2. Kant’s hybrid justification of remorse

There are three ways of justifying the pain of remorse which are helpful in understanding Kant’s account. One is retributive: we should have painful feelings in response to past wrongful actions because we deserve them. Retributivist justifications are often called ‘backward-looking’ because they refer only to past wrongs. Pure retributivism would be the view that only retributive considerations play a role in justifying remorse. A second is forward-looking: we should feel remorse because it improves us and motivates us to act more morally in the future. A purely forward-looking justification would hold that only forward-looking considerations play a role in justifying remorse. A third is hybrid, drawing on multiple independent justifying reasons. I argue that if we exercise interpretative care in assembling the sometimes-contradictory remarks in Kant’s texts, we have a good case for attributing a hybrid justification to Kant, which has a retributive basis and forward-looking constraints.

Kant’s account of reasons for remorse has received little focused discussion in the literature. The most prominent discussion is indirect, in the context of Kant’s account of conscience, in a debate between Thomas Hill and Allen Wood about whether conscience punishes. A resolution of this debate cannot on its own answer the question about Kant’s reasons for remorse, since Kant’s reasons for punishment are themselves matters of controversy. However, a review of the debate will situate the present inquiry in the literature and provide helpful context.

On Thomas Hill’s interpretation, a ‘bad conscience “hurts,” and … Kant treats this pain as analogous to the suffering imposed on lawbreakers by the system of criminal justice’ (Reference Hill2002: 352). He points out that Kant sees conscience as an inner court in which we ‘think of ourselves as playing several roles: that of accuser, defender, and finally a judge’, who

Hill reads Kant as holding that conscience ‘enforces the law, passes sentence’, and ‘makes us suffer for our misdeeds’ (Reference Hill2002: 353).

By contrast, Allen Wood holds that conscience does not punish (Reference Wood2008: 187). He claims that ‘Kant denies that it is even possible to punish oneself’, citing MM, 6: 335, which is a discussion of social contract theory and the justification of bodily punishment according to principles of right. There Kant states that ‘it is impossible to will to be punished’, but just as clearly states that there is indeed a sense in which ‘I draw up a penal law against myself as a criminal’: ‘pure reason in me (homo noumenon) … subjects me, as someone capable of crime and so as another person’ to punishment (MM, 6: 335, boldface added). Kant’s point here is not that I do not punish myself, but that my self-punishment involves a multiplicity of perspectives on myself. Kant also addresses this concern about self-punishment at MM, 6: 485 (quoted above), where he faults ‘monkish ascetics’ for aiming at penance (büßen) rather than the experience of remorse (moralisch zu bereuen), in part because it involves ‘punishment [Strafe] chosen and executed by oneself’^t, while punishment ‘must always be imposed by another’. But as Hill emphasizes in his remarks quoted just above, Kant holds that agents must take on multiple perspectives in the experience of conscience, just as we saw they must do in drawing up penal law against themselves.Footnote ⁷ Kant discusses this in detail at MM, 6: 438:

If punishment must be inflicted by another person, and conscience contains a judge who we must represent as a different person even though that person is an aspect of ourselves, then if that judge inflicts painful emotional punishment on us, we must represent it as inflicted by another person even though it is inflicted by an aspect of ourselves. Kant’s criticism of monkish ascetics is not that it involves self-punishment as such, but rather that it involves self-punishment which is not mediated by the inner judge of conscience – that it is about doing penance by imposing unpleasant experiences ‘such as a fast’ on oneself instead of experiencing wahre, moralische Reue, and is thus a self-deluding substitute for the kind of self-punishment inflicted by one’s inner judge.

The most definitive passage supporting the view that conscience punishes is one that neither Hill nor Wood cite in their texts quoted here, at MPT, 8: 260:

It is reasonable to assume that the virtuous have consciences which function properly, so if their consciences punish, then it is the proper function of conscience to punish.

