The Postulate of Immortality in the Critique of Practical Reason (and Beyond)

Abstract

It is widely claimed that the second Critique’s argument for the postulate of immortality is relevantly different from the first Critique’s argument for the postulate. It is also widely claimed that after the second Critique, Kant distances himself from its particular version of the argument, and even the postulate altogether. It is the purpose of this article to challenge these claims, arguing instead that (a) there is overwhelming textual evidence showing that Kant did not abandon the postulate; (b) the second Critique does not contain a substantially different argument for the postulate than how it is argued for in the first Critique; and (c) the philosophical objections levied against the second Critique’s argument, including its putative substitution of holiness for virtue, become moot once its argument is better understood.

1. Introduction

Kant’s doctrine of the Highest Good is among the most controversial aspects of his ethical theory. Interpreters disagree about its fundamental structure, its overall function, its compatibility with other aspects of the Critical philosophy, and whether or not Kant modified it in any substantial way through the 1780s and 90s. Among the long list of disputed topics is: whether or not Kant changed his argument for the postulate of immortality, whether or not he really needed this postulate (hereafter the Postulate of Personal Immortality or PPI), ¹ and whether or not he eventually abandoned it.

Perhaps the most common view with respect to the above is that while in the Critique of Pure Reason, Kant presents the afterlife as the domain in which God distributes happiness in proportion to moral worth, in the second Critique, he instead ties PPI to a duty we have to attain a level of moral perfection beyond what we are capable of in this life (Allison 1990: 171–3; Guyer 2000: 352; Surprenant 2008; Tizzard 2020; Palatnik 2022). It is likewise claimed that either already in the second Critique or perhaps thereafter, Kant distances himself from an earlier ‘theological’ or post-mortem conception of the Highest Good to a ‘secular’ or this-worldly conception (Reath 1988; Kleingeld 1995; Guyer 2000). Hence, while it is usually accepted that Kant initially represents the Highest Good as a post-mortem distribution of ‘happiness in exact proportion with the morality of rational beings’ (A814/B842), many argue that he migrated to a ‘secular’ or this-worldly position where morality and happiness are instead linked within the physical order (Reath 1988; Engstrom 1992; Kleingeld 1995; Guyer 2016; Wood 2020). ²

I will argue that this story, which has been a mainstay in Kant's scholarship for decades, is largely incorrect. As such, my goal in this article is to defend the following theses. First, Kant does not abandon PPI, as is so often claimed. Second, the second Critique’s argument for PPI only appears different from the first. Not only are they essentially the same but there is one moral argument for PPI used by Kant across the Critical period. Third, the various problems that have been identified with the putative new argument of the second Critique can be dissolved or at least better addressed once that argument is more fully understood.

I will thus proceed as follows. While the main aim of this article is to reconsider the second Critique’s argument for PPI, I will first go over the standard textual evidence for the claim that Kant abandoned PPI in the 1790s (section 2). With that textual matter out of the way, I will then turn in section 3 to the second Critique’s argument. In section 4, I will discuss some of the problems often tied to the putative new argument of the second Critique, and show how they are resolved through the reading I will here present. Lastly, I have added a fifth section in response to a question raised when this article was presented as a keynote address at a meeting of the North American Kant Society.

2. Does Kant abandon PPI?

Although Kant never explicitly asserts that he no longer sees a need for PPI, interpreters point to various passages, trends, and philosophical considerations to suggest that by the late 1780s or early 1790s, he came to at least lean against his earlier view that it is integral to his understanding of the Highest Good. For example, in his influential 1988 paper, ‘Kant’s Two Conceptions of the Highest Good’, Andrews Reath contends that in the years following the second Critique, the ‘theological version [of the Highest Good] is de-emphasised’ and ‘the secular version becomes more prominent’ (Reath 1988: 607n). Likewise, in her similarly influential 1995 paper, ‘What do the Virtuous Hope For?’, Pauline Kleingeld asserts that by the second Critique ‘the immortality postulate is no longer invoked for the purpose of a post-mortem reward’, instead locating ‘happiness in the sensible world’ (Kleingeld 1995: 101) by way of ‘a just legal system, a free public sphere, the moral development of citizens’ and so forth (Kleingeld 1995: 103). ³

Paul Guyer is, of course, also a key advocate for this view, claiming, for example, that in the second Critique, Kant ‘dramatically separates the postulation of immortality from the postulation of God’ (Guyer 2005: 289) and then after 1788, ‘simply dropped the postulate of personal immortality’ (Guyer 2020: 146). ⁴ Since, however, Kant never outright makes this claim, but instead continues to discuss PPI during the 1790s, Guyer’s strategy is to offer examples where Kant’s discussions of PPI appear in the subjunctive or are otherwise contextualised, then generalise from these examples to his conclusion that Kant no longer really supports PPI after the second Critique. ⁵

Now, I am not going to argue against there being some passages where Kant may very well be just discussing one or another claim about PPI hypothetically. I will, though, have more to say about this shortly. But first, let me turn to the evidence that Kant did not withdraw his claim that PPI is integral to the Highest Good, and, in fact, remains integral precisely in the same way as presented in the first Critique. ⁶

Let me begin with a cluster of comments about PPI in the third Critique. First, in §91, a section devoted to the nature of practical faith, Kant writes that our use of the concept of the Highest Good is ‘commanded by practical pure reason’ and the conditions for its possibility must therefore be assumed. He then writes that these conditions include ‘the existence of God and the immortality of the soul’ (CPJ, 5: 469). A page later, he adds that ‘God and immortality, [serve] as the conditions under which alone we can, given the constitution of our (human) reason, conceive of the possibility of [the Highest Good]’ (5: 470).

