kt1:595-670] begins by ex­posing the il­lusion inherent in a fourth type of proof, originated by Kant himself nearly twenty years earlier, in Kt15. Especially relevant in this respect is Section III of Kt15 [155-63(355-66)], where Kant describes the systematic relation between the four types of arguments in terms of a perfect 2LAR.

          This special form of argument, based on the ‘complete deter­mi­nation’ of em­pirical objects and sometimes referred to as the ‘possibility proof’ [e.g., Wo78:40-79], is discussed separately in Kt1 not only because it cannot be clas­sified as ‘traditional’, but also because Kant believes it forms the theo­ret­i­cal basis for—i.e., ‘the only possible ground for the demonstration’ [cf. Kt15’s title] of—the other three proofs [cf. Go71:84-5n]. Since the character of this type of proof is not as well known as that of the other three, I shall begin by dis­cus­sing it in more detail than the other three. In preparation for this account, let us examine the logical relationship between the four types of proof.

          Kant summarizes the difference between the three traditional proofs by pointing out the different theological perspectives they assume: ‘ontotheology ... considers God merely in terms of concepts ... Cosmotheo­logy presupposes ... the existence of a world in general. And finally, physicotheology makes use of experience of the present world’ [Kt26:1003(31-2); s.a. Kt1:633,660]. These three forms of ‘speculative theology’ cor­respond directly and unambigu­ously to the logical, transcendental, and empirical perspectives, respectively.[2] Kant’s own possibility proof, as we shall see, corresponds in a similar fashion to the specula­tive/hypo­thet­ical perspective. Thus we can clarify the sys­tematic relationship between these types of proof by plot­ting them onto the 2LAR model of Kant’s four perspectives,[3] as fol­lows:

 

 

Figure AIV.1:

The Four Types of Theoretical Argument for God’s Existence

 

          Kant defines the relationship between these four arguments as a perfect 2LAR in Kt15: 155-62(355-65), where he explains how ‘the four imaginable arguments’ can be ‘reduced to two principle sorts’ [162(365)], as represented in Figure AIV.1 by the vertical and horizontal axes. He explains the logical structure of their relationship in Kt15:155-6(356): ‘All argu­ments for the exis­tence of God can be taken but either from the intellectual conceptions of the merely possible [+], or from the conceptions of the existing [-] taken from ex­perience.’ The former type either treats ‘the possible as a ground’ and ‘the exis­tence of God as a consequence’ (-) or else treats ‘the possible as a conse­quence’ and ‘the Divine existence as a ground’ (+). The latter type argues either ‘from that, whose existence we experience’ to ‘the existence of a first and inde­pendent cause’ (-) or else ‘from what experience teaches are immediately in­ferred’ to ‘his existence and his attributes’ (+). In Wa63a:267 Walsh clarifies this fourfold distinction (without referring to Kt15), by stating a simpler version of the two 1LARs that generate this 2LAR: (1) the ontological and pos­sibility proofs are based on logical (+) grounds, whereas the cosmological and physicotheological proofs are based on causal (-) grounds; and (2) the physi­cotheological and possibility proofs begin ‘from a truth about the world’ (+), whereas the cosmological and ontological proofs begin ‘from an idea’ (-).[4] Us­ing this model [i.e., Fig. AIV.1] as a map to guide our steps will enable us to steer our way swiftly through Kant’s discussion of the errors committed by each type of ar­gument.

          Kant hints at his possibility proof as early as 1755. For he claims in Kt11 that all

 

reality must be united in a single being.... God ... is given as the absolutely neces­sary principle of all possibility.... [For] in God, the source of all reality, there [must exist] everything that is in a real notion.... Among all beings God is the only one in whom existence is prior, or, if you prefer, identical with possibility. And no notion of that possibility remains as soon as you separate it from his exis­tence. [394-6(224-5)]

 

He fully develops this notion in Kt15, where he argues [83(247)]: ‘All possi­bility presup­poses something actual, wherein and whereby all that is cogitable is given. There is therefore a certain actuality, whose annulling would annul even all internal possibility in general.’ Since ‘all that exists is thoroughly de­termined’ [85(250)], there must be a ‘necessary Being [who] compriseth the last real ground of all other possibility’ [86(253)]. As Wood puts it, Kant’s possibility proof contends ‘that the idea of God takes its rational origin from the fact that it is presupposed by any attempt to think of individual things in general as thoroughly determined, and hence as absolutely possi­ble.’[5]

          Kant explains ‘the principle of complete determination’ in Kt1:599-600 by saying: ‘if all the possible predicates of things be taken together with their con­tradictory opposites, then one of each pair of contradictory opposites must be­long to [any given thing]. This principle ... concerns, therefore, the content [of our knowledge of things], and not merely the logical form.’ Although we can never ‘know every possible [predicate]’ of a thing, and so can never exhibit its complete determination in concreto [601, t.b.], the idea of complete determina­tion, when applied to ‘the concept of an individual object’, becomes ‘an ideal of pure reason’ [602]. This ideal is ‘the concept of an individual being’ who ‘possesses all reality’ (i.e., ‘the concept of an ens realissimum’); hence it is ‘the concept of a thing in itself as completely determined’.[6] As such it ‘serves as basis for the complete determination that necessarily belongs to all that exists.’ But reason presupposes

 

only the idea of such a being, and this only for the purpose of deriving from an un­conditioned totality of complete determination the conditioned totality, that is, the totality of the limited. The ideal is, therefore, the archetype (prototypon) of all things, which one and all, as imperfect copies (ectypa), derive from it the material of their possibility ... [605-6]

 

This means the possibility of all individual things, if we are to think of them as thoroughly determined objects of knowledge, must be derived from ‘limitations of a greater, and ulti­mately of a highest, reality’: viz., ‘the possibil­ity of that which includes in itself all reality’, the ens realissimum [606; s.a. 609]. This, Kant concludes, is ‘the concept of God’, which, taken as an ‘ideal of pure reason ... is thus the object of a transcendental theology’ [608].

          Kant does not want to criticize this argument too severely, for, as we saw in IV.2, his affirmation of God as an ‘idea’ is in some respects dependent on the notion of God developed here, but proceeds on hypothetical rather than logical (i.e., speculative) grounds. In this sense Wood is correct to say [Wo78:59] ‘it was precisely Kant’s critical doctrines ... which made it possible for him to accept the concept of God as an ens realissimum.’ The problem with the spec­u­­lative form of the argument is that it ‘hypostatises’ the idea of God by assuming it can be used to determine the nature of God as a real individual object of knowledge, whereas its proper (practical) use is only to clarify the concept of God, ‘without requiring that all this re­ality be objectively given and be itself a thing. Such a thing is a mere fiction’ when hyposta­tized as an object (i.e., a constituent element) in systemt.[7] Kant blames this illusion on a ‘transcendental subreption’ [Kt1:611], according to which a person substitutes ‘dialectically for the distributive unity of the empirical perspective, the collective unity of experience as a whole’ [610]—i.e., of the transcendental perspective. Despite its tendency to confuse these two per­spectives, Kant’s possibility proof does establish, as it were, the grounds for speculation: it demon­strates ‘that reason requires’ the concept of an ens realis­simum [620] as ‘the primal concept in thought’ [Kt69:302], but it establishes no conclusions whatsoever as to its actual existence. Although it is not techni­cally a theological proof, it sets the stage for each of the three traditional proofs by establishing, from the speculative/hypo­thetical perspective, a defi­nition of the theoretical concept of God that is assumed by the three traditional arguments.

          The general characteristic of ontological arguments, according to Kant, is that they attempt to prove the existence of ‘something the non-existence of which is impossible.’[8] In their simplest form such arguments suggest that the very concept of God as ‘an absolutely necessary being’ [Kt1:620] necessitates God’s existence. Kant criticizes this simple form as ‘arguing from the logical possibil­ity of concepts to the real possibility of things’ [624n; s.a. 626]. If the term ‘necessary’ refers to a judgment, then its object may be quite intelligible, as, for example, in the case of geometrical judgments such as ‘a triangle has three an­gles’. ‘But’, warns Kant, ‘the unconditioned [logical] necessity of judgments is not the same as an absolute [real] necessity of things’ [621]. On the contrary, the former ‘is only a conditioned necessity of the thing’. Thus, the above judgment does not imply ‘that three angles are absolutely necessary, but that, under the condition that there is a triangle (that is, that a triangle is given), three angles will necessarily be found in it’ [622]. The rather obvious point being made here is that if we ‘reject the predicate [e.g., ‘is omnipotent’] while retain­ing the subject [e.g., ‘God’], contradiction results ... But if we reject subject and predicate alike, there is no contradiction; nothing is then left that can be contradicted.’[9] ‘The concept of necessi­ty’, Kant therefore concludes, ‘is only to be found in our reason, as a formal condition of thought; it does not allow of being hypostatised as a material condition of existence’ [648].

          The only option for the proponent of the ontological argument, Kant con­tends, is to ground it in the concept of God as ens realissimum, established by the possibility proof [Kt1: 614,624]. The argument would then assert that in this case real existence is ‘contained in the concept of a thing that is possible’ because ‘“all reality” includes existence’ [624]; a God who is the ens realissi­mum, the most real being, must exist. It is in response to this particular ver­sion of the ontological argument, and not to all versions whatsoever, that Kant makes his famous proclamation that existence is not a predicate. There are in­stances, of course, where existence is a predicate: any empirical acknowledg­ment of a thing’s existence does add some­thing to our concept of a thing—it adds an intuition, enabling us to make a synthetic judg­ment about that thing (e.g., that it exists). But the ontological argument, we must remember, presup­poses the logical perspective, where ‘is’ is used only in analytic judg­ments, and so, never serves as a predicate.[10] Viewed from the logical perspective, ‘is’ serves either a real func­tion by indicating ‘merely the positing of a thing’, or a logical func­tion by acting as ‘the copula of a judgment’ [Kt1:626], but it can never convey reality itself [s.a. 667]. This is why Kant says ‘“Being” is obviously not a real predicate’ [626, e.a.]: nothing is really real from the logical perspec­tive, for when one’s attention is limited wholly to concepts, ‘the real’, as a characteristic attributed to a concept, ‘contains no more than the merely possi­ble’ [627]. ‘Hence the conception of an absolutely necessary being, while doubtless an indispensable idea of reason, is for human understanding an unattainable prob­lematic conception’ [Kt7:402].