Wood acknowledges that, on Kant’s view, ‘conscience can “judge us punishable”’, but he claims that Kant thinks conscience does not inflict punishment upon us itself, and that ‘our happiness or misery is left for the ruler of the world to decide’, citing MM, 6: 439n, 440, and 460 in support. But the purpose of Kant’s discussion of God at MM, 6: 439n and 440 is to explain that we must think of the ‘internal judge’ who ‘pronounces the sentence of happiness or misery’ (MM, 6: 439n) as God, while emphasizing that this idea is given to us ‘subjectively, by practical reason’ (MM, 6: 439) and is merely another perspective on ‘morally lawgiving reason … present in our inmost being’ (MM, 6: 440). This makes it clear that anything we must think of God as doing to us in the context of conscience must also be understood as something we do to ourselves.Footnote ⁸ MM, 6: 460–1 is a dramatic passage in which Kant talks about the limits of vengeance:

The kind of punishment at issue here is external punishment we inflict on bodily distinct others, which Kant says cannot be done out of vengeance, because only God can punish out of vengeance. Implicit but obviously present here is the assumption that we must not think of ourselves as like God when we punish bodily distinct others. If we follow Wood in applying this passage to conscience, it would seem to imply that, since we must think of the inner judge as God, and we can think of God as punishing out of vengeance, and this inner judge is an aspect of ourselves, we can think of ourselves as punishing ourselves out of vengeance. We may be hesitant to make too much of Kant’s remarks about vengeance here, since Kant’s God does not always sound vengeful (L-Th, 28: 1086). But this passage also refers to desert, so it implies that God punishes retributively even if we bracket the reference to vengeance, and thus, if we follow Wood in applying this passage to conscience, that we punish ourselves retributively too.

However, since Kant does not refer to conscience at MM, 6: 460, more direct evidence that conscience is retributive would be helpful. Kant offers more direct evidence in his discussion of the negative duty to promote others’ ‘moral well-being’:

This remark seems to imply quite straightforwardly that to experience pain from the pangs of conscience – to suffer remorse – is to deservedly suffer inner reproach.

However, Kant also makes many remarks implying that forward-looking considerations play a role in justifying remorse. In the ‘monkish ascetics’ passage discussed earlier, he states that moralisch zu bereuen requires a ‘view to improving’ (MM, 6: 485). In the Religion, he criticizes ‘remorseful self-inflicted torments [reuige Selbstpeinigungen] that do not … originate in any genuine disposition toward improvement’ (Rel, 6: 77). In the Anthropology he warns against regarding our ‘record of guilt as … simply wiped out (through remorse^t [Reue]), so that [we are] spared the effort toward improvement’ (Anth, 7: 236).

In fact, in some passages, Kant suggests such a complete reliance on forward-looking considerations in justifying remorse that he seems to advocate a purely forward-looking approach, such that the only reasons to feel remorse are forward-looking. In the Mrongovius Anthropology lecture notes (1784–5), Kant criticizes ‘idle desires’ connected with the ‘wish that something would not have happened which, however, now is impossible’ as a ‘distraction’ which is ‘senseless and harmful’. He gives ‘remorse’ (Reue) as an example of such desires, and says that it ‘is good merely insofar as it impels us to cancel the consequences thereof and to act better in the sequel’ (Anth-Mr, 25: 1335, boldface added). His point seems to be that when we feel remorse, we should try to eliminate painful thoughts connected with the wish that we had not acted badly and focus on acting better. But retributive (backward-looking) justifications of remorse would seem to be intrinsically bound up with the thought that past bad actions should not have happened, and so the Anth-Mr, 25: 1335 remarks suggest that there is no role for retribution. The Starke 2 Anthropology lecture notes (1790–1791) express a similar thought: ‘fruitless remorse [Reue] is useless, because it attacks our powers too much’ (Anth-S2, 34). The strongest suggestion of a purely forward-looking account comes in a pair of passages in the Menschenkunde Anthropology lecture notes (likely 1780–1):

If remorse is a delusion with no value in itself if it does not motivate us to improve or make amends, then the justification of remorse is purely forward-looking, and there is no role for retribution. If we attribute this view to Kant, then we have a clear conflict with the view suggested at MM, 6: 394 and 6: 460 (quoted above).

