The question of whether or not Kant’s Philosophy of RightFootnote ¹ can be understood as part of his critical moralFootnote ² philosophy has always been disputed. One of the main arguments in favour of the independence of Right from morality relates to the coercive power that is, according to Kant, conceptually associated with Right: ‘Right and authorization to use coercion […] mean one and the same thing.’Footnote ³ Accordingly, proponents of the independence thesis hold that if Right and the power to coerce were analytically linked, then Right as the epitome of heteronomy (i.e. external lawgiving) could not possibly appeal to moral autonomy (or the categorical imperative) as its grounding principle.Footnote ⁴ However, does the authority to coerce really call into question the dependence of Right on Kant’s moral philosophy? In this chapter, I will propose the alternative, opposite view that – on Kantian grounds – coercion as a normative problem only becomes explicable against Kant’s critical moral philosophy. For it is the moral personality of the coerced that calls for a normative justification of coercion in the first place (1). That said, coercion that does not violate one’s moral personality becomes morally irrelevant since moral personality is conceptionally restricted. Thus, the very embedding of Right in Kant’s critical moral philosophy provides a two-way solution: an explanation for why coercion is a normative problem in the first place; and a justification as to what extent coercion is legitimate. However, proponents of the independence thesis aim to explain coercion as an analytical implication of the notion of equal, relational freedom and thus miss this crucial dual dependence on Kant’s critical moral philosophy. Consequently, they are neither able to offer a Kantian justification for the normative bindingness nor for the enforceability of Right. (2). Conversely, tracing coercion back to the limits of moral personality does not only explain why coercive force is grounded in moral personality (and autonomy as its constitutive principle). Even more importantly, it requires us to reconsider whether Kant is able to argue consistently against the external enforceability of internal perfect duties (e.g. the prohibition of suicide) (3).

5.1 Coercion as a Moral Problem and Unlawful Treatment as a Mere Means

Why is somebody legally entitled to coerce someone else? – Before explaining why this is fundamentally the wrong question, I would first like to clarify what I mean when I use the term coercion with reference to Kant. From Kant’s perspective – echoed by contemporary legal theory – coercion can mean two things: first, physical coercion in the sense of vis absoluta (i.e. the factual restriction of action); and, second, psychological coercion in the sense of vis compulsiva (i.e. the pathological motivation by conflicting sensual inclinations, e.g. fear of punishment). When Kant analytically derives the authority to coerce from the concepts of right and wrong in § D of the Introduction into the Doctrine of Right, he speaks of coercion in the former sense:

Since Kant is concerned with factual resistance to a wrongful act here, coercion under § D can be taken to mean only physical coercion in the sense of vis absoluta. This is to be distinguished from psychological coercion in the sense of vis compulsiva, which consists in the prospect of physical coercion in response to potential wrongdoing. In connection with a universal law, such psychological coercion can even be considered the epitome of external, juridical lawgiving as referred to by Kant in the Introduction into the Metaphysics of Morals.Footnote ⁶ To compel someone to do something in this latter sense means nothing other than to act as an external, juridical legislator – that is, as someone who declares the duty to the obligor and links it to a pathological incentive.Footnote ⁷ This chapter will focus on coercion in the sense of vis absoluta and disregard vis compulsiva.

My main argument is that the coercive power of Right is less a problem than evidence for a derivational reading of Kant’s legal philosophy. This is due to the fact that the authority to coerce does not need to be derived from the categorical imperative at all if one understands Right as an integral part of Kant’s critical moral philosophy. While on the surface this might sound surprising, it becomes clear when we realize that Kant’s actual issue is to demonstrate the inadmissibility of coercion. The central question for Kant is therefore not Why and when may I coerce someone?, but rather Why and when may I not coerce someone?.

Coercion (in the sense of using physical force) only becomes a normative problem when someone can legitimately claim not to be coerced. Therefore, use of force or coercion in any form towards stones, plants, or animalsFootnote ⁸ is always normatively permissible. This is the case because, according to Kant, stones, plants, and animals lack moral personality, that is, they are not suitable subjects of moral (i.e. legal and ethical) entitlements and duties.Footnote ⁹ Coercion does not become a normative problem until a person is involved.Footnote ¹⁰ This is already implied in Kant’s definition of the ‘concept of right, insofar as it is related to an obligation corresponding to it (i.e., the moral concept of right)’ in § B of the ‘Introduction to the Doctrine of Right’,Footnote ¹¹ according to which Right concerns the intersubjective relationship of persons. Consequently, the question of the permissibility of legal coercion, which is raised in §§ C–E, arises only for them. Kant mapped this out clearly as early as 1784 in the Feyerabend lecture on natural law:

Unlike things, persons enjoy the status of being subjects of rights and duties and as such cannot be coerced to do something or used by others without further justification.Footnote ¹³ It follows that the plurality of persons generates the problem of legitimate coercion in the first place. Yet why should it be that a person has this special standing not to be coerced? As I will outline in this chapter, Kant sees the reason for this in the fact that a person is an autonomous being and as such an end in herself or himself. Or to put it differently: without presupposing Kant’s critical moral philosophy and the autonomy theorem on which it is based, coercion – at least for KantFootnote ¹⁴ – ceases to be a normative problem.Footnote ¹⁵ This raises two questions: what is the justificatory relationship between the doctrine of Right and Kant’s critical moral philosophy? And how can legal coercion be explained with reference to this?