In fact, we see PPI also at work in Kant’s oft-cited discussion of the Righteous Atheist, where it should not be overlooked that Kant describes him as denying not just God but also the immortality of the soul. For it is out of this dual denial that there arises the risk of moral despair due to ‘all the evils of poverty, illness, and untimely death’ (CPJ, 5: 452) that so saturate our world. Accordingly, consider the following: what if Kant were trying to make the case only for the postulate of God? What would that do? For such evils are treated as items of empirical evidence and so whether or not one adopts the postulate of God, the facts cited have not changed: the world still has within it the same level of poverty, illness, and untimely death. As such, what then is the contribution of this postulate on its own? How does it make the moral fortitude of the Righteous Theist any more durable than the Righteous Atheist? The answer, of course, is that there is an implied hope here, but it is a hope that involves, according to Kant, both postulates – and not just one – given that he begins this discussion by describing the Righteous Atheist as denying both God and the immortality of the soul. ⁷ This, moreover, is reflected in Kant’s many discussions of the Highest Good in the third Critique, where he writes, for example, that we cannot expect it to be realised ‘either on the part of our own physical capacity or on the part of the cooperation from nature’, and so requires both ‘the assumption of God and immortality’ (CPJ, 5: 471n). Let me now turn to a number of shorter works from the early 1790s.

First, and perhaps most ironically, let me discuss the Theory/Practice essay of 1793. I note the irony here because this text has been used in support of the claim that Kant loosened the connection between morality and happiness in the Highest Good since he seems to describe their relation more loosely as ‘universal happiness combined with and in conformity with the purest morality throughout the world’ (TP, 8: 279). ⁸ What, however, is overlooked is that Kant then clarifies in a footnote that the Highest Good ‘contains no prospect of happiness absolutely, but only of a proportion between it and the worthiness of a subject’ (8: 280n); and for our purposes, what is crucial here is that our commitment to the Highest Good ‘exacts from reason belief, for practical purposes, in a moral ruler of the world and in a future life’ (8: 279).

Second, again in 1793, this time in the Real Progress essay, Kant discusses the Highest Good as ‘the ultimate purpose of practical reason’ (PM, 20: 294) and that our concept of this ultimate purpose is ‘framed’ by the postulates of freedom, God, and the immorality of the soul. He characterises the latter as ‘the continuance of our existence’ in order for there to be ‘consequences which are commensurate to the moral behaviour of such beings’ (20: 295). A few pages later, Kant goes on to discuss the nature of belief and presents what he calls his ‘Credo’ or ‘articles of confession of pure practical reason’. They are, of course, the Highest Good and the three postulates, stating ‘I believe in a future eternal life, as the condition for an everlasting approximation of the world to the highest good’ (20: 298).

Third, in the 1796 Proclamation of Peace in Philosophy (PPP) essay, Kant references the postulates, which are to be affirmed by ‘morally-practical reason’. Of immortality, he describes it as ‘a state in which man’s weal or woe is to be allotted to him in proportion to his moral worth’ (PPP, 8: 418). He also helpfully articulates the function of the postulates as follows: ‘the being who is alone able to carry out this proportionate distribution is God; and the state in which this consummation can alone be assigned to rational creatures, in full accordance with that purpose, is the assumption of a continuance of life already founded in their nature, i.e., immortality’ (8: 419).

I do not see in any of these the sort of subjunctive framing, qualifying, or bracketing Guyer attributes to his sample passages from the Religion. Moreover, contrary to the claim that we can find no argument for PPI after 1788, consider that in many of the passages I have here cited, the same argument as the one used in the first Critique is repeated: to create a domain wherein each individual will face the ‘consequences which are commensurate to the moral behaviour of such beings’ (PM, 20: 295); with ‘man’s weal or woe … allotted to him in proportion to his moral worth’ (PPP, 8: 418).

Now, as my fourth source of evidence, let us return to the Religion. For the sake of space, I am going to offer just two brief points. The first is to explain why we find some discussions of immorality framed as subjunctives. The quick answer is because of the purpose for which Kant wrote the Religion, namely, to explore the question of ‘unity’ or ‘compatibility’ (R, 6: 13) between rational religion and historical faith. Hence, each part of the text draws particular tenets from the latter, holding them up to our ‘moral concepts’ (6: 12). Where unity or compatibility is found, the tenet is affirmed as a proper ‘vehicle’ for rational religion; when not found, the tenet is either just put to the side as inconsequential – the virgin birth is such an example given by Kant (6: 80) – or as morally hazardous, a debasement of religion as ‘cult’ or ‘religion of rogation’ (6: 13, 51, 84, 174). Accordingly, much of the text has a subjunctive flavour to it: were a religion to claim P, would P be compatible with morality? That, for example, is how the General Remark of Part One opens: with the supposition to be investigated therein, i.e., whether or not divine aid is necessary for us to undergo a ‘change of heart’. ⁹

The second point I want to make regarding the Religion provides an application of its aforementioned separation of genuine moral religion vs mere ‘cult’ or a ‘religion of rogation’. In his painfully inaccurate treatment of Judaism, Kant argues that it is not a genuine religion, but just an ‘old cult’ (6: 127), as it lacks three necessary conditions for a proper moral religion. He claims that Judaism’s prescriptions are not universal, but apply only to members of the faith (6:126–7); that it does not concern the ground of willing but only ‘external observance’ (6: 126); and it supposedly lacks a doctrine of the afterlife. On this, he writes that for Judaism ‘all rewards or punishments, are restricted to the kind which can be dispensed to all human beings in this world’, whereas ‘no [genuine] religion can be conceived without faith in a future life’ (ibid.). ¹⁰

Before moving on to the argument for PPI in the second Critique, let me discuss one more textual issue, specifically, the common view that Kant’s phrase ‘highest good in the world’ indicates a shift to a secular, this-worldly account of the doctrine.