          For the sake of criticizing their general character, Kant reduces all cosmo­logical argu­ments to a single form: ‘If anything exists, an absolutely necessary being must also exist. Now I, at least, exist. Therefore an absolutely necessary being exists’ [Kt1:632]. The minor premise is an empirically observable fact whose inner certainty is not open to serious doubt. But the major premise depends on the assumption of a series of conditioned existents, caused by other conditions, and ending in some necessary uncondi­tioned existent [cf. 649-51]. This proof, Kant says, appears to be ‘the most convincing’ of all the proofs, for it ‘sketches the first outline of all the proofs in natural theology’ [632]; yet for this very reason it rests on even more ‘pseudo-rational principles’ than the others [634]. He criticizes rather briefly four such ‘deceptive principles’ [637-8]: (1) the category of causality does not ‘enable us to advance beyond the sensible world’ [cf. 663-4]; (2) rather than leading to a ‘first cause’, the em­pi­rical series of conditions may well be infinite; (3) the unconditioned cannot be regarded in terms of ‘real necessity’, since the ‘removal of all the conditions’ would eliminate the only context in which the con­cept of real necessity has any meaning;[11] and (4) ‘logical possibility’ and ‘transcendental possibility’ must not be confused. But his major criticism is that this ar­gument, which ‘professes to lead us by a new path’, ends up requiring for its validity the on­tological argu­ment [637], with its con­cept of a necessary being as the ens realissimum [636].

          Kant’s criticism here is not that the cosmological argument is based on the ontological, but that the former must be used in conjunc­tion with the latter in order for its conclusion to establish knowledge of God (just as an appearance must be conceptualized before it can become an object of knowl­edge): ‘although the cosmological proof presupposes an experience in general, ... it soon aban­dons this guidance and relies on pure concepts alone.’[12] Kant aptly describes the mutual interdependence of these two types of argument in Kt1:640: ‘The whole problem of the transcendental ideal amounts to this: ei­ther, given [real] absolute necessity, to find a concept which possesses it [as in the ontological argument], or, given the concept of something, to find that something to be absolutely necessary [as in the cosmological argu­ment].’ But neither of these tasks can be fulfilled without conflating the transcendental and logical perspec­tives, and thus giving rise to dialectical illusion. Once he realized the perspecti­val illusion inherent in this type of argument, Kant came to re­gard its object as ‘not a rational entity’, but ‘a fictitious logi­cal entity’ [Kt7:468]. He therefore concludes in Kt1:630 [t.b.] that ‘we can no more extend our stock of [theoreti­cal] insights by mere ideas, than a merchant can better his position by adding a few noughts to his cash account.’ If, how­ever, we leave the logical and transcendental perspec­tives and allow our particu­lar experiences of the world to serve as a major factor in the argument, perhaps more satisfactory re­sults can be obtained.

          As long as we regard them as purely theoretical proofs intended to es­tablish certain truths about God, such arguments as we have dis­cussed so far, Kant warns, ‘will never descend from the schools and enter into every-day life or be able to exert the smallest influence on or­dinary healthy intelligence’ [Kt7:476]. The final type, however, is somewhat closer to home, for the ‘physico­theological’ (better known as ‘teleological’) argument makes use of our ‘de­ter­m­inate experience ... of the things of the present world’ [Kt1:648], and so proceeds entirely ‘by the empirical road’ [656]. Kant includes here arguments from particular causes to a ‘supreme cause’ [s.e. 649-51], which would nowa­days be considered by many to be cosmolog­ical.[13] However, the main empha­sis of the argument [cf. 653-4,660] is that the immeasurable ‘variety, order, purposiveness, and beauty’ of the world require us to ‘assume some­thing to support it’ [650]—viz., ‘a supreme Author’ [652]. Kant has a deep respect for this argument, ‘the oldest, the clearest, and the most accordant with the common reason of mankind’ [651], and believes it would be ‘utterly vain to attempt to diminish in any way [its] authority’.[14] But it operates ‘in accordance with the principles of analogy’, so it can never establish any­thing greater than ‘probability’ [654], though it can help strengthen the other arguments [651-2]:

 

The physico-theological proof, as combining speculation and intuition, might there­fore perhaps give additional weight to other proofs (if such there be); but taken alone, it serves only to prepare the understanding for theological knowledge, and to give it a natural leaning in this direction, not to complete the work in and by it­self. [665]

 

Thus it is doomed to failure from the outset if its proponents believe it can estab­lish ‘apodeic­tic certainty’ [652]; for no ‘experience [can] ever be adequate to an idea’ [649].

          Kant is very careful to clarify the constructive intentions of his criticism of this argu­ment: ‘It cannot hurt the good cause, if the dogmatic language of the overweening sophist be toned down to the more moderate and humble require­ments of a belief adequate to quieten our doubts, though not to command un­conditional submission’ [Kt1:652-3]. The physicotheolog­ical argument can serve this function best when it is used in conjunction with the other arguments, for the most it can prove on its own is the need for ‘an architect of the world who is always very much hampered [just as the human architect is] by the adaptability of the material in which he works, not a creator of the world to whose idea everything is subject’ [655; cf. Wo78:140]. Thus, proponents of such an argument who wish to define the concept of God in a manner sufficiently high to match the ‘loftiness’ of the idea of God must fall back upon the a priori proofs, which alone are concerned with ‘the existence of an abso­lutely necessary being’ [Kt1:657, e.a.]. In the end the physicotheological argu­ment, if it is intended to be a rigorous proof, will therefore suffer from the same illusions as those suffered by the arguments it seems at first to circum­vent.[15]

          Each argument fails because each is based on the same general assumption: each treats the idea of God as an ideal—i.e., as an object that can be instan­tiated ‘not merely in concreto, but in individuo, that is, as an individual thing’ [Kt1:596]. Yet such an assumption is mistaken [Kt7:463; s.a. 466]:

 

No intuition corresponding to the conception of a being which has to be sought beyond nature is possible for us. So far, therefore, as that conception has to be de­termined theoretically by synthetic predicates, it always remains for us a problem­atical conception.... The particular conception of a supersensible being cannot pos­sibly be subsumed in any way under the universal principles of the nature of things ...

 

To succeed in doing so, Kant says, would require ‘attributing omniscience to yourself’ [480]. If God is conceived as a being who transcends our categories of thought, no argument that depends upon these categories—i.e., no theo­retical argument—will be able to bring us any closer to the goal of knowledge of the existence of such an object from the standpoint of systemt: all such attempts ‘are altogether fruitless and by their very nature null and void, and ... do not lead to any theology whatsoever.’[16] For the problem is not so much with these various attempts to answer the question of God’s existence, as with the ques­tion itself. As Barth points out in his elaboration of Kant’s theology, ‘to speak of existence or non-existence is per se not to speak of God’ [Ba72:275]. For to answer the question of God’s ‘existence’—i.e., to solve the problem once and for all—would require us to be able to say: ‘This, not that, is the necessary be­ing’ [Kt1:640]. And this is why to regard God as an ideal object, one that ‘designates an individual as being among the things that are possible’ [639-40], always leads us into illusory thinking.

          Thus Kant’s criticisms are not intended so much to close the books on all theoretical ar­guments for the existence of God[17] as to curb the pretensions of those who mistakenly be­lieve such arguments can prove what no argument can ever prove, that the transcendent can be known to exist as such, without ever having to become immanent. Before we can regard it as pointing towards a real object, and so to act as the starting point of theology, our idea of God requires us to adopt a rational faith—much as does the thing in itself, in order to be justified as the starting point of Critical philosophy.[18] Kant readily admits that the ar­guments can supplement our faith by providing good reasons ‘to postulate the existence of an all-sufficient being’, but warns that ‘in presuming so far as to say that such a being necessarily exists [i.e., with real, not just logical neces­sity], we are no longer giving modest expression to an admis­sible hypothesis, but are confidently laying claim to apodeictic certainty.’[19] That he regards these arguments as evidence of our insatiable desire to reach out towards an unattainable knowledge of the transcendent—evidence that can be used to draw tentative conclusions about God’s nature [see V.2-3]—is implied in Kt1:651, where he observes that the postulate of God

 

is in conformity with the demand of our reason for parsimony of principles; it is free from self-contradiction, and is never decisively contradicted by any experience; and it is likewise of such a character that it contributes to the extension of reason’s empirical perspective, through the guidance which it yields in the discovery of order and purposiveness.

 

          Kant has nothing against the proper use of rational arguments in order to clarify our con­    cept of God, and so to ‘prepare the path for a religious response to Him’ [Co67:126]. In­deed, he says: ‘If what we have in view is the coming to a decision [an act]—if, that is to say, the existence of some sort of necessary be­ing is taken as granted ...—then [such arguments] must be allowed to have a certain cogency’ [Kt1:615; cf. Wa63b:128]. For as Collins ex­plains in Co67:113: ‘A reversal in perspec­tive takes place [in Kant’s treatment of rational the­ology], so that now the critically aware mind comes to view the entire theoreti­cal approach to God as a development out of a fundamental moral and religious belief in Him, rather than the converse relationship.’ This is correct, as long as we re­gard this ‘reversal’ as referring far more to the tradition, than to a sharp differ­entiation between Kant’s early and more mature position. For even in his first theological essay, in spite of its more positive attitude towards theoreti­cal proofs in general, Kant concludes [Kt15:163(366)] by warning us of their relative insignifi­cance: ‘It is absolutely necessary to convince one’s self of the exist­ence of God; but it is not just so necessary that it should be demonstrated.’ Accordingly, his criticisms of such proofs, in both Kt15 and Kt1, are directed only against those ‘incautious’ thinkers who ‘are easily de­ceived ... into taking the limits by which the human mind is circumscribed as limits within which the very essence of things is contained’ [Kt19:389]. To do so is to ignore the supreme principle of Critical philosophizing, the principle of perspective.