Let me start with the first question:Footnote ¹⁶ If we interpret Kant’s legal philosophy against his critical moral philosophy, then someone only has rights because and to the extent that she is an end in herself. A violation of rights is nothing more than a treatment of that person as a mere means; and, although often incorrectly denied, any treatment of another person as a mere means constitutes a breach of a legal duty.Footnote ¹⁷ A legally valid contract, for example, requires that the other party not be treated merely as a means, that is, that his contractual interests are met. Likewise, to infringe upon another’s property for one’s own enrichment without consent means to treat this person as a mere means. This justificatory nexus can be identified for the first time in Kant’s Feyerabend lecture on natural law.Footnote ¹⁸ It also finds its way into the Grundlegung with Kant’s analysis of perfect duties against others. In the case of fraud (‘a false promise’) and ‘assaults on the freedom and property of others […] he who transgresses the rights of human beings intends to make use of the person of others merely as means’.Footnote ¹⁹ Kant’s examples show that legal duties – for here, without exception, we are dealing with perfect external duties – essentially consist in not using others merely as a means. He made this explicit in the Vorarbeiten on the Doctrine of Right:

External freedom as Kant describes it here is obviously nothing other than the innate right to freedom presented in the Doctrine of Right: ‘freedom’ as the ‘independence from being constrained by another’s choice’.Footnote ²¹ This is made even clearer in the Vorarbeiten for Theory and Practice, according to which ‘freedom as a human being according to the innate right [consists in] not being subject to the choice of others merely as a means’.Footnote ²²

This justificatory nexus of violating one’s right by treating a person merely as a means highlights an important clarification of the concept of freedom in Kant’s Doctrine of Right. External freedom is the ‘object’ protected by Right. It represents what the innate right consists in and what the legal duties corresponding to that innate right refer to. However, the legal validity, that is, the bindingness of legal obligations protecting this very freedom, depends on the obligee and the obligor both being ends in themselves and thus follows from their shared moral autonomy. It follows that external freedom is not a specifically human capacity (e.g. physical freedom of movement, or psychological freedom as in the ability to determine oneself according to one’s purposes). Instead, external freedom is nothing other than ‘freedom in the external use of choice […] insofar as it [sc. choice] is determined by laws of reason’.Footnote ²³ External freedom refers to human choice in external actions (i.e. the faculty to perform external actions as one pleases) insofar as one’s choice is subject to autonomous laws of reason. In other words, the choice or the exercise of one’s choice is morally indifferent if the choice is not subject to a moral law and is therefore not regulated in a certain way (e.g. restricted to a specific scope of action). Consequently, human choice is only free choice in that it can be determined by laws of reason, and it is precisely this determinability by the laws of reason in which moral autonomy finds its expression.

In consequence, external freedom as the ‘object’ protected by law (i.e. independence from another’s necessitating choice) is the normative demand that the laws of Right specify for the external use of choice because human beings are morally autonomous and thus necessarily ends in themselves.Footnote ²⁴ A person exists ‘as an end in itself (which is the supreme limiting condition of the freedom of action of every human being)’,Footnote ²⁵ and ‘hence so far limits all choice’Footnote ²⁶. This is why, for Kant, coercion by others is prima facie normatively problematic, as by coercing someone the coerced would ‘serve merely as a means to any end of another’.Footnote ²⁷ Thus, legal subjects being morally autonomous and therefore ends in themselves is not only crucial for the validity of Right, but also gives rise to legal coercion as a normative problem in the first place.Footnote ²⁸

Before turning to legal coercion in more detail, let me address – albeit briefly – three possible objections to this reading of Kant: first, one might object that Kant had abandoned this justificatory nexus in the Doctrine of Right since he does not explicitly refer to moral autonomy or the end-in-itself-formula of the categorical imperative in the Introduction to the Doctrine of Right. However, this objection overlooks the fact that stipulating legal subjects to be ends in themselves is implicit in Kant’s notion of moral personality, which is – as shown above – central to the Doctrine of Right. This is most evident in Kant’s treatment of criminal law when he states that ‘a human being can never be treated merely as a means to the purposes of another or be put among the objects of rights to things: his innate personality protects him from this’.Footnote ²⁹ A person is not a thing and therefore cannot be treated arbitrarily. This would disregard his or her moral ‘personality, by which alone they are ends in themselves’.Footnote ³⁰ Consequently, as Right refers to the ‘practical relation of one person to another’,Footnote ³¹ Kant implicitly traces legal freedom back to the legal subjects being ends in themselves and being morally autonomous.Footnote ³² This is the case as it follows from ‘[m]oral personality’ ‘that a person is subject to no other laws than those he gives to himself (either alone or at least along with others)’.Footnote ³³

A second objection might be rooted in the view that the moral value of being an end in itself is in fact normatively too ambitious and thus an ill-suited foundation for Right. This would be the case if one assumes the end-in-itself-formula of the categorical imperative ultimately to require us to treat everyone else as an end. However, Right is obviously not about promoting ends in themselves in this positive way. Instead, Right can be deduced analytically from the concept of freedom of choice in its external use. Therefore, rightful behaviour is possible without taking into account the special interests and wishes of others (as external duties of virtuedemand),Footnote ³⁴ which is a prerequisite for treating them as an end. ‘If I do not contribute to another’s happiness, I do not infringe upon his freedom, but let him do what he wants.’Footnote ³⁵ Thus, Right is only about the moral value of being an end in itself in the negative sense of not treating others as a mere means.

