1. Introduction: the role of moral faith
Attempting to grasp the proper role that the practical postulates play in Kant’s system is akin to walking a tightrope. On the one hand, the reader must accord them their rightful place as the objects of subjectively necessary belief. Introduced as part of the solution to practical reason’s dialectic in the Critique of Practical Reason, Kant argues we must believe in our freedom, immortality, and the existence of God to affirm the real possibility of the highest good—the final end of pure practical reason. Without the reality of these ideas, we could not explain the synthetic connection between the highest good’s two elements, virtue and happiness. It would remain unintelligible how a virtuous life might also, nonaccidentally, be a happy one. From this perspective, the practical life of the human being who necessarily seeks both ends is thrown into question; and in its search for an answer, practical reason runs the risk of straying into its dialectical use and inadvertently corrupting its own foundations. Rescuing us from this state of conflict, the postulates safeguard the moral disposition. They are thus a crucial element in Kant’s system, bearing a great deal of weight.
In filling out this account, however, Kant’s interpreter must be careful not to inadvertently undermine the theoretical edifice she is trying to support. Here it helps to separate the postulates into two groups associated with different interpretive difficulties. On the one hand, the concept of freedom is so integral to Kant’s picture that its status as a mere postulate, an object of pure practical rational belief, can seem far too weak. His interpreter must puzzle over the fact that Kant seems to offer no explicit argument for freedom as a postulate, and reconstruct one that is consistent with freedom’s logically reciprocal relationship to the moral law. On the other hand, Kant’s assertion that we must necessarily believe in our immortality and the existence of God can seem too strong, appearing to be completely foreign to the spirit of his enlightenment project. Kant, of course, explicitly distances himself from those of his predecessors who would ground the authority of moral command in the will of a divine being (KpV5:125–26).1 But when discussion turns to the postulates of immortality and God, it is common for his interpreters to wonder whether antiquated ideas have taken up an unfortunate position in Kant’s otherwise progressive system, representing a conscious or unconscious need to incorporate into his view elements of the religious culture that dominated his time.2
Even those who are sympathetic to the role that Kant assigns to religious belief struggle to square it with the rest of his system. As with many other issues in Kant scholarship, the field of interpretive options can be divided in two, in accordance with whether commentators emphasize (or overemphasize) the role of reason or sensibility as Kant’s primary explanatory concern, thereby giving rise to a rationalist or empiricist take on the issue at hand.3 While rationalists take the postulates to answer to concerns about the consistency or seeming irrationality of moral behavior, empiricists take the postulates to answer to concerns about the possibility of psychological motivation, or what moves the agent insofar as she is a sense-dependent being. To pave the way for my own account, I will begin by showing how both routes are attended by characteristic difficulties whose source is an unnecessarily dualistic take on Kant’s view of moral psychology. To successfully cast the practical postulates in the role Kant envisages for them, we must abandon this framework for a more unified account of human mindedness. The alternative view I suggest could be called “sense-dependent rationalism” to capture its emphasis on the sensible form of rational life we occupy, as opposed to rational norms in isolation from sensibility (pace the rationalist) or psychologically determined motivations that float free of reason (pace the empiricist). As I will argue, we must head in this interpretive direction in order to show how the postulates function as subjectively necessary but nevertheless rationally grounded elements of human moral life. After suggesting a broad framework for interpretation that can apply equally to all three postulates, I will show its particular application by turning to the postulate of immortality.
2. Leaving dualism behind: dispensing with the empiricist vs. rationalist framework
Turning now to some paradigmatic examples of rationalist and empiricist interpretations of the practical postulates, we can see how these particular readings are framed by broader interpretive commitments. The empiricist, for example, confines the postulates to the subjective, sensibly determined side of human morality and, in particular, the role it plays in explaining motivation. Paul Guyer’s account serves as a prominent example, for he insists that “the entire doctrine of the postulates of practical reason is stated within the limits of human psychology” (2000, 367). On this reading, belief in freedom, God, and the afterlife is helpful insofar as it provides added motivation to do what we otherwise would, unflinchingly, were we less flawed creatures.4 Disconnected from our rational capacities, belief in immortality and God “have no recommendation except that they are effective in motivating creatures like us to act in the way and toward the end that reason demands” (367).
Though Kant can seem to speak in terms that support this account, it runs afoul of his most basic insight concerning moral value: namely, that willing is good in virtue of its form and not its object. All willing needs an object, so the mere fact that one is aiming at an end need not corrupt the moral principle. But if this material end is incorporated into the determining ground of the will, Kantian morality will no longer be autonomous in character. If, for example, belief in God is tied to the reward of everlasting happiness, and this latter hope is what actually motivates the subject to do the right thing, her action is dependent on a material as opposed to formal ground. Rather than being determined through the form of its principle, willing is determined by reference to its object—happiness. And since Guyer takes the postulates to motivate the subject by appealing to our subjective psychology insofar as it is insulated from reason, it is difficult to substantiate his reading without appealing to this material determining ground of the will. If human beings need to anticipate a divinely bestowed everlasting happiness in order to do the right thing, even some of the time, purely formal, autonomous willing seems beyond our grasp.
