2. Retribution, deterrence, and dilemma

Traditional, retributive interpretations of Kant’s penal theory rely on several striking passages. Kant claims punishment ‘can never be inflicted merely as a means to promote some other good for the criminal himself or for civil society. It must always be inflicted upon him only because he has committed a crime’ (MM, 6: 331). Doing otherwise, Kant argues, uses a person as a mere means, violating her innate right; thus, he rejects wholly deterrent penal theories (MM, 6: 331). Moreover, Kant claims the law of retribution is ‘always the principle for the right to punish since it alone is the principle determining this idea a priori’, such that ‘whatever undeserved evil you inflict upon another within the people, that you inflict upon yourself’ (6: 332, 363). The law of retribution, he continues, is the only method that ‘can specify definitely the quality and the quantity of punishment’ (6: 332). Since the law of retribution implies like for like, there is no appropriate penalty for murder other than death (6: 333). Hence, Kant’s defence of the death penalty and his notorious claim that ‘every murderer – everyone who commits murder, orders it, or is an accomplice in it – must suffer death’ (6: 334). By predicating just punishment on legitimate wrongdoing and proposing that the proportional law of retribution is the only appropriate way to specify punishment, these passages have traditionally suggested a thoroughly retributive penal theory.

Kant’s claim that punishment is necessary to treat criminals and their victims as ends in themselves puts a unique spin on these retributive commitments, but many modern interpreters still reject this traditional interpretation. One such interpreter, Sarah Holtman (Reference Holtman and White2011: sec. 2), argues that the traditional interpretation conflicts with some of Kant’s other claims and does not cohere with his coercive conception of justice. This conception turns on the circumstances of justice, in which our actions necessarily affect one another and thereby threaten our freedom.Footnote ³ To eliminate this threat, justice ought to unite our choices with others’ choices so that, together, we can live freely (see Holtman Reference Holtman, Pinheiro Walla and Ruhi Demiray2020: sec. 4). Thus, Kant articulates the universal principle of right: ‘Any action is right if it can coexist with everyone’s freedom in accordance with a universal law, or if on its maxim the freedom of choice of each can coexist with everyone’s freedom in accord with a universal law’ (MM, 6: 230). Conversely, actions that cannot coexist with everyone’s freedom are unjust. Since injustice coercively hinders another’s freedom, the universal principle of right authorises hindering such hindrances with coercive law, which includes a state’s penal system. This kind of state-sanctioned coercion, Kant argues, is the only solution to the problems the circumstances of justice pose.

By justifying the state on these premises, Kant limits the scope of state power to external relations. This conforms with his argument that moral motivation is a mechanism wholly internal to an agent’s volition. Since the state cannot influence these internal practical deliberations, it is limited to external carrots and sticks. Juridical duties, then, stem from ‘incentives and aversions’ (MM, 6: 219). Deterrence, as an external aversion, can now enter the Kantian picture as a means to secure external freedom. Although deterrence threatens to treat persons as mere means and contravene Kant’s moral commitments, we will see in the next section that Kant’s justification for the coercive state avoids this problem. As such, Holtman concludes that Kant’s conception of the state undermines the traditional interpretation since it makes punishment’s purpose and justification partly deterrent (see Holtman Reference Holtman and White2011: 109-10; see also Byrd Reference Byrd1989: sec. 1). This observation grounds modern interpretations that claim tenable Kantian penal theories must mix deterrent and retributive elements (Holtman Reference Holtman and White2011: 110). Holtman’s positive account substantiates one such mixed penal theory, which she calls a civic respect theory.Footnote ⁴

Holtman argues that by emphasising the criminal’s moral and juridical personality, we can account for her intentions and state of mind (i.e., her subjectivity), deter future crime, and make the criminal understand the wrong she inflicted on her contemporaries by employing the law of retribution (Holtman Reference Holtman and White2011: 111-2). This respects the criminal’s victim, and our respect for the criminal herself limits the extent of deterrent punishment. This limitation implies that – under special circumstances – a criminal’s subjectivity can excuse her from or reduce her punishment.Footnote ⁵ These considerations, Holtman claims, are grounded by Kant’s conception of civic respect, which entails four basic criteria of just punishment.Footnote ⁶ The first two are retributive in that punishment should (1) acknowledge the victim’s or victims’ loss and (2) express the state’s and citizenry’s respect for criminals (Holtman Reference Holtman and White2011: 121-2). Criterion (2) is retributive since Kant insists that respecting criminals necessitates retributive punishment. The remaining criteria are deterrent since punishment should (3) deter future crimes and (4) account for the criminal’s subjectivity, which may have made her insensitive to deterrence (Holtman Reference Holtman and White2011: 121-2).Footnote ⁷ These criteria are compelling, but applying them to certain cases reveals a problem, the exposition of which will clarify Holtman’s criteria and my classification of them.