Take for example TP, 8: 279, which, again, is touted as offering evidence for the this-worldly interpretation. There, he refers to the Highest Good as ‘in the world’, but in the very same sentence, he goes on to describe what we must believe with respect to its possibility, including ‘a moral ruler of the world and in a future life’. Similarly, in Real Progress, Kant likewise refers to the ‘highest good in the world’ and describes our mandate to ‘conform to this final end’ as dependent on ‘a future eternal life’ (PM, 20: 298). In fact, even in the second Critique, Kant locates the Highest Good as ‘in the world’, and does so, quite remarkably, at the opening of the section on the postulate of immortality. He writes that ‘[t]he production of the highest good in the world’ requires and is realised through a post-mortem ‘endless progress’ (CPrR, 5: 122).

So, how could the Highest Good be ‘in the world’ while still in need of the postulate of immortality? Moreover, how could the Highest Good be ‘in the world’ if Kant quite clearly tells us that our allotted ‘weal and woe’ is distributed post-mortem (PM, 8: 418)?

Now, perhaps the claim is that from the temporal perspective, the Highest Good’s distribution of happiness still happens within this life, even though immortality must still be postulated in order for us to become worthy of happiness. That is perhaps God knows what we will or will not accomplish post-mortem and adjusts our happiness in this life accordingly. Though that strikes me as particularly bizarre, and moreover, out of line with Kant’s rejection of the idea that we should treat the joys and sorrows of this life as indicative of our moral status, a point that he makes as far back as his 1756 essays on the Lisbon earthquake (VUE, 1: 419–27, 431–61, 465–72) and forward through his 1791 Theodicy essay (MPT, 8: 255–71). ¹¹ So, then, what is the alternative?

It is very simple: Kant does not mean by the word ‘world’ the causal order of nature. Rather, he uses ‘world’ more broadly, as was typical during his time, to cover all that is, rather than just the planet Earth. Baumgarten, for example, defines ‘world’ (mundus) as the unity of all things, the highest-order totality, a totality ‘that is not part of another’ (Baumgarten 2013: §354), ¹² and Kant follows in kind, where ‘world’ likewise reflects ‘a whole which is not part [of any greater whole]’ (ID, 2: 387) or ‘absolute totality’ (2: 391) of all form and matter, all that is sensible and intelligible (2: 387f). ¹³ Likewise, in the Critique of Pure Reason, Kant (a) contrasts a more mundane use of ‘world’ as the totality of phenomenal objects against ‘nature’ as that totality under a system of laws (A418/B446), and then (b) contrasts the former use of ‘world’, against a ‘transcendental sense’ of the term, such as when dealing with ‘cosmological ideas’, where ‘world’ extends more broadly to the ‘sum total of existing things’ (A419/B447). ¹⁴

Hence, if one goes back to those passages where Kant describes the Highest Good as obtaining post-mortem, and even that our ‘endless progress’ post-mortem is ‘in the world’ (CPrR, 5: 122), we may understand ‘world’ in a manner inclusive of not just the physical cosmos but also the sort of highest-order totality indicated by Wolff, Baumgarten, and Kant’s pre-Critical use of ‘mundus’, and, in fact, as also coordinate with the ‘world-concept’ that Kant identifies with the purview of philosophy as the ‘science of the relation of all cognition to the essential ends of human reason’ (A839/B867), the ‘science of the relation of all cognition and of all use of reason to the ultimate ends of human reason’ (JL, 9: 24). ¹⁵

To summarise, we find in both Kant’s pre-Critical and Critical writings explicit discussions of how he uses the word ‘world’, discussions that clearly point beyond its contemporary use, and to a sort of totality where it would make sense for Kant to have referred to the Highest Good as ‘in the world’, while in the very same sentence describing it as obtaining post-mortem. We, therefore, have very solid reasons to not read Kant’s use of ‘world’ in ‘highest good in the world’ as so many have over the past five decades. Both philologically and by Kant’s own hand, the intended meaning reflects not just ‘this’ world, the physical/causal order, or ‘the unity in the existence of appearances’ but rather, a highest-order totality, the afterlife included.

3. Kant’s argument for the postulate of immortality in the Critique of Practical Reason

As mentioned earlier, it is commonly held that Kant substantially modified his argument for PPI in the second Critique. While it is elsewhere presented as the domain in which God distributes happiness in proportion to moral worth, it is commonly alleged that in the second Critique, Kant deviates arguing instead for PPI because of what is needed with respect to our duty to moral perfection. We find this interpretation in Allison, Wood, Guyer, and a host of others since gaining dominance in the 1990s and early 2000s. But, as I will now argue, this widespread claim is incorrect.

A cornerstone of Kant’s account of the Highest Good is that it is not merely a synthesis of morality and happiness but more specifically, it synthesises the two as an ‘exact distribution of happiness in proportion to moral worth’ (hereafter, the Principle of Proportionate Distribution or PPD). Although some of the literature on the doctrine prefers to loosen this principle, such as the ‘satisfaction of individuals’ morally permissible ends’ (Reath 1988: 603) or ‘maximally lawful happiness’ (Guyer 2000: 340), Kant does not deviate from PPD. I have already mentioned the footnote in Theory/Practice where Kant clarifies that he still intends PPD, despite how some have treated the passage. We likewise find PPD across numerous texts from both the 1780s and 90s. This includes, of course, the first and second Critiques (A809/B837, 5: 110), but also the Orientation essay (8: 139), Religion (6: 5), the Proclamation Essay (8: 418), and so forth.

