Ambiguous Sovereignty: Political Judgment and the Limits of Law in Kant’s Doctrine of Right

Abstract

Kantian legalism is now the dominant scholarly interpretation of Kant and an important approach to legal and political philosophy in its own right. One notable feature is its construal of the relationship between law and politics decisively in law’s favour: Law subordinates politics. Political judgment is constrained by and only permissibly exercised through law. This paper opposes this subordination through a close analysis of an ambiguity in Kant’s conception of sovereignty. Understanding this ambiguity requires seeing that, for Kant, law cannot subordinate the sovereign’s political judgment because it is a condition which makes a legal system possible. Kant was not an exacting legalist, but a practical reasoner struggling with the pressures politics imposes on finite beings. Through discussion of a leading contemporary approach, and close engagement with its historical source, the paper raises new considerations about the relevance of political judgment in the relationship of law and politics.

This paper addresses the relationship between law and politics in the context of legalistic readings of Kant’s political philosophy, as found in its most mature statement The Doctrine of Right. This ‘Kantian legalism’ refers to the recent interpretation and development of Kant’s political philosophy that is best exemplified in Arthur Ripstein’s Force and Freedom, Sharon B. Byrd and Joachim Hruscka’s Kant’s Doctrine of Right: A commentary,¹ and subsequent work influenced by these two texts. There the relationship between law and politics is decisively in law’s favour: Law subordinates politics. Political judgment is constrained by and only permissibly exercised through law. In other words, political judgment has its place, but this place is within the legal order. This is now the dominant approach to Kant’s political philosophy in English and it is not only a scholarly project. Kantian legalism is now proposed as a liberal and republican approach to legal and political philosophy in its own right.² Ripstein is as engaged in the philosophy of law as well as historical scholarship.³ Moreover, Kantian legalism is applied to contemporary questions in the philosophy of law and politics.⁴

I have a good deal of admiration and sympathy for Kantian legalism, so my project in this paper is not oppositional, but cautionary. I cannot endorse the full-throated subordination of politics to law, nor do I think Kant himself would either. Kantian legalism leaves out something distinctly political in Kant which makes law possible. This is the role of political judgment, which I highlight through a close analysis of Kant’s conception of sovereignty. Only a political, rather than legalist reading explains how an ambiguity in Kant’s conception of sovereignty is resolved through judgment. The substantive consequence of this for our reading of Kant is a reappraisal of the nature of his commitment to the rule of law. It should also give us pause before reading Kant’s political philosophy as exacting legalism. I suggest it is better read as a sustained work of practical reasoning which embraces the limitations of all human thought and struggles against the pressures that politics imposes on us as finite beings. It is only judgment, not law, that gives human beings a chance to manage these pressures. This thought is left out of Kantian legalism because it doesn’t attend to what is ambiguous and political in Kant’s concept of sovereignty. Through this discussion of a leading contemporary approach to legal and political philosophy, and close engagement with its historical source, I hope to raise new considerations about the role of political judgment in our thinking about the broader, general, and perhaps perennial question of the relationship of law and politics.

Kantian Legalism

Legalism is an approach to political and legal philosophy characterised by the subordination of politics to law.⁵ Political problems are resolved by subjecting them to legal processes, and rightful political action takes place within the constraints of law. More than this, the idea is that rightful political action, especially state action and power, ought to be exercised through law and legal processes. I introduce and use the term ‘Kantian legalism’ here to refer to readings of Kant’s political philosophy, especially in The Doctrine of Right, along these lines.⁶ These have become especially influential within Kant scholarship in the last decade.⁷

Kantian legalists differ on the details, but they all agree that, for Kant, the freedom and equality of all citizens is realised through law and the legal state; the Rechtsstaat. The freedom and equality of citizens cannot be achieved in the state of nature. This is because in the state of nature the relations between individuals are in some way problematically unilateral when it comes to rights claims such as property and contract. The thought is that if individuals are free and equal, then there is a problem to explain how the ability of any individual to place others under obligations through rights claims (such as to not interfere with an object by claiming that object as my property) can be compatible with the equal freedom of all. The solution to this is law grounded on the omnilateral or general united will. The sovereign of the legal state represents this will while making laws and thus the law is not unilateral, and the law can then underly the actions of individuals. The state is structured to achieve this end which Kant calls a rightful condition. As Thomas Sinclair puts it: ‘The most important means by which this is achieved is that there exist legislative, judicial, and executive roles in the Kantian state that are defined by obligations to create and sustain through law a rightful condition’.⁸ It is thus law and legal processes which secure the freedom and equality of citizens according to Kantian legalism. The effect of this on politics is to subordinate it to law. Free relations between individuals are guaranteed only insofar as politics is subjected to legal processes which are omnilaterally legitimated.

Kantian legalism thus contains substantive commitments, such as to the rule of law. But it is also a philosophical approach that holds a particular view about the purpose of Kant’s political and legal philosophy. Kantian legalism holds that the principles Kant derives are not only realised through law, but that Kantian political principles are the principles of legal systems. The Doctrine of Right is taken to be the elaboration of the philosophical principles that comprise the ideal legal system that would fully realise freedom and equality. The task of politics is to apply these a priori principles in the real world and, when this is done, we achieve a rightful condition. Importantly, rightful politics then becomes a matter of specifying these a priori principles in practice by making law, and all other political action needs to take place within the constraints of this law. Rightful politics becomes a fundamentally legal matter.

Unsurprisingly, then, law is central to Kantian legalism, but what of political judgment? I have signalled that my challenge to Kantian legalism will be to stress the relative importance of political judgment as against that of law in The Doctrine of Right, so what does Kantian legalism say about judgment? As I see it, judgment has three roles within Kantian legalism but each of these roles shares a common dependence upon the existence of a legal order, in virtue of which each of these kinds of judgment might be described as legal and not political. The first role of judgment is as verdict: The decision of a court is ‘a judgment’.⁹ Verdict is the judgment, ‘an individual act of public justice’, by which the judge applies positive law in a particular case.¹⁰ That is, the verdict of a judge resolves Kant’s practical syllogism: ‘These are like the three propositions in a practical syllogism: the major premise, which contains the law of that will; the minor premise, which contains the command to behave in accordance with the law, that is, the principle of subsumption under the law; and the conclusion, which contains the verdict (sentence), what is laid down as right in the case at hand’.¹¹

Verdict is perhaps the paradigmatic example of legal judgment, but the second role for judgment in Kantian legalism is administrative judgment. State officials are able to make judgments that individuals would not because they hold a legally defined public office that authorises them to make certain judgments. As the office is legally defined and the law is conceived as omnilateral, what would be problematically unilateral for the private person to do is not problematic for the public office holder.¹² The final and third role for judgment in Kantian legalism is legislative judgment. This is the kind of judgment involved in omnilateral public law-making, in which a priori principles of law are specified in positive legislation.