There are, however, remarks in which Kant indicates that there are multiple reasons for remorse, and this would imply a hybrid theory. In Kant’s critique of Johann Schulz’s moral theory, he attributes a number of theses to Schulz with the clear implication that Kant does not accept them. One of these theses is that ‘Remorse [Reue] is merely a misunderstood representation of how one could act better in the future, and in fact, nature has no other purpose in it than the end of improvement’ (RS, 8: 110). If Kant rejects this, he thinks that remorse has another purpose distinct from improvement, and it is reasonable to assume that this is retribution, in light of the foregoing. Support for this view is also offered by Herder’s Metaphysics notes, which attribute to Kant the statement that ‘[i]f remorse [Reue] about the past prevents all attention to the future, it is absurd’ (Met-Her, 28: 90). This suggests that it is appropriate for remorse to be both forward-looking but also backward-looking to some degree. A final and definitive text supporting a hybrid justification appears in the Collins Ethics lecture notes. There Kant writes that preachers attending the dying ‘must … see to it, that people do indeed feel remorse ^t for [bereuen] the transgression of self-regarding duties, since these can no longer be remedied, but that if they have wronged another, they genuinely try to make amends’ (Eth-C, 27: 354). Kant here gives us a case where he thinks the imminence of death means that there is a forward-looking way to respond to some but not all of one’s past wrongs. He states clearly that we should feel remorse for the actions to which we cannot respond in a forward-looking way, and based on MM, 6: 394 and 460, it is fair to assume that we should feel such remorse because we deserve it. His point about the actions to which we can respond in a forward-looking way seems to be that we should feel remorse about them but channel it into improvement and making amends.

Though there are tensions among the passages we have considered, together they give us sufficient evidence to attribute the following view to Kant. We must retributively inflict remorse on ourselves for past wrongs. But this self-retribution has two forward-looking constraints. First, remorse should be channelled into improved behaviour: it should lead us to act better in general (which we have seen at Rel, 6: 77, MM, 6: 485, RS, 8: 110, and Anth-Me, 25: 934 and 1086), and it should prompt us to make amends to the particular people we have wronged (which we have seen at Eth-C, 27: 354 and Anth-Me, 25: 934). Second, remorse must be moderated insofar as moderation is required to avoid damage to our rational agency. We saw this recently in his admonitions against remorse’s ‘distraction’ (Anth-Mr, 25: 1335) and ‘attacks [on] our powers’ (Anth-S2, 34). We have a duty to cultivate conscience, which implies that we have an ability to rationally shape conscience (MM, 6: 401), and the reasons Kant gives us for feeling, channelling, and moderating remorse give us principles for shaping it. We should be careful about which actions we cause ourselves pain over, to avoid a ‘micrological’ conscience ‘burdened with many small scruples on matters of indifference’, and a ‘morbid conscience’ which ‘seeks to impute evil in [one’s] actions, when there is really no ground for it’ (Eth-C, 27: 356). We are obligated to ‘sharpen’ conscience if it is too dull (MM, 6: 401), but we must not make it too sharp: to brood over remorse (über Reue zu brüten) can ‘make one’s whole life useless by continuous self-reproach (Vorwürfe)’ (Corr, 11: 333), and an ‘excess of remorse (Kummers) over … transgressions of duty’ can prompt suicide (Eth-V, 27: 642).

This section has set out Kant’s basic moral reasons for requiring and constraining remorse. We might think of these reasons as outlining the abstract structure of his theory of remorse. But we need more concrete details to answer some important questions that remain. How much remorse does Kant think we should feel, and for how long? And what is the mechanism by which we channel remorse into self-improvement? Kant engages these questions most directly in his discussions of the ‘descent into the hell of self-cognition’, which has connections to his concept of conversion.

3. The descent into the hell of self-cognition

Kant says he borrows the phrase Höllenfahrt des Selbsterkenntnisses from his acquaintance Johann Hamann (CF, 7: 55).Footnote ⁹ A version of this phrase appears in the Collins Anthropology notes (dated to 1772–3), where Kant says that one of the reasons the science of anthropology is neglected, despite being very interesting, is that ‘one conjectures he would not find much to rejoice at if he were to undertake the difficult descent into Hell toward the knowledge of himself [die schwierige Höllenfahrt zur Erkentniß seiner selbst]’ (Anth-C, 25: 7). Kant uses it again in a famous passage in the Doctrine of Virtue:

Following a number of commentators, I think Kant’s reference to removing an evil will, and developing the predisposition to good means that we should link this passage to the conversion in the Religion (Rel, 6: 73), which Kant also calls a ‘justification [Rechtfertigung] of a human being who is indeed guilty but has passed into a disposition well-pleasing to God’ (6: 76).Footnote ¹⁰ Kant does not explicitly discuss remorse in the interpretatively challenging discussion of the conversion in the Religion, but I will argue that we can connect that discussion to a crucial discussion of the Höllenfahrt in Conflict of the Faculties (CF, 7: 10, 54–9) in a way that sheds light on both these texts and on Kant’s theory of remorse.