And yet, something does, at least prima facie, look different in the second Critique. Rather than presenting PPI for the sake of the distribution of happiness, the section devoted to PPI instead opens with it as required for the possibility of our ‘complete conformity of dispositions with the moral law’ (CPrR, 5: 122). The argument thus appears to be that since ‘complete conformity’ is not a possibility for us in this life, given the principle ought implies can, it is therefore ‘necessary to assume’ the possibility of ongoing post-mortem moral progress (ibid.). But if that is so, what then of PPD? What then of the distribution of happiness? Does it now take place in this life? Or maybe it still obtains in the afterlife, but this point is just not being overtly used by Kant in this instance when arguing for PPI. Or maybe all the pieces just no longer really fit together, leading Kant, as some claim, to rethink the Highest Good more broadly. Let me, however, propose a more straightforward option, one that shows that the second Critique is actually consistent with the texts both before and after it – that is, one that not only maintains consistency across all three Critiques but also explains why he continues to use PPI for the sake of the distribution of happiness in the 1790s, just as he did in the first Critique.

The question at hand is what (if anything) has been overlooked in the common reading of the postulate in the second Critique? The answer, quite simply, is that Kant does in fact reference our post-mortem happiness, describing it therein specifically in terms of the classic Christian notion of the beatific state. Within the section devoted to PPI, Kant describes the beatific state as one of perfect happiness, one’s ‘complete well-being independent of all contingent causes in the world’ (CPrR, 5: 123n), and then in the following section, where Kant discusses the postulate of God, he describes it as the state of ‘well-being proportioned’ to our moral perfection or holiness (5: 129). In the latter as well, he tells us that the beatific state ‘cannot be attained at all in this world’ but rather requires both God and a reality separate from the order of nature (ibid.).

Hence, while the opening language of the second Critique’s section on PPI does seem, when considered in isolation, to advance a new and distinct argument for the postulate, it is important to not overlook Kant’s ongoing discussion of post-mortem happiness. And perhaps even more importantly, it is important to not overlook the broader framework here, namely, that the purpose of the Dialectic of Pure Practical Reason is to answer whether or not practical reason’s quest for its unconditioned condition suffers the same fate as the unconditioned condition for theoretical reason. Kant articulates precisely this problem at the opening of the second Critique’s Dialectic, raising the question as to whether or not the Highest Good, as ‘unconditioned totality of the object of pure practical reason’ (CPrR, 5: 108), like its theoretical counterparts, is ultimately just transcendental error or illusion.

To that end, Kant generates the Antinomy of Practical Reason and then, like the solutions to the dynamical antinomies in the first Critique, turns to the phenomenal/noumenal distinction to at least create the space for the possibility of the object at issue. However, unlike freedom and necessary being in the first Critique, where, at best, Transcendental Idealism provides for their logical possibility, Kant argues that in light of the needs of practical reason, needs which find their objective ground in the moral law (CPrR, 5: 143n), we may positively affirm the Highest Good. Although Kant grants that it is logically possible for the Highest Good to obtain within nature, he tells us, and in fact tells us numerous times not only in this text but in many others, that ‘our reason finds it impossible for it to conceive’ (5: 145) how it could obtain within nature. As noted above, he says that specifically about the beatific state as well, for not only can we not think it plausible for PPD to obtain amidst all the exigencies of this life but the sort of perfected happiness of the beatific state likewise needs to be understood as so separated as well. ¹⁶

Now, with the Highest Good, PPD, and the beatific state as the backdrop or context for Kant’s argument for PPI, let me put these pieces together. First, Kant proposes that we have the Highest Good as a necessary end (CPrR, 5: 108, 124, 143). Second, he describes the Highest Good in terms of an ‘exact distribution of happiness in proportion to moral worth’ (5: 110, 129, 145). Third, he tells us that God’s role is to judge our worth and distribute happiness accordingly (5: 124, 128, 140). Fourth, he tells us that our role is to achieve a moral status that makes us worthy of happiness (5: 124, 129, 144). Fifth, he specifies that the sort of happiness God distributes is the beatific state (5: 123n, 129), which, of course, is the classic Christian conception of the post-mortem state for those who have become ‘well-pleasing to God’. In fact, Kant writes that the ‘doctrine of Christianity’ contains ‘a concept of the highest good … which alone satisfies the strictest demand of practical reason’ (5: 127–8), and that includes both the moral demand upon us and the promise of a beatific state to be enjoyed in everlasting life.

So, where does this leave us? No doubt, there is an element of truth to the reading given by Guyer and others, where Kant presents PPI as providing for the possibility of a level of moral improvement beyond the limits of this life. But that element is nevertheless part of a larger picture, one that flows out of the essential function of the doctrine of the Highest Good and, most importantly, involves both PPD and Kant’s characterisation of the beatific state as what is offered by God. Hence, the reason why eternal striving matters has more deeply to do with the state of perfect happiness offered by God. Moreover, Kant does not leave room for some ersatz alternative to the beatific state, say, for example, the enjoyment of an eternal day on a tropical island, because one might then feel that one need only achieve a level of moral improvement necessary for that ersatz joy.