Each of these roles for judgment is premised on the legal order. That is, they are judgments that take place within established institutional and legal constraints – principally the separation of authorities. The judge only passes verdict on laws passed by the legislature, state officials only judge within the constraints of their offices as defined in law by the legislature, and the legislature does no more than pass law as one of three state authorities. Rightful political judgment is, in this sense, legal judgment according to Kantian legalism. Every political judgment can be given effect through law because rightful political judgment always operates within the legal order of the Rechsstaat.¹³ It is this feature of Kantian legalism which my argument is intended to challenge. Rightful political judgment need not be given effect through law, nor operate within the legal order on my reading of The Doctrine of Right. In particular, the rightful political judgment of the sovereign need not be legal. Indeed, the legal order itself depends on the sovereign’s political judgment, not the other way around.

Ambiguous Sovereignty

In order to see the important role of political judgment in The Doctrine of Right, we need to look carefully at the role of the sovereign. There is a core and systematic ambiguity in Kant’s concept sovereignty which can best be understood as being resolved by political judgment and in relation to a political conception of sovereignty, not through legal judgment and a legal conception. Now, to say that there are ambiguities in the concept of sovereignty would not strike political philosophers as being anything new.¹⁴ The most frequently discussed ambiguities in the notion of sovereignty involve either the relation of the state to the sovereign, or that of the people to the sovereign state, and one finds these in The Doctrine of Right too.¹⁵ But there is a different ambiguity in Kant that relates to the relationship of law and politics, and it concerns the scope of the sovereign’s authority in relation to the other authorities of the state.¹⁶

The Three Authorities

As noted above, Kant distinguishes three state authorities – sovereign or legislative, executive and judicial – and three persons who each hold one of these authorities – sovereign or legislator, regent and judge. Each authority has different powers in the Kantian state. Sovereign authority gives general or universal laws for the state.¹⁷ The laws are ‘the same (…) for all and all for each’ meaning that the laws are given by all (united) and to all (general). For this reason, sovereign authority belongs to the general united will of the people, though vested in the person of the legislator.¹⁸ The executive authority provides coercive physical power and other incentives to enforce the laws. It is held by the person of the regent, who is also called the governor, the directorate or government.¹⁹ The regent provides directives and decrees, not laws, and governs the state. The regent is the only authority which can exercise coercion and punish, but, as an agent of the state, they are subject to the laws of the sovereign.²⁰ The judicial authority provides verdicts on applications of the law to particular cases. As with the executive authority, the judicial authority is constrained by the laws of the sovereign and may not deviate from them.²¹ These three authorities are essential for the establishment of a state and are ‘dignities’, meaning they are ‘eminent estates without pay’ or what we could call political or civil statuses.²²

Sovereignty as Public Authority

Kant’s discussion of the three authorities has been noted for its obscurity.²³ One reason for this may be that when considered in relation to these three authorities of the state, Kant’s concept of sovereignty appears to be ambiguous between two conceptions. The first is the conception of sovereignty as public authority, which refers to the three authorities of the state taken together:

The three authorities in a state, which arise from the concept of a commonwealth as such (res publica latius dicta), are only the three relations of the united will of the people, which is derived a priori from reason. They are a pure idea of a head of state, which has objective practical reality. But this head of state (the sovereign) is only a thought-entity (to represent the entire people) as long as there is no physical person to represent the supreme authority in the state and to make this idea effective on the people’s will.²⁴

In this passage, the sovereign – as a thought entity – represents the people as united in a commonwealth and so contains within it the three authorities.²⁵ This thought entity is then represented by a physical person, the sovereign, and so this physical person also represents public authority in general. This person, Kant goes on to say, could be an autocrat, or an aristocratic or democratic assembly.²⁶

Sovereignty as Legislature

The second, opposing, conception of sovereignty is sovereignty as legislature. Instead of conceiving of sovereignty as containing all three authorities, this conception conceives of sovereignty as only one of the authorities: ‘the sovereign authority (sovereignty) in the person of the legislator’.²⁷ This sovereign authority only has part of the authority of the state, that of legislation. As a consequence, Kant describes the sovereign as interacting with the other state authorities: ‘The sovereign may take the regent’s authority away from him…’ and ‘the sovereign must also have in his power, in this case of necessity (casus necessitatis), to assume to role of the judge (to represent him) and pronounce a judgment…’²⁸ This sovereign holds not public authority in general but only one part of public authority, that of legislation.²⁹

Ambiguous Sovereignty in Kantian Legalism

This ambiguity between sovereignty as public authority and sovereignty as legislature is carried over into Kantian legalism. I say ‘carried over’ because it is not actively theorised by Kantian legalists, but nonetheless remains present in their texts. Jacob Weinrib is an exception in noting the ambiguity, but his purposes lead him away from discussing it further.³⁰ More often the ambiguity is carried over without comment. As an example, Ripstein states that ‘the three branches: legislature, executive and judiciary. Together, they comprise the sovereign’.³¹ Yet he also writes ‘the sovereign (legislature) also has the power to tax and to spend monies on the creation of public spaces…’³² I suspect the ambiguity is untheorized simply because the legalist approach calls for precise definition and delineation of important concepts, and Kant’s imprecision on this point can be chalked up to poor drafting, his self-confessed rushed exposition, or The Doctrine of Right’s troubled publication history.³³ Yet it is because I noticed that this ambiguity is carried over, even into the very phrasing and formulation of these texts, that I first began to suspect it had systematic importance.