In Kant’s view, we all begin our lives as radically evil, which means that we are willing to comply with the moral law only insofar as we are also able to satisfy the incentives of self-love (Rel, 6: 36). This is a way of incorporating evil into our maxims. Kant claims that since this evil is in our ‘maxims in general (in the manner of universal principles as contrasted with individual transgressions)’, it entails ‘an infinity of violations of the law’ and thus an ‘infinity of guilt [Schuld]’ for which we must expect ‘infinite punishment [Strafe] and exclusion from the Kingdom of God’ (Rel, 6: 72).Footnote ¹¹ He holds that elimination of radical evil requires us to undergo a conversion in which we transform our disposition so that we choose to satisfy self-love only on the condition that we do our duty. But he thinks that the mere fact of becoming a better person through conversion does not negate our desert of punishment (Rel, 6: 72). Instead, he thinks conversion introduces a puzzle about when we can appropriately be punished. He thinks that it should not be thought of as ‘fully exacted before’ the conversion (Rel, 6: 73). On the other hand, ‘after his conversion … the punishment cannot be considered appropriate to his new quality (of thus being a human being well-pleasing to God)’ (Rel, 6: 73). Kant’s proposed solution is as follows:

While it is important to emphasize that Kant is cautious about using the historical story of Christ as an epistemological foundation in theology throughout his corpus and seems to rely on it only as metaphor, it is clear that he is telling us that the conversion involves a pain that is appropriately understood as excruciating and is like the pain of a mortal wound in some (perhaps symbolic) sense. And since Kant presents this pain as answering to the demand for the infinite punishment he claims we deserve, it is also presumably meant to be understood as infinite. The sense in which it is infinite is unclear, however.Footnote ¹²

Is this pain to be understood as infinite in intensity? It is not clear that Kant’s account of the intensive magnitude of sensation permits infinitely intense sensations. At A176/B217, Kant says that the intensive magnitude of a sensation ‘can be raised from 0 up to any greater degree’, but this presumably does not entail that we can arrive at an infinite degree. And even if Kant’s theory of sensation allows for the possibility of infinitely intense pain, human physiology and psychology probably set limits. But despite these limits, Kant is clearly prescribing a very intense pain, and perhaps its intensity is meant to symbolically represent an infinite intensity.

Is this pain to be understood as infinite in duration? This passage is not clear about how long the pain Kant prescribes should endure, but he goes on to describe the ‘the suffering which the new human being must endure while dying to the old human being’ as occurring ‘throughout his life’ (Rel, 6: 74), and this makes it clear that at least some aspect of conversion pain is experienced as long as we live. Further, Kant’s appeal to the claim that we can only make ‘endless progress’ toward complete conformity with the moral law in his argument for immortality as a postulate of pure practical reason (CPrR, 5: 122; also see Rel 6: 75) might lead us to speculate that at least some of this suffering must continue forever, even beyond earthly life, and thus must be eternal.Footnote ¹³

One way to get clearer about the nature of conversion pain is to attend to the distinction Kant draws toward the end of the passage above between the sacrifice ‘already involved in the emergence from the corrupt disposition’ and the ‘long train of life’s ills’ to which this ‘emergence’ is an ‘entrance’. This suggests that Kant is prescribing some kind of first stage of pain of finite duration, followed by a second stage that involves ‘life’s ills’, which in connection with the remark about suffering throughout life (Rel, 6: 74, quoted above) can be understood to continue as long as one’s life continues.Footnote ¹⁴ Perhaps pain in the first stage symbolically represents infinitely intense pain, and pain in the second stage is literal eternal pain.

In light of the discussions earlier in this article, it seems natural to think that remorse for our past evils would fit somewhere into the pain of conversion, but Kant does not explicitly address the pain of remorse in this discussion of the conversion. His clearest idea in this passage is about pain that results from resistance to the inclinations which ‘lead to sin’, a resistance which we must demand of ourselves in a new way after the conversion.Footnote ¹⁵ That is, while our inclinations persist after the conversion, we have rejected the evil maxim of conditioning compliance with law on satisfaction of our inclinations and have adopted a new maxim of conditioning satisfaction of inclinations on compliance with law, and resisting inclinations can be painful. Since having inclinations which do not necessarily conform to law is simply part of what it is to be a sensibly conditioned rational agent, we must assume that our inclinations will persist as long as we are such agents, and so it makes sense to think that the pain deriving from resisting those that dispose us to act immorally would continue as long as we continue as such agents. This kind of pain is thus a good candidate for what I claim is a second stage of conversion pain.