Guyer is of course aware of this objection. In response, he insists that Kant’s morality is complex enough to incorporate representations like the postulates, which motivate without serving as competing or corrupting incentives (2000, 368). I agree that Kant’s position can accommodate this possibility but would deny that it can do so as Guyer construes it. For he insists that Kant explains both the need for, and possibility of, the practical postulates through his “profoundly dualistic conception of human nature” (365). I would contend that it is precisely this dualism which rules out the subtleties Guyer depends on to substantiate his view. On his reading, reason and sensibility are viewed as wholly independent capacities whose different laws and interests pull in opposing directions, making it difficult for either to attain its ends. This complicates the philosophical explanation of human morality: to fully account for the possibility of our acting in accordance with practical reason’s laws, Guyer thinks we must also say something about how sensibility interferes with these laws, and—more importantly—how, in virtue of this interference, we must cater to sensibility’s own whims to allow reason to get where it wants to go.5 Now consider how this clean separation of the interests and operations of reason and sensibility creates difficulties for the psychological complexity Guyer wants to find in Kant. He claims that representations like the practical postulates of freedom, immortality, and God can motivate moral behavior without corrupting it, while also insisting that the point of these representations is to cater to our particular subjectivity, our sensibility as it functions independently of reason. Indeed, Guyer goes so far as to insist that these representations have absolutely no rational merit beyond the base requirement of being noncontradictory. They motivate, instead, by appealing to our sensible nature, speaking to sensibility on its own terms, in virtue of its own interests.6 Having conceded this much, it’s hard to see how the practical postulates could appeal to a nonrational capacity and its amoral, subjective end of personal happiness without corrupting the objective interests of reason which run in precisely the opposite direction. Put another way, one cannot establish reason and sensibility as wholly separate capacities in conflict and declare that the postulates are only effective for the latter, while also suggesting that we can frame the postulates as representations which motivate without appealing directly to sensibility’s particular laws and interests.7
While the empiricist undermines the moral law by offering an account of the postulates that is too subjective, driven by the material end of happiness, the rationalist goes to the opposite extreme, turning the postulates into logically necessary objects of rationally determined belief. As we will see, this too undermines the moral law. Allen Wood’s account from Kant’s Moral Religion serves as a paradigmatic example.8 Wood seizes on one of Kant’s remarks from his Lectures on the Philosophical Doctrine of Religion, where he claims that the argument for moral faith is an “absurdum practicum,” such that “if I deny this or that I would have to be a scoundrel” (LR 28:1083). Picking up on the close analogy to a theoretical absurdum logicum Kant draws in the lecture, Wood suggests we read Kant as suggesting that we implicitly commit ourselves to certain practical beliefs in aiming at the highest good:
Assume that I deny either the existence of God or of a future life. Now if I deny either of these, then I cannot conceive the highest good to be possible of attainment. If I deny that I can conceive the highest good to be possible of attainment, then I presuppose or imply that I will not pursue the highest good, or commit myself not to pursue it.(Wood 1970, 29)
Wood’s rationalist reading of the postulates thus proceeds on the basis of a tight analogy with theoretical reason. He assumes that the postulates are necessary objects of belief in a logical sense, which concerns the apparent rationality of one’s actions. Consider his thoughts on the faithless subject who pursues moral ends without believing in the possibility of the highest good: “in this case I will have to admit that I am acting ‘irrationally’ and that according to my own beliefs I should (in a logical, but not a moral sense of ‘should’) give up my pursuit of the highest good and my obedience to the moral law, and become a Bösewicht [scoundrel]” (1970, 30). Such an agent would effectively find persuasive logical grounds to abandon the moral law.
We can conduct a general diagnosis of the issues facing Wood’s account by invoking the dualistic framework already attributed to Guyer. As the rationalist version of this framework, Wood’s account runs into trouble because it attempts to interpret the postulates by appealing to reason alone, construed as wholly independent from sensibility. But as others have pointed out, claiming there is an objectively or logically necessary connection between recognizing the force of the moral law and believing in the postulates undermines the objective validity of the moral law, effectively releasing the faithless from moral obligation.9 If one thinks there are no grounds for accepting an objectively necessary condition of the moral law, then the moral law itself, by extension, will fail to obligate as an unconditionally valid principle of practical reason. So, tying the postulates too closely to the moral law will cause the slightest doubt with respect to the former to destroy the force of moral obligation.10 Against Wood’s rationalist reading, we must therefore heed Kant’s insistence that the ideas of freedom, immortality, and God constitute only subjectively necessary objects of belief. Fully respecting this aspect of Kant’s view requires showing how the postulates speak to a subjective need grounded in our sense-dependent reason whose fulfillment nurtures the interest in morality that we always already have and cannot lose through any kind of logical force.11 Indeed, given Kant’s insistence on the moral law’s ability to move “even the most hardened scoundrel” (G 4:454) to the approval of good conduct, we should doubt whether the seeming irrationality of committing oneself to the moral law really is a live possibility, even in the absence of morally based faith. As I will go on to claim, the issue is rather the hopelessness that one’s moral vocation seems to present insofar as the highest good is thought to be impossible.12 The universal, objective grounds motivating moral action cannot cease to be compelling, but one can easily begin to doubt whether the world has been set up to facilitate such action, and this has the power to stop us in our tracks. To give a proper account of the postulates, we must therefore acknowledge that their necessity speaks to holistic, world-directed concerns that are intimately bound up with subjective motivation—though, pace Guyer, this motivation is not to be construed as speaking merely to sensibility.