If confronted by the ambiguity, however, a Kantian legalist might attempt to argue that with a more precise delineation of terms the ambiguity could be eliminated. There are two approaches that are most promising for this task, but neither is successful. The first is to draw on the distinction between a public office and the person who holds it, as discussed in Section I in the context of administrative judgment. Using this distinction, one could argue that there is no ambiguity in Kant’s concept of sovereignty, it is just that Kant sometimes refers to the office and at others to the person. However, the ambiguity transcends this distinction. I have highlighted above how when conceiving of sovereignty as public authority, the sovereign is a thought entity that comprises the three authorities taken together which is itself represented by a physical person. By contrast, when sovereignty is conceived as legislature, the sovereign authority as an office is represented in the person of the sovereign (legislator). In other words, Kant conceives of both sovereignty as public authority and sovereignty as legislature as both an office and a person and so the ambiguity remains.³⁴

The second approach is to argue that the concept of representation can be invoked to defuse the ambiguity.³⁵ As again discussed in Section I, for Kant and for Kantian legalists, the sovereign represents the general united or omnilateral will. One might think there is some way to explain the ambiguity away by these means. One might think, for example, that because the sovereign as legislature is closely connected to the general united will, that this is what needs to be represented by a physical person, and this physical person is the sovereign as public authority; however, this would be to miss the point. The question is, what is the idea or thought entity of sovereignty that is represented? Is it the conception as one legislative authority, or the conception of three public authorities together?

The Separation of Authorities

If we can’t explain away the ambiguity, we can still try to explain it. To do this, we need to look more closely at the conception of sovereignty as legislature. Understood as one authority among three, Kant argues that it ought to be part of a separation of authorities. Kant’s clearest statement of this is in the following passage:

Accordingly, the three authorities in a state are, first, coordinate with one another (potestates coordinatae) as so many moral persons, that is, each complements the others to complete the constitution of a state (complementum ad sufficientiam). But, secondly, they are also subordinate (subordinatae) to one another, so that one of them, in assisting another, cannot also usurp its function; instead, each has its own principles, that is, it indeed commands in its capacity as a particular person, but still under the condition of the will of a superior. Third, through the union of both each subject is apportioned his rights.³⁶

As argued by Paul Guyer and others, Kant argues both for a separation of functional authorities, and for a separation of the persons vested with those authorities.³⁷ But Kant goes further than this in the claims that the three authorities are ‘subordinate to one another’ and none may usurp the function of either of the others. In the capacity or ‘function’ of each dignity, none of the three persons may interfere with either of the other two in the ‘function’ of their dignity. This is a true separation. As Ripstein rightly points out, it does not take the form of instrumental checks and balances as is common in separation of powers doctrines.³⁸ Rather, each defines the limits of the others’ authority with regards to apportioning rights. We might say that it is only in virtue of their co-ordinate separation that each and all three together apportion rights. This requires the three authorities to create, apply and maintain a system of laws.

The separation of authorities is the core feature of what Kant calls his a priori ‘state in idea’ or ideal legal state. It is justified because it is necessary for the state to establish rightful law. It serves to create the conditions in which rightful law can be established, that is, in which law can be both rightful and effective in the external, empirical world.³⁹ To be rightful the law must be given by the general united or omnilateral will. This is Rousseau’s general will but modified for Kant’s philosophical purposes.⁴⁰ For Kant, the general will reflects his shared concern with Rousseau for autonomy; that all moral laws, including laws of right, be somehow self-given. Kant rejects Hobbes’ unilaterally willing sovereign and replaces it with this omnilateral will which is given by all and to all.⁴¹ The sovereign authority represents the omnilateral will and it is the sovereign’s will that wills the laws. However, the omnilateral will is not sufficient for establishing a system of laws. As well as omnilateral willing of laws, particular willing is also required to apply law in particular cases, both in the giving of verdicts where a case is submitted under the law, and in the application of coercive or incentivising force to uphold the law. This particular willing is unilateral; it is neither given by all or to all. Without particular willing, the omnilaterally willed laws could not take effect in the world. The particular wills necessary for the state are embodied in the executive regent and the judge, and it is only because they are the particular wills of particular persons that they can perform their functions within the state. Only the sovereign represents the general will directly, not the regent or the judge.

The omnilateral and particular willing of a system of law is made possible by the separation of authorities. The sovereign represents the omnilateral will and so can only will laws the same for all and for each. The sovereign cannot, for this reason, deal with particular cases.⁴² For the sovereign to act in these cases would be to act ‘beneath its dignity’ as the sovereign’s subjects have particular wills and the sovereign is a superior not an equal of its subjects.⁴³ One might even say that in intervening in a particular case, the sovereign would be relinquishing their right of representation of the omnilateral will.⁴⁴ Certainly, they would be overstepping the bounds of their authority and threaten the rightfulness of the system of laws. By separating a regent to govern using coercion or other incentives ‘directed to particular cases’ and the judge to give verdicts in particular cases, Kant makes possible the particular willing that makes possible an effective system of law.⁴⁵ The laws remain legitimate because both judge and regent are subordinated to the omnilaterally willed law and may not make law themselves. Hence, the regent can will only decrees ‘given as subject to being changed’ and not laws.⁴⁶ Moreover, the judge must be a separate person from the regent because both the regent and judge are particular persons with particular wills. This means that the regent cannot judge in their own case, except with the danger of them thereby committing a wrong.⁴⁷ In so far as the regent is to be held accountable under the laws of the sovereign, they must therefore not judge in their own case.⁴⁸ One might think there is a problem here in that the judge could judge in their own case. This is what Kant describes happening in a case of equity. Perhaps because he is aware of this point the scope of the judge to give verdicts on non-lawful grounds of equity is severely curtailed.⁴⁹ However, since inappealability of the highest judge is inherent in the concept of judicial authority, there is in the extreme case no route of appeal and no-one else could have standing to make this judgment.⁵⁰ Kant’s justification of the separation of authorities is thus designed to allow the state to establish a system of laws which is legitimate because the laws proceed from the general united will, but also effective because it can address particular cases and subsume these under law.