Consider another passage that conflicts with traditional interpretations. In it, Kant acknowledges that the interplay between criteria (3) and (4) implies the right of necessity, ‘an authorization to take the life of another who is doing nothing to harm me, when I am in danger of losing my own life’ (MM, 6: 235). This right is a right in name only, he notes, since it contradicts the universal principle of right; after all, my freedom of choice cannot coexist with another’s if I use it to kill another (6: 235). Nonetheless, he claims we should find one who exercises the right of necessity subjectively unimpugnable and, therefore, unpunishable. Kant clarifies by asking us to imagine that, in a shipwreck, Jones pushes Smith off a floating plank of wood, killing him to save herself. While Jones is objectively culpable, she is subjectively unpunishable because a law punishing her for killing Smith ‘could not have the effect intended’ since ‘the punishment threatened by the law could not be greater than the loss’ she would suffer if she followed it (6: 235). In other words, Jones’ subjectivity made her insensitive to deterrent reasons; as such, punishing her could not have a deterrent effect. According to criteria (3) and (4), then, excusing Jones from punishment is sound. Since the state can only use coercion to achieve its ends, and coercion cannot be effective in the cases to which the right of necessity applies, the state cannot punish people like Jones despite their objective culpability.

This analysis suggests that, in addition to retributive proportionality and guilt, a necessary condition for just punishment is that the threat of punishment can guide a person’s actions. In other words, a threat of punishment must be able to deter the persons it threatens.Footnote ⁸ However, this discussion does not illuminate Kant’s reference to objective culpability and subjective impunity. Presumably, Kant’s conception of imputation grounds these categories. He defines imputation as ‘the judgment by which someone is regarded as the author (causa libera) of an action, which is then called a deed (factum) and stands under laws’ (MM, 6: 227).Footnote ⁹ When a court determines imputation, it has the authorisation to enforce a deed’s rightful consequences; otherwise, imputation is merely evaluative (6: 227). Kant asserts that if a person does what duty requires, we cannot impute her action’s good or bad effects (6: 228). If a person acts meritoriously, we can only impute her action’s good effects (6: 228). However, if a person acts wrongly, she is ‘morally culpable’, and we can only impute her action’s bad effects, the rightful consequence of which is retributive punishment (6: 227-8). In any case, ‘Subjectively’, the degree to which we can impute an action’s effects must ‘be assessed by the magnitude of the obstacles that had to be overcome’ by the agent (6: 228). As such, one is objectively culpable if she does wrong but can limit the imputation of her wrongdoing’s bad effects by appealing to subjective, contingent factors. This feature creates conceptual space for subjective impunity. With this conception of imputation, we can specify what constitutes responsibility and clarify the categories on which the right of necessity relies.Footnote ¹⁰

Clearly, imputation and the right of necessity are substantively compatible with Holtman’s civic respect theory. Nonetheless, it is unclear how they are mutually compatible with the passages that traditional retributive interpretations emphasise, which – per Kant – justify her first and second criteria. Importantly, despite its role in the right of necessity, imputation seems to ground the law of retribution by structuring and justifying the courts’ enforcement of the rightful consequences of a person’s actions. If this is the case, however, it creates a dilemma. To grasp it, consider Kant’s defence of the death penalty. Informed by the categorical imperative (CI), innate right, and the universal principle of right, the general will of a state’s citizens constructs juridical law. Since the general will represents the criminal’s objective self (i.e., her pure practical rationality), the criminal’s own idealised and responsible perspective justifies the law’s punishment (MM, 6: 335).Footnote ¹¹ As such, the criminal cannot object to her sentence by appealing to her subjectivity and is liable to whatever punishment the law of retribution specifies. Since the criminal’s objective self necessarily accepts her punishment, by carrying out the death penalty in appropriate cases, we treat her as an end in herself.Footnote ¹²

Kant’s conclusion may be shocking, but if we take his position, then respecting the criminal as the author of her actions necessitates punishing her according to the law of retribution. For him, this notion grounds and substantiates Holtman’s first and second criteria. However, Kant’s justification for these criteria implies his conclusion about justifying punishment to the criminal’s objective self, which seems unable to accommodate Holtman’s deterrent criteria and, thus, the right of necessity. Let us make this explicit. If the death sentence is the only way to treat a person who unjustly kills another as an end in herself, then excusing someone because of her contingent circumstances wrongs her. In excusing her, a court would effectively refuse to impute her action’s effects, which disrespectfully implies that she is not the author of her actions. However, if persons like Jones do author their actions, they are objectively culpable for killing in the cases to which the right of necessity applies, the rightful consequence of which is death. This raises a question: in such cases, how does a criminal’s subjectivity override her objective culpability and make her subjectively unimpugnable? This brings any mixed Kantian penal theory’s deterrent and retributive elements into conflict, creating problems for theorists like Holtman. Although the criteria she articulates are compelling, Kant’s justifications for them do not appear to be mutually compatible.Footnote ¹³