One might object, on Wood’s behalf, that there is an important motivational component to his view insofar as the agent is supposed to rest uneasy with the logical conclusion that rejecting the postulates logically requires her to give up her pursuit of the highest good. Because she wants to be moral, the defense goes, she has a vested interest in not regarding herself as a moral failure, and thus in doing what it takes to avoid this.13 But I would deny that this account of the motivational component captures Kant’s purposes in introducing the postulates. To see this, we can attend to remarks he makes about their motivational power in section 87 of the Critique of Judgment. I take his discussion to reveal that the issue at stake in showing the real possibility of the highest good is fundamentally the intelligibility of our moral vocation, and the hopelessness that one would feel if the highest good were thought to be impossible. This strongly suggests that the motivational aspect of the postulates concerns not a movement inward to our self-image and desire to be good, but a movement outward to the structure of the sense-dependent world, and the way that even firm moral convictions are tested by its limits. Kant is here focused on the particular postulate of God, and it would be illustrative to remain with it for a moment. The worry is that the heterogeneity of moral and natural laws gives rise to despair about the possibility that virtue and happiness could be synthetically connected in the manner required by the highest good. Kant insists that the righteous man lacking faith will be faced with the following scenario:
Deceit, violence, and envy will always surround him, even though he is himself honest, peaceable, and benevolent; and the righteous ones besides himself that he will still encounter will, in spite of all their worthiness to be happy, nevertheless be subject by nature, which pays no attention to that, to all the evils of poverty, illnesses, and untimely death … The end, therefore, which this well-intentioned person had and should have had before his eyes in his conformity to the moral law, he would certainly have to give up as impossible.(KU 5:452)
As Kant goes on to insist, this realization would weaken the moral disposition. But the issue here is not that the agent is worried about the logical consequences of his faithlessness—he logically must be or will be a scoundrel—rather, he is dealing with the morally relevant consequences of his sense-dependent intellect and the way these subjective factors shape his view of the cohesiveness of moral life. Because the natural laws that determine the sensible world do not in any way speak to morality, he reaches an impasse. How could the highest good be possible? This picture suggests the subjective motivational component related to the postulates is focused on overcoming the limits of sense-dependent reason and the way they inhibit our ability to form a concept of the real possibility of the highest good. Lacking this rational content, our moral disposition will suffer. So, we need the postulates, not to reassure us of our moral worthiness, but to fully make sense of the moral world we inhabit. All this implies that there are both rational and motivational aspects to the postulates that dovetail more perspicuously than either Guyer or Wood recognize in their respective accounts. By answering to concerns about the possibility of a cohesive moral world, the postulates represent a rational achievement, based in our subjectivity, that has the power to motivate us towards the good.
This result suggests we need a different approach to account for the practical postulates, one which does not impose a dualistic understanding of reason and sensibility onto Kant’s position. Rather than tearing apart our rational, objective and sensible, subjective grounds for acting, I will argue that we should understand them as inseparably connected. This new interpretive path is grounded in a general principle that has the potential to reshape much of our understanding of Kant’s practical philosophy. It can be expressed as follows: the cognitive content of a practical representation and its capacity to motivate the subject should not be understood as fundamentally separable from one another. A central tenet of the reading under development is that Kant’s equation of the will with practical reason (cf. G 4:412–13) entails that our ability to will an end cannot be sharply separated from our ability for conceptually articulated thinking about that end. Hence acts of the will must be understood through reference to the cognitive standing of their representational content as much as through reference to the efficacy that marks them as acts of the faculty of desire. Against Guyer’s dualism, we must resist the Humean line of thought that associates motivational “oomph” with sensible impulse, positioning sensibility as an independent source of motive. It is certainly true that our sense-dependence complicates the picture by introducing potential material ends that do not necessarily agree with the laws of practical reason, and our subjective grounds for acting can fail to track what reason requires. But none of this entails that sensibility must float free of reason’s influence, reflecting a Janus-faced conception of agency divided between impartial reasoning and unbridled sensible desire. Our imperfect rationality expresses itself through poorly executed acts of reasoning, not isolated moments of utter irrationality shaped by empirical psychology.14 In making this cognitivist turn, however, we must be equally careful not to completely overlook that the postulates, as practical, are meant to speak primarily to a motivational issue that applies specifically to sense-dependent beings like us. Against Wood’s dualism, we cannot conclude that moving away from a psychological reading of the postulates reduces them to logically necessary offshoots of morality that we accept on pain of absurdity, to square up our rational beliefs. Catering only to rationalistic worries of this kind fails to capture the role the postulates are meant to play in Kant’s system. Rather than serving as the logical yoke we must bear if we want to keep our moral dignity intact, we should think of the postulates as providing an important cognitive and motivational resource that we need on subjective grounds, owing to the structure of our sense-dependent reason. Although a nonrational psychologized desire for happiness plays no part in explaining the practical postulates, our sense-dependence and the limits it places upon cognition, and thereby motivation, certainly does.