This crucial argument that the separation of authorities is the necessary condition of rightful law-making is widely accepted in the literature, even if the reading of the argument Kant gives may differ somewhat to the one I have presented here. Byrd and Hruschka argue that each of the authorities is necessary for the establishment of a rightful condition under universal law, and that the separation of authorities in accordance with the state in idea is part of what is necessary for fulfilling the ideal constitutional principles of a state that rules through rightful law.⁵¹ For Ripstein, each of the authorities is necessary for the possibility of omnilateral law-making as each solves a normative problem of the state of nature.⁵² The legislature solves the problem of unilateralism in property acquisition by underwriting such acquisitions with omnilateral law. The executive provides assurance that legal rights will be respected. The judiciary resolves the indeterminacies that arise in the application of universal laws to particulars. In the absence of a state and the three authorities, each of these problems can only be resolved unilaterally and hence threatens equal freedom.⁵³ It is not my purpose in this paper to suggest either of these readings is wrong. When it comes to the separation of authorities, I believe Kantian legalists are right to stress its role in setting the conditions for omnilateral law-making. Where I do diverge is on the second feature of the complex relations between the three authorities in Kant’s ‘state in idea’ which is the hierarchy of authorities. The arguments I present concerning the hierarchy can be applied to any account of the separation of authorities that retains the concern with necessity of the separation of authorities for rightful law-making.

The Hierarchy of Authorities

As well as a separation of the three authorities, Kant also establishes a hierarchy of the three authorities with the sovereign at the top. Beyond the direct functions of the state authorities with regard to the establishment of law, the sovereign legislature holds powers over the other two authorities which are not matched with symmetrical powers. This creates a hierarchy, not equality, between the three separate authorities. Consider the following passage, which I referenced in extract above:

The sovereign can also take the regent’s authority away from him, depose him, or reform his administration. But it cannot punish him (and the saying common in England, that the king i.e. the supreme executive authority, can do no wrong, means no more than this); for punishment is, again, an act of executive authority, which has the supreme capacity to exercise coercion in conformity with the law, and it would be self-contradictory for him to be subject to coercion.⁵⁴

The first sentence shows that the sovereign has authority over the regent – i.e. the person who holds executive authority – in a right to remove, depose or reform them. The sovereign legislature’s authority here does not manifest in a right to exercise executive authority against the regent, but to strip their authority rendering them no more than a private citizen.⁵⁵ The sovereign has the authority to decide who can exercise executive authority and, to that extent, how they exercise it.

The Doctrine of Right also gives the sovereign legislature authority over the judge or judicial authority. First, the sovereign has power of appointment of judges: ‘neither the head of state nor regent can judge, but can only appoint judges as magistrates’.⁵⁶ Second, in certain circumstances or ‘cases of necessity’ the sovereign has the power to ‘assume the role of judge’ and decree a different sentence to that pronounced by the judge.⁵⁷ Finally, the sovereign also has the right to grant clemency of which Kant says: ‘This right is the only one that deserves to be called the right of majesty’.⁵⁸ The right of clemency is limited. Firstly, it is restricted to cases of ‘wrongs done to himself’ i.e., to the sovereign, and never ‘crimes of subjects against one another’.⁵⁹ One plausible interpretation of this restriction is that the right of clemency only applies in the case of a crime against the state, not a private person, or crimes of public law, and not private law. Secondly, even then the right of clemency may not be invoked if it ‘could endanger the people’s security’.⁶⁰ Nonetheless, the right of clemency is a second right of the sovereign legislature to annul the verdict of the judge. There are no symmetrical powers for the judge over the sovereign legislature. Nor are there such powers for the regent.

Kant has a principled reason for endorsing this hierarchy of authorities as well as the separation of authorities. This is the necessity of the state as the condition for possible law-making. To see why this is the case, it will be helpful to consider the example of Kant’s ‘population-wide murder conspiracy’, before moving on to the justification in general.⁶¹ It is worth quoting in full:

Accordingly, every murderer – anyone who commits murder, orders it, or is an accomplice in it – must suffer death; this is what justice, as the idea of judicial authority, wills in accordance with universal laws that are grounded a priori. – If, however, the number of accomplices is so great that the state, in order to have no such criminals in it, could soon find itself without subjects; and if the state still does not want to dissolve, that is, to pass over into a state of nature, which is far worse because there is no external justice at all in it (and if it especially does not want to dull the people’s feeling by the spectacle of a slaughterhouse), then the sovereign must also have it in his power, in this case of necessity (casus necessitatis), to assume the role of judge (to represent him) and pronounce a judgment that decrees for the criminals a sentence other than capital punishment, such as deportation, which still preserves the population. This cannot be done in accordance with public law but it can be done by executive decree [Machtspruch] that is, by an act of the right of majesty, which as clemency, can always be exercised only in individual cases.⁶²

In this case, the judge is confronted by a murder conspiracy encompassing the whole, or nearly whole, population. By law the rightful judgment is to execute each of the citizens. The reason for this is that for the crime of murder, no punishment is adequate but that the perpetrator die.⁶³ Earlier in the section from which this passage is taken, Kant also states that the law of punishment in general is a categorical imperative.⁶⁴ So it is not only that by positive, or what Kant calls subjective, law that the judge ought to pronounce this verdict, but by objective law, grounded a priori and established by the sovereign legislature.⁶⁵ Following this, the executive regent ought to carry out the executions. This course of action would be the three authorities performing their respective functions in accordance with the separation of authorities. The sovereign establishes the lawful punishment for murder in accordance with a priori principles; the judge applies the law through their verdict; and the executive regent carries out the extreme coercive function of executing the entire population. Yet Kant still holds that the sovereign must intervene via a ‘Machtspruch’ to prevent this from happening. How can this be justified and made compatible with the justification for the separation of authorities?

The key is to see how the sovereign still represents the omnilateral will throughout this case. Kant’s reasoning is that carrying out the original verdict would lead to the risk that the state would ‘pass over into the state of nature, which is far worse because there is no external justice at all’. In the unlikely empirical circumstances Kant imagines, the normally desirable fact that the particular willing of the regent and the judge cannot be directed at anything other than applying a general law in a particular case has created a situation in which the state, lawfully and rightfully, might rule itself out of existence, thus destroying any possibility of rightful lawgiving. Without population there cannot be a state and without a state there cannot be rightful laws. The judge’s particular will is constrained by omnilateral law to will the destruction of the conditions that make such willing possible: But this is a contradiction in willing. For this reason, the sovereign acts in a particular case whilst still representing the omnilateral will because that act secures the conditions for the possibility of the omnnilateral willing of rightful laws. The sovereign assumes the role of the judge, exercises their authority to pronounce a sentence that maintains the population, such as deportation to another province, and carries it out through executive decree. The separation of authorities is violated, but the sovereign does not act from their particular will, and so is rightful and justified in doing so.