At this point, we might ask whether a different kind of mixed Kantian penal theory avoids this problem. Matthew Altman develops one such theory, which elaborates on H. L. A. Hart’s (Reference Hart and Gardner2008) argument that the justification of particular punishments is retributive despite the justification for the institution of punishment being deterrent. His theory is sophisticated (see Altman Reference Altman2021b), but since we are focusing on Kant, we will evaluate Altman’s theory as a Kantian penal theory. The Hart-inspired strategy behind Altman’s approach is most compelling when he recognises that Kant folds the institution of criminal punishment into a coercive state schema that aims to maintain a rightful civil condition. Although this move generates deterrent considerations, because the state is meant to protect and enhance autonomy, its penal system must respect criminals’ autonomy and the imputation to which they are, as such, subject (see Altman Reference Altman, Himmelmann and Serck-Hanssen2021a: 1684-5). However, as with Holtman’s theory, if we apply Kant’s justifications to Altman’s mixed theory, then we face the problem I just articulated. Appealing to the different levels at which deterrence and retribution are relevant is insufficient to resolve it because, within Kant’s text, nothing mediates these considerations in the cases to which the right of necessity applies. Altman might have some reply to this line of thought, but it is not clear that Kant does.Footnote ¹⁴

We have a dilemma, then, if we agree with Kant that retributive punishment – including the death penalty – is the only appropriate response to crime and that accounting for persons’ subjectivity is morally desirable. On one horn, objective culpability denies subjectivity’s relevance and necessitates punishment according to the law of retribution, such that treating criminals otherwise uses them as a mere means and trivialises the wrong done to the criminal’s victim or victims. On the other horn, punishment’s purpose is partly deterrent, which creates excusing conditions under which a criminal is subjectively unimpugnable and objectively culpable. If we impale ourselves on the first horn, we lose the penal theory’s deterrent and excusing features, which are juridically necessary and morally desirable. If we impale ourselves on the second horn, we no longer treat the criminal as an end in herself, wronging her and disrespecting her victims. We will elaborate on this dilemma, but as it stands, two explanatory desiderata are necessary to resolve it.

First, we need to show how subjectivity is relevant to a criminal’s culpability and penal sentence in the cases Kant identifies and what distinguishes these excusing subjective considerations from non-excusing subjective considerations in other cases. Second, we need to show why we should consider subjectivity in penal sentencing and explain the interplay between it and objectivity in a way that accommodates excuses in the cases we discussed. Doing so is necessary for Kant to maintain his commitment to justifying punishment to the criminal without arbitrarily deciding when a criminal’s subjectivity is relevant. I will argue that the followability requirement, which O’Neill articulates, is the theoretical resource necessary to satisfy these desiderata. I must connect Kant’s moral and political philosophy to make this argument. Helpfully, this move clarifies the dilemma’s structure.

3. Morality, justice, and honour

The foundation of Kant’s theory of justice, to which I alluded, is the innate right of humanity. He introduces the concept by writing that ‘Freedom (independence from being constrained by another’s choice), insofar as it can coexist with the freedom of every other in accordance with a universal law, is the only original right belonging to every man by virtue of his humanity’ (MM, 6: 237). To make sense of this passage, we might appeal to Kant’s moral philosophy. On his account, a law is definitionally unconditional, and a universal law must arise a priori from practical rationality. These considerations ground the CI, the supreme law of morality. Roughly, humanity is the rational capacity to autonomously recognise the CI, which makes a being an end in itself.Footnote ¹⁵ This intrinsic normativity grounds persons’ particular ends, the making of which demonstrates a person’s dignity. A person’s dignity necessitates that we treat her as an end in herself, which entails that our maxims must be universalizable, thus grounding persons’ self-conception as lawmakers in a kingdom of ends. These moral considerations, Kant seems to claim, generate a person’s innate right.