However, Kant’s focus on the pain of resisting inclinations creates an interpretative conundrum. Kant thinks that to understand the infinite punishment we must receive for past evils as ‘adequately executed in the situation of conversion itself’ (Rel, 6:73, quoted above), we must see suffering in this life as punishment for our past evils. But the pain of resisting inclinations is a puzzling thing to regard as punishment for past evils, for three reasons. First, since having inclinations which do not necessarily conform to law is simply part of being a sensibly conditioned rational agent, dealing with the pain of resisting inclinations is presumably part of the daily affairs of such agents. How can an ordinary kind of unpleasantness required to maintain rational agency be regarded as punishment? Second, it is pain we must grapple with in order to do the right thing in the present, so it seems to lack the backward-looking reference which seems essential to regard pain as deserved in light of past wrongs. The third reason takes longer to explain but is the most important. Consider some features of Kant’s view of the ontological and moral status of inclinations. The ‘natural inclinations do not have us for their author’ – they are ‘conatural to us’ and we ‘cannot presume ourselves responsible for them’ (Rel, 6: 35). They themselves are ‘innocent’ and bear ‘no direct relation to evil’ (ibid.). They only lead to sin if we adopt a fundamental maxim of conditioning our compliance with law on satisfaction of our inclinations and that fundamental maxim does not have a ‘ground … in the natural inclinations’ (Rel, 6: 34). The ‘inclinations only make more difficult the execution of the good maxims opposing them’ (Rel, 6: 57n). Now consider two sensibly conditioned rational agents: agent 1, a hypothetical agent who begins life in the right way, by prioritizing duty over inclination, and thus does not deserve to suffer, and agent 2, who begins life in the wrong way, by inverting that priority, and thus does deserve to suffer. Kant’s view of the ontological and moral status of inclinations would seem to imply that both 1 and 2 would have to contend with the pain of resisting inclinations. If God or Reason say that it is just for 2 to have this pain because of his past evils, then 1 can reasonably object that it is unjust for him to have this pain too. We might suppose that there is an implicit background metaphysics which respects the deserts of 1-type agents by ensuring that they are never incarnated as sensibly conditioned agents, but that would seem to imply that we are in some sense responsible for being assailed by inclinations, and (as explained) Kant appears to reject this.Footnote ¹⁶

Kant’s idea that the pain of conversion is justified by our past evils would thus make more sense if he had a more plausible locus of punitive pain. I think what I claimed above to be the first stage of conversion pain can provide such a locus if the pain in that stage is remorse. Evidence to support this idea appears in Conflict of the Faculties, where Kant explicitly discusses the concepts of conversion, the Höllenfahrt, and remorse. Care is required to put this discussion forward as evidence about Kant’s own view, because he offers it in the context of a critique of the Pietist conception of conversion.Footnote ¹⁷ In the Preface to this work, he says that Pietists hold that forgiveness requires ‘an overwhelming remorse [Zerknirschung]’, which he also calls ‘a deep remorseful sorrow^t’ (ein tiefer reuiger Gram) (CF, 7: 10). However, he faults their view not because of the deep remorse it prescribes, but because they hold that the ‘human being’ cannot ‘attain this by himself’, and think ‘the remorseful^t [reuvolle] sinner must especially beg this remorse [Reue] from heaven’ (ibid.). He goes on to argue that it is ‘obvious that anyone who still has to beg for this remorse^t [Reue] (for his transgressions) does not really feel remorse for ^t [wirklich nicht reuet] his deeds’ (ibid.). So he is not denying the claim that we ought to feel deep remorseful sorrow – he is rather asserting (contra Pietism) that we can attain it on our own. Later, in the ‘General Remark: On Religious Sects’ (CF, 7: 48–61), he explicitly states that this remorse is part of conversion. At 7: 54, he asks ‘how is rebirth (resulting from a conversion by which one becomes another, new man) possible[?]’ and writes that