A third objection to the justificatory nexus outlined above could be that Right according to Kant abstracts from the ends the legal subjects are pursuing and thus cannot be derived from the categorical imperative (in whatever formula). While it would go beyond the scope of this article to discuss the complex question of the derivability of the general law of Right from the categorical imperative,Footnote ³⁶ we can nevertheless point out that Right restricts legally legitimate actions to the extent that at least one permissible maxim can be found for them. This is the case despite the fact that Right does not specifically prescribe any maxims, this being the function of the doctrine of virtue. Thus, if at least one permissible maxim can be found for a rightful action, then that action does not amount to treating other legal subjects merely as means.Footnote ³⁷ Irrespective of the disputed derivability of Right from the categorical imperative, Right hence reflects the fact that legal subjects are ends in themselves.

5.2 Coercing without Treating merely as a Means

How does all this help us to understand the legitimacy of legal coercion? As I have tried to show, Right protects our independence from someone else’s necessitating choice, precisely because as persons we are ends in ourselves and may not be used merely as a means. And since applying coercive force is the paradigmatic case of using someone merely as a means, any form of coercion is prima facie impermissible. Right therefore generally prohibits any form of coercion, since it contradicts the claim of autonomous rational beings to be respected in their moral personality and not to be treated as a mere means.Footnote ³⁸ Moral personality, however, does not amount to inviolability and does not render any coercion illegitimate. This is the case because the moral laws which govern personality, and which are an expression of persons being ends in themselves, also restrict moral personality. When it comes to the use of choice in external relations, these restrictions necessarily follow from the plurality of autonomous rational beings. Precisely because human beings form a community with other persons, their own use of choice is subject to the normative stipulation that it must be reconciled with the others being ends in themselves. The minimum conditions for this – that is, the necessary conditions under which beings that are ends in themselves can coexist according to a universal law – are formulated by Right. It guarantees each legal subject a use of choice according to a universal law. At the same time, Right restricts each legal subject’s external freedom to this very use. Therefore, any use of choice beyond this restriction can no longer be understood as an expression of the moral personality of the legal subject. The legitimacy of legal coercion is just the flipside of this, as Kant nicely put it in the Vorarbeiten to the Metaphysics of Morals:

To the extent that Right restricts the use of choice according to a universal law, it only reflects that moral personality – and thus the scope or sphere in which persons are in fact ends in themselves that must not be treated merely as a means – is conceptually restricted. In § C of the Introduction of the Doctrine of Right, Kant puts this in a nutshell by saying that ‘freedom is limited to those conditions [sc. the conditions according to the universal law of right] in conformity with the idea of it [sc. freedom] and that it may also be [in fact] actively limited by others’.Footnote ⁴⁰ In other words, anyone who commits a violation of rights is in this respect (i.e. insofar as he violates the rights of others) no longer a moral person worthy of protection from interference. Consequently, his or her moral personality, which is to be legally protected, is not affected by opposing coercive acts. As a corollary, the need to normatively justify the legitimacy of coercion never does arise in the first place.Footnote ⁴¹ Insofar as the concept of Right can be analytically derived from the concept of free external choice of moral persons according to universal laws, the permissibility of external coercion corresponds directly to this:

Just to reiterate: coercion or coercive force against non-persons (stones, plants, animals) is morally indifferent in relation to them;Footnote ⁴³ this is the case because moral personality calls for a normative justification of coercion in the first place. However, moral personality is not unlimited but conceptually restricted due to the plurality of persons interacting with each other. It follows that this restriction marks the limit of morally permissible or impermissible coercion. For Kant, Right and coercive power both follow from the fact that moral personality – and thus the sphere within which interference with others amounts to treating them merely as a means – is conceptually restricted. This could be called a scalable ‘sphere of protection’ that comes with being a moral person and that gives rise to legal duties which protect one’s moral personality. Beyond this sphere of protection, coercion does not violate the ‘moral freedom [of the other], for to that extent I do not derogate his freedom’.Footnote ⁴⁴ This is how Kant can analytically equate Right and coercion in § E of the Introduction to the Doctrine of Right.Footnote ⁴⁵ Thus, there is no need for a positive derivation of ‘a legal entitlement to coerce’ from the categorical imperative. On the contrary, the permissibility of coercion or use of coercive force is the default case since coercion is morally irrelevant if it does not violate one’s moral personality. According to this view, coercion is always permissible as long as the legally protected freedom of the other – namely his or her moral personality in legal terms – is not impaired.

However, one might think that the proponents of the independence thesis would come to similar conclusions, since they too hold the view that coercive power is nothing other than the flip side of relational freedom: ‘[T]he entitlement to coerce is simply the entitlement that others exercise their freedom consistent with your own.’Footnote ⁴⁶ This reading of Kant, however, presupposes a concept of freedom as independence from being constrained by the choice of another person, without being able to explain the categorically binding nature of this idea. By invoking the independence of law from morality, proponents of the independence thesis cannot refer to the aforementioned notion of autonomy and persons being ends in themselves. They are left with having to resort to prudential reasons. It would go beyond the scope of this chapter to go into detail on this and outline the (in my view) shortcomings of the various independentist readings of Kant’s legal philosophy and the justifications of legal coercion that go hand in hand with them.Footnote ⁴⁷ However, the most common line of reasoning seems to be as follows: independentists understand freedom of choice pursuant to Kant as the ability to set and pursue one’s own ends. This ability presupposes only practical freedom,Footnote ⁴⁸ but not transcendental freedom. This is why Kant’s doctrine of Right, which reconciles and guarantees freedom of choice according to a universal law, is said to be independent of Kant’s critical moral philosophy.Footnote ⁴⁹ Since every person naturally wants to realize his or her own ends, no person can ‘reasonably’ reject legal restrictions of freedom because Right defines the conditions that enable us to realize our own ends in accordance with everyone else.Footnote ⁵⁰ As a corollary, legal coercive power just means exercising the freedom to which everyone is entitled who wants to freely realize his or her purposes.