The kingdom of ends is Kant’s ideal moral and political situation, which Holtman describes as a moral community united by its recognition of the CI, its fundamental respect for humanity as an end, and the moral and sociopolitical perspective that each person ‘takes on her own choice of maxims in pursuit of her particular ends and on all related actions’ (Holtman Reference Holtman, Pinheiro Walla and Ruhi Demiray2020: 89; see also G, 4: 448). Ideally, we are united by shared standards and acknowledgement of each other’s value as persons: we follow the CI and legislate autonomously, thereby creating a community bound by a general will subject to moral constraints. However, we do not live in a perfect world, and limited resources can produce distributive conflicts even if two persons’ ends are universalizable. Even if they do not conflict, these persons’ choices inevitably affect one another, constraining their external freedom. Such problems, which the circumstances of justice imply, threaten our innate right as lawmakers. Kant argues that resolving these problems necessitates a state that can appropriately regulate these external relations (see MM, 6: 311; see also Holtman Reference Holtman, Pinheiro Walla and Ruhi Demiray2020: 95).

According to this conception of the state, its purpose is to protect persons’ innate right. To be authoritative, a state’s coercive laws should represent the general practical will of its citizens so that each has sufficient reason to follow them, thus protecting and facilitating their lawmaking capacity. Therefore, through the general will, the law is self-legislated. As such, when criminals break the law, they threaten their victims’ innate right and make themselves liable to punishment due to self-legislated standards. This commitment clarifies Kant’s argument involving a criminal’s objective self in the preceding section and shows that – because criminals’ practical reason sanctions the law – the deterrent component of his juridical system avoids treating them as mere means.Footnote ¹⁶ By incentivizing its citizens to follow the law through punishment, the state protects persons’ innate right, confirming punishment’s partly deterrent purpose and creating space for subjective impunity. Simultaneously, innate right constrains punishment’s extent since civic respect requires treating criminals appropriately. However, we should note that appropriate treatment, due to the law of retribution, necessitates punishment and seems to deny subjective excuses. We may view the dilemma, then, through the lens of Kant’s joint moral and political commitments. These connections provide a productive way to think about the dilemma and coherently substantiate the concept of innate right, which circumscribes and justifies a coercive state that protects its citizens’ dignity.

To put this approach to use, let us consider crimes of honour, which raise the same problem as the right of necessity. Per Kant, one commits a crime of honour to protect one’s honour as a citizen and person, making one partly excusable. These crimes are exemplified, he thinks, by an unwed mother who kills her illegitimate child and a duelling soldier who kills someone who insulted his military honour. Kant argues that, in each case, although the killers are objectively culpable because they violate rightful laws prohibiting killing, a court cannot sentence them to death because of the discrepancy between the people’s and state’s conceptions of honour (MM, 6: 337). By describing the people’s conception as subjective and the state’s conception as objective, Kant seems to obliquely reference his account of imputation and the cases to which the right of necessity applies. While the strength of the incentives involved in each type of case is different, in both, the criminal’s subjectivity creates an excuse because – despite her objective culpability – punishing her under such conditions would contravene the law’s partly deterrent purpose. Due to their society’s subjective, contingent conception of honour, criminals of honour are compelled to think their crime is necessary to protect their innate right. Since these societal demands mirror the operative pressures in the shipwreck case, we should excuse criminals of honour from the death penalty.Footnote ¹⁷

Kant’s conclusion about crimes of honour raises the same kind of question as the right of necessity: how do subjective incentives of honour excuse objective culpability? In these cases, however, there is an additional layer of complexity since Kant does not clearly articulate what is threatening the criminal of honour’s innate right. After all, having an illegitimate child or being called a coward does not necessarily violate one’s innate right. Here, conceptions of honour, their grounds, and their content become relevant.Footnote ¹⁸ Kant implies that innate right grounds a specific conception of honour that makes duties of justice accessible to all citizens (MM, 6: 236-7). As such, what it means to have rightful honour, which ‘consists in asserting one’s worth as a human being in relation to others’, will vary according to cultural standards (6: 236). Just as Kant specifies his theory of justice to its appropriate circumstances, we must specify laws to their communities, which entails substantiating rightful honour with their contingent cultural standards.