This passage clarifies Kant’s view of remorse in five ways. First, it appears to have the same basic structure as the critique of Pietism in the preface: it is obviously critical, but the target of criticism is not the view that we ought to experience profound remorse, but is instead the view that we cannot attain it on our own. Second, it provides support for the idea that Kant himself endorses the view that we ought to experience such remorse, because he clearly indicates that such remorse is (or is part of) the Höllenfahrt des Selbsterkenntnisses, which is established as an element of Kant’s own moral psychology by the way he refers to it at MM, 6: 441 (quoted above). Third, Kant’s reference to the kind of remorse which Pietists might acquire by begging God for it as Buße rather than Reue is in line with the distinction we observed earlier, such that Buße is not wahre, moralische Reue. Fourth, if remorse is self-retribution inflicted by a part of ourselves which we must also represent as God, then it is comprehensible to regard it as punishment that we deserve for our past evils, which I argued above is a problematic way to regard the pain of resisting inclinations. Fifth, if we suppose that Kant’s own view is that we should experience remorse like this under our own power (presumably through the faculty of conscience), then Kant’s smelting metaphor gives us a detailed phenomenological picture of the mechanism he thinks we use to improve ourselves with remorse, and of how the experience of remorse comes to an end.Footnote ¹⁸ We force the pain of remorse to rise to a great intensity, which leads to a kind of heartbreak, and with it, a conversion in which we adopt the disposition to satisfy self-love only on the condition that we do our duty, at which point the remorse abates and is driven ‘completely from [our] heart[s]’ (CF, 7: 55, quoted above). Kant presumably sees the ability to drive remorse from our hearts when we complete our conversions as (at least part of) the solution to the problem about excess remorse which he warns us about at Eth-V, 27: 642.

I want to be cautious about suggesting that Kant thinks there is a causal process here, such that remorse causes the conversion, and the conversion causes remorse to cease, since it is clear that Kant thinks of this phenomenology as the manifestation of noumenal freedom. However, if we have an empirical character which is governed by causal laws, along with a noumenal character which is not governed by causal laws (see e.g. A539/B567), we can entertain the possibility that there are empirical-psychological laws according to which the phenomenal experience of remorse sometimes causes our phenomenal experience of conversion without suggesting that remorse and conversion are not manifestations of noumenal freedom. We must, however, assume that there is enough diversity in such laws, or in the antecedent conditions determining their instantiation, to accommodate experiences of remorse which do not conform to the ideal Kant seems to prescribe here, since actual experiences of remorse clearly differ widely, and it seems safe to say that many do not conform to Kant’s ideal. This is in one way unproblematic, since Kant regularly emphasizes that his moral theory involves ideals which are (at best) challenging to actualize. It is problematic in other ways, however.

4. The hazard of suicide

Kant’s remarks on conversion and the ideal experience of remorse are apparently meant to describe something he thinks it is possible for us to experience, but also something we deserve to experience and that we must experience to purge ourselves of radical evil. (See e.g. MM, 6: 441, quoted above.) This suggests that he is asking us to crush our hearts with a remorse whose intensity symbolizes crucifixion, or a mortal wound, until we complete our conversions. What does this mean for remorseful people who have not had the conversion experience, or are not confident they have? Kant’s theory would seem to imply that they must continue to punish themselves with ever-more intense remorse, until they are confident that they have made the Höllenfahrt and converted themselves. But it seems safe to say that many might seek to follow this advice without ever having an experience that gave them confidence that they were converted and could therefore justifiably give themselves a respite from remorse. We have seen a number of remarks from Kant in which he acknowledges that remorse which is protracted or intense can damage rational agency, and even dispose people to suicide. People who strive for the Höllenfahrt without ever becoming confident they have achieved it are likely to experience remorse that is protracted and intense.