 

2. Recent Attempts to Revive the Ontological Argument

          The foregoing account of Kant’s arguments against the traditional theoretical proofs for the existence of God largely ignored the huge volume of secondary literature that exists on this subject—one of the most discussed aspects of Kant’s entire philosophy.[20] The variety of criticisms of Kant’s claims is almost as great, though worthy of a brief and somewhat random overview. To make this task more manageable, I shall concentrate in this section on just a few of the more influential attempts that were made in the twentieth century to save the ontological argument from Kant’s devastating criticisms.

          Writing at the very beginning of the century, Paulsen points out the insufficiency of Kant’s version of the ontological argument [see Pa02:223], claiming that Kant sets up a ‘spurious and vulgar representation of God as a particular being’, not because this is what pro­ponents of the ontological argu­ment believe in, but because it was easy to refute. Ironi­cally, claims Paulsen, Kant himself ends up supporting the idea of God as a transcendental unity, which is what most proponents of the ontological argument would claim the argument is actually about! This is to some extent a valid point, though Paulsen may be underestimat­ing how many dogmatically-minded philosophers and philosophically-minded religious be­lievers really do, for all practical purposes, adopt the vulgar idea of God as a particular being. For this reason, the general aspect of Kant’s refutation is not simply a straw man.

          One of the most detailed accounts of Kant’s position is given in Wo78:95-145; but Wood’s treatment typifies such accounts by failing to see the importance of the principle of perspective in Kant’s arguments. As a result he ends up concluding that Kant’s ‘famous but badly unargued’ criticism ‘of the traditional theistic proofs is on the whole unsuccessful’, in spite of the merely historical fact that his ‘attack succeeded in bringing these proofs into disrepute’ [123,149]. Wood judges the relative value of the proofs in the reverse order as Kant: for Wood the ontological argument is to be highly regarded, with the cosmological a close second, and the teleological is thought to be nearly worthless. But Wood gives away the rationalist bias that lies behind his interpretation (and is continually hinted at by his use of rigorous logic as the sole means of judging the meaning and validity of an argument) when he says ‘the ontological proof ... might well succeed in the context of a more rationalist episte­mology’ [123]. Kant, of course, would agree, but counter that an argument is not valid simply by virtue of its consistency with a mistaken set of presuppositions: Kant’s crit­icisms of the traditional arguments are directed far more to such presupposi­tions than to any of their particular formulations, which may account for Wood’s inability to grasp their force.

          Concerning the ontological argument in particular, Wood acknowledges that ‘Kant’s view has been, and still is, widely accepted’; yet he casts doubt on its validity, for ‘no one as far as I can tell has ever presented a really persuasive argument for it’ [Wo78:110]. He therefore believes he is justified in giving the ontological argument more valid­ity than Kant allows. Apparently Wood is looking for a formal (i.e., logical) argument in defense of Kant’s criticism. He fails to see that such an argument is not available, and is indeed irrelevant, since Kant’s whole ap­proach is rooted in the principle of perspective. Once this principle is recog­nized as the foundation for Critical philosophy, the perspectival differ­ence between logical and transcendental reality obviously implies that ex­istence cannot be a real predicate in the re­quired sense: no proof is required as long as this doctrine is properly explained in terms of its perspectival context. Indeed, the negation of the proposition is unmistakably self-contradictory when its perspectival terms are fully explicated: to assert that logical existence can serve as a real (transcendental) predicate is to conflate the logical and transcendental perspectives.

          Malcolm’s earlier attempt to sneak the ontological argument past Kant’s criticism [Ma60:145-56] suffers from this same mistake. After describing the terms ‘impossible’ and ‘necessary’ as logical terms [145], he argues: ‘God’s exis­tence is either impossible or necessary.... Assuming [the former] is not so, it follows that He necessarily exists’ [146]. In order for this to be an ontological argument at all, ‘God’s existence’ must here refer to God’s real existence. Mal­colm’s acceptance of this assumption indicates his failure to recognize that real existence and logical necessity or impossibility belong to different perspectives, and so cannot be used in a single argument without falling victim to a category mistake. It is quite true both that God’s real existence is either really impossible or really necessary and that necessary existence is a logically necessary con­stituent of the concept of God. (The for­mer dichotomy is supported by the fact that a rejection of the latter leads direct­ly to Findlay’s proof for God’s non-exis­tence [Fi63:96-104].) But nothing in Malcolm’s argument—or anyone else’s for that matter—can bridge the gap between these two truths: the choice we face in considering the question of God’s reality cannot (validly) be decided simply by clarifying what is logically involved in the mere concept ‘God’, as Malcolm claims [Ma60:149]; for the conclusion of such arguments, ‘God’s real exis­tence is logically necessary’, represents what might be called a ‘perspectival incongruity’.

          Smart is quite right to say ‘there can never be any logical contradiction in denying that God exists’ [Sm55:34]—provided ‘exists’ is viewed from the empirical or transcendental per­spectives, but not the logical. Malcolm fails to appreciate Smart’s nonlogi­cal perspective when he argues that such a denial is itself a logical contradiction [Ma60:154-6]. The contra­diction would be to hold that the concept of God de­notes a perfect being whose real existence is not necessary; to claim that the object to which such a concept refers might not exist in the empirical world is in no sense logically contradictory. This is essentially Plantinga’s point in Pl61, though the extent of his awareness of the perspectival source of Malcolm’s confusion [e.g., 171] is shrouded by the fact that he couches his rebuttal in unnecessarily complex logi­cal ter­minology. Coburn [Co71] responds to Plantinga by charging that he neglects Kant’s claim that existence is not a property [20-2,27-8], but then ironically (having clarified this important point) turns around and disagrees with Kant’s claim [26]. Coburn’s disagree­ment, however, is once again rooted in a failure to realize that for Kant ‘exists’ is being taken solely from the logical perspective.[21] Kant does not deny that empirical existence is a prop­er­ty (this being Coburn’s view); his claim is that existence is not a property of a concept.

          Engel [En63] claims Kant intends a threefold distinction between real predicates, nonreal predicates, and the logical use of the copula, ‘is’. He claims commentators are wrong to con­nect ‘nonreal’ with ‘logical’, because the latter applies only to the use of the copula. Engel’s basic idea of a threefold distinction is perceptive and has considerable poten­tial, but his account of it suffers from a subtle conflation of perspectives. The copula is the root of logic and is thus the defining feature of the logical standpoint (a standpoint Kant never critiques as such); Kant’s distinction between ‘real’ and ‘logical’ in Kt1, however, falls entirely within the theoretical standpoint. Engel’s failure to base his threefold distinction on these clear perspecti­val distinctions leads him into a hopelessly confused interpretation, concluding (not surprising­ly) with a negative assessment of the validity of Kant’s position.

          Another major proponent of such a priori arguments in the twentieth century has been Hartshorne, who argues: ‘If “God” stands for something conceivable, it stands for something actual’ [Ha41:135]. He thinks this form of the ontological argument avoids Kant’s ‘slavishly accepted criticism’ [HR53:97] by allowing that the concept of God might be unintelligible (thus justifying atheism), but insisting that if it is to be intelligible, then ‘existence’—and not just logical, but real existence—must be a necessary part of the concept (thus justifying theism). Hartshorne, however, fails to grasp the essential perspectival point of Kant’s criticism. For Kant’s rebuttal would simply be that Hartshorne has never stepped out of the logical perspective, so his conclusion cannot refer to God as a real object, but only to our concept of God. That is, if he has proved anything, it is that the concept of God must stand for something conceived as actual; its actual actual­ity, so to speak, necessarily remains uncertain as long as we limit ourselves to the logical perspective; for this perspective, by definition, abstracts from all existing objects.

          Hartshorne’s other formulation of the argument is also covered by Kant’s criticism. ‘The necessary being’, argues Hartshorne, is ‘that individual which existence implies’ [Ha41:132]. This is technically not an ontological argument at all, but more like a cosmological argument. Though it may perhaps be too severe a criticism, one is tempted to say of Hartshorne what the young Kant says in Kt14:(67) of someone who uses ‘mixed ratiocination’ (i.e., syllogisms with more than three elements): ‘It is a pity to see the trouble that an able man takes trying to improve a useless thing. The only useful thing one can do with it is to annihilate it.’ The difference, as we saw in the previous section and in IV.1-3, is that Kant does see a certain degree of usefulness in such arguments, as long as they are not regarded as proofs establishing irrefutable conclusions.

 

3. Clarifications of the Nature and Status of Kant’s Moral Argument

          In IV.4 I have already considered several commentators’ views on the nature and status of Kant’s moral argument [s.e. notes IV.15,17,18]. But that treatment only scratches the surface of the secondary literature on this subject, so in this section I shall respond to a wider range of possible objections. For instance, Kant’s attempt to put the fact of human obligation in place of the traditional proofs as a rational ground for faith in God is regarded by many as ‘artificial’ [e.g., We26:65] and con­trived, since his own arguments in the first three stages of systemp demonstrate that moral laws are valid quite apart from any postulation of God’s existence. Kant attempts to clarify this point in Kt35:(40): ‘Moral laws can be right without a third being [i.e., without a God in addition to the moral agent and the moral law], but in the absence of such a being to make their performance necessary they would be empty.... Knowledge of God is, therefore, necessary to make the moral laws effective, but it is not necessary for the mere apprehension of those laws.’ In this sense, the relationship between duty and the postulate of God is analogous to that between intuition and conception, at least to the extent that we could say: God without moral laws would be dumb, moral laws without God would be powerless.