If this subjective conception of honour is left unregulated or unreformed by ‘barbarous and undeveloped legislation’, Kant claims that it generates a discrepancy between one’s duty and what one thinks is one’s duty (MM, 6: 337). This produces crimes of honour that criminals think are necessary to protect their innate right but which, objectively, are unnecessary (6: 336).Footnote ¹⁹ This context elucidates Kant’s examples; as a lawmaking end in yourself, you have a duty to prevent others from mistreating you, and – in the circumstances of justice – this duty is specific to the context in question. However, being imperfect persons, agents are likely to mistake some things for duties when they are not. This phenomenon can occur at a cultural scale, such that people may have a subjective conception of honour that suggests having a child out of wedlock or being called a coward mistreats you as an end in yourself. In each of these cases, because the cultural beliefs Kant cites are so endemic, if the persons in question were to disregard them, they would face perceptions of indignity that make them subject to social ostracisation or violence. These dangers emerge even though rightful, objective honour does not entail – and must condemn – these cultural, subjective conceptions of it.Footnote ²⁰

Therefore, conflicts between subjective and objective conceptions of honour create crimes of honour; they are especially problematic due to the sociopolitical necessity of each conception. In these crimes of honour, one’s innate right seems under threat, and because there is no way to defend it other than by breaking the law, subjective conceptions of honour incentivize crimes of honour while the objective conception condemns them. Thus, there is an analogy between crimes of honour and the right of necessity. In both cases, from the killer’s perspective, the killing is necessary to protect her personhood, but because her action does not accord with the law, she is objectively culpable. Due to her subjectivity, whether there was an external law forbidding her action or not, the killer would have committed it. As such, punishment is purposeless on deterrent standards. In each case, a person does wrong to protect her innate right, which the law and deterrent punishment are supposed to protect. These features generate an excuse and make punishment inappropriate.

In crimes of honour and the cases to which the right of necessity applies, then, the killer is objectively culpable but unpunishable with the death penalty, which would – according to Kant – otherwise be appropriate (MM, 6: 336-7). Since a court cannot ignore the subjective conception of honour and its incentives due to honour’s crucial role in society, just as it must acknowledge a person’s subjectivity if the right of necessity applies, it must acknowledge the state of mind honour creates (6: 336-7).Footnote ^{21 ,} Footnote ²² Here, imputation is relevant because it grounds subjective excuses. Recall, however, that it also makes persons liable to retributive punishment by determining an action’s rightful effect, including the death penalty. As such, imputation is necessary to justify applying the law of retribution and the notions of subjective impunity and objective culpability. Kant seems to notice this tension by writing that refusing to execute criminals of honour seems ‘either cruel or indulgent’ since the law of retribution and respect for their personhood seems to demand it (6: 336). Nonetheless, he concludes that we should not execute them. Without explicitly analysing how imputation generates this tension, Kant leaves obscure how a criminal’s subjectivity can override her objective culpability, preserving the dilemma and undermining his argument.

Although connecting right and morality makes this problem especially apparent, this move also furnishes the resource necessary to resolve it. This resource is a constraint on practical rationality that grounds these excusing features and coheres with Kant’s emphasis on the general will and moral community. I contend that O’Neill’s followability requirement is the key to this dilemma. Kant argues that even the CI’s strongest demands are followable if one is appropriately motivated by respect for the law (G, 4: 439-40). Moreover, the CI implies publicity, requiring that its demands – and agents’ particular maxims – are cognitively accessible (see Zinkin Reference Zinkin2016). These two commitments ground the followability requirement. As O’Neill notes, if the laws that generate reasons are not followable, how we could universalise them, how they would work in a kingdom of ends, or how they could function as juridical law is unclear. If such reasons were not followable, they could not be rationally authoritative, making their status as reasons dubitable since – to be reasons – rational beings must recognise their normative authority (O’Neill Reference O’Neill1998: 51). Since the CI does have rational authority over us, and a state’s laws must have authority to effectively protect persons’ innate right in the circumstances of justice, the reasons that they generate must be followable.

O’Neill argues that two considerations substantiate followability, namely, a reason’s scope and the features that make it followable. This prompts two questions: ‘By whom must practical reasoning be followable?’ and ‘What does it take for practical reasoning to be followable?’ (O’Neill Reference O’Neill1998: 52). A reason’s scope allows us to distinguish the kingdom of ends, to which only moral laws apply, from the moral communities living under the circumstances of justice, to which juridical laws apply. In each case, to make the general will effective, the community must engage in inclusive practical reasoning. This kind of reasoning forces lawmakers to ‘constantly assume the intelligent cooperation and interaction of many [diverse] others’, which will ensure that others can follow their reasoning (p. 54). Accordingly, when we reason to promote followability, we cannot operate upon specifics about the persons in question; rather, we should consider persons’ general capacities, motivational sets, circumstances, and consequent vulnerabilities (p. 56). As such, for any form of practical reasoning to give morally considerable beings reasons, be they instrumental, moral, or juridical, this reasoning must satisfy the followability requirement.