The population-wide murder conspiracy is not the only place in which Kant explicitly makes clear his commitment to justifying the sovereign’s right to maintain the existence of the state. Kant also argues that the sovereign, in addition to rights of taxation, may further levy forced loans, even in cases where that deviates ‘from previously existing law’ when ‘the state is in danger of dissolution’.⁶⁶ In a similar vein, Kant’s justification of the right of the state to tax where it is necessary to maintain ‘the poor, foundling homes and church organizations’ is based on his view that ‘The general will of the people has united itself into a society which is to maintain itself perpetually; and for this end it has submitted itself to the internal authority of the state in order to maintain those members of the society who are unable to maintain themselves’.⁶⁷ Beyond The Doctrine of Right, in Toward Perpetual Peace, Kant discusses the possibility of permissive laws that allow the sovereign to postpone putting laws into effect lest that see the state ‘devoured by other states’.⁶⁸ Again, Kant justifies the sovereign acting in defence of the state. This indicates that, although the population-wide murder conspiracy is an incredibly unlikely and exceptional circumstance, this authority of sovereign is invoked in the normal course of politics.⁶⁹ However, the population-wide murder conspiracy is the most important example in the present context because it is in that case that Kant says explicitly that the sovereign is justified in violating the law and the separation of authorities in order to maintain the state.

This line of reasoning can be expanded to cover the other cases in which the sovereign violates the separation of authorities and thus justifies the hierarchy overall. The sovereign intervenes on the executive and judicial authorities when doing so is necessary for maintaining the state, as the condition which makes rightful lawgiving possible, under empirical conditions that may not always be favourable to this. For example, the authority to strip the regent of executive authority means that should the regent, who is of course a finite human being and therefore fallible, begin to threaten the state, then the sovereign is able to rightfully intervene. Therefore, the sovereign’s authority allows them to maintain the state in cases where the application of law through separated authorities is not sufficient to the task, or even where the application of the law itself is the source of the threat.

Political Sovereignty

The hierarchy of authorities thus rests on the principled need to maintain the existence of the state as the condition of possible law-making. But how are we to understand this principle and in turn understand the nature of sovereignty in The Doctrine of Right? In the literature, the hierarchy of authorities, and therefore sovereignty, is most often understood in legal terms, but this approach runs into problems. Byrd and Hruschka argue the sovereign is “‘higher’ than both the highest executive and the highest judge and thus stands above them, because the second and third powers are subject to, or under the law”.⁷⁰ This is true, but it does not seem to me to account for the features of the hierarchy as just described. The fact that executive and judicial authority must be exercised in accordance with law seems to be more of an expression of the separation of authorities with regard to law-making than an account of the hierarchy. The subordination of the regent and judge to law is perfectly compatible with the three authorities being ‘coordinate and subordinate’.⁷¹ That describes the separation of authority with regard to law-making, rather than the hierarchy of authorities. Indeed, in the passage cited by Byrd and Hruschka in support of this understanding of the hierarchy, Kant does say ‘the regent is subject to the law’.⁷² However, he also says ‘The sovereign can also take the ruler’s authority away’ indicating that the authority of the sovereign over the regent consists in more than the subordination of executive authority to law.⁷³

Ripstein also argues that ‘the legislative will takes priority’ because ‘the exercise of judgment and the enforcement of rights’ by the judge and regent respectively ‘must be done in accordance with law’.⁷⁴ But he goes further to argue that ‘The only way a judge or enforcer can be empowered consistent with right is through an act of legislative will’.⁷⁵ This is suggestive that not only is the sovereign higher because the powers of the regent and judge must be exercised under law, but the hierarchy also consists in the power of the sovereign to appoint the regent and the judge through legal means. This, I think, does address the hierarchy of authorities, because these powers of appointment are not reciprocal. This idea is expanded by Marie Newhouse who argues ‘the legislative authority is first among equals’ because it has ‘the capacity to make constitutional laws that allocate authorities among institutions within the empirical state’.⁷⁶ However, the legalistic understanding of the hierarchy still faces problems. One is that it makes it hard to understand how the sovereign’s judgment in the population-wide murder conspiracy case can be rightful. Recall that the sovereign does not act through ‘public law’ and, indeed, acts contrary to established positive legislation.⁷⁷ The same problem applies at least to the case of forced levies which deviate ‘from previously existing law’.⁷⁸ Whilst it might be possible to understand the sovereign’s powers of appointment in legal terms, if one is available, an understanding of the hierarchy that can explain all its facets together is preferable.

A second problem with understanding the hierarchy as an assertion of the legal supremacy of the sovereign over the regent and judge is it causes one to wonder in what sense a separation of authorities really obtains at all. In his analysis, Guyer argues that the hierarchy of authorities overrides any claim Kant might have to a separation of authorities. Relying on a similarly legal understand of the hierarchy, Guyer presents us with a useful challenge.⁷⁹ Guyer claims ‘there is no clear sense in which the legislature is ever subordinate to the executive’ and ‘As long as the entire judicial process – finding of fact, application of law, enforcement of law – is governed by the laws of the legislature, the authorities involved in the judicial process remain subordinate to the legislature. There is no obvious sense in which the judiciary is coordinate to the legislature’.⁸⁰ Hence, Kant’s claim that the three authorities are coordinate and subordinate to each other is a ‘rhetorical flourish’.⁸¹ Yet given that Kant’s coordinate and subordinate claim occurs within the main text of The Doctrine of Right, hence as part of Kant’s system of right and not, for example, in the remarks or in the preface, it would be surprising if the claim were entirely rhetorical.⁸² How, then, can we explain the hierarchy but not reduce Kant’s subordinate/coordinate claim – and hence the separation of authorities – to mere rhetoric?