          Michalson takes this argument a step further in Mi86 (and repeated in Mi90:22-5). He claims the moral proof is based on certain ‘non-moral elements’ related to what Kant calls the ‘needs of reason’, requiring ‘symmetry, proportion, and coherence in the universe’ [Mi86:505]. Kant presents these as undefended assertions that he seems to think everyone will take as self-evident. The argument of Kt4, ‘culminating in the postulate of God’s existence, is in effect Kant’s explicit demonstration of his implicit confidence in the rationality of the universe’ [509]. Michalson’s claim, in a nutshell, is that the moral argument rests not on the moral law so much as on Kant’s assumption of ‘a universe that makes sense’ [509]. Thus he observes [512]: ‘The believer does not assent to belief in God in order to be moral, but only in order to satisfy reason’s demand for symmetry.’ All of this is quite accurate; but I see no sense in which it detracts from the central arguments Kant is advancing [see IV.4], especially when we keep in mind that Kant’s demand for symmetry is part and parcel of the architectonic implications of his central Copernican hypothesis. Michalson’s observations are rendered harmless once we recall that the moral argument in systemp is not meant to be part of a ‘philosophy of religion’, but only the bare foundation for a moral theology.

          Nuyen [Nu94:122] challenges Kant’s moral argument by attempting to show that ‘the promotion of the highest good can be a moral duty for a rational nonbeliever.’ The ‘religious postulates’[22] are therefore not necessary, mainly because [130] ‘given the limits of human cognition, we do not know whether the task of trying to attain the highest good is ultimately possible. Thus, it is not incoherent for the nonbeliever to pursue the highest good.’ Kant would agree, but would add that it would be foolish to do so, given the empirical evidence against morality producing happiness. Nuyen’s argument amounts to the view that a person can simply ignore the fourth stage of systemp, though in so doing he or she will not have any basis for hoping in the ultimate fulfillment of the highest good [131]. Ironically, this is precisely Kant’s point: morality is hopeless for the ‘rational nonbeliever’! Kant, however, would regard acting without hope as irrational. Nuyen is certainly right that the nonbeliever can promote the highest good, because this duty does not require ‘that we attain the highest good.’ But he neglects to point out that there is no reason to obey duty in this case. Nuyen himself admits ‘there is something absurd about’ such a scenario, due to the ‘uncertain’ outcome of such action [132]. But the latter is not Kant’s focus. His claim is stronger, that we must at least be able to conceive of the possibility of its fulfillment; and this, the rational nonbeliever cannot do, so his or her continued adherence to the moral law becomes irrational. Nuyen concludes with another tacit confession of the weakness of his objection to Kant’s position [132]: ‘In the end, it could well be more rational to be religious.’

          In response to the view defended by Walsh and others [see note IV.15], Friedman argues that Kant’s moral argument ‘does not merely enable the moral believer to defend his belief as something which he requires. The argument establishes that there is a God’ [Fr86a:8]. But this ignores the subjective nature of the moral argument: it establishes that moral persons can be convinced there is a God. Friedman goes on to acknowledge this point [9]: ‘Don’t look to the world, Kant is saying, where the objects in question cannot, in principle, be found; look to the subject ... In a practical argument I am the evidence.... The I of the practical argument is a universal I.’ Indeed, this raises an important point about certainty: ‘I am not less certain of God for not knowing Him.’ That is, moral and theoretical certainty are different types of certainty, not different degrees. Friedman points out that here Kant and Kierkegaard are quite distinct from each other [10-1]. ‘Kierkegaard believes ... [t]he human subject is not certain of God.’ The remainder of Fr86a is an excellent discussion of Kierkegaard’s understanding of faith and uncertainty. Unfortunately, Friedman leaves the reader hanging by never returning to assess Kant’s view from a post-Kierkegaardian perspective.

          In Wo89 Wood accuses Kant of ‘violating an important moral principle’ [417]: ‘Kantian moral faith’ is itself immoral ‘simply because it is faith, that is, belief held in the absence of sufficient proof or evidence’ [419]. Wood bases his entire argument on what he calls ‘Clifford’s Principle’, the claim ‘that a belief can be morally justified only if it is epistemically justified’ [420]. Yet this is not so much a ‘moral principle’, as Wood applies it, but a direct challenge to Kant’s emphasis on the primacy of practical reason. As a result of rejecting this fundamental Kantian principle, Wood ends up using unKantian notions of ‘belief’ and ‘evidence’. He totally ignores Kant’s sharp distinction between theoret­ical belief (intended to produce knowledge) and practical belief (arising out of a pursuit of moral goodness). Thus, his claim that ‘Good evidence produces belief irresistibly’ [431] applies only to theoretical belief (whose aim is to produce knowledge); moral beliefs are notorious precisely for being incapable of finding evidence sufficient to produce ‘irresistible’ belief. Kantian moral faith deals with objects (God and immortality) that cannot possibly be epistemically justified, because they are transcendent. Clifford’s Principle, however, properly applies only to beliefs that have empirical objects, objects that could have some epistemic justification. Wood fails to distinguish between these two and therefore too readily assumes that Kant’s moral faith falls under (and is thus made ‘immoral’ by) Clifford’s Principle. Kant would actually agree with this principle insofar as it condemns a person for acting in a way that ignores clear empirical evidence to the contrary; but he would regard actions based on such unjustified theoretical belief as reason for calling a person not ‘immoral’, but foolish.

          Wood’s main point is ‘that Kant’s moral argument is not capable by itself of producing rational conviction that God exists’, because its conclusion ‘is not that God exists, but rather that there is a rational inconsistency between disbelief in God and a disposition to promote the highest good’ [Wo89:434; see note IX.28, above]. A good theoretical argument, by contrast, ‘cannot fail to convince [even an atheist] that there is a God.’ Instead of proving that God exists, Kant’s argu­ment shows [435] ‘that belief in God would be [a] very desirable thing for a moral agent to have.’ Wood thinks Kant is forced in the end to justify his moral faith in God’s existence by making a Kierkegaardian appeal to ‘emotional attachments’ [432,436]. Such a ‘clandestine’, ‘nonrational’ basis for faith supposedly amounts to ‘heteronomy’ [437], thus introducing the allegedly immoral element into Kant’s moral theory. What Wood neglects is that ‘nontheoretical’ is not the same as ‘nonrational’ for Kant, and that practical evidence can be—or even is—properly subjective without being heteronomous. Indeed, Wood has unwittingly turned Kant on his head: by rejecting the primacy of practical reason, he fails to recognize that for Kant an insistence on ‘objective evidence’ and theoretical justification [436] is far more likely to condone an ‘immoral’ disposition than a proper focus on the subjectivity of moral-practical reason. Wood’s position does away altogether with practical reason as a standpoint distinct from and superior to the theoretical. As a result, he ironically ends up dubbing what is genuinely practical as ‘heteronomy’ and places his trust in a kind of theoretical claim to certainty that Kant would regard as heteronomous.

 

4. Kant’s Alleged Atheism: Can a ‘Fiction’ Replace All Theoretical Proofs?

          The conclusion of IV.2 alludes to those interpreters who erroneously think Kant’s occasional references to God as a ‘fiction’ imply that Kant does not actually believe in a real God at all.[23] The notion that Kant is really a closet atheist is based on a total failure to appreciate the perspectival subtleties in his System, by treating systemt as Kant’s final word on such matters. Friedman [Fr86b:515], for example, asserts that Kant’s God is not real, but a mere ‘thought thing’; he shows how Hegel related this criticism directly to Kant’s criticism of the theoretical arguments [516]: ‘Kant’s God does not have the “reality” that a metaphysical or speculative argument might bestow on him.’ While this is in a sense true, it is misleading. Kant’s view is that this metaphysical type of ‘reality’ is a false hope and thus an improper or even dangerous foundation for religion.

          As argued in Chapter V, God’s transcendence prevents us from successfully ‘proving’ God’s existence as an external reality; the moral argument confirms God’s reality, but only as a force we experience as immanent in human reason, through the moral law [see Fig. V.3]. When Kant calls God a ‘fiction’, he is referring only to the status of the concept, ‘God’, as viewed as an object from the speculative perspective of systemt. When the hypothetical per­spective replaces the speculative, this fiction attains the more secure status of a regulative ‘idea’ [see IV.2]. Along these lines Hicks [Hi74:382] says: ‘Knowledge of God is precluded ... [yet] the thought of God is indispensable because of the fragmented structure of human knowl­edge.’ This is correct, but does not imply that Kant’s God is a mere thought; the limitation is on our side, not God’s. When Kant uses terms such as ‘fiction’, he is seeking to pro­tect God from those who would storm heaven, so to speak, with their uncritical knowledge-claims.

          This problem becomes more pronounced when Kant makes statements such as ‘the Idea [of God] proceeds entirely from our own reason and we ourselves make it’ [Kt1:442; s.a. Kt6: 443]. In Kt8:168-9(157n) Kant clarifies that, although ‘it does indeed sound dangerous, it is in no way reprehensible to say that every man creates a God for himself, nay, must make such a God according to moral concepts ..., in order to honor in Him the One who created him.’ That this ‘making’ refers to a transcendental condition of necessitation that makes honoring God possible, not to an empirical act of fabrication that would make such honoring a farce, is confirmed by Ward: ‘We make God only in the sense that we make every rational structure in the world’ [Wa72:163]. In a 1794 letter to Beck Kant again confirms that this way of talking about God does indeed apply to all our knowledge: ‘we can only understand and communicate to others what we ourselves can make’ [AA11:496(Zw67:216)].