Applying the abovementioned cognitive-motivational principle to Kant’s account of the practical postulates, I will show that they provide motivation through rendering the subject’s practical rational representations themselves more determinate. This cognitive work enables a more efficacious representation of one’s moral vocation, one that is resistant to doubt and uncertainty. The upshot of this argument then, is that belief in the postulates contributes to the moral disposition from within the practical cognitive system itself. Belief motivates by adding to the determinacy, the representational content, of the subject’s concept of the highest good and its real possibility, and it is this cognitive achievement, this sharpening of the concept, which accounts for pure practical reason’s increased efficacy.
As I will argue, the postulates respond to limits on our cognition that can be directly attributed to our sense-dependence and the way it affects both the theoretical and practical use of reason. It is these limits which condition doubt or uncertainty about the intelligibility and thus possibility of the highest good. Belief in each of the postulates allows us to transcend a particular limit by extending the content of our thought in a certain direction. Each of these extensions involves determining an idea of reason through one of the categories of relation—substance, causality, and community—and providing a more determinate representation of how the highest good could be reached. In the case of immortality, which will be my focus here, this added determinacy is attained by thinking a key condition on the possibility of the highest good through the category of substance. Through this cognitive work, we think the kind of being capable of realizing the completely moral disposition commanded by the highest good. And since, on my nondualistic picture, the cognitive content of a representation is immediately related to its subjective efficacy—its capacity to motivate—this added determinacy strengthens the moral disposition.
So, we will see that belief in the postulates of faith does indeed spring from a subjective need, but one that is grounded a priori in the structure of sense-dependent reason itself. Kant’s account of the postulates of faith is thus situated well within the limits of his metaphysics of mind: though based on a subjective need, we need not appeal to anything other than the fact of our sense-dependence to account for the role that the ideas of freedom, immortality, and God must play within the Kantian system. Interpreters therefore need not turn to empirically based psychological factors about human emotions to account for Kant’s position. In light of this reading, it will also be much more difficult to argue that the postulates of faith represent an undesirable “tacked-on” element of Kant’s moral picture that is spurred by extra-philosophical religious concerns.
3. The practical postulates: Theoretical propositions grounded in pure practical reason
Each postulate requires its own unique argument to explain its place in Kant’s system. But we can nevertheless turn to the concept of the practical postulate in general and enumerate a few necessary features common to each account. The first important point to grasp is that Kant thinks of the postulates as propositions of speculative reason asserted on practical grounds. Thus, although the fact of their necessity is determined by moral laws, their assumption itself is an act of speculative reason in general, what Kant calls “a theoretical proposition, though one not demonstrable as such insofar as it is attached inseparably to an a priori unconditionally valid practical law” (KpV 5:122). Kant thus distinguishes between what justifies our belief in the postulates, or what the postulates reference as grounds, and the speculative activity of judging them to have objective reality. While the former are purely practical cognitions of the moral law and its necessary object—the highest good—the latter are theoretical or speculative propositions in their own right.
Second, though the postulates are practically grounded theoretical propositions, they are not the objects of theoretically determinate cognitions. Theoretical reason is forced to grant “that there are such objects” (KpV 5:135) because of a need of practical reason, but it cannot further determine them for its own purposes, restricted as it is to objects of possible experience. Kant thus does not violate the conditions of theoretical cognition set out in the first Critique in understanding the postulates as objects of speculative reason in general. He remains committed to the idea that freedom, immortality, and God are supersensible ideas of reason that could not be presented in sensible intuition as objects of theoretical cognition. We cannot know how any of these ideas are theoretically possible, but we can and must assume their reality on practical grounds. Insofar as they are grounded by practical need, as opposed to sensible experience or a priori cognitions about the conditions of the possibility of sensible experience, the practical postulates do not extend theoretical cognition. Hence, they are more appropriately characterized as the objects of belief.