In short, what we have in the second Critique is still essentially the same argument for PPI as in the first Critique. Both posit an afterlife as the domain in which happiness is distributed. But in the second Critique, Kant fills out more detail than in the first, including how to conceptualise post-mortem happiness and what sort of moral state is incumbent upon us, given PPD. If it is granted that this is not a substantially different argument for the postulate from the one given in the first Critique, then it further overcomes the common query as to why Kant seems to revert back to his previous argument in the 1790s. For there is, in fact, no reverting back at all, as it has been his argument all along. And, as I argued in section 2, the textual evidence overwhelmingly indicates that Kant did not abandon the postulate of immortality in the 1790s. The doctrine and its postulates remained stable throughout the Critical period. I would think that it is a preferable position than to think that in such an important work as the second Critique, Kant temporarily shifted to a new argument to only then run into such troubles that everything thereafter is a muddle between repeated discussions of PPI as it was given in the first Critique while also allegedly abandoning PPI for a this-worldly alternative. All this gets ironed out to Kant’s advantage as a coherent thinker under my reading, though a thinker that is more set for his times than the sort of secularism that tends to be preferred by contemporary academics. ¹⁷

4. What is the goal of our moral improvement?

The next topic I would like to discuss is the question of what sort of moral status is required of us in order to merit the post-mortem happiness Kant refers to by way of the beatific state. It has often been argued that he waffles with respect to this status, sometimes positing a moral status that is not achievable by us even through eternal effort, sometimes positing an achievable status, but one that does not depend upon the immortality of the soul. Such concerns led Henry Allison to criticise the argument for the immortality of the soul in the Second Critique as ‘artificial in the extreme’ (Allison 1990: 172). Likewise, these tensions inform Guyer’s claim that Kant ‘retracted’ the thesis that ‘immortality is necessary for us to have the opportunity to perfect our virtue’ (Guyer 2020: 157).

Now, on the one hand, Allison and Guyer are right to criticise an argument that would postulate the immortality of the soul in order for us to realise something that either is not in principle possible for us or does not even require our immortality. However, as discussed in the previous section, PPI is not tendered specifically for the sake of our being able to achieve a particular moral state. Rather, it is more fundamentally posited as the domain in which happiness is to be distributed by God. Accordingly, if the more direct rationale for the postulate is as I have proposed, let us use this to revisit what Kant writes in the second Critique with respect to what we must achieve to become so worthy.

First, Kant tells us that our moral goal is the ‘complete conformity of dispositions with the moral law’, identifying this complete conformity with ‘holiness’, and a ‘perfection of which no rational being of the sensible world is capable’ (CPrR, 5: 122). From there, he rejects two options. The first is a more ‘lenient’ goal that would instead conform ‘to our convenience’ (ibid.). The second is ‘full acquisition of holiness of will’, which contradicts our understanding of what it is to be a finite rational agent (5: 122–3). Both, he writes, are to be rejected because both would ultimately hinder our moral efforts (5: 123). ¹⁸ Accordingly, once we attend to the sentences that follow Kant’s initial characterisation of the goal, we find him clarifying his point in a way that halts the standard objection debated in the literature (Allison 1990; Guyer 2016; Hahmann 2018; Tizzard 2020; Palatnik 2022) that he illicitly substitutes the first Critique’s goal of virtue with now the goal of holiness.

So, what then precisely is our goal in moral improvement? And what standard is used by God in determining our moral worth? Kant tells us the answer(s) to these questions can neither be too ‘lenient’ nor ‘too strict’, and ‘too strict’ is presented specifically as ‘full acquisition of holiness of will’. Accordingly, he answers the two questions by locating the answers between these two poles. With respect to the question of God’s standard for determining our moral worth, Kant explains that God intuits our moral disposition, judging whether or not we have made an ‘immutable resolution’ to strive morally. But the way in which we work toward this goal, and what we do when sustaining it, is to engage in a process of moral improvement from ‘lower to higher moral stages’ (CPrR, 5: 123).

In other words, the way we strive to become worthy of happiness is to dedicate ourselves to moral improvement and that dedication, rather than any specific stage of virtue, is what God judges. This is an important point for Kant, as he is committed to the idea that divine justice cannot be diluted by mercy, but rather they must align. ¹⁹ Hence, in light of PPD, and in light of the beatific state being what God distributes, it is not as if God allows an asymptotic approach to perfect virtue to stand in for perfect virtue (something that the Beck translation of the second Critique in particular I think intimates). ²⁰ Rather, Kant’s point is that whether or not we are worthy of the beatific state is in virtue of whether or not our moral resolution is so totally immutable that it would not even falter through eternity.

It is, therefore, not that we need an eternity of striving to become worthy of the beatific state, but rather we would have to have the sort of immutable resolution that, given the immortality of the soul, would be reflected in ‘an endless progress toward’ ‘the complete conformity of the dispositions with the moral law’ (CPrR, 5: 122). Since, however, God does not need to wait until the end of eternity, so to speak, but just knows our moral disposition through ‘a single intellectual intuition’ (5: 123), if what God sees when judging us is the sort firmness of resolution needed, then he can both offer the beatific state in accordance with PPD but also know that in the agent’s eternal participation in that state, there will be no faltering.

What may, though, gum up the works is that if God’s judgement is based upon our ‘immutable resolution’ rather than a degree of success in one’s eternal moral striving, it may seem like there is, in fact, no need for the postulate of immortality after all. This is a point Guyer has recently made, writing that such a resolution ‘can be achieved at any time, not something that has to be deferred to some future life’ (Guyer 2020: 159). He, of course, is correct. If Kant sees the moral status of our disposition as the basis for our worthiness to be happy, then he cannot base the need for PPI upon a moral achievement that is only possible through eternal striving. ²¹ However, rather than this yielding a problem in Kant’s argument for the postulate, as Guyer contends, I have proposed that the argument for PPI remains always, including in the second Critique, based instead upon the need for a domain distinct from the physical causal order through which God distributes happiness. Only derivatively does eternal striving become a consideration. The argument for PPI is more deeply based upon the need for such a domain to make possible a state of ‘complete well-being independent of all contingent causes in the world’ (CPrR, 5: 123n) and then, given the positing of such an eternal state of happiness, an account of our moral work therein is also given its place.