          This is the proper context for Kant’s many otherwise uninterpretable assertions about God in Kt9,[24] such as that the idea of ‘God is the moral-practical self-legislative reason’ [21.145]; ‘God is not a thing existing outside me, but my own thought’ [21.153]; this idea is ‘the product of our own reason’ [22.117(201)], which ‘is not a hypothetical thing but pure practical reason itself in its personality’ [22.118(201-2)]. Ignoring the perspectival context of such comments, some interpreters take them to imply that Kant is seeking to identify God with practical reason. Ward, for example, claims that in Kt9 ‘Kant makes explicit a doctrine which is implicit in many earlier works, that God and practical reason are identical.’[25] Since practical reason also belongs to human persons, this means human beings are also ‘identical with God’ [Wa72:164]. Likewise, Sullivan [Su71:123-4] interprets such claims to mean: ‘The source of this moral obligation, this “God,” is not really anything other than reason itself.’ Collins correctly charges such a view with failing ‘to give due weight to Kant’s distinction between the being of God and our idea of God (only the latter being assimilated to the moral imperative)’ [Co67:119n]. McCarthy says aspects of Kant’s theory in Kt8 already lead the reader to ‘suspect an imminent dissolution of God’ [Mc82:204]. Elsewhere he calls attention to ‘passages of [Kt9] with an atheistic drift’ [Mc86:63]. These raise the question of whether ‘God is reason’ [97-8]; but the text provides no decisive answer [98], except that somehow ‘God escapes dissolution.’ Copleston also regards the evidence as inconclusive, but believes Kant still upholds, even in Kt9, God as ‘a Being distinct from man’ [Co60b:392]. As evidence he quotes Kt9:21.15: ‘the concept of God is the concept of an obligation-imposing subject outside myself’ [e.a.]. Given that Kant himself criticizes the ‘sophistry’ of those who transform ‘an objective question of what the thing is into the subjective question of what the word, by which we signify the thing, means’ [Kt8:89n(84n)], he would be unlikely to commit a similar fallacy by denying God’s reality on the basis of the radically subjective nature of our idea of God. As Schrader points out, such a strict identification of ‘God with the moral law’ is really ‘a Fichtean position’ [Sc51a:231n], not to be confused with Kant’s.

          A related problem has to do with the issue of whether God is a merely subjective and internal reality (e.g., a mere idea) or whether God has some objective and external exis­tence as well (e.g., as a ‘Being’). Peters thinks that even when Kant argues in Kt9 that ‘the categorical imperative leads directly to God, and affords surety of His reality’ [quoting Ad20:790-1], he is relating this only to the idea of God, not to any real God existing outside the human mind [Pe93:98; s.a. Su71:121-2,128]. The basis for this view lies in numerous comments Kant makes, such as [Kt9:21.144; s.a. 22.108] that ‘God must be represented not as a substance outside me, but as the highest moral principle in me.’ Ward regards this type of claim as being ‘in open contradiction’ [Wa72:161] with Kant’s other claims, such as that our ‘concept of God is that of a Being as supreme cause of the world-beings and as a person’.[26] Indeed, Kant elsewhere unambiguously affirms the need for God to exist ‘outside of mankind’.[27] Likewise, Kant acknowledges in Kt23:405 that the inner ‘voice’ we associate with our aware­ness of God could come either from our ‘own reason’ or ‘from an other’ (or both) and that the habit of personifying the former in terms of the latter is acceptable within Critical limits. I therefore argue in XII.2 that in making the former type of claims Kant is stressing the immediacy of the categorical imperative and not intending to give a dogmatic denial of the possibility that God may also have some existence independent of human beings.[28]

          Once we understand that Kant’s reason for limiting our understanding of God to the status of a practical, subjective belief is to keep us humble, his position seems quite com­pat­i­ble with theism—albeit, in a refined, Critical form. Thus Baillie rightly says [Ba39:161] that for Kant ‘what is revealed to us is not theoretical knowledge but practical guidance, and ... what is asked of us in return is not intellectual assent but willing obedience.’ Kant gives no decisive answer to the question about God’s external reality precisely because it is a theo­ret­i­cal question. He does ask the question, sometimes very clearly, as in Kt9:22.120(Su71:119): ‘There is a God in the soul of man. The question is whether he is also in nature.’ As suggested in V.3 [s.e. Fig. V.3], Kant’s answer would have been: theoretically we do not and cannot know, so it is best to assume a negative answer; but practically and there­fore in our everyday (moral) life, we can act as if a moral God lies at the foundation of the natural world. As Collins rightly affirms, God ‘becomes the God of our religious hope and desire’ only when viewed as both ‘the author of nature and [of] moral order’ [Co67:152]. As such, God is not merely a substance, but is Substance itself. And Kant’s motive for restricting his affirma­tions about God to the internal, subjective realm is to heighten (not diminish) the sense of submissiveness we ought to have towards this voice within. This, as we shall see in Chapter XII, is the key to the unity of Kant’s entire philosophical System.

          A softening of the ‘atheism’ charge leads many interpreters to regard Kant as a deist. Perhaps chiefly responsible for this trend is Heine, who caricatures Kt1 as ‘the sword that slew deism in Germany’ [He59:268], while seeing Kt4 as a feeble attempt to revive deism.[29] Its popularity may be due more to Greene’s claim in his strategically located introductory essay to Kt8, that this work is ‘a deistic classic’ [Gr34:lxxvii; s.a. lxvi]. Webb implies that Kant was a deist for most of his life when he says that in Kt9, Kant ‘was prepared to repudiate ... the deism which had been so predominant in his youth—the deism which taught a merely transcendent God’ [We26:200-1]. Ironically, Vleeschauwer sees Kt9 itself as ‘a public confes­sion of deism’ [Vl62:177]. Following this trend, Zweig infers from Hamann’s 1759 letter to Kant that Kant equates ‘deism’ with ‘sanity’ [Zw67:35n]. Yet in this letter Kant actually sets himself up as an arbitrator between Hamann the Christian and Berens the deist. Zweig’s placement of Kant on the side of Berens is not justified from the content of the letter, which shows us a typical example of Kant acting in his Critical position as a middle man.

          Fortunately, a growing rank of scholars now rejects such inter­pretations. Despland, for example, argues that in his philosophy of religion ‘Kant ... moved beyond the classical deist position’ [De73:198; s.a. 199-201,228,262; No73:431; Wi80:515]. Michalson agrees that ‘Kant’s flexible attitude towards [the parerga in Kt8] helps to distinguish him from the main­stream deism of his day’ [Mi89:265; see AVII.1], while McCarthy says the same about Kant’s use of symbols [Mc86:69; see V.2]. As Collins puts it [Co67:117], ‘Kant regards reli­gious deism and the varieties of nature-based theism as incomplete, preliminary forms of reli­gious life.’ Indeed, as I demonstrate in Part Two, Kant moves beyond these to form a moral theism[30] that is thoroughly compatible with his Critical principles. Kant’s theistic outlook is ac­knowledged so consistently throughout his writings that I would even call into question the assumption that Kant ever seriously defended a deistic position. Hare concurs with this view, arguing ‘that Kant is not a deist’ even ‘in the ordinary sense’ of the term [Ha96:41,45].

          Admittedly, Kant usually expresses his rejection of deism in very cautious terms. This is un­derstandable, given the dominance of deism in the philosophical climate of his day. Nev­ertheless, some texts reveal his dissatisfaction with deism so clearly that all debate on this question ought to be a thing of the past. In a 1789 letter to Jacobi, for instance, Kant ap­proves of his friend’s refuta­tion of ‘the syncre­tism of Spinozism and the deism of Herder’s God’ [AA11:74(Zw67:158)]. And in Kt2:356-7, Kant says that if theism and anthropomor­phism are both abandoned, then ‘nothing [would] remain but deism, of which nothing can come, which is of no value and which cannot serve as any foundation to religion or morals.’

          The only way deism can plausibly be attributed to Kant is by defining the term in a way that is foreign to him and arguing that, because his Critical theology defends positions that others associate with deism, we are justified in calling Kant a deist. This is Wood’s strategy in Wo91:1 [s.a. Wo89:435], where he regards Kant’s ‘idiosyncratic’ definition of ‘deism’ as merely ‘a device to deflect reproach from Kant’s own heterodox religious views.’ He argues that Kt8’s title is a direct reference to deism (in its ordinary sense)—though he admits that ‘the title by itself does not necessarily imply that Kant em­braces the deistic position that reli­gion can get along without revelation’ [Wo91:2; cf. VI.2, above]. Yet he merely assumes that Kant’s numerous expressions of openness to revelation and other orthodox Christian beliefs [cf. VIII.2-4 and AVII-AVIII] ‘are meant to appease those who think [religion] cannot [get along without revelation].’ (I refute this all-too-common assump­tion in notes VI.28, VIII.4, 27, and 45.) Wood gives an accurate account of Kant’s reasons for rejecting most mani­festa­tions of ecclesiastical faith [3-7] and rightly points out that Kant supports ‘rational or natural religion’ [7]; but he neglects the fact that for Kant some empirical manifestation is necessary in order for religion to be genuine, and that this will inevitably involve an appeal of one kind or another to revelation, even though ‘bare reason’ is incapable of saying in advance which revelation is true [see VII.3.A and VIII.3.A]. Thus, when Wood defines a ‘deist’, quite cor­rectly, as ‘someone who believes in a natural or rational religion rather than a religion based on supernatural revelation’ [10, e.a.], he inadvertently excludes Kant from the ranks of deism. For Kant is always careful never to deny revela­tion a priori (as any good deist must), but al­ways to put revealed religion in its proper context, allowing for its possibility and even pro­priety, as long as it remains in partnership with rational religion. In the end [e.g., 11-2], Wood’s own deism shows through, for he demands knowledge, rather than being satisfied with rational belief—the latter being the only basis for the religion Kant wishes to defend. So when Kant confesses his ignorance, Wood reads this as a rejection rather than a humble accep­tance.