Third, practical justification for the postulates is obtained insofar as the objective reality15 of these ideas of reason must be assumed to answer the question that spurs sense-dependent practical reason’s dialectic—namely, how is the highest good possible? This is the subjective practical need of which Kant speaks. But, importantly, the justification to assume the objects that fulfill this need does not originate merely from the need itself. To take this reading would be to psychologize a key aspect of Kant’s moral metaphysics. The justification for assuming the postulates is better understood as follows: because reason recognizes the highest good as its necessary end in connection with an apodictically certain moral law, it must assume the objective reality of the conditions required to realize this end.16 In other words, because the moral law commands us to realize the highest good, and the moral law is unconditionally valid, the highest good cannot be an empty concept—it must be really possible, and we can assume the objective reality of any ideas we recognize as required for its real possibility. Thus, in postulating the conditions necessary to realize the highest good, the will, as Kant says:
… obeys an inflexible command of reason that has its ground objectively in the character of things as they must be appraised universally by pure reason and is not based upon, say, inclination, which is by no means justified in at once assuming, for the sake of what we wish on merely subjective grounds, that the means to it are possible or that its object is real.(KpV 5:143)
But, importantly, though there is an objective and universal ground for assuming the possibility of the highest good, the particular postulates that we make in order to show this possibility are bound up with the nature of sense-dependent reason, and so not objectively but subjectively universal in character. Kant is very clear about this a few pages on from the passage just cited. After reiterating the objective need to promote and thus grasp the possibility of the highest good, he qualifies:
But as for how we are to represent this possibility … reason cannot decide this objectively. Now a subjective condition of reason enters into this, the only way in which it is theoretically possible for it to think the exact harmony of the realm of nature with the realm of morals as the condition of the possibility of the highest good.(KpV 5:145)
Here, Kant acknowledges that how we think the various aspects of the highest good are to be realized—through, e.g., our immortality, or the existence of a God who orders the universe according to moral laws—is bound up with the limits on and resources available to our sense-dependent discursive reason.17 This is not to say that we arbitrarily choose which postulates to assert: they are rationally necessary, but insofar as they are connected to our particular form of reasoning they are also subjective in character. In what follows, I will examine this idea in more detail, showing how the particular limits of sense-dependent reason necessitate the postulate of immortality.
To summarize, the practical postulates are thus theoretical propositions that affirm the objective reality of supersensible objects on the basis of practical grounds that do not, as practical, extend the limits of theoretical cognition. Insofar as freedom, immortality, and God are recognized as conditions on the possibility of a necessary, objectively valid end, pure practical reason is licensed to assume their objective reality. Ideas that would otherwise be transcendent for speculative reason thereby “become immanent and constitutive inasmuch as they are grounds of the possibility of making real the necessary object of pure practical reason (the highest good)” (KpV 5:135). In this way, pure practical reason is assured of the possibility of the highest good, and the moral disposition is protected.
4. The postulate of immortality
4.a Nonholiness as subjective limitation
Beyond this general account, each postulate answers to a particular worry that arises from the subjective limits of sense-dependent reason. To fill out my reading in more detail, I will focus on the postulate of immortality, which has recently received sustained attention.18 The subjective need for this postulate enters the picture because the command to realize the highest good contains the command to realize a disposition that is in complete conformity with the moral law. This containment relation holds because the highest good is both the supreme and complete good of pure practical reason. That is, it does not merely represent morality as our most noble or worthy end, the highest good also represents a complete totality of all ends insofar as they take the moral law as their determining ground. So, the command to realize the highest good entails that all willing, insofar as it is directed at the highest good, takes the moral law as its determining ground. At this point, the limits of reason assert themselves and a subjective need emerges. For the complete conformity of one’s disposition with the moral law is holiness, and Kant is adamant throughout the practical works that holiness is “a perfection of which no rational being of the sensible world is capable at any moment (Zeitpunkte) of his existence” (KpV 5:122). All that is possible “for a rational but finite being” is “an endless progress from lower to higher stages of moral perfection” (KpV 5:123; emphasis mine). The appeal to immortality thus emerges because something about our sense-dependence precludes the possibility that the moral law could completely determine our conduct.
My account will turn on the claim that this holiness is precluded simply insofar as the subject is dependent on sensible desires that could potentially serve as material determining grounds of the will. The possibility of the highest good is thus thrown into question on grounds that are inseparable from the nature of practical discursive reason itself. This is a crucial step in the present argument for the postulates. Kant and his readers often appeal to our inscrutably evil nature to explain certain elements of the practical picture, but we must take care to note that this reference to our fallenness is couched less in literary and psychological trope than it is grounded in a metaphysical account of our finitude. It is the distinction between intuitive and discursive intellects, understood in its practical guise, which ultimately explains the need for the postulate of immortality.