From this, O’Neill concludes that, for some course of deliberation to satisfy the followability requirement for persons in a particular circumstance, the reason that it constructs must be intelligible, cognitively accessible, and practicable. In other words, reasons must represent ‘real possibilities for those who are to be offered reasons’ despite their circumstantial vulnerabilities (O’Neill Reference O’Neill1998: 57-8). Otherwise, we fail to treat them as ends in themselves. In the next section, I will argue we can use the followability requirement to develop the interplay between subjective and objective considerations such that, although a priori concepts like the law of retribution and imputation always hold, we can qualify their import by appealing to the moral and juridical features that followability makes salient. By justifying our sensitivity to such features, especially the circumstances under which agents act, we can resolve the dilemma by satisfying the two explanatory desiderata I articulated in the preceding section.

4. Interpreting Kant’s penal theory

To begin, let us recapitulate the dilemma. If we connect Kant’s moral and political philosophy in the way I suggest, Kant’s assertion that a person can possess subjective impunity conflicts with his retributive commitments in a particularly poignant way. On one horn, the right of necessity and Kant’s analysis of crimes of honour imply that punishment’s partly deterrent purpose necessitates excusing some persons from punishment due to their subjectivity. On the other horn, Kant’s retributivism claims that respecting a criminal as an end in herself requires punishing her, precluding subjective excuses and deterrence. Imputation informs these claims and the interplay between Kant’s notions of objectivity (which he equates with law) and subjectivity (which he equates with phenomenal contingency), which complicates the problem. To address it, we must satisfy two desiderata: we need to show how subjectivity interacts with culpability to explain the right of necessity and crimes of honour and why, as such, our penal theory must acknowledge excuses in these cases. Now, by reexamining imputation with the followability requirement’s resources, we can resolve this justificatory dilemma and clear the ground for mixed Kantian penal theories.

Recall Kant’s assertion that:

In the preceding discussion, we assumed subjectivity’s relevance, but the followability requirement grounds this assumption. Reconsider the shipwreck example. On Kant’s account, by killing Smith, Jones violates moral and juridical law because her circumstances force her to choose between immorality and her life. Although she should not kill Smith, given the state of mind Jones’ circumstances create, even if she knows that her action invites the death penalty, not killing Smith is not a real motivational possibility. As such, the reasons juridical laws generate are unfollowable, such that sentencing Jones to death for her wrongdoing is inappropriate. However, she remains morally culpable for violating the CI since respect for the law should have motivated her to act appropriately.

This observation is crucial since imputation, which Kant uses to substantiate responsibility, is the linchpin holding the dilemma together. Upon reflection, we should see that human beings’ general capacities and motivational sets, the latter of which include a strong inclination toward self-preservation, generate vulnerabilities that – due to the followability requirement – lawmakers must consider when evaluating wrongdoing. Given these facts, for most people, the reasons proscribing killing will be unfollowable in the cases to which the right of necessity applies. Although Jones’ action is an appropriate subject of imputation since she is responsible for it, her subjectivity makes the relevant reasons unfollowable. Just punishment, to be justifiable, must assume failure to comply with an authoritative reason; since juridical laws cannot produce followable reasons in the cases to which the right of necessity applies, they do not give people who exercise the right of necessity authoritative reasons. As such, the right of necessity makes the persons who exercise it subjectively unimpugnable. To acknowledge this reality, we must accommodate subjective excuses.

There are two problems with this conclusion. First, it might seem incompatible with Kant’s metaphysical claim that beings, insofar as they are rational and act under the idea of freedom, can always appreciate and be motivated by the CI’s authority. This authority grounds respect for the law, an incentive that takes priority over all others such that when one recognises it, one cannot help but act on it (G, 4: 403). In the circumstances of justice, we assume persons are free and rational; after all, the followability requirement would be unnecessary if we were not reasoning with moral personalities. Therefore, someone who is in a case to which the right of necessity applies, if she is sufficiently rational, can recognise and fulfil her obligation to refrain from murder. So, do her circumstances really make the law unfollowable? To answer this question, we must consider the relation between followability, authority, and action-guidingness. In doing so, we may resolve a second problem with this initial conclusion, namely, that it translates the followability requirement from the moral domain to the juridical domain without the kind of nuance that is necessary to make sense of the right of necessity.