However, on Kantian grounds this independentist argument is flawed in two ways: first conceptually, since it conceives of external freedom as some capacity, whereas it is actually – as shown above – the ‘object’ protected by Right.Footnote ⁵¹ Second, it is normatively flawed since the categorical bindingness of legal standards and restrictions on the legitimate use of coercive force cannot be explained in this way. This is the case because according to the independentist reading, compliance with legal standards is (if authors do not resort to postulating quasi-moral principles governing Right)Footnote ⁵² only hypothetically required, namely insofar as it is conducive to one’s own interest in self-preservation. This is best illustrated by the example of the nation of devils which Kant refers to in Towards Perpetual Peace and which is, ironically, often cited by independence theorists in an attempt to defend their position.Footnote ⁵³ The devilsFootnote ⁵⁴ are exactly the kind of beings that possess the qualities that are said to be necessary to be a suitable subject of rights and legal obligations: they are rational in the sense that they can set and pursue their own ends. However, they lack moral personality because they are not transcendentally free and thus not autonomous.Footnote ⁵⁵ Such beings would certainly be able to act in accordance with the prescripts of Right. However, even if legality is achievable for a nation of devils, devils would never have a concept of legal validity or legal obligation:

Devils can indeed be subjected to an external coercive mechanism that – despite egoistic self-interests – guarantees prudentially lawful behaviour. However, this is at best a legal order in the technical sense. It does not suffice for the moral concept of Right that Kant has in mind, in which Right corresponds to a moral obligation. This is what Kant himself says when he talks about the natural guarantee safeguarding perpetual peace. For, according to Kant, the natural guarantee merely answers the following question:

Legality within a nation of devils merely proves that the prudent devils will opt for a legal order out of cleverly calculated self-interest. This is how ‘nature comes to the aid of the general will grounded in reason […] precisely through those self-seeking inclinations’.Footnote ⁵⁸ Devils, however, have no concept of moral obligation because they are not persons. This lack of moral personality not only puts an end to any form of categorical moral obligation (be it ethical or legal), but also to the normative problem of justifying coercion.Footnote ⁵⁹ A devil is not a person and therefore needs to be considered as a thing in moral terms. To force him to do something, to injure him, or even to kill him are simply normatively irrelevant acts. Or to put it differently: any independentist reading of Kant that seeks to justify the binding nature of Right independently of Kant’s critical moral philosophy fails even to explain why coercion is a normative problem in the first place.Footnote ⁶⁰ It may be unwise for a devil to coerce others unilaterally and without limits. However, this is neither categorically forbidden nor in need of normative justification.

5.3 With Kant beyond Kant, or: Are Internal Perfect Duties Externally Enforceable?

As we have seen, the permissibility of legal coercion is just the flipside of the moral personality being conceptually restricted: if you try to hit me and I violently push your arm away, I am not treating you merely as a means since your wrongful action is not an expression of your moral personality that I must respect. According to Kant, however, Right is restricted to intersubjective relations among persons, ‘insofar as […] actions […] can have (direct or indirect) influence on each other’.Footnote ⁶¹ Consequently, coercive power also finds a corresponding limitation and depends on the injury of another person in this intersubjective relationship. Thus, Kant claims that only external legal obligations (i.e. legal obligations towards others) are coercively enforceable. But is this claim actually warranted?

If coercion is legitimate if and only if it does not amount to treating others merely as a means, then this seems to endorse more coercive action than Kant claims. For if coercive power ultimately results from the fact that moral personality is conceptually limited in a way that renders all coercive acts permissible that do not affect the moral personality of the coerced, then this also seems to hold true for the realm of perfect duties against oneself: the legal internal duty of honeste vive and the internal perfect duties presented in the Doctrine of Virtue.Footnote ⁶² These duties do not only reflect a conceptual restriction of our moral personality in a specifically legal sense, namely the ‘right of humanity in our own person’.Footnote ⁶³ They, too, are in principle externally enforceable insofar as they prohibit external actions.Footnote ⁶⁴ This holds for internal legal duties (not to prostitute oneself, not to sell oneself into slavery, etc.) and the perfect duties against oneself in the Doctrine of Virtue (prohibition of suicide, prohibition of harming oneself, etc.).Footnote ⁶⁵ If legitimate coercion represents the flipside of our moral personality being conceptually restricted, and if internal perfect duties reflect such a conceptual restriction, then these duties should also be enforceable.

Kant himself opposes a conceptual equation of perfect duties and enforceable duties, as it has been implemented in contemporary natural law philosophy. Rather, he sees a requirement to substantiate the authority to coerce based on the concept of law. In the lecture on Metaphysics of Morals Vigilantius, Kant complains that until his time it had been

If we take Kant at his word, any restriction of the authority to coerce would depend on whether such a restriction can be derived from the idea of Right that ‘the choice of one can be united with the choice of another in accordance with a universal law of freedom’.Footnote ⁶⁷ In the case of perfect duties towards oneself, however, such a restriction seems questionable for two reasons.