There are a number of distinctions tracing the difference between a discursive and intuitive intellect that Kant implicitly or explicitly draws upon in discussing the need for the postulate of immortality. Unpacking these will shed some light on the nature of discursive reason and its associated limitations. First, Kant’s appeal to the holy will invokes the distinction between subjectively necessary and subjectively contingent acts of practical reasoning. In the Groundwork, this distinction is introduced as Kant determines the two most basic types of practical reasoning: that which proceeds infallibly in accordance with reason and that which depends in part on the subjective conditions of sensibility (G 4:412). The former type of willing is characteristic of a holy being who is completely self-sufficient and chooses only what universal laws of morality command. The latter type describes the nonholy will of a being whose capacity to reason, though self-determining, is not completely self-sufficient and self-producing. What makes the crucial difference here is that the holy will is homogenous and completely unified, producing everything from reason alone, while the nonholy will is heterogenous, depending on reason and sensibility, and so subject to other determining factors—namely, sensible needs and desires—that are not necessarily in conformity with reason. This entails that, while the holy will necessarily acts in complete conformity with the moral law, the nonholy will only contingently does, if it does so at all (G 4:412–13). Because sensible desire is also necessarily incorporated into the activity of the will as its matter, it can influence practical reason’s activity. So, although our practical rational activity is subject to a necessary law expressive of the form of practical cognition, it is quite contingent whether this moral law actually determines the will.19 The finite subject’s dependence on the deliverances of sensibility thus opens up the possibility that sensible matter might be wrongfully incorporated into the will. Put even more abstractly, though reason strives for unity, the heterogenous elements required for its activity entail that it is essentially a combinative, synthetic power, and this foundational aspect of its character accounts for the possibility that such combination might be badly executed. Since we cannot escape this possibility, our moral disposition is by nature contingent, and the best we can hope for is still “only virtue,” a disposition accompanied by consciousness “of a continuing propensity to transgression or at least impurity, that is, an admixture of many spurious (not moral) motives to observe the law, hence a self-esteem combined with humility” (KpV 5:128). This vulnerable commitment to our moral vocation cannot be overcome, and so the complete moral disposition required for the highest good to be possible seems out of reach.
We can add to the nature of this contingency by attending to a second distinction: namely, that between a moral disposition which is complete and holistically expressed, and a moral disposition which is incomplete and drawn out.20 The latter captures our form of practical reasoning, which, as discursive, unfolds through discrete acts of judgment that involve universal formal constraints expressed through principles and concepts, but also particular instances of temporally determined sensible matter. Understood as a judgment incorporating sensible matter that takes place in a particular context, at a particular moment, such activity cannot be expressive of a complete disposition that encompasses the subject’s practical determining grounds considered as a totality. This is a fundamental difference between the discursive, sense-dependent, nonholy will, and the nondiscursive holy will. It is not just the presence of sensible desire that rules out our holiness, making it possible for us to act in a way that does not agree with the moral law. Each act of will—contingent in itself—also unfolds through an act of subsumption that cannot be representative of a complete disposition, whether moral or immoral. Regardless of whether or not our desires conflict with the constraints of morality, our dispositions are necessarily incomplete because our sense-dependence entails that their realization is spread out over time, constituted by a manifold of particular acts. Unlike the nondiscursive holy will whose perfect unity and self-sufficiency can represent itself as a complete whole,21 our discursive practical rational activity unfolds through the application of principles and concepts to the manifold sensible factors constitutive of a particular practical context. Instead of perfect, infinite, eternal self-realization, we have acts of reasoning that are by nature autonomous—determinable according to laws of pure reason—but nevertheless not wholly self-sufficient and very much in progress. Our willing is essentially ampliative, gaining further determination, strength, and improvement through the particular acts of subsumption that bring sensible matter to the unity of universal, formal principles. Given the seemingly infinite array of sensible particulars to incorporate, and the precariousness of our subjectively contingent moral success, this ampliative work is never done. Spread out over time in this fashion, our mode of practical reasoning or willing precludes the complete moral disposition.22
Our sense-dependent discursive practical activity is thus by nature both contingent and incomplete. The act of incorporating sensible desires into the will, basic to our practical form of life, fixes us as beings who must realize our moral dispositions over time through discrete acts that can each and all fail to live up to the dictates of practical reason. Hence Kant’s conclusion that for finite, sense-dependent beings, “virtue is always in progress and yet always starts from the beginning” (MS 6:409). As autonomous beings interested in the moral law, we recognize the moral law’s necessity as the only legitimate determining ground of the will. But our sense-dependence entails that the progress we make towards virtue is necessarily faltering and incomplete. Thus, when Kant rules out our holiness on subjectively contingent grounds having to do with our discursive, sense-dependent nature, we should not only think of the account of evil that he develops in Book 1 of the Religion. Lest we mistake this metaphysically grounded picture for a scripturally based commitment to original sin, we must recognize the extent to which humanity’s evil nature is, for Kant, bound up with the fundamental structure of our mindedness. We fail to be holy not just because we encounter wayward inclinations that can become the ground of evil action, but because the character of holiness implies a kind of perfection and completeness that can never be exhibited by beings who exercise their practical rationality in judgment. In fact, as we are now in a position to see, the inclination-centered explanation of evil that talks only of the precedence humans give to selfish motivations depends upon this more general metaphysics of mind for its foundation.23
4.b Immortality and the category of substance
Given this subjective limitation, the complete conformity of moral disposition that is precluded through a critique of our practical reason must be accounted for in another way. Since we are not holy, Kant concludes, we can only attain the requisite disposition through a moral progress that seems to us endless, but which conceived as a totality can demonstrate the required commitment to morality. For this endless progress to be possible, we must assume immortality of the soul, what Kant describes as “the existence and personality of the same rational being continuing endlessly” (KpV 5:122). The concept of immortality is thus what enables us to represent the possibility of a complete disposition that transcends the capacity of our sense-dependent discursive intellect, but is nevertheless a necessary requirement on the possibility of the highest good.