The concern that we might lack confidence about whether we are converted is a serious one, at least in the context of Kant’s moral epistemology. There are two puzzles here. First, there are (perhaps surprisingly) no behavioural criteria for conversion. Kant thinks that conversion does not guarantee subsequent moral perfection – conversion only gives us ‘hope … to find [ourselves] upon the good (though narrow) path of constant progress from bad to better’ (Rel, 6: 48). I think Kant does assume that conversion entails that one’s actions will improve in the long run, but it may be a very long run, since reason requires us to postulate immortality to ensure that we have enough time to make the necessary improvement (CPrR, 5: 122). Years or decades of actions that violate duty after an experience that seemed like a conversion do not entail that that it was not a conversion. If it was a conversion, then perhaps it is sufficient to respond to modest moral failures with moderate experiences of remorse, rather than the intense remorse of the Höllenfahrt, because we can hope that we are already on the path from bad to better. But if it was in fact not a conversion, then intense remorse is demanded. How can we tell? The second puzzle is that even if, after a seeming conversion, we seem to be fulfilling all our duties, and fulfilling them for the right reasons, we still cannot be confident that we are converted. This is because of Kant’s scepticism about ever really being certain about our underlying motives:

In the context of conversion, this passage implies that even if we seem to have had a conversion experience, we may still be radically evil, and we may need to continue to strive for intense remorse.

The concern that some may be driven to suicide by remorse is also a serious one, and it is recognized not only by Kant, but also in contemporary clinical psychology.Footnote ¹⁹ There may even be a kind of tragic rationality in responding to the belief that one deserves infinite punishment by excising oneself from the world. Perhaps this is why Kant makes the troubling remark that suicide from ‘excess of remorse [Kummers] over [one’s] transgressions of duty’ is not a ‘crude’ kind of suicide ‘which should be an object of general hatred’, but is rather a suicide which ‘could betray a worth of the soul’, like suicide for ‘the conservation of [one’s] honour’ (Eth-V, 27: 642), despite the fact that he sees suicide as a grave violation of duty.

We may even have an example of suicide motivated by remorse in Kant’s own correspondent Maria von Herbert. In a 1791 letter to Kant, she writes that when she revealed a protracted but harmless lie (apparently connected with a previous relationship) to the man she loved, ‘his love … vanished’ (Corr, 11: 273). She asks Kant for ‘solace, or for counsel to prepare [her] for death’, making it clear that she has been contemplating ‘tak[ing her] own life’ (ibid.). In 1792, Kant writes back to encourage her not to commit suicide, recommending ‘composure’, and advising that ‘life, insofar as it is cherished for the good that we can do, deserves the highest respect and the greatest solicitude in preserving it and cheerfully using it for good ends’ (Corr, 11: 334). But he denies her a ‘moral sedative’: he remarks that even a harmless lie is ‘a serious violation of duty to oneself and one for which there can be no remission’, and lectures her about some of the details of his theory of remorse discussed above, sternly advising that her ‘bitter self-reproach’ for the lie should not be Reue over ‘imprudence’ (Unklugheit) but Reue ‘grounded in a purely moral judgement’ (auf bloßer sittlicher Beurtheilung …Verhaltens gründet) about her behaviour (Corr, 11: 331–3). He goes on to attempt to temper his admonition with a remark part of which we have already considered:

This remark obviously leaves open the possibility that we have a reason to think continued self-torture is deserved if we are not sure of having reformed, that is, if we are not sure our conversion is complete. Von Herbert appears to have been quite familiar with Kant’s work, and thus may well have been familiar with his scepticism about ever really being certain about our underlying motives. She wrote to Kant again in 1793, expressing a ‘sense of constantly reproaching [vorwerfen] herself’ (Corr, 11: 401), and a continued desire to ‘shorten [her] so useless life’ (Corr, 11: 402), and once again in 1794, stating that while she still wished to die, she felt that ‘if people take morality and friends into account [they can] with the greatest desire to die still wish for life and try to preserve it no matter what’ (Corr, 11: 486). Kant did not reply to either of these letters, and von Herbert committed suicide in 1803.