First, as shown above, Right and coercive power according to Kant both follow from the fact that by virtue of being persons, our choice in external actions is subject to moral laws. By designating certain acts as impermissible, Right also describes morally permissible coercive acts, for those who commit a legal wrong are not affected in their moral personality by opposing coercive acts. However, this line of reasoning applies quite generally to perfect duties, be they external or internal: self-harming actions that violate perfect duties against oneself are just as little an expression of one’s own moral freedom in the external use of choice as actions that harm others and violate external legal duties. In either case, the moral personality is unaffected by the opposing coercive actions. The ‘right of humanity in one’s own person’, which is the basis of all perfect duties against oneself,Footnote ⁶⁸ sets the morally permissible external use of choice apart from the morally impermissible external use of choice. By recognizing perfect duties against oneself, Kant thus identifies certain forms of behaviour or actions as not morally worthy of protection and thus as potentially subject to coercion – regardless of whether someone else is harmed by them or not.

Second, restricting legitimate coercion to the fulfilment of external duties (and exempting perfect internal duties) would be self-contradictory. For if the right of humanity in our own person prohibits certain behaviour, it cannot at the same time designate it as legally worthy of protection vis-à-vis other persons. If it did, then this would mean that I would be allowed to commit an act that is at the same time forbidden. This can be illustrated by the example of the prohibition of suicide: if I prevent a suicidal person from killing himself by coercion (e.g. by grabbing his arm), then he would not have the right to resist my intervention. He would only have this right if he could claim that he had been wronged by my intervention. However, the suicidal person neither has a right nor any other moral entitlement to kill himself; otherwise we could not say that he has a perfect duty not to commit suicide. It follows that my use of coercive force in this scenario is not wrong. This follows trivially from the principle of contradiction to which Kant refers in § C of the Introduction into the Doctrine of Right: a legal wrong is committed by me only where the other person has a legal right. Tertium non datur.

One can go even further in this respect: according to Kant’s concept of Right, I even have a right to prevent the other person from committing suicide, since my choice is compatible with the choice of the suicidal person according to a universal law of freedom. This is the case because this very law of freedom prohibits suicide. However, it would be wrong to justify my coercive power by saying that the suicidal person legally has an obligation towards me to preserve his moral personality. He certainly does not. Yet the suicidal person does have an obligation towards me to tolerate the intervention. For my conduct is in complete harmony with the Universal Principle of RightFootnote ⁶⁹ and is in this respect an expression of my innate right to freedom.Footnote ⁷⁰ This seemingly odd conclusion becomes clear by focusing on the debitum and distinguishing what is owed to whom: in the case of perfect duties against oneself, there is no external legislation because the obligor owes the fulfilment of the duty only to himself. Thus, in the example previously outlined, the suicidal person owes the preservation of his life to himself, but not to me. Consequently, I have no moral authority to oblige him to stay alive under the threat of coercion, that is, to bind him by means of external legislation. Or to put it differently: with regard to the debitum ‘preservation of his life’, I have no moral standing and therefore no entitlement to enforce this duty. However, since the prohibition of suicide is a perfect duty, it simultaneously defines the scope of my innate right and thus also of the corresponding duty of neminem laede. With regard to the latter debitum ‘respect for my innate right’, the suicidal person does indeed owe it to me to tolerate my physical intervention. The reason for this obligation does not lie in the perfect duty against oneself (prohibition of suicide), but in the fact that resistance to my intervention would violate my innate right.

Seen in this way, the morally owed behaviour (debitum) and the normative reason for morally legitimate coercive actions (physical coercion) diverge when viewed against perfect duties to oneself. Kant certainly did not aim to argue for a prohibition of suicide that could be enforced by others. But if for Kant the power to coerce begins where moral personality ends, then it cannot be limited to external legal obligations. Instead, use of coercive force in accordance with internal perfect duties is also legitimate, even though – strictly speaking – it is not the internal duty that is being enforced, but the innate right of the coercer.

6.2 What Is Kantian Legislation?

In the general introduction to the Metaphysics of Morals, Kant tells us that legislation, whether ethical or juridical, consists of both an objective and a subjective element. It requires, first, a law that objectively ‘represents an action that is to be done as necessary’ and, secondly, an incentive that subjectively ‘connects a determining ground for the faculty of choice […] with the representation of the law’ (RL 6:218.13–17). Contemporary moral philosophy tends to neglect the second, subjective part. But it is important to realize that legislation would be incomplete without a motivational element. An objectively valid law would not be a law for us – it would lack authority – if there were no connection to our faculty of choice or Willkür. That law would remain ‘merely theoretical’ (RL 6:218.20), that is, the state would not be able to hold us to it because we would be unable to act on it. So, how do juridical and ethical legislation differ?

As to the first element, this is a highly contentious question. The minutiae of the debate lie beyond the scope of this chapter, but it is worth noting that at least some juridical laws will be positive laws not grounded in a priori reasoning. Trivially, the categorical imperative does not decide the question whether we should drive on the left side of the road or the right. That is a matter of convention. However, once a state has settled the question one way or the other, the law is binding for anyone driving a vehicle on its territory. The Highway Code can be coercively enforced. Other legal obligations can, perhaps, directly or indirectly be derived from the principle of morals. But, again, the details need not concern us.