The key, I believe, is the distinction between the two conceptions of sovereignty distinguished in Section II. Over the previous two sections, we have been looking closely at Kant’s conception of sovereignty as legislature. We have seen that this sovereign is embedded in complex relations with the regent and the judge with the purposes of establishing rightful law and ensuring the sovereign has the authority to maintain the state. I now want to bring the conception of sovereignty as public authority into the discussion. Recall that on this conception all the authorities and all the authority of the state adhere to the sovereign. What we can now see is that there is a certain fuzziness to the conception of sovereignty as legislature that blurs the distinction with the conception of sovereign as public authority. The sovereign legislature does not hold all authority in the state, but it does hold some authority over all other authorities in the state. In other words, Kant’s sovereign legislature does not hold all state authority, but is not simply one authority among three either.

Narrowing the gap between two conceptions of an ambiguous concept is not to eliminate the ambiguity, but in this case it does help us understand it. I suggest that the best way of understanding this ambiguity is not as a failure of precise definition of a legal concept, but as the use in practical reasoning of a political concept – hence I call this interpretation the political conception of sovereignty. According to this reading, the sovereign in the Kantian state holds all the authority of the state. However, they are also under an obligation to alienate their authority over judicial and executive matters onto other persons. The regent and the judge hold and enact the authority of the sovereign in executive and legislative matters. This is what I take Kant to mean when he describes the regent as an ‘agent of the state’ or the ‘organ of the sovereign’.⁸³ The authority of the regent and the judge is ultimately the sovereign’s but ought to be actioned by different persons. This also means that in cases such as the population-wide murder conspiracy and the others discussed above, the sovereign has the authority and right to intervene on the executive and judge in their functions because the sovereign is only interfering with rights to executive and judicial authority that they already hold.

On the political conception of sovereignty, both the conception as legislature and the conception as public authority hold it certain respects. The sovereign is regarded as holding all public authority but is also regarded as obligated to rule as a sovereign legislature. Rightful law is established through a separation of authorities, and so the sovereign is obligated to alienate executive and judicial authority onto other persons in order to rule rightfully. In contrast to Guyer’s claim, then, there is a sense in which the regent subordinates the sovereign. The sovereign is ideally subordinate to the regent and the judge with regard to their functions in law-making. Kant’s claim that the three authorities are coordinate and subordinate to each other thus holds in the form of an imperative on the sovereign. Given that the separation of authorities comprises part of the ideal state in idea, this imperative structure of the claim is not out of place.

However, as we have seen, under some empirical circumstances maintaining or attaining such a separation of authority becomes impossible without thereby imperilling the state upon the existence of which the possibility of rightful law depends. So, the sovereign is also under another obligation to maintain the state. These two obligations are not inherently contradictory but can become so under certain empirical circumstances. Kant’s ambiguous and political conception of sovereignty seems to leave us with a threat of contradiction, but it is here that the role of political judgment emerges most clearly, and most decisively. Kant’s sovereign is ambiguous because they are active in the politics of the state, not a static legal institution which frames political life. The sovereign is tasked with judging when the empirical circumstances permit rule through rightful law, and when they require the sovereign to exercise their extra-legal authority. Once we understand this extra-legal or political quality of Kant’s sovereign, we can also see how the sovereign’s judgment in the population-wide murder conspiracy case can be understood as rightful, and do so in a way consistent with other instances where the hierarchy is either expressed or invoked in The Doctrine of Right. This gives us reason to favour a political over a legal reading of Kant’s conception of sovereignty.

As an example of political sovereignty in action, consider the right to punish. This is a coercive right, and as such a right of executive authority.⁸⁴ What the political conception implies is that the right to punish is a right of the sovereign, because all rights of the state ultimately adhere to the sovereign. However, given the obligation to establish rightful law and hence to rule through a separation of authorities, the exercise of this authority ought to be carried out by the regent. The sovereign then rules rightfully as a sovereign legislature without executive or judicial functions. This is ‘to make the kind of government suited to the idea of the original contract’ or, as Kant would put it elsewhere, to rule in ‘the spirit of a representative system’ or in a republican ‘manner of governing’.⁸⁵ Fulfilling this obligation may require reforms of the state, which Kant holds must ‘be carried out only through reform by the sovereign itself…’⁸⁶ The sovereign is required to judge when doing so is compatible with the ongoing existence of the state. However, even once this step to separate authority is taken, the sovereign cannot absolve themselves from affairs of state other than giving law. Recall the previously discussed right of the sovereign to depose the regent if the regent becomes a threat to the state. It is easy to imagine how a rogue regent with the right to punish through the exercise of coercion might be such a threat and why Kant would want the sovereign to act against them by deposing them. According to the political conception of sovereignty, the political judgment as to when this is necessary falls to the sovereign.

What is interesting about this is that it allows a certain flexibility for the sovereign to approximate as closely as possible to rightful rule, even in circumstances which make this very difficult. Kant is very hesitant about allowing the sovereign to exercise executive authority under any circumstances. He writes ‘In that case the sovereign behaves through its minister as also the regent and so as a despot’.⁸⁷ He is also concerned that any form of coercive resistance to the regent would be self-contradictory as the executive authority is the supreme coercive authority. If another authority could resist them, they would no longer be supreme. But the sovereign still has a course of action open to them which they can take if they judge it necessary to maintain the state. They can strip the regent of executive authority and appoint another in their place. The previous regent would then return to being only a private citizen. Were they to continue to try to exercise coercion this would be a clear wrong, and the new regent would be acting rightfully in coercively putting down what now could only be described as a rebellion. These are the kind of fine political judgments that the political sovereign is tasked with whether or not a separation of authorities has already been established. The challenge for the sovereign is to come as close to rightful ruling as the circumstances allow, which may even involve backwards steps if the circumstances demand it.⁸⁸

Political Judgment and the Ideal Legal System

Once we have the political quality of Kant’s concept of sovereignty in view, the ambiguity in sovereignty eventually resolves into a contradiction between two obligations on the sovereign, which is itself resolved through the political judgment of the sovereign. One might say that this amounts to clearing up the ambiguity and restoring consistency to Kant. In one sense, this is correct. In so far as Kant’s political philosophy should enable us to make consistent political judgments, the ambiguity is not problematic. But if the text is supposed to be the blueprint for an ideal legal state or system of law as Kantian legalists would have it, then the ambiguity still presents a problem. Kant does not define the circumstances in which sovereign judgment may subordinate the ordinary processes of rightful law and the separation of authorities, he leaves it to the judgment of the sovereign. We might be able to say that if the sovereign is wrong about whether there is a threat to the state, then they have exercised judgment poorly. We might also be able to say that if the sovereign is not acting in good faith and is using the threat to the state as a pretext for usurping other authorities, then this is wrong. But in both cases, what we don’t have is a set of conditions and circumstances by which the rightfulness is to be judged without either relying on empirical judgment about how bad a threat to the state is, or on suppositions about what the sovereign is really thinking.