The followability requirement implies that to engage in inclusive practical reasoning in the circumstances of justice, we must consider the general characteristics of those for whom that reasoning is authoritative. In these circumstances, persons are not deliberating with pure practical reasoning; they all desire, as ‘a natural necessity’, happiness (G, 4: 415). This desire can obscure the moral law or occasionally make it motivationally inaccessible because of the vulnerabilities it generates (4: 405). Since no external incentives can make one respect the moral law, and juridical laws in the circumstances of justice are limited to just such incentives, legislating as if we can make persons recognise an internal moral incentive represents a scope error. Inclusive practical reasoning about justice must, therefore, be legislated externally. In the cases to which the right of necessity applies, juridical laws’ external incentives fail due to the criminal’s subjective vulnerabilities. Since their incentives fail to effectively guide action, juridical laws fail to be followable; because they are not followable, they cannot purport to give people who exercise the right of necessity authoritative reasons. Since the state cannot give people who exercise the right of necessity authoritative reasons for action, it cannot punish them for violating the laws that purport to give them those reasons. Moreover, given its scope and aims, a state can punish juridical failings but not moral failings. As such, a state’s courts must conclude that people like Jones are subjectively unimpugnable.

That said, despite the subjective impunity that juridical followability justifies, because the persons who commit these wrongs still possess the motivational capacity to act upon the CI and necessarily act under it, they are objectively culpable.Footnote ²³ After all, these criminals’ objective, purely rational selves would have acted as the general will, which expresses their practical rationality in the juridical domain, obligated them to act. To see this argument from a different angle, consider Arthur Ripstein’s (Reference Ripstein2005) argument that Kant’s account of legitimate coercive threats grounds the right of necessity. As we saw in section 2, Kant assumes that for a punishment to be just, the threat of that punishment must be able to motivate. Ripstein argues that this feature of Kant’s penal theory is due to the conceptual claim that, for one to make a threat and not merely utter it, the threat must be capable of guiding the action of the entity one is threatening (Ripstein Reference Ripstein2005: 421). Ripstein is careful to note the conditions under which we can impute a criminal action’s rightful consequences: a person must have acted criminally, and the state’s laws must successfully threaten punishment for the relevant type of criminal action (Ripstein Reference Ripstein2005: 423). In the shipwreck case, although Jones acts criminally, the law fails to threaten punishment for her action because it cannot guide her action. Thus, she is excused.

However, Ripstein needs to explain why threats must guide action. On both his account and mine, to secure the freedoms ensured by persons’ innate right, a state’s laws must be motivationally efficacious. If its laws are not motivationally efficacious, then the state lacks the authority to punish. For its laws to be motivationally efficacious, a state must consider the persons to whom they apply and what it would take, in virtue of these persons’ general characteristics, for its laws’ incentives to motivate. In other words, we may justify the conceptual requirement that a threat be action-guiding – the feature on which Ripstein’s account turns – with the followability requirement. In this sense, juridical followability determines what it takes for a state’s laws to be action-guiding under certain circumstances and, as such, authoritative. By determining which laws generate authoritative reasons, the followability requirement determines who the state may punish. This discussion satisfies the first half of our first explanatory desideratum by showing how a criminal’s subjectivity can excuse her from rightful punishment and clarifying the categories of penal subjectivity and objectivity.Footnote ²⁴

We can satisfy the second half of this desideratum by identifying the distinction between followable reasons in punishable crimes and unfollowable reasons in unpunishable crimes. When Jones kills Smith, she displays an incapacity to recognise countervailing reasons, making them unfollowable. If a person commits tax fraud or batters another person, could she argue that she – like Jones – is subjectively unimpugnable because she could not recognise an authoritative, followable reason to abstain? I suggest that the circumstances under which these crimes occur are different in kind from crimes of honour and the cases to which the right of necessity applies. In a minimally functional state, citizens will not think tax fraud, battery, and other crimes are necessary to protect their innate right; as such, they cannot appeal to followability.Footnote ²⁵ Moreover, because these criminals are involved in the inclusive practical reasoning that generates authoritative laws, they cannot claim the reasons for them to follow the law do not obtain. Rather, such criminals unjustifiably exempt themselves from these reasons’ authority, making them punishably culpable. With that, we satisfy our first desideratum.

To satisfy our second desideratum, we must explain why and how the interplay between objectivity and subjectivity is salient in Kant’s penal theory. In other words, we need to explain why a state must accept the preceding account of juridical followability. To do so, let us elaborate on the followability requirement by answering its questions about the scope and features of the reasons with which we are concerned. To answer the first question, we may stipulate that the penal system’s scope is its state’s citizens. We can thus avoid some complications and clarify who is subject to the reasons in question. To answer the requirement’s second question, regarding what it takes for a law to be followable in the circumstances of justice, we may begin with the ideal. In the kingdom of ends, there would be no state because the circumstances of justice do not necessitate it; juridical laws would not be necessary since everyone does their duty and harmonises their ends.Footnote ²⁶ Even if juridical laws were necessary, coercion would be unnecessary to enforce them. After all, Kant argues that justice justifies coercion as a hindrance to hindrances of freedom that are inconsistent with universal law; since the CI definitionally forbids such hindrances of freedom, they will only occur in nonideal circumstances of justice. As we have seen, these circumstances necessitate a state to create external laws that protect persons’ innate right. As such, a state’s laws must account for its society’s particular nonideal circumstances so that the reasons they generate are followable. Therefore, laws must be sensitive to contingency and grounded in the general will.