As to the second element, it is uncontroversial that there the perspectives of law and ethics diverge sharply. This is how the issue is introduced in the general introduction to the Metaphysics of Morals:

There is a designated ethical incentive (Triebfeder), variously called ‘respect’ (or ‘reverence’) for the moral law (Achtung), ‘moral interest’, or ‘moral feeling’. That much is clear from the Groundwork and the Critique of Practical Reason. An action that coincides with what the moral law says (an action that is, in that thin sense, right) has moral worth (it is morally valuable or good) only if it proceeds from the appropriate inner attitude, from the right maxim. From the viewpoint of ethics, motives other than respect are illegitimate. When the moral law speaks, we must obey it; we must do what it says because we appreciate its unconditional authority. Juridical legislation, however, is independent of any particular incentive as long as there is an incentive that enables agents to do their duty.Footnote ² The details of this will become clearer as we proceed. For now, let us note that while the juridical sphere does not privilege the duty motive, it does not exclude that motive either.Footnote ³ In fact, there is a sense in which it is our ethical duty to obey the laws of the state.Footnote ⁴ Kant says that, though there are many ‘directly ethical duties’, internal legislation makes ‘the remaining duties’ – that is to say, juridical duties – ‘one and all indirectly ethical’ (RL 6:221.1–3).Footnote ⁵ Note that the conception of ethics now in play is not restricted to duties of virtue but rather encompasses all categorical imperatives. It coincides with the discipline Kant normally calls ‘moral philosophy’.

6.3 Why is External or Coercive Legislation Needed?

In a juridical state, external legislation relies on coercive institutions. It arises as human beings exit the state of nature. The state of nature, Kant tells us, is a ‘a state devoid of justice’ (RL 6:312.24–5), if not entirely devoid of rights. It is possible to acquire property in that state, but any such acquisition can only be ‘provisional as long as it does not yet have the sanction of public law, since it is not determined by public (distributive) justice and secured by an authority exercising this right’ (RL 6:312.30–3).

Kant emphasizes the importance of provisional rights. They generate the exeundum, that is, it is by virtue of these rights that we are forced to enter the juridical state, which turns provisional right into actual or ‘conclusive’ right. Accordingly, the juridical state – commonly equated with the ‘state’ simpliciter – is characterized as ‘that relation of human beings among one another that contains the conditions under which alone everyone is able to enjoy [theilhaftig werden] his right […]’ (RL 6:305.34–306.1).

Note, however, that the notion of a ‘provisional right’ is problematic. On the one hand, Kant wants to break away from the Hobbesian paradigm. There are, he tells us, actual, practically relevant rights in the state of nature. On the other hand, these rights are not conclusive because they are not secured by coercive means and thus, in a sense, not really rights at all. Only rights that can be claimed are rights in the proper sense of the word. As such, only conclusive rights are rights; and conclusive rights exist only in the juridical state.Footnote ⁶

6.4 How Does the State Secure the Rights of Its Citizens?

Let us turn to the coercive measures put in place by state institutions to secure the rights of individuals. Once again, the right to property can serve as an example. The state must, among other measures, promulgate and enforce laws to the effect that those who steal other people’s property will not only not be allowed to keep their spoils.Footnote ⁷ Stolen goods must be returned to the original owner. If restitution turns out to be impossible, the owner deserves compensation.Footnote ⁸ In addition, they will be punished, that is, they will suffer an ill that is proportional to the crime. These measures are intended as a clear signal that it is in no one’s interest to break the law. Crime will be severely discouraged.

To make external legislation work, Kant has to assume that human beings take a natural and unavoidable interest in their own well-being.Footnote ⁹ As this interest is general – it aims at whatever gives us pleasure – it needs to be made determinate. Initially, we do not have a very good sense of what is in our interest. Prudential deliberation helps us discover what promises satisfaction long-term. Though this is often difficult to achieve, the threat of punishment makes this task a little easier by impressing upon us, as citizens of the state, that breaking the law will not be to our advantage.Footnote ¹⁰

6.5 Elements of Kant’s Theory of Punishment

It is apparent from what has been said so far that Kant’s theory of punishment is not purely ‘retributive’. As Sharon Byrd has argued, three elements need to be distinguished with care.Footnote ¹¹ There is, first of all, the general threat contained in the law as promulgated by the state. The purpose of this threat is deterrence. The second element is the execution of this threat, namely the act of punishing an individual, which is legitimate only if the law was in force (promulgated, publicized) at the time it was broken. Retrospective legislation would fall foul of the rule of law. Thirdly, every law has to specify a sentence, which for Kant must be informed by the principle of retribution (ius talionis). Everyone who has broken the law is thus meant to receive the same fixed and fair punishment.Footnote ¹²

From the perspective of juridical legislation, the particular motivation of citizens is not only irrelevant when they comply with the law; it is also irrelevant when they – consciously, wilfullyFootnote ¹³ – break it. The state does not punish criminals because they acted on this or that morally objectionable motive, for example out of hatred, jealousy, or greed. Criminals are punished because they have broken the law, that is, because they have committed a crime. And there was no good reason, no justification, to do so. As long as certain conditions of rational agency (such as conscious deliberation or premeditation) are met, what matters is that unconditional law was violated. From the point of view of the law, motivation – which can never be ascertained with certainty – cannot make an action worse or better, let alone excuse it.