Kantian legalism favours a discourse of legal precision, set categories, and fine distinctions, and I suspect would reject this tolerance of the ambiguity inherent in political sovereignty.⁸⁹ More importantly, however, the interpretation I have given suggests that Kant’s political principles are principles of political judgment, not principles of ideal legal systems or even of legal judgments. This is the core of my challenge to Kantian legalism. Whilst it would be wrong to say of The Doctrine of Right, and Kant’s political philosophy in general, that it is not concerned with the elaboration and justification of a system of a priori legal principles which are to serve as the normative ideal for all positive legal systems, my point is that it is concerned with much more besides. The Doctrine of Right is a system of external duties that concern interaction between agents from which all rights derive.⁹⁰ Such external duties are not exhausted by those a priori legal principles that ground a system of law but concern all external duties including those that cannot be captured within a legal system. This renders The Doctrine of Right an inherently political text; it aims to orientate the agent in a world shared with others by establishing grounds for external duties and principles for practical judgment.⁹¹The Doctrine of Right is thus best understood as offering a sustained attempt at political and practical reasoning of which the ambiguous concept of political sovereignty is, I suggest, a key part. Kant’s aim is not legal precision, but principles that guide the judgment of political agents. In Kant’s practical reasoning in The Doctrine of Right the perspective of the finite, which is to say the human, judging agent is never lost. If sovereignty is political, ambiguous, and active, rather than legal, precise, and static, then it is not possible to eliminate judgment from our politics. Nor is it (always) possible to constrain it through law. Judgment is needed at each stage of political thinking and acting, never allowing us to rest assured that law grounded on a priori principles justifies everything we do. The political sovereign is active within the Kantian state and must exercise judgment in contending with all the vagaries of empirical political circumstance.

There are no explicit conditions that could frame a constitutional provision determining when the sovereign exercises their judgment, because the framing of such an ideal constitution is not Kant’s primary concern. Indeed, the means by which he could do this are ruled out. As we have already seen, positive law (constitutional or otherwise) does not constrain the sovereign’s judgment. Recall that Kant claims that the sovereign’s usurpation of the judge in population-wide murder conspiracy case ‘cannot be done in accordance with public law’.⁹² This is an explicit statement that when the sovereign usurps the role of the judge, the sovereign does not act through law by changing it or introducing a loophole. It is perhaps because Kant invokes ‘justice, as the idea of judicial authority’ as the grounds for his policy of an-eye-for-an-eye that such legal solutions are unacceptable in this case.⁹³ This is all the more likely if we also accept my earlier characterisation of this law of punishment as objective law. Kant instead turns to a different solution, one that takes us beyond the law, to a ‘Machtspruch’ or executive decree.⁹⁴ As we saw in Section IV, this also applies to other cases in which Kant argues for the sovereign to act extra-legally in defence of the existence of the state, such as that of forced loans, or in violation of the separation of authorities, such as deposing the regent. We might add to this last point that Kant’s rejection of ‘mixed constitutions’ and his claim that a ‘moderate constitution, as a constitution of the inner rights of a state, is an absurdity’ show that he is also not concerned with other legal institutions constraining the sovereign’s judgment.⁹⁵ The other authorities do not hold rightful authority over the sovereign and so cannot constrain the sovereign’s judgment.

The core difference between the reading of The Doctrine of Right presented here and that of Kantian legalists, is the claim that Kant’s primary concern is not framing the ideal constitution because his political principles are principles of political judgment not of ideal legal systems. I am not arguing that an ideal legal system is unimportant for Kant. Indeed, the necessity of creating and maintaining a system of law is Kant’s justification for the separation of authorities. Moreover, I also accept that ideals play a key role in Kant’s political philosophy. Nor am I claiming that Kantian legalism sees no place for political judgment.⁹⁶ Instead, I am making a claim about the relationship of this ideal of a legal system to political judgment. For Kantian legalists, political judgment is legal in the sense that it is constrained within the established legal order. State authorities are empowered by law to make judgments that would otherwise be wrongful. This leaves open the matter of the political judgments reached by different state authorities, but it requires that these judgments be grounded in law or given expression through law. It is this understanding of political judgment in Kant that I reject. Political judgment in Kant is not constrained by the ideal legal system, but rather the latter depends on the former. The sovereign’s judgment is political.⁹⁷ It is, as Kant terms it in the population-wide murder conspiracy example, a ‘Machtspruch’. It is an exercise of power, unconstrained by law upon which the possibility of the legal system and of rightful law-making depends.

Kant’s ideal legal system or the ideal ‘state in idea’ is then a regulative ideal that orientates for the sake of practical action.⁹⁸ Ideals in Kant’s practical philosophy are the realisation of ideas of reason. A useful example of what this means can be found the Critique of Pure Reason where Kant says that the idea of wisdom has as its counterpart ideal the perfectly wise person or sage.⁹⁹ An ideal is an individual thing entirely determined by the idea. Whether this is the ideal sage by the idea of wisdom, or the ideal legal system by the idea of rightful law. Such ideals are impossible for us to realise, but they serve to orientate the action of practical moral agents.¹⁰⁰ Thus in so far as wisdom is a virtue, we ought to strive to realise sagacity in our own person, even though being perfectly wise is impossible. The same is true of Kant’s ideal legal system in so far as the sovereign ought to rule through rightful law, with the attendant separation of authorities, so they are constrained by the idea of law in the form of an obligation. But for this argument to support a legalistic constraint on judgment it would require that this idea of law must be decisive in the sovereign’s judgment. Political judgment must be made the mere application of the idea of law to empirical circumstances.