Lastly, let me note that the reading I have here offered not only aligns the second Critique’s argument for PPI with the argument for PPI prevailing through the Critical period but it also aligns Kant’s account of moral worthiness in the second Critique with the Religion. For in both, God judges our underlying moral disposition, what Kant in the former text refers to as our ‘immutable resolution’, and in the latter, the ‘change of heart’ or ‘single and unalterable decision’ to prioritise morality over self-interest in the supreme maxim of one’s moral disposition (R, 6: 48). Likewise, both texts distinguish between God’s insight into our fundamental disposition versus how we pursue our moral improvement through the ‘gradual reform’ associated with the cultivation of virtue (6: 47). That is, those who pursue (and succeed in bringing about) a ‘change of heart’ will strive to improve themselves through cultivating greater moral strength and a greater alignment of their character with the demands of morality. ²² Recognising this dual perspective (how we, from the inside of our lives, understand our moral vocation versus how we represent God’s judgement of us) helps unravel Kant’s presentation of PPI in the second Critique, and is, moreover, an important element in the Religion, discussed both in the General Remark of Part One (6: 44–52) as well as in the ‘three difficulties’ enumerated by Kant in Part Two (6: 66–78).

5. God as the ‘author of nature’

This section is intended to address one particular challenge raised on the occasion of this article’s initial presentation to the North American Kant Society. Namely, despite all the evidence that Kant does not abandon PPI as the vehicle for PPD, how should we account for the passages where he refers to God as the ‘author of nature’ or ‘cause of nature’ in the context of PPD? After all, do these phrases not suggest a ‘natural’ or this-worldly locus for the Highest Good? And yet, since they are used not merely in the context of PPD, but even amidst discussions of PPI as the vehicle for PPD, finding a way to understand ‘nature’ in this context is salient to both my interpretation as well as to the consistency of Kant’s own representations of the Highest Good.

We will proceed as follows through this section. First, we will consider how ‘nature’ was used during Kant’s time, particularly by Baumgarten and G. F. Meier, who serve as key sources for much of Kant’s philosophical terminology. ²³ Second, we will look at Kant’s references to God as author/cause of nature in the Canon of the first Critique, a text where it is generally acknowledged that Kant held to a post-mortem conception of the Highest Good. Third, we will then look at three passages from the second Critique, two where Kant refers to God as the author/cause of nature amidst discussions of the afterlife and then one more complex passage. To begin, we have from Baumgarten’s Metaphysica, the following definition:

The NATURE (cf. §431, 466) of a being is the collection of those of its internal determinations that are the principles of its alterations, or, in general, of the accidents inherent to it. Hence, to the nature of a being pertains its (1) essential determinations (§39), (2) essence (§39), (3) faculties, (4) receptivities (§216), and (5) all the powers with which it is equipped (§197). (Baumgarten 2013: §430)

And when Baumgarten discusses the human soul, we can see ‘nature’ used as follows:

Aside from this proper sense of nature in the human soul, and hence of what is natural for it, there are other improper and synecdochic senses … Thus, for example, those that are inborn are always called natural in contradistinction to those that are acquired; and those which are natural are always contradistinguished from those which are artificial, related, chosen, or, in the end, obtained through habit, etc. (cf. §710). (§758).

Likewise, when discussing the immortality of the soul, Baumgarten also uses ‘nature’. For example: ‘Hence the human soul, which naturally preserves its nature after death (§780, 781)’ (§782).

As one more item relevant to the historical use of the term, let me offer a passage from Meier’s Metaphysik where he discusses a more narrow vs wider use of ‘nature’ in German:

By nature is sometimes understood nothing at all other than the nature of bodies, and hence the theory of nature [Naturlehre] or physics is straightforwardly called the theory of nature [die Lehre der Natur]; although in such no other nature is further treated than of the nature of bodies. By the theory of nature [Naturlehre] in the wider sense, can be understood the science of the natures of things in general. And in such would have to be treated the nature of God, of spirits, and of all other things. The science of the nature of bodies is the theory of nature in the more narrow sense.

The whole of nature, or the nature of the whole world, consists in the collection of all natures of all parts of the world, taken individually and together … Consequently, the whole nature of our actual world is the collection of all beings, essential parts, attributes, powers, faculties, capacities, with which all substances, all elements, all thinking substances, all spirits, all materials and bodies are equipped. (Meier 1756: §398–9). ²⁴

Our question is thus whether Kant uses ‘nature’ in its wider sense, and more specifically, whether this is its use when referring to God as the author/cause of nature in the context of the Highest Good. With regards to the former, let us begin with his distinction between natura considered formaliter and materialiter. Of the former, he writes that it ‘signifies the connection of determinations of a thing in accordance with an inner principle of causality’ (A418n/B446n). The latter – material nature – might initially be taken to be closer to what is more commonly understood as ‘nature’, but he means by it the ‘subsisting whole’ of the ‘thoroughgoing connection’ (A419/B446n) of the aforementioned ‘inner principle(s)’.