6.6 Why Do Criminals Break the Law?

Still, punishment must be proportionate to the crime. The state is not allowed to impose draconian sentences even if threatening such measures would prevent transgressions more effectively.Footnote ¹⁴ Moreover, even the best of states cannot ensure that all crimes are duly punished. Anyone inclined to break the law will be acutely aware of this. As a consequence, there is only so much a state can legitimately do to discourage crime.Footnote ¹⁵ Deterrence can never be perfect – or else, no law would ever be broken, no right would ever be infringed and, a fortiori, no citizen would ever be punished.Footnote ¹⁶ The crucial question, then, is this: what goes on in the mind of citizens when they decide to commit a crime?

On Kant’s view, human beings do not break the law because they want to defy it; such a desire would be diabolical (see RGV 6:37.21). Immoral action in general, and illegal action in particular, are governed by the principle of happiness (see KpV 5:22). Human beings break the law because they expect to profit from the illicit deed, the state’s best efforts notwithstanding. And this is possible only if the prohibited course of action appears to be in their interest. This impression may well be wrong. But prudential calculation is notoriously complicated, in part because we cannot predict the future with certainty. We can misjudge the situation, for example because we do not expect to be found out or convicted, or because punishment cannot be sufficiently severe to make it the case that crime is not worth our while. Also, some people may be more risk averse than others, and so more likely to take chances with compliance.Footnote ¹⁷

6.7 What Makes Self-Interested Behaviour Punishable?

So, if crimes are committed out of self-interest, if the crime is taken to be prudentially adequate at the time it is done, the state cannot, it would seem, punish the criminal for that – even though the threat of punishment appeals to the human desire to be happy.Footnote ¹⁸ Consider the following quotation from the Critique of Judgement. Kant is engaging in a thought experiment. What would be the result if the human will could only be determined by the expected agreeableness or disagreeableness of prospective options, which are commensurable and can be ranked on a single scale?

We cannot punish criminals for their folly. In fact, in the second Critique Kant himself indicates that imprudence and punishability are different in kind – so different, in fact, that they are quite distinct even in the judgement of the agent who is justly punished (KpV 5:37–8). There must therefore be something (broadly) immoral about crime for judicial punishment to be appropriate.

Roughly, criminals are punished because they have wilfully broken a law promulgated and valid at the time it was broken.Footnote ¹⁹ The Kantian state does not punish criminals because of any particular kind of moral deprivation, a task that is left to the Deity (see TL 6:460.33). And as with complying with the law, the state does not pay attention to the particular motive of the criminal as long as it is clear that he was responsible for the act, namely that the decision to break the law was based on a free choice. But even though the law does not concern itself with the specific motive of the agent, there is an assumption that (i) the agent had no legitimate motive at his disposal and (ii) that there was a legitimate motive to which he did have access, a motive that would have been sufficient to produce an action in conformity with external legislation. The law thus takes account of the agent’s mindset without paying attention to specific motivations, which were one and all illicit. This idea is reflected in the following definition:

But this raises further questions about the conditions that need to be given for an agent to be conscious of a prospective transgression in a way that renders it punishable by right.

6.8 How Can an Act That Is Judged Prudentially Adequate Be Punishable?

Consciousness of a transgression as such cannot merely consist in being conscious of what one is doing, such as appropriating someone else’s property. One must also be conscious of its wrongfulness. One must be conscious that one ought – and therefore can – do otherwise.Footnote ²⁰ This qualification is vital. Without this consciousness, it would be manifestly inappropriate to hold criminals to the legal standards set by the state. To abide by the law, it needs to be a live option. And for that we need conscience, whose task it is to watch the legality of action in particular.Footnote ²¹ As a result, internal legislation is needed to back up external legislation. After all, all juridical duties are indirectly ethical.

Only external compliance is required in the sphere of right; but at times the motive of duty is the only motive that can secure external compliance. The motive of duty is not required from a juridical point of view, which is motivationally neutral – and thus does not preclude our acting on the duty motive as such. So, it is contingently demanded that we act for the sake of the law – not as such, but because it is the only way we can secure conformity with law (or ‘legality’).Footnote ²² That is why it is so important that all duties are indirectly ethical.Footnote ²³

Note that this does not mean that the state can demand that we act in accordance with duty out of duty on any particular occasion;Footnote ²⁵ rather, it must assume that we face a meaningful choice, that we can refrain from breaking the law even if we judge it to be in our interest. Agents do not think they are justified overallFootnote ²⁶ as they violate the law (though they do think they are prudentially justified, or at least likely to give it a good shot).Footnote ²⁷ In some cases (in cases in which the criminal act is, despite the state’s best efforts, judged to be in the interest of the agent) the motive of duty is needed to make sure that obeying the law is more than just a theoretical option for the agent – a theoretical option already dismissed on prudential grounds. There is no free choice without the duty motive. The duty motive is thus a background condition of moral responsibility.Footnote ²⁸

6.9 Hegel’s Dog

Without the availability of the duty motive, Hegel’s criticism of Feuerbach’s deterrence theory of punishment would pose a problem for Kant’s view as well:

This is a common concern. Readers and commentators are worried that Kant cannot espouse a deterrence theory of punishment – not even if, to use Sharon Byrd’s memorable phrase, deterrence in threat of punishment is coupled with retribution in its execution – because this kind of external manipulation would amount to treating human beings as mere means rather than persons. But on the reading given above, Kant’s theory of punishment is dependent on the personality of those who think that breaking the law might be in their interest. It is their autonomy – and the availability of the duty motive that is tied up with it – that makes them a fit object of a legal punishment that is neither ‘natural’ nor unduly moralized.Footnote ²⁹

6.10 Backups in Law and in Ethics

Our account of the conditions of juridical punishment has an interesting implication for Kant’s moral philosophy as a whole. So far, we have argued that, appearances notwithstanding, Kant needs to rely on the duty motive in his legal theory. Respect needs to be available to motivate action in accordance with the law when – on, one would hope, rare occasions in a well-ordered state – agents judge that such action is not in their interest and are therefore inclined to break the law. Without the availability of moral interest, juridical punishment would be illegitimate.