          In a similar fashion Sokol assumes Kant ought to reject theism, if he wishes to advocate the ‘secular humanism’ implied by his ‘Principle of Autonomy’ [So86:424]. Sokol admits Kant does enter ‘into the territory of the theist’ [430]; but he regards this as an ‘unjustified step’ from ‘the standpoint at which he began, a secular humanism’ [431]. The key difference is that ‘[t]he secularist ... can have no higher motivation for acting morally ... than acting out of respect for the moral law ... But the theist recognizes another authority’—i.e., God. What Sokol neglects is Kant’s dependence on the principle of perspective: Kant’s whole point, as shown in Chs. VI-VIII, is that the so-called ‘secular humanist’ perspective that is appropriate for moral philosophy, and is admittedly deistic, must be transformed into the perspective of religious theism, if morality itself is to attain its highest ends. (See note VIII.30 for more on Sokol’s interpretation.)

          One of the most significant factors in understanding how Kant can be a theist without proving that God exists is his conviction that logical argumentation has definite (and rather narrow) limits of proper applicability. He demonstrates this throughout the Dialectic of Kt1, but most notably in the Antinomies, where he demonstrates that logical arguments can be concocted to prove either of two opposing theses. The lesson he wants us to learn from this, I believe, is that philosophers must resist the temptation to treat pure logic as their god—i.e., as a sufficient method for obtaining truth on any and every issue. For Kant, ‘reason’ includes more than just logical cognition; logic is, in fact, a derivative of willing; and even aesthetic and/or teleological judging (as the standpoint of genuine critique) have architectonic priority over abstract logical argumentation. Indeed, to depend on purely logical forms of argument (see e.g., my comment on Plantinga in AIV.2) can sometimes be unethical, inasmuch as it can lead a person to support conclusions that contradict the clear demands of practical rea­son.

          Kant’s refusal to give priority to the theoretical application of logic in the grounding of theology can be regarded as an antidote to the kind of Humean theology advanced by Flew in Fl66. For Flew is a good example of the approach to philosophy that Kant regards as harm­ful to the best interests of humanity: an uncritical trust in the all-sufficiency of logic plus empirical ev­idence to prove or disprove any and every point at issue [see e.g., 2.16,49-52], and a neglect of the perspectival difference between human beings (beings necessarily bound by perspec­tives) and God (a perspectiveless Being) [see e.g., 2.40]. For instance, he argues [2.21] that in order to form a concept of God, there must be ‘some definite and positive statements which are literally true, if the whole enterprise is to be capable of getting off the ground at all.’ He cites ‘God exists and possesses such and such basic characteristics’ as ‘the very least’ that is required. I argue in Chapters IV-V, by contrast, that for Kant belief in God coupled with a symbolic description of basic divine characteristics is all philosophical theology can ever hope to demonstrate. Perhaps for this very reason, Flew never engages fully with Kant’s position: to do so would be to threaten the viability of his own approach. Instead, he argues primarily with orthodox (mostly Catholic) theologians, such as St. Thomas, who share his own bias for the power of theoretical reason. Whenever he touches upon the views of theologians who do take a Kantian (perspectival) approach seriously, he dismisses them without argument, charging them with being ‘an abstraction’ [2.18], ‘rash’ [2.19], or ‘uncongenial’ to the true Christian [2.44; cf. 1.16]. Flew reveals the full extent of his logico-empirical bias when he concludes his book [9.29] with the claim ‘that the universe itself is ultimate; and, hence, that whatever science may from time to time hold to be the most fundamental laws of nature must ... be taken as the last words in any series of answers to questions as to why things are as they are.’

          Kant’s critique of all theoretical proofs of God’s existence is chiefly responsible for giving him the reputation of being the ‘destroyer’ of metaphysics in general and theology in particular [see I.1]. But we have now seen that his position is disastrous only for types of theology that claim to establish knowledge of God’s existence. Kant’s critique puts such proofs in their proper place by showing how they can be pointers to an unknowable reality, even though we remain necessarily ignorant of whether or not what they point to actually exists. Such pointers must ultimately be grounded in moral belief in order to be convincing. In this way he actually protects theistic religion from philosophical sophisticates by insuring that the unconditioned reality we call ‘God’ cannot be apprehended as such by our limited powers of understanding and reason.

 

  [1].  Kant does not include in his discussions of the possibility of proving God’s existence those powerful arguments against such a belief, based on the problem of evil. But he does deal with the problem at length in Kt64 and more briefly elsewhere. For a discussion of his position, see AVI.1.

  [2].  Unfortunately, Kant does not draw attention to this clear correspondence in his dis­cussion of their relationship in Kt1:618-9. Instead, he suggests that the ontological argument should be discussed first because it is transcendental, whereas the others contain an ‘empirical factor’. But this is true only in the loose sense of ‘transcendental’ [see KSP1:114], whereby all the arguments have a transcendental aspect. The transcendental perspective as such is concerned with experience in general (as in the cosmological argument); the perspective dealing only with the relation between concepts (as in the ontological argument) is the logical [see Kt1:624n].

  [3].  Cf. Fig. III.4. Collins adds ethicotheology to the three types of theoretical the­ology and presents the relationship between them in the form of a four­fold diagram [Co67:102; s.a. De73: 138]—a procedure I also follow in V.3 [s.e. Fig. V.4]. This results, however, in a less perfect 2LAR than that given in Figure AIV.1, because ethicotheology (or ‘moral theology’, as Kant sometimes calls it [e.g., Kt1:660,842]) belongs to an entirely different standpoint than the others. Whereas the three types of theoretical theology correspond directly to the three types of theoreti­cal proof, ethicotheology corresponds only indirectly to the possibility proof (by virtue of the correlation between the hypothetical perspective and the prac­tical standpoint).

  [4].  That the cosmological argument begins not with evidence from induction but with the concept or ‘thinkability’ of God’s existence is well explained by Dell’Oro in De94:61,68. Likewise, she correctly identifies the possibility proof as starting from ‘the real’ [133].

  [5].  Wo78:62; s.a. En29:48-62,118-21. Greene’s summary of Kant’s position, as arguing ‘that it is impossible that nothing should exist, hence that something must neces­sarily exist’ [Gr34:xliiin], fails to highlight its most essential features. Webb’s synopsis in We26:32 is more helpful: ‘Briefly stated, [Kant’s possibility proof] is that something is possible, that any possibility presupposes a real being, and that in this way the exis­tence of a real being antecedent to any mere possibility of the existence of anything else is demonstrable from the very fact of possibility.’

  [6].  Kt1:604. The fact that our way of determining an object requires in addi­tion to intuition a power of conception leads us to assume, Kant notes, that the ens realissimum is not only a thing in itself, but also ‘a supreme under­standing’ [611n]. These two characteris­tics correspond to the two aspects of divine reason discussed in V.1: intellectual intuition and intuitive understanding. On a different but related note, Dell’Oro points out that in Kt15 Kant associates the ens realissimum with ‘the Principle of Identity’, regarding these as together constituting the ‘principle grounding all positive truths’—the principle of negative truths being the Principle of Contradiction [De94:135].

  [7].  Kt1:608. That this relates only to the speculative perspective and therefore does not commit Kant to regarding God as a mere figment of our imagination will be argued in AIV.4.

  [8].  Kt1:620. Webb calls the possibility proof ‘a modified form of ... the Ontological Argument’ [We26: 30]. Yet such a claim could be mis­leading, since Kant carefully distinguishes be­tween the two: ‘Either the existence of a most real being is deduced from its concept [as in an ontological (+-) proof], or the determinate concept that we must form of it is deduced from the necessary existence of some being [as in the possibility (++) proof]’ [Kt69:302-3; s.a. 331-2]. Treash claims Kant’s new proof reconstructs the usual analytic a priori argument on synthetic a priori grounds [Tr79:13-4].

                By contrast, I believe Kant’s remarks suggest, rather appropri­ately (given its position in Figure AIV.1), that the possibility proof can best be interpreted in terms of what I have called the ‘analytic a posteriori’ perspective [III.2; s.e. Fig. III.4]. For both proofs are analytic (deductive), but the lat­ter alone stipulates its concept on the basis of an a posteriori fact (an existing be­ing), as is characteristic of all analytic a posteriori reasoning [see KSP1:134-9]. Dell’Oro has a golden opportunity to acknowledge this status in her discussion of the views of Campo and Lamacchia [De96:124-5]: the former points out that the proof ‘is cast in a strongly formalistic language’ and yet it ‘appears to proceed a posteriori’, while the latter says the proof ‘shifts from the domain of the purely thinkable [i.e., the analytic] to the domain of existent things [i.e., the a posteriori].’ Unfortunately, instead of grasping the paradox, she merely accepts their assessment that the proof is ‘fundamentally equivocal’. Kant lends support to my position when he says his ‘method’ in Kt15 is ‘to ascend from what is immediately certain by observation [i.e., a posteriori] to the more general [i.e., analytic] judgment’ [97(267)]. This account of the epistemological status of the possibility proof could form the basis for a Kantian response to the defense of St. Thomas’ a posteriori ap­proach in Fo95. It may also explain why Kant warns in Kt69:303 that this proof makes progress in meta­physics only ‘through the back doors.’ Later, he warns again that in this proof, ‘the concept of a necessary being is by no means the concept of a being that is in any way determinate’ [326n], as it would be if it were based (like the ontologi­cal argument) on analytic a priori reasoning.