The idea of immortality helps us here because it functions as a sense-dependent analogue to the kind of representation that could grasp a complete and perfectly moral disposition. As my earlier analysis suggests, the possibility of representing ourselves as completely moral thus emerges as an impossibility as much because of our powers of representation as because of our moral shortcomings. The idea of immortality is, in other words, the best possible way for the discursive intellect we possess, limited as it is, to represent this requirement. The representation has two aspects: first, that of moral perfection; and second, that of completeness of disposition, both of which are accounted for through the idea of an endless progression, ensuring we have the infinite capacity to exercise the particular, sense-dependent acts of judgment constitutive of our moral improvement. Given that our moral lives unfold in this gradual way, through painstakingly slow increments, we need an infinite amount of time, of exercises of judgment, to get where we’re going and realize a complete and morally perfect disposition. To cement this idea, Kant imagines how our disposition would look to the intuitive intellect not dependent on sensible conditions like temporality:
The eternal being, to whom the temporal condition is nothing, sees in what is to us an endless series the whole of conformity with the moral law, and the holiness that his command inflexibly requires … is to be found whole in a single intellectual intuition of the existence of rational beings.(KpV 5:123)
Through the idea of immortality, we come as close as possible to representing ourselves as free from the subjective limitations conditioned by our sense-dependent discursive intellect. The idea of endless progress thus functions analogously to the type of representation enjoyed by a nondiscursive being, who grasps the moral disposition of a finite agent as a complete whole unified by that agent’s respect for the moral law.
To arrive at this representation, pure reason must abstract from the flux and inconstancy of particular sensible determinations to reveal the underlying idea of a permanently existing subject who grounds or unifies these particular determinations. This suggests we think our immortality through the unschematized category of substance, thus uncovering practical reason’s positive use of the categories. Kant hints towards this extended use of the categories at the end of the “Transcendental Deduction” in the first Critique. Having established that the categories can only yield theoretical cognition through their application to sensible intuition, he notes:
So that one may not prematurely take issue with the worrisome and disadvantageous consequences of this proposition, I will only mention that the categories are not restricted in thinking by the conditions of our sensible intuition, but have an unbounded field, and only the cognition of objects that we think, the determination of the object, requires intuition; in the absence of the latter, the thought of the object can still have its true and useful consequences for the use of the subject’s reason, which, however, cannot be expounded here, for it is not always directed to the determination of the object, thus to cognition, but rather also to that of the subject and its willing.(B166; final emphasis mine)
We should thus expect the categories to find further use in the practical sphere, and it is precisely in the immanent extension of practical reason through the postulates that this use is found. For each of the ideas of immortality, freedom, and God, practical reason finds itself licensed to employ a category to determine an idea of reason from the practical perspective, and thereby show the possibility of the highest good. The relevant categories here all come from the same subheading: the categories of relation, which are used to specify various metaphysical relationships—between inherence and subsistence, cause and effect, and the reciprocity between agent and patient (B106)—to show how the highest good, or some part of it, could be brought about.24 In the case of immortality, we discover the specific need to determine ourselves through the category of substance to reach an analogue of nondiscursive representation that allows us to think ourselves as capable of possessing the kind of completely moral disposition required for the highest good to be possible.25 Our subjective need to account for the possibility of the highest good thereby “lead[s] through the postulates of practical reason to concepts that speculative reason could indeed present as problems but never solve” (KpV 5:132). So, while speculative reason could only think the ideas of immortality, freedom, and God, practical reason finds itself entitled to postulate their reality as objects.
We find more evidence that the categories must be involved here by turning to these particular “problems” from the first Critique, which the second Critique is able to solve. Kant traces these parallels in a section entitled “On the Postulates of Pure Practical Reason in General” (KpV 5:132). Remaining with the postulate of immortality, we find that it provides the solution to the problem that speculative reason attempts to treat in the “Paralogisms” in the “Dialectic” of the first Critique (KpV 5:133): namely, that of establishing the nature of the soul. The category of substance is explicitly mentioned here. As Kant says, reason was there unable to make progress on this problem in the first Critique “because it lacked the mark of permanence by which to supplement the psychological concept of an ultimate subject, necessarily ascribed to the soul in self-consciousness, so as to make it the real representation of a substance” (KpV 5:133). Confined as it was to the sensible representation of the self, speculative reason could not legitimately represent the thinking subject as a substance.26 This mark of permanence indicative of substance is instead furnished by practical reason “by the postulate of a duration required for conformity with the moral law in the highest good as the whole end of practical reason” (KpV 5:133). On the basis of the need to achieve a completely moral disposition, we finally find the grounds to postulate our immortality and determine ourselves through the category of substance.