Let us consider the general will. In the circumstances of justice, a state’s laws are followable for two primary reasons. The first is the state’s ability to create a civil condition by using coercion to solve the problems that the circumstances of justice create. The second is the state’s foundation in practical reason, which Holtman (Reference Holtman, Pinheiro Walla and Ruhi Demiray2020) and Herman (Reference Herman2021) emphasise by predicating a richer conception of morality on the state and the general will that grounds its authority. As we saw when discussing Ripstein’s account, for laws to have authority, they must be motivationally efficacious; if they are not efficacious, then the state’s laws are not followable. If the state’s laws were unfollowable, it would not be a real solution to the problems that the circumstances of justice create, undermining its authority. As such, what makes reasons of justice followable is one’s rational capacity to recognise the law’s authority and the state-sanctioned coercion necessary to make justice practicable. The general will is necessary to realise these conditions since, otherwise, inclusive practical reason could not ensure the laws’ reasons are followable or justify their coercive force to the citizens they represent. To use this kind of reasoning so that the state can effectively protect citizens’ innate right, we must acknowledge the followability requirement, the satisfaction of which necessitates sensitivity to contingency, including – inter alia – persons’ capacities, circumstances, and values.

As I noted in the preceding section, to make its laws accessible, a society needs a conception of rightful honour whose content is specified to its particular circumstances. While there is an objective standard the state and citizens ought to approximate, there are several ways they can meet or aspire toward these standards that inclusive practical reasoning will substantiate. In this sense, a society’s subjective conception of honour constitutes one of the contingencies to which a state must be sensitive. As with anything in a nonideal context, though, we should expect that a society’s conception of honour will be imperfect. Nonetheless, it will be motivationally efficacious. This efficacy, conjoined with a subjective conception of honour’s imperfections, can create social pressures that make some laws, if not unfollowable, followable only at a potentially devastating cost to one’s social standing and inclusion in the local sociopolitical community. Thus, because of her community’s subjective conception of honour, the criminal of honour thinks – or is forced to think – that her crime protects her innate right, making the criminal’s actual duties unfollowable. As such, while the criminal of honour is objectively culpable for her wrongdoing, we should excuse her from the death penalty for the same reasons we excuse someone who exercises the right of necessity.

In sum, although a subjective conception of honour is necessary to make the universal principle of right followable and further substantiate innate right, the circumstances of justice that make it necessary generate problems. Given the circumstances in which they occur, in crimes of honour and the cases to which the right of necessity applies, juridical reasons become effectively unfollowable. Although the criminals are objectively culpable, due to the followability requirement, treating them as ends in themselves necessitates considering their subjective vulnerabilities and excusing them from punishment on that basis. Since a coercive state is necessary to protect persons’ innate right and may do so through deterrent laws, the general will that justifies these laws must recognise that when deterrent threats become unfollowable, the state cannot carry out the punishment they threaten. As such, the followability requirement grounds excuses for criminals of honour and necessity. Given these observations, we can defend Holtman’s four criteria – and other mixed Kantian penal theories – by arguing that respecting persons necessitates acknowledging their vulnerabilities and self-authorship, such that both penal retribution and deterrence are necessary to make our circumstances habitable.

By answering the followability requirement’s second question in this way, we satisfy the second desideratum by demonstrating why the interplay between objective and subjective considerations is salient.Footnote ²⁷ In so doing, we resolve the dilemma and make coherent the disparate strains of thought articulated in section 2. While we are morally culpable for wrongdoing, there are subjective factors the state must consider to effectively and justly protect persons’ innate right with deterrent punishment. The followability requirement determines when a criminal’s subjectivity is relevant, such that punishing someone when they are subjectively unimpugnable defeats part of the punishment’s purpose and wrongs the criminal. Since criminals’ vulnerabilities sometimes make juridical reasons unfollowable, excusing them from punishment despite their wrongdoing respects them as rational beings and citizens. Even if we excuse some criminals because the threat of some punishment was unfollowable, we can retain Kant’s respect-based arguments and, thus, his retributive commitments to guilt and proportionality. By using the followability requirement to make Kant’s justifications for penal retribution and deterrence compatible, my argument clears the ground for mixed Kantian penal theorists like Holtman. With these additional resources, they may further develop the complex, compelling, and difficult application of both deterrent and retributive criteria to our penal institutions.