Kant also describes natura considered formaliter as any lawful order ‘of whatever sort’, including ‘also the moral order (hence not always the physical order)’ (EAT, 8: 333n). Additionally, in the lectures, he is reported to have said that ‘nature means the summation of all inner principles and all that which belongs to the existence of the thing’ (L-Met, 28: 221). Further, Kant also uses ‘nature’ with respect to both our ‘sensible nature’ and ‘supersensible nature’ (CPrR, 5: 43). Hence, just as when Baumgarten refers to God as the ‘author of universal nature’ (Baumgarten 2013: §983), inclusive of the inner principles of the soul, and when Meier (quoted above) uses ‘the whole of nature’ to include ‘all substances, all elements, all thinking substances, all spirits’, so too this seems Kant’s usage as well.

The next question is whether when Kant refers to God as the author/cause of nature, ‘nature’ again has this wide meaning. Since there is general consensus that in the Canon of the first Critique, Kant saw PPD as obtaining post-mortem, we can surmise that this was his intent when he used ‘author/cause of nature’ amidst discussions of the immortality of the soul. We have, for example, A810-1/B838-9, where God is referred to as the ‘cause of nature’ for the sake of happiness ‘in exact relation with morality’ (A810/B838) in ‘a world that is future for us’ (A811/B839). Likewise, at A815/B843, when discussing the attribution of omnipotence to God for the sake of PPD, he writes that such a capacity is needed ‘so that all of nature and its relation to morality are subject to it’.

Now, when looking at the second Critique, the meaning might be more disputed given that some claim that Kant was starting to migrate towards a this-worldly account of the Highest Good. However, we have already raised independent reasons for rejecting this putative transition. But further, we can again find very similar minglings between the postulate of immortality and references to God as the author/cause of nature therein.

One such example comes from the second paragraph on the postulate of God where God is characterised as the ‘cause of all nature distinct from nature, which contains the ground of this connection, namely the exact correspondence of happiness with morality’ (CPrR, 5: 124). This discussion appears right after Kant opens his discussion of the postulate of God by contrasting its role relative to the postulate of immortality, each having a particular role with respect to PPD. Quite simply, after having postulated the immortality of the soul, Kant then explains that we must also postulate the ‘existence of God’ (ibid.), referring to God as the ‘cause of all nature’, specifically for the sake of the ‘exact correspondence of happiness with morality’, and this correspondence requires both the postulates of God and immortality.

A second example is even stronger, where Kant is comparing ‘[t]he doctrine of Christianity’ to his analysis of the Highest Good. He there again discusses the beatific state as ‘attainable only in eternity’ (CPrR, 5: 129) and God as the ‘holy author’ (5: 128) who makes the Highest Good possible. Hence, again, divine authorship is tied to the system of reality inclusive of the afterlife.

The third example is, however, more complex, but when understood properly helps relieve Kant of the apparent ‘flip-flops’ (Mariña 2000: 340) between the post-mortem and this-worldly conceptions of the Highest Good in the second Critique. When taken out of context, the passage offers prima facie evidence for the this-worldly interpretation, but once understood in context, such evidential value is defeated:

[I]t is not impossible that morality of disposition should have a connection, and indeed a necessary connection, as cause with happiness as effect in the sensible world, if not immediately yet mediately (by means of an intelligible author of nature), a connection which, in a nature that is merely an object of the senses, can never occur except contingently and cannot suffice for the highest good. (CPrR, 5: 115; my emphasis).

Prima facie, it looks as if Kant sees happiness as an ‘effect in the sensible world’ rather than post-mortem, despite the postulate of immortality and the beatific state. However, the passage here does not reflect Kant’s actual view, but rather one that he is rejecting. Presented very concisely, the path of argumentation in the Dialectic is as follows: we must posit the Highest Good as a synthesis between morality and happiness (CPrR, 5: 107–13); this leads to an antinomy raising concern about its possibility (5: 113–4); the Critical Resolution creates the modal space for its possibility by proposing that the synthesis can have a ground outside of the sensible world (5: 114–9); practical reason has a prerogative to such space (5: 119–22), which is then filled in by the postulates of immortality (5: 122–4) and God (5: 124–5, 132); then Kant turns to the moral/practical warrant for assent (5: 132–5, 146).

The passage at issue at 5: 115 appears in the Critical Resolution, where Kant’s primary goal is to show that we can open the modal space, which in the sections of the Dialectic then to follow is filled through the postulates of God and immortality. Nevertheless, after explaining how this space is made possible through appeal to the phenomenal/noumenal distinction (broadly akin to the resolution to the dynamical antinomies in the first Critique), Kant then reviews and rejects a number of options as to how it might be filled before moving forward to his preferred solution. This includes that morality and happiness are reconciled within the physical, which Kant dismisses as utterly ‘strange’, given the obvious evidence to the contrary, that it could ever come in ‘precise proportion to virtue already in this life’ (CPrR, 5: 115). ²⁵ He then targets the idea that the Highest Good is satisfied through the inner pleasure found in virtue, which is rejected because we cannot know if we are in fact virtuous (5: 116), and so are at risk of subreptively using a state of inner pleasure as a sign of virtue when it is merely a self-satisfaction coming from a pathological motive (5: 116–7). The third option targets an inner calm from having mastered one’s inclinations, which Kant refers to as a ‘negative satisfaction’ or ‘contentment’ (5:11 7) in substitution for the sort of positive happiness that Kant regards as genuinely proper to the principle of happiness that we are looking to reconcile with virtue (5: 118).

Accordingly, understood in context, the Critical Resolution overcomes the preceding antinomy by positing a supersensible ground through which morality can lead to happiness. However, the question then becomes what specific relation between morality and happiness will be adequate for the Highest Good. Kant then considers the three scenarios just mentioned (this-worldly happiness, a feeling of reward from moral conduct, and contentment with oneself for having mastered one’s inclinations), rejecting each before then turning to the solution he then affirms in the postulates of immortality and God.