Owen Ware argues that in the Religion and Metaphysics of Morals, Kant struggles against his scepticism about the knowledge of our underlying motivations which could demonstrate the completeness of our conversion. Ware thinks that while Kant holds that we lack knowledge that our conversion is complete, he arrives at ‘subjectively sufficient grounds’ for the belief that our conversion is complete, grounds which yield what Kant calls ‘conviction’’ (Ware Reference Ware2009: 694, citing A822/B580). Ware argues that a key threat to knowledge of our underlying motivations is that when we evaluate ourselves, we are always inclined to see ourselves as better than we really are, or as Kant puts it in the Groundwork, that we like to ‘flatter ourselves by falsely attributing to ourselves a nobler motive’ (G, 4: 407). He argues that Kant circumvents this problem in his account of conscience by appointing God as an authoritative and distinct person within us: representing the verdict of conscience as handed down by God provides the verdict with an authority that self-evaluation lacks. Ware thinks this implies that if we are diligent about submitting not only our particular actions but also our maxims for the scrutiny of conscience, and conscience does not hand down a verdict of ‘guilty’, we can have conviction that our conversion is complete. This is a plausible and interesting account of some of Kant’s aims in his theory of conscience. But as discussed earlier (and as Ware acknowledges), we must recognize that what we represent as God in conscience is also an aspect of ourselves which we ourselves set up as authoritative, however rationally rarefied it may be. As Kant puts it, ‘[a]lthough it certainly sounds questionable, it is in no way reprehensible to say that every human being makes a God for himself’ (Rel, 6: 169n). Would it be unreasonable for someone uncertain about the completeness of his conversion to worry that he might have unwittingly incorporated a bit of self-flattery into the God he has made and that the verdict of his conscience therefore lacked authority, so that even if his conscience did not demand continued self-retribution, he might deserve it anyway? This objection might be seen as asking more of Kant’s ethics than it can reasonably provide, were it not for the serious damage to rational agency that threatens us if the authority of Kantian conscience is unstable in rational reflection. It seems to me that it is unstable and that its authority depends as much on a kind of faith as it does on reason, a faith which is least likely to be found in the people who need it most. Maria von Herbert may be an example of such a person, and her brother, Baron Franz Paul von Herbert, may be another.

The historical evidence that Franz Paul had expertise in Kant’s philosophy is as strong as it is with respect to Maria. The von Herbert family were lead paint manufacturers in Klagenfurt, and Franz Paul left the family business to study Kant’s philosophy in Jena and Weimar, making the acquaintance of a number of Kantian philosophers including Karl Leonard Reinhold. According to Wilhelm Baum, Reinhold and von Herbert had many conversations, and Reinhold wrote in a letter that von Herbert was his ‘house and table companion’ for four weeks (Baum Reference Baum1996: 489). Upon Franz Paul’s return to Klagenfurt, he established a discussion circle to debate Kant’s philosophy, despite strong local political opposition (Langton Reference Langton1992: 481, Baum Reference Baum1996: 499–504). Franz Paul committed suicide in 1811, after leaving this gnomic but sadly evocative passage in his will in 1810:

This passage is of course too oblique and brief to count as a critique of Kant’s idea of conscience, but it seems fair to say that it manifests concern about the reflective instability of that idea. It would probably be unfair to suggest that Kant’s philosophy played a role in the death of either of the von Herbert siblings. People commit suicide for a variety of reasons, and the von Herberts’ consciences were not the only painful things in their lives. Maria had lost her love, and Franz Paul’s activities were the subject not only of political opposition but also active police scrutiny (Baum Reference Baum1996: 499–504). Further, for all we know, the von Herberts may have had a genetic predisposition to major depressive disorder and may have suffered from chronic lead exposure, which is also positively correlated with major depression. But it does seem reasonable to think that we have textual evidence that remorse and anxieties about conscience may have played a role in their deaths and that neither found anything in Kant’s philosophy to keep them alive. It is probably unreasonable to hope that any philosophy can prevent suicide. But it does not seem unreasonable to expect an account of how to moderate remorse from a philosopher who advocates heart-crushing remorse despite acknowledging that it can prompt suicide.

It may be objected that Kant does address the problem of moderating extreme feelings in his account of rationally regulating feeling to avoid what he calls affect (e.g. MM, 6: 407–9). ‘Affect’ refers to feeling that we allow to develop without moderation by reason, which can be overwhelming in a way that is damaging to agency, because it makes ‘reflection … impossible or more difficult’ (6: 407). Kant holds that we have what he calls a ‘duty of apathy’ to suppress affect, which we can fulfil through our ‘inner freedom … to bring all [our] capacities and inclinations under [our] control and so to rule over [ourselves]’, and avoid being ‘governed by [our] feelings’ (6: 408). Perhaps we should regard remorse so intense that it might overwhelm us and drive us to suicide as affect, and expect ourselves to draw on this inner freedom to diminish its intensity. But it would seem that we cannot regard even terribly intense remorse as affect, because of the role Kant gives such remorse in conversion – such remorse is an aspect of the infinite punishment we deserve, and is part of the descent into the hell of self-cognition which is necessary for conversion.