  [9].  Kt1:622; s.a. Kt69:304. Kant points out this characteristic of the ontological argument as early as Kt11:394-5(224): ‘In framing for ourselves a notion of a certain be­ing which we call God, we have framed it in such a manner that existence is in­cluded in it. If, then, the precon­ceived notion be true, it is also true that God ex­ists.’ Along these lines, Webb points out that the problem with the ontological argument is that ‘if I conceive [of God] at all, I conceive [of God] as being real; whether [God is] real or not, does not affect the conception’ [We26:32].

                England explains in En29:54 why Kant did not regard his own possibi­lity proof as falling under the same condemnation: ‘We cannot pass from a concept to the affir­mation of existence. But ... we can and must pass from the inner possibility of things in general to ... the affirmation of the existence of that which makes possibility itself possible.’ In other words, for Kant ‘the merely possible ... is a pure abstraction from actu­ality. The possible presupposes something actual’ [48]. But England concludes his discussion of Kt15 by accusing Kant of committing an ‘essen­tial contradiction’ [En29:62] by ‘maintain[ing] on the one hand that God is the one and only (unicus) being ...; and, on the other hand, that contingent things enjoy an individual existence different from the being of the one ultimate substance.’ Yet this criticism merely belies England’s failure to recognize Kant’s reliance on the principle of perspective. For the two views England claims are contradictory Kant would regard as arising out of two distinct perspectives—or rather, out of the dis­tinction between our ‘perspectiveless’ immediate experience (the ‘one’) and the per­spectives of our empirical knowledge (the ‘many’). That Kant never abandons such a position (as England suggests), but merely modifies it, is demonstrated by the evidence given throughout Part Four.

[10].    Cf. Wo78:107. In a section of Kt15 entitled ‘Existence is no Predicate ...’, Kant notes that from God’s Perspec­tive ‘no predicate at all can be want­ing’ [72-3(230-2)]. He then distinguishes between two human perspectives: the logical, which erroneously attempts to derive existence ‘from merely possible concepts’, and the empirical, from which existence is properly viewed ‘not so much [as] a predicate of the thing itself, as rather of the thought that one has of it.’ In the next sec­tion Kant further explicates this perspectival distinction: ‘Existence is the absolute [i.e., real, empirical] Position of a Thing and is thereby distinguished from every Predicate, which as such is always laid down but relatively to another Thing’—i.e., as ‘a mark’, functioning as a ‘respectus logicus’ [73(232); s.a. 82(245)]. Thus he recommends the proposition ‘God is an existing thing’ be rephrased as ‘Something exist­ing is God’, so that ‘the thing itself together with all the predicates is abso­lutely posited’ [74(234)]—i.e., as real from the empirical perspective. He makes the same perspectival distinction in a slightly different way in Kt15:75(235-6) [s.a. 79(241)]: ‘in an existing [thing] nothing more is posited than in a mere possible [thing] ..., but by some­thing existing more is posited than by a mere possible, for that refers to absolute position of the thing itself.’ (The term ‘thing itself’ in these passages must be regarded as referring to what Kant was later to call the ‘phenomenon’ [see KSP1:IV.3], since he had not yet formulated his mature theory of the thing in itself. This is especially clear in Kt15:124(308), where Kant says ‘things themselves ... constitute the raw materials of nature’; for the ‘thing in itself’ in systemt constitutes, as it were, the raw material of our knowledge. His dis­tinc­tions between ‘the logi­cal perspective’ and ‘the real perspective’ [83(247)] in gen­eral, and between ‘logical necessity’ and ‘real necessity’ [82(245-6)] in particular, are among the most significant contributions of Kt15.

[11].    In Kt1:645 Kant adds that ‘absolute necessity is a necessity that is to be found in thought alone.’ That is, absolute necessity cannot be real necessity, since it is con­fined to the logical and/or hypothetical perspectives.

[12].    Kt1:642-3; s.a. 639. Likewise, in Kt15:159(360) Kant says ‘whatever may be done to [cosmo­logical proofs], they never can become any thing else than conclu­sions from conceptions of possible things, but not from experience ...’.

                Wood defends an interpretation similar to the one presented here [Wo78:125]: ‘Kant’s criti­cism of the cosmologi­cal proof is not that the ontological serves it as a tacit premise, but rather that if we assume that the cosmological proof works, then we are committed to holding that the ontological argument works as well.’ After analyzing the logic be­hind the argument, Wood concludes that Kant’s position ‘does not follow’ [129]. However, Wood fails to take into con­sideration what was uppermost in Kant’s mind in making this claim: viz., the systematic and perspectival relationship be­tween these arguments in their purest forms. Given this relation­ship [see Fig. AIV.1], Kant’s position is not only consistent, but architectonically correct.

[13].    In Sw74:11-2 Swinburne condemns Kant for dividing all theoretical arguments into three distinct types on the grounds that teleological arguments in particular take a variety of forms, establish sometimes quite different conclusions, and may carry varying degrees of validity. Such a complaint is misplaced, however, inasmuch as Kant’s divisions are not intended as historically accurate classifications of the aggregate (i.e., empirical variety) of actual proofs. Rather, he is making an architectonic observation about how each perspective of systemt gives rise to an ideal type of proof. That the structure of some real proofs makes them difficult to classify does not invalidate the point Kant is making, any more than a scientist’s inability to determine the cause of specific event in nature would invalidate the principle of causality [see KSP3].

                Swinburne’s unsympathetic reading of Kant [see e.g., Sw74:17n,58n] is rather ironic, since his own theory of ‘prior probability’, focusing on the ‘explanatory power’ of a hypothesis, adopts the same (hypothetical) perspective as Kant’s possibility proof. Without recognizing any histori­cal connection with Kant, he even supports a position akin to the ‘primacy of practical reason’: scientific explanation is ultimately grounded in personal explanation [115-6]. What he apparently fails to realize is that, from the standpoint of scientific explanation, personal explana­tion must be regarded as regulative, not constitutive of knowledge. A careful back­ground study of Kant would have helped Swinburne to clarify his position in a number of such in­stances. Kant, however, rejects any attempt (such as Swinburne’s) to use theoretical arguments to prove ‘it is probable that there is a God’ [Kt69:299]; for ‘the expression probability is wholly absurd in such an application.’ This is because God, as ‘supersensible’, ‘transcends all cognition possible for us. Thus there is no way to achieve progress to the supersensible through the sensible ..., and so there is no approach to certainty and thus no assent whose logical value can be termed probability.’

[14].    Kt1:652. In Kt15 Kant likewise praises the ‘physicotheological method’ as ‘extremely sen­sible’ and ‘more natural than any other’, since ‘it furnishes a more intuiting conception of the supreme wisdom’ [117(297-8)]. Here the curious phrase ‘intuiting conception’ should not be regarded as a pre-Copernican reference to intellectual intuition, but rather as a precursor of his mature theory of schematism. Since schematism appears in the third stage of systemt [see Fig. III.6], Kant’s allusion to it here supports the placement of the physicotheological proof on the -+ pole of the map in Figure AIV.1.

                ‘The method has its faults’, Kant continues in Kt15:118(298), ‘though they are in­deed to be imputed but to the procedure of those who have used it.’ Kant’s main complaint is against those who, in using this method of proof, portray God as cre­ating individual things merely for their use by mankind, apart from their contextual role in nature, as determined by universal laws [see 118-22 (298-305)]. Thus, he agrees in Kt15:123(306-7) that ‘the cohesion of useful laws with a necessary unity [i.e., the ‘order and fitness’ in nature] ... afford a proof of a wise Author; though the dependence on him from this standpoint must be represented from another per­spective.’ For although physicotheology produces ‘a very beautiful argument, ... it is never capable of the strictness of a demonstration.... There is only one God and only one argument, by which it is possible to perspect his existence ... [viz.,] that the negation of the Divine existence is absolutely nothing’ [162(365-6)]. The main difference between Kt15 and Kant’s more mature theological reflection is that this new perspective is regarded in Kt15 as established by the possibility proof, whereas in his mature System it is established, as we saw in IV.4, from an entirely dif­ferent (‘moral’) standpoint. Recognizing the close architectonic connection between the hypothetical perspective and the practical standpoint [see KSP1:II.4] can help us to appreciate the high degree of affinity between his early and later writings on theol­ogy, despite their subtle differences.

[15].    Kt1:657-8. Wood complains that Kant ‘makes no real case at all that physicothe­ology, by its very nature, involves any hidden dependence on ... a priori argu­ments’ [Wo78:132]. But Kant’s point is not that such a dependence necessarily holds; rather it is that it holds if physicotheology is to operate with a sufficiently well-determined concept of God. For the validity of such a concept can be defended only by the a priori arguments. The relationship in question is parallel to that be­tween judgments and their corresponding concepts and intuitions. A judgment could, in principle, be made quite apart from any intuition and/or concept, but only at the cost of losing all sense and meaning.

[16].    Kt1:664. Walsh interprets Kant’s claim that the concept of God is itself ‘devoid’ of mean­ing by explaining that, although it is possible to use cate­gories beyond the limits set by sensible intui­tion in order to think the concept of God, to do so we would need to use pure categories, which ‘we simply do not know how to apply’ [Wa63a:265]. In other words, when Kant says the concept of God is unthinkable, he really means that it presents us with a thought that can­not be schema­tized, and thus, can never become an object of empirical knowledge. As Wiebe argues in Wi80:531, Kant’s view is really quite similar to the medieval theological view that ‘God, even if he is the first in the order of being [cf. the ground of all possibility] is still ... last in the order of knowing.’ The extent to which this similarity holds becomes most evident in XII.2, where we examine Kant’s view of the ontolog­ical priority of God as an immediately ex­perienced reality. Unfortunate­ly, Wiebe ignores the significance of the difference between Kantian and medieval theology, as entailed by Kant’s shift from a natural (i.e., theoretical) to a moral (i.e., practical) starting point.