Lastly, let me note an interesting implication of our discussion in this section: when one considers Kant’s discussions of the question of how to understand the unity of nature and freedom insofar as it involves the Highest Good (cf. e.g. Guyer 2005, Esser 2016, Sweet 2022), this unity (given the contentions of this article) must also involve PPI. For if, as has been argued herein, Kant never abandons PPI, instead of having God somehow align the physical with the moral (or at least it is represented as if God does so), the capstone for the connection of nature and freedom, fact and value, is and ought, is ultimately, for Kant, the afterlife. ²⁶

This does not mean that we ignore nature in its more ordinary sense, for God may be understood as making this world fit for our basic survival (CB, 8: 115; TPP, 8: 360) and moral striving (R, 6: 100; IUH, 8: 30; TPP, 8: 362). That is, the unity of nature and freedom of course also involves the physical world, but it is not in this world – in our current lives – where God’s agency ultimately aligns morality and happiness. That, rather, happens post-mortem, and thus we bring nature and freedom into systematic unity through the Highest Good, not by a this-worldly aligning of morality and happiness but rather by understanding this world as the domain wherein we strive to be worthy of happiness, which God metes out – to those who are worthy – post-mortem.

Perhaps more profoundly, we may also emphasise that it is not just that this world is where we strive to become worthy of post-mortem happiness, but also that God has made this world fit for such striving, creating an order of nature amenable to our physical and moral characteristics, allowing us to evolve both physically and culturally, ultimately towards an ‘ethical community’ where humanity actively works together towards their common moral improvement (R, 6: 94f). ²⁷

6. Conclusion

My goal in this article has been to demonstrate that the prevailing understanding of the postulate of immortality over the past five decades is substantially mistaken. Contrary to what the majority of Kantians claim, neither (a) does Kant withdraw the postulate in the 1790s, nor (b) is there a substantive change in his argument for the postulate in the second Critique.

With regards to (a), I have provided numerous passages where Kant continues to affirm the need for the postulate of immortality. I have also provided evidence that those who treat Kant’s phrase ‘in the world’ according to contemporary English parlance, overlook his more expansive use of the word Welt, a use that is in line with other expansive uses of Welt during and prior to his day. In effect, not only do we see Kant’s repeated affirmations of the postulate of immortality during the 1790s as the ‘state in which man’s weal or woe is to be allotted to him in proportion to his moral worth’ (PPP, 8: 418) but also once Kant’s use of the phrase ‘in der Welt’ is more fully understood, central evidence used in support of a this-worldly account of the Highest Good is at least seriously challenged, and – I hope the readers of this article will agree – ultimately routed.

With regards to (b), I have argued that the backdrop for Kant’s argument for the postulate of immortality is the need to posit a domain compatible with the sort of happiness Kant takes God to be distributing, i.e., the beatific state, a state of ‘complete well-being’ (CPrR, 5: 123n) that ‘cannot be attained at all in this world’ (5: 129). It is thus because of PPD that our moral worthiness must align with the state of complete well-being. What leads readers to focus on the issue of moral striving is that Kant must then unpack the nature of this moral worthiness, which after eliminating both something too ‘lenient’ as well as ‘full acquisition of holiness’ as something unattainable for us (5: 122), he posits instead as an ‘immutable resolution’ that must, so to speak, stand the test of eternity.

Thirdly, once we read the second Critique’s argument for PPI as having PPD as its backdrop, this not only aligns the text with Kant’s standard PPD-based argument for PPI but we also no longer have to struggle through the allegation that Kant goes back and forth between different rationales for PPI, eventually succumbing to so many problems that he finally abandoned it in favour of a this-worldly Highest Good. Not only is this not supported textually, but once we see the putative problems as artefacts of the secondary literature rather than intrinsic to the text itself, like it or not, Kant remains committed to the so-called ‘theological conception’ of the Highest Good, as found not merely in the first Critique, but across such later works as the Religion and the Theory/Practice, Real Progress, and Proclamation essays.

Finally, let me briefly comment on why something that may seem so marginal is really of great importance for Kant scholarship. It is my view that Kant’s philosophy of religion is grounded upon and flows out of his doctrine of the Highest Good. It is the principle according to which ‘morality leads inevitably to religion’ (R, 6: 8n). But even more so, it underlies the evaluative project of the Religion, for when Kant holds up doctrines of historical faith to moral concepts, central to this is the Highest Good. That is, without a proper understanding of the Highest Good, one will likely end up with a flawed understanding of how Kant’s philosophy of religion fits within his broader philosophical system, as well as considerable confusion over its many details.

Accordingly, getting Kant right on the postulates is vital to getting him right on the Highest Good, which in turn is vital to getting him right on what is going on in the Religion, and even why, as Kant tells us in the B-Preface to the Critique of Pure Reason, part of what drove him to develop Transcendental Idealism was to determine the limits to knowledge ‘in order to make room for faith’ (Bxxx). Many Kantians, I expect, will not be personally amenable to the outcome of this project. They may prefer to present Kant in a manner more appealing to contemporary academics, who tend to have a more secular outlook, if not even a hostility to religion. I can sympathise. But I see it as a duty of stewardship to read Kant in line with his own times, and moreover, to take Kant at his word, as genuinely committed to this ‘room for faith’, a faith that he tells us is both sincere (MPT, 8: 267) and certain (A829/B857). An ‘as-if’ approach to the postulates, a this-worldly account of the Highest Good, a semiological reading of the Religion might, to many, feel more in line with contemporary attitudes, or is perhaps more personally appealing, but as I hope to have shown in this article, at least with respect to PPI, such was not Kant’s intent.