Now, the idea that the motive of duty should serve as a backup has been ascribed to Kant’s moral philosophy before – if to his ethics, rather than his philosophy of law. Scholars like Richard Henson and Allen Wood maintain that we need to act from duty or respect for the moral law only when inclination-based motivation fails to point in the right direction. For the most part, action can thus be determined by benign inclination. They reject motivational rigorism on Kant’s behalf.Footnote ³⁰ On these views, which can be seen as a response to broadly virtue ethical concerns, Kant’s ethics does not include a requirement that obligatory action always be done for the sake of duty. Rather, the moral motive – or an action that has moral worth – is required only in cases of conflict. Even from the point of view of ethics, there would then be nothing objectionable about acting on inclination as long as what one wants to do coincides with what one morally ought to do.Footnote ³¹ But doing away with Kant’s motivational rigorism in this manner means legalizing the sphere of ethics, which, in Wood’s case, goes hand in hand with de-ethicizing the sphere of right. We would argue that both tendencies should be resisted.

In the end, however, external legislation is an abstraction. The ethical perspective is more fundamental and it is, importantly, inevitably the perspective of the agent. From a first-person point of view, all moral decision-making is ethical, that is, the legal perspective exists only from a third-person point of view. As an agent, it would be weird for me to say: I care only about conforming with the law of the land; in one area of practical normativity, there is no need for me to do my duty from duty. I cannot say: I ought to do it, juridically, because the state forces me to do it – if only the state’s forcing me to do it makes me want to do it, and what I want cannot be regulated by oughts. I ought to obey the law whether I want to do it – whether I judge it to be in my interest – or not. We are experiencing juridical obligations as categorical imperatives, not as the hypothetical imperatives the legislator employs to enforce them.Footnote ³² And that is possible only because they fall within the sphere of ethics. In fact, our view can even accommodate Willaschek’s ‘paradox of juridical imperatives’.Footnote ³³ His worry is that ‘juridical laws cannot find expression in categorical imperatives, after all, because juridical laws do not require obedience for their own sake’.Footnote ³⁴ The answer is that even if from the point of view of the state they do not, from the point of view of the agent they do – which is why they are proper categorical imperatives after all.Footnote ³⁵ The paradox can be resolved in true Kantian fashion by distinguishing two standpoints from which we can view juridical obligation.Footnote ³⁶

6.11 Objection: What about Kant’s ‘State of Devils’?

Our view faces one obvious objection. What, we may wonder, about Kant’s dictum that establishing a state is a problem that can be solved even for a people or nation of devils (ZeF 8:366.15–17)? Surely, the state must sustain the threat and execution of punishment? And devils, for the lack of respect for the law, cannot do their duty from duty? So, how can it be right to punish them?

It is clear from the historical context of this passage that Kant is exaggerating. First of all, Kant is reacting – maybe overreacting – against August Wilhelm Rehberg’s allegation of utopianism.Footnote ³⁷ The state does not rely on the divine nature of its citizens; on the contrary, even devils can live together in a well-ordered state. Secondly, the devils he has in mind are not endowed with a diabolical will in the sense defined in the Religion (see, again, RGV 6:37.21). They do not violate the law for the sake of violating the law. For a nation of truly diabolical devils, the task of establishing the state would be insoluble. Rather, the devils envisaged in On Perpetual Peace are a race of egotistical maximizers of self-interest.

So, rational, less than diabolical devils can make use of a state in a Hobbesian fashion, because it is in their interest. However, they do not have a state with the same kind of practical normativity as ours. There would be no exeundum either. This is Kant’s advice:

Would those devils that violate the laws of the land be punishable? Yes and no. They would not be punishable if by that expression we mean the infliction of just punishment. Making good on the threat of punishment would not be deserved. But any pain inflicted on such creatures does not count as undeserved either. It might thus still be in the interest of devil citizens to live in a state that is, by and large, to everyone’s advantage.

Human beings are importantly different from this breed of devils in that in their case acts of punishment are morally, not just instrumentally, justified.

6.12 Conclusion

So, the sphere of right and the sphere of ethics are less distinct than one might think. The former relies on the latter via the notion of legal obligation as indirectly ethical. The state cannot tell us: You have to do your (juridical) duty from duty. But the state can tell us: You have to do your (juridical) duty, period. It just happens to be the case that when committing a crime seems means–ends rational, the motive of duty – moral interest or respect for the law – is the only incentive available. And we cannot act without an incentive. If we then refrain from breaking the law, we do so from duty; if we do not, we should and could have decided not to break the law from duty. That is why, when we commit a crime, we deserve to be punished.

Let us note in conclusion that there are other ways in which the sphere of right rests on internal legislation. Most prominently, the exeundum does not depend on external legislation since it is only by virtue of our leaving the state of nature that external legislation arises in the first place. As Onora O’Neill puts it, ‘this duty to leave the state of nature and to enter a civil society is necessarily unenforceable, since it is a duty to establish the possibility of enforcement’.Footnote ³⁸ So, our conclusion should not come as a surprise.Footnote ³⁹