However, we have already seen that the sovereign is also under an obligation to maintain the state. Moreover, we have seen that this obligation is grounded in Kant’s ideal ‘state in idea’. The ideal is thus complex and arises from to two imperatives, both binding on the sovereign, that are not inherently contradictory but might become so under particular empirical circumstances. Put another way, the sovereign is stuck between two orientations. Again, think of the population-wide murder conspiracy. It is only under the empirical circumstances in which there is such an enormous conspiracy that the sovereign is forced into a judgment between two obligations. The first is to allow the verdict to go ahead in accordance with objectively rightful law, the second is to usurp the function of the judge in order to maintain the state. Faced with this dilemma, the a priori idea of rightful law cannot be decisive in the sovereign’s judgment since there is nothing a priori that allows for a judgment between the two obligations.

This dilemma is a conflict of duties. Strictly speaking, Kant denies that a true conflict of duties is possible, but in the introduction to The Metaphysics of Morals he discusses the potential for conflicts in what he calls the grounds of obligation.¹⁰¹ Following Jens Timmermann’s analysis, such grounds of obligation are best conceived of as pro tanto duties which can be defeated by other grounds of obligation judged to be stronger.¹⁰² To find the stronger ground of obligation, the judger cannot simply look to a priori principles, but must judge according to the moral conditions of their experiential circumstances. This is a moral judgment that can only be made by the judging agent in the conditions they find themselves. The conflict ‘arises in moral practice’.¹⁰³ Applying this idea to the case of the sovereign exercising political judgment, the sovereign must judge which of the grounds of obligation – the idea of rightful law, or the maintenance of the state – is stronger in the circumstances of their moral practice. In this political instance of a conflict of duties, the sovereign will have to make a kind of prudential judgment: Will ruling rightfully, through the law and the separation of authorities, threaten the state? In the case of the population-wide murder conspiracy the answer is clear, it will do so and so the ground of obligation to maintain the state is stronger. In other cases, the answer won’t be so clear, and the sovereign may have to accept some risk to the state in order to rule rightfully.

The ideal legal state thus cannot provide definitive orientation to political judgment, and nor can the idea of law itself be decisive. Kant’s ideal for the legal state is part of his practical reasoning in The Doctrine of Right, but it is not his only focus. His focus is on the rightfulness of political judgment, and the principles that guide it. He finds that the application of law, and the formulation of a legal ideal is not sufficient for this task. The picture of moral and political judgment that emerges in Kant is a good deal more complicated than that.

The Limits of Law

Kant’s political sovereign carries a heavy burden of responsibility.¹⁰⁴ They are bound to make political judgments, but neither a priori principles, nor regents, judges, legal institutions nor positive law provide certain guidance. This is the limit of law in Kant’s political philosophy: Law can never fully subordinate political judgment. In this limit, Kant recognises that the force of empirical political circumstances on us finite beings mandates that judgment is exercised at every step in politics. However, he also holds that this judgment has a moral quality and a moral force. Prudent judgment is grounded not in instrumental reasoning, but in the obligation to maintain the state. Kant morally vindicates the role of prudence at the heart of his politics; exercised in accordance with rightful principle, prudence is not amoral or immoral, but is, to the contrary, deeply moral. This is what he means when he claims that a moral politician is ‘one who takes the principles of political prudence in such a way that they can coexist with morals’.¹⁰⁵ Rightful political judgment requires the sovereign as a moral politician to exercise political judgment to navigate the twin obligations to rule through rightful law-making and to maintain the necessity of the state. To emphasise the role of political judgment in Kant’s politics is not to dismiss the importance of moral judgment. It is rather to argue that there is a form of moral judgment distinctive to politics, and that the responsibility for this judgment falls on the sovereign above all.¹⁰⁶

To conclude this paper, I want to set out where my arguments leave me in relation to Kantian legalism and to the literature on Kant’s political philosophy more broadly. The active, political, prudential and judging Kantian sovereign I have argued for is in stark contrast to any legalist reading of the concept. As an interpretation of the ambiguity inherent in the notion of sovereignty, the political conception emphasises that, for Kant, establishing, maintaining and improving a rightful constitution in the world is not a legal project but a political one.¹⁰⁷The Doctrine of Right is the key text for understanding how he understands right and law and, in the central concept of sovereignty, Kant grounds law on the political judgment of the sovereign. Political judgment is thus central to The Doctrine of Right, not the marginal concern it often appears to be in Kantian legalism.

As well as this difference with Kantian legalism, my view also contrasts with approaches to political judgment in Kant which follow Hannah Arendt in drawing on material from Kant’s Critique of the Power of Judgment, and as well as from other of Kant’s political, anthological, or historical writings.¹⁰⁸ A full picture of the role of political judgment in Kant’s political philosophy might well require these kinds of accounts. But my purpose here, and my challenge against the Kantian legalist, is to show that the principles derived in The Doctrine of Right itself, are principles of political judgment, not only of legal systems. It is not only in his historical, anthropological, and pedagogical writings, and well as other texts on politics, that judgment is key for Kant. Though these writings may take up the question of the refinement of political judgment, it is in The Doctrine of Right that the central importance of political judgment is established. By contrast with Arendt who declared Kant never wrote a political philosophy – a point Ronald Beiner refined by saying that, since Kant supposedly sees no place for prudence, he has no account of political judgment¹⁰⁹ – for Kant, as I read him, political judgment determines the relationship of law and politics. My claim thus does not amount to the well-established understanding that Kant thought that moral principles, such as those found in The Doctrine of Right, require supplementation by judgment, something that Kantian legalists are as clear about as any other group of interpreters.¹¹⁰ Rather my claim is that the moral principles of The Doctrine of Right should be understood as apt for judgment. That is, as principles of political judgment.

Kant’s complex ideal of the ‘state in idea’ or Rechtsstaat arises from two obligations, to rule rightfully through law and to maintain the state, which can become contradictory under certain empirical circumstances. Whilst a legalist reading would attempt to resolve such a possible contradiction through the precise delineation of principles in philosophy, I have argued that Kant held that for this task the sovereign’s practical political judgment is central. Thus, contrary to Kantian legalism, politics is not subordinated to law, rather law depends on political judgment. The task of politics is not only to apply law in particular circumstances; the task of politics judge what is necessary to secure the conditions in