4.c Determinacy and Motivation
Finally, having developed the cognitive gains made possible by the postulate of immortality, we can turn to the issue of motivation. In section 2 I argued that the motivational help provided by the postulates cannot just be conceived of in terms of the subject wanting to maintain her self-conception as a morally interested and therefore dignified being. This is, no doubt, a necessary presupposition for any argument about the possibility of a moral disposition, but it should be thought of as a baseline requirement that cannot waver. The motivational help that Kant speaks of in connection with the postulates is better understood using the principle I articulated at the end of section 2, which states that the cognitive content of a practical representation and its capacity to motivate the subject should not be understood as fundamentally separable from one another. My suggestion is that we need not look beyond the conceptual content of the postulates and their relation to the highest good to explain how they help motivate us towards moral action.
Returning to section 87 of the third Critique, Kant vividly communicates the role of the postulates through his discussion of Spinoza, a philosopher who he takes to be sound of morals but lacking faith in immortality and God because of his theoretical beliefs. Kant is clear that without this faith, “the moral way of thinking has no way to persevere in its collision with theoretical reason’s demand for a proof (of the possibility of the object of morality), but vacillates between practical commands and theoretical doubts” (KU 5:472). Howsoever committed Spinoza may be to the moral law and its unflinching commands, he is plagued by questions about the possibility of the highest good that theoretical reason must ask but cannot answer. These questions essentially reveal the limits of sense-dependent reason: because theoretical reason is restricted to the realm of temporally conditioned appearances, it has no determinate conception of how transcendental freedom could be possible, or of how a temporally extended being could attain a completely moral disposition, or of how moral laws could be synthetically connected to temporally determined natural laws so as to produce happiness in accordance with virtue. Without further resources, these limits inevitably give rise to doubts about the possibility of the highest good. Kant thus concludes of the faithless Spinoza:
… if he would remain attached to the appeal of his moral inner vocation and not weaken the respect, by which the moral law immediately influences him to obedience, by the nullity of the only idealistic final end that is adequate to its high demand (which cannot occur without damage to the moral disposition), then he must assume the existence of a moral author of the world, i.e., of God, from a practical point of view, i.e, in order to form a concept of at least the possibility of the final end that is prescribed to him by morality.(KU 5:452–53; final emphasis mine)
With this passage, Kant suggests that it is the cognitive work of filling out a really possible concept of the highest good that serves to strengthen and protect the moral disposition. Without it, the hopelessness of our moral vocation in being directed towards an impossible end cannot help but weaken our sense of the intelligibility or cohesiveness of moral life, hampered as it is with theoretical doubts. Discursive human reason thus finds a way to respond to such doubts using the concepts available to it. The objective necessity of the moral law and its final end specify conditions of possibility that the judging subject then meets through the valid use of the categories applied to the concepts of freedom, immortality, and God. Kant often describes this activity using the language of conceptual determination: the freedom of the rational will is, for example, not merely thought by practical reason, but “even determined with respect to the law of its causality and cognized assertorically; and thus the reality of the intelligible world is given to us, and indeed as determined from a practical perspective, and this determination, which for theoretical purposes would be transcendent (extravagant), is for practical purposes immanent” (KpV 5:105). For the concept of God, practical reason likewise yields “a precisely determined concept of this original being” (KpV 5:139), attributing to it the marks of a being capable of synthetically connecting laws of morality and nature to produce happiness in proportion to virtue. Finally, as we have seen through my detailed treatment of the postulate of immortality, practical reason also finds the ground, the mark of permanence, required to determine the conscious subject through “the real representation of a substance” (KpV 5:133). By adding determinate representational content to these ideas and unifying them in the concept of the highest good,27 freedom, immortality, and God help prevent the hopelessness that would prevail if pure practical reason’s final end were taken to be empty.
To conclude, we are finally in a position to see the full significance of my emphasis on how we think about practical representations and the way they motivate. Against other commentators, I have argued that the postulates are crucial because they allow us to transcend cognitive limits that render opaque the real possibility of our moral nature and the system of ends it aims us towards. The desiderative element germane to this picture is thus nothing over and above the efficacy that the subject’s practical representations possess in virtue of the increased determinacy of their content. It is this added content that helps protect the moral disposition, staving off the dialectical illusion natural to practical reason’s use. This conclusion has important consequences for Kant’s view of human nature that, I think, his interpreter would do well to recognize in respect to a number of different issues. For rather than a dualistic vision that depicts sensibility as external to reason and in need of soothing on its own terms, we see evidence of a unified cognitive power.