[17].    Rohatyn, for example, takes Kant’s criticism to be an actual ‘disproof’ of God’s existence [Ro74:30]! But far from ruling out the use of analogy [31], Kant’s moderate treatment of the proofs prepares the way, as I argue in V.2-4, for a modern view of religious language as symbolic.

[18].    See KSP1:V. Kant draws a similar analogy in Kt4:101, where he says the distinc­tion ‘between the attributes of divine existence as independ­ent of all temporal con­ditions and that of a being in the world of sense ... is precisely that between the ex­istence of a thing-in-itself and that of a thing in appearance.’ But because both of them are transcendent and unknowable, the ideal and the thing in itself cannot, from any human perspective, be objectively differentiated. In both cases ‘the ab­sence ... of all definition [leaves us] merely with the conception of a not-sensible something con­taining the ultimate ground of the world of sense’ [Kt7:466]. This is what leads Kant to call the idea of God ‘a merely transcendental object, of which, as it is in it­self, we have no concep­tion whatso­ever’ [Kt1:725-6]. (In both cases the conceptual form of the object requires no faith; faith is needed only to posit a relation to some transcendent reality [see KSP1:V.2]. Of course, we use analogies to speculate about how they might be differentiated [724-5], as Kant does when he calls the ideal ‘the thing in itself as fully determined’ [604]. This implies that only the thing in itself (as well as its manifestation as the transcendental object) is ‘inscrutable’ in the strict sense: the ideal ‘can never be said to be inscrutable. For since it ... is in no sense given as thinkable object, it cannot be in­scrutable in the manner in which [the tran­scendental] object is’ [642]. Rather, our difficulty in com­prehending it has to do with the fact that it is based on ‘a mere idea’ [see IV.2].

[19].    Kt1:640. On the other hand, Kant would be no more sympathetic to those recent philosophers who regard theoreti­cal proofs as supplying not certainty but ‘probability’ [see AIV.2 and note AIV.13].

                  Copleston supports a position similar to Kant’s (though without acknowledging Kant’s agreement) when he says in Co74:11-2: ‘the existence of the One ... is affirmed by ... “philosophic faith”. The traditional metaphy­sical arguments for the existence of God make explicit for reflective thought this basic movement of the mind or spirit; but they cannot compel it. They presuppose and rest on philosophic faith. This is why it is possible for a man to admit the force of, say, Kantian criticism of the ar­guments and yet feel that there is something in them. For they represent a real act of transcending towards that which eludes man’s grasp.’ Many of the original proponents of the traditional arguments saw them in much the same way. Anselm’s famous defense of the ontolo­gical argument, for example, was not intended to convince unbelievers that knowledge of God’s exis­tence is necessary or even possible. On the contrary, Anselm clarifies his position at the outset: ‘For I do not seek to un­der­stand so that I may believe, but I believe so that I may understand. For I believe this also that “unless I believed I shall not understand” [Is. 7:9]’ [An65:115]. Thus Alston rightly interprets Anselm’s intentions as attempting to prove only an existence ‘in the under­standing’ [Al60:105], which says nothing about a real God unless it is accompanied by faith.

[20].    Some typical examples of the many summary accounts of Kant’s position on the theoretical proofs, aside from those discussed later in this section, are Co60b:294-307, Co67:102-11, Fo93, and Gr34:xlii-xliii [s.e. the three notes]; s.a. Ke23:522-40. In addition, almost any of the numerous introductory and/or critical interpretations of Kt1 listed in Part Two of the KSP1 Bibliography contain a section or chapter on the proofs. De94 is a somewhat awkward, though impres­sively focused study of Kant’s view of the proofs, convincingly arguing that there is a high degree of continuity on this issue throughout his writings; for a summary and detailed response, see Pa98. MacKinnon [Ma90a: 363] refers to C.C.J. Webb’s Problems in the Relation of God and Man and R.C.S. Walker’s The Coherence Theory of Truth as excellent attempts to defend the ontological arguments; unfortunate­ly, limitations of time have prevented me from consulting these works.

[21].    This tendency to misread Kant’s critique of the ontological argument is widespread. See e.g., Vi70, who argues that Kant is wrong to think existence is not a predicate, but constructs arguments that assume the empirical perspective rather than the logical [s.e. 363,365n]. For a list of several other articles related to the ontological argument but not discussed here, see De94:72n.

[22].    Nu94:124 and passim. This should read ‘theological postulates’, because the postulates on their own do not require religion. Deism suffices if we never progress beyond systemp.

[23].    See e.g., Kw83:111 and Va81; s.a. note VI.5. Perhaps the best known advocate of this position was Vaihinger, whom Kemp Smith reports as believing ‘that all our concepts of noumenal being, including that of Divine Existence, are but fictions’ [Ke23:609; see Va81].

[24].    When reading Kt9 we must always keep Webb’s reminder [We26:194] in mind, that ‘we have before us rather the juxtaposition of different points of view than a synthesis of them in a unified system.’ This is why Chapter XII is preceded by a chapter that attempts to locate the gap Kt9 is to fill in Kant’s System. Only by keeping in mind its place in the System can we hope to weave Kant’s diverse and conflicting comments into a coherent synthesis.

[25].    Wa72:164. Ward softens his claim [164] by explaining that Kant in­tends ‘to identify God with the unknowable, unitary, self-positing source of the rational structures of reality; and, in particular, with this source as it is expressed in the morally practical reason of the individual ratio­nal agent.’ To call God the source of practical reason is not as radical as to say God is practical reason. Yet Ward goes on to promote the stricter claim once again, saying that this ‘identification of God and practical reason ... is not explicitly stated in any prior works’ [165] and implies ‘God is identical with my real, or rational, self, and with the selves of all rational men.’

[26].    Kt9:21.19(Wa72:161). England highlights the same contradictory tendencies, aptly regarding them as different ‘point[s] of view’ [En29:199]. He cites Kt9:21.417 as an example of Kant’s view that God is ‘not a thing existing outside myself’ and Kt9:21.323,325,327-8 as some of the many passages where Kant writes freely as if God is an actual external reality. England also notes [En29: 199-200n] that Kant sometimes declares the idea of God to be a mere ‘Dichtung’ (‘fic­tion’) [e.g., Kt9:21.390,609], yet elsewhere explicitly denies this to be the case [e.g., 21.358,341]. And he observes that, although ‘Kant speaks of God as “a mere Idea of reason”’ [En29:200, quoting Kt9: 21.410], ‘this “mere Idea” is said to have the greatest inward and outward reality.’ (England’s references are to Reicke’s version of Kant’s Werke, not to the AA edition.) Such apparent contra­dictions may be what led Jaki to call Kt9 ‘a classic morass of subjectivism’ [Ja81:11]; yet such an assessment can be avoided by appealing to the principle of perspective [see below and Ch. XII].

[27].    Kt8:6(5-6). One of Kant’s clearest statements in this regard comes in Kt2:361, where he says the ideas of reason depict ‘the relation between that which lies beyond [the boundary of our knowledge] and that which is contained within it.’ Kant goes on to explain that reason is ‘constrained to look beyond this boundary’ and that in so doing it leads us ‘to something which is not itself an object of experience but is the ground of all experience. Reason does not, however, teach us anything concerning the thing in itself ...’. Likewise, in Kt7:474 Kant argues that freedom gives us the basis for determining the idea of God in a way that constitutes (practical) ‘knowledge’ not only of ‘the supersensible within us’, but also of ‘the supersensible without [i.e., outside of] us’. Such passages might imply an ‘imaginative projection’ [Da93c:340-4]; but they surely confirm the link made in note V.21 between Kant’s theory of God’s nature and Tillich’s notion of God as the Ground of Being.

[28].    Ward sometimes portrays Kant’s position as being more self-contradictory than it actually is, as in Wa72:163, when he quotes Kt9:22.117 as saying ‘God is “the product of our own reason”.’ For Kant actually says this about our idea of God, not about God as such. Nevertheless, Ward does aptly summarize ‘[t]he whole point of Kant’s argument’ [Wa72:164] as being ‘that one should come to recognise that the dichotomy “real/fictional” simply cannot apply to the concept of “God”. The concept is necessary but practical.’

[29].    See IV.1 for more on Heine. Gulick questions the ability of a ‘purely intellectual abstraction’ such as ‘God’ to ‘motivate moral behaviour’ [Gu94:103] and claims that ‘God as architect and creator is a lifeless deistic being serving the philosophical quest for unity and order.’ But surely the symbol of an architect connotes a God who is alive [see note IV.24]!

[30].    S.e. IV.4, V.4, and VI.2; cf. Lo81:303. McCarthy [Mc86:63] affirms that ‘those who would interpret Kant’s position in [Kt8] as ultimately deism’ must admit after reading Kt26 ‘that Kant himself regarded his God-concept as theistic.’ McCarthy later [90-1] develops this view in a way that is strikingly similar to the position I defend in VI.4, though I had already published the article version of that chapter when I first came across Mc86: ‘Christianity is not ... simply morality. It is moral theism.... [It] is beyond both the plain theism of natural theology and the deism of transcendental theology. If this definition reduces the life of religion to morality, it does not make religion thereby merely equivalent to morality.... [Rather, it] elevates Christianity above morality to moral religion.’ McCarthy’s account of Kant’s ‘moral theism’ differs from mine in a number of ways, however, as when he claims [104] Kant’s ‘theos is remote, transcendent, and impersonal.’ See V.2-4 and XII.2,4 for a refutation of this claim.

 

Back to the Table of Contents for this book.

Back to the listing of Steve Palmquist's published books.

Back to Steve Palmquist's home page

Send comments to the author: StevePq@hkbu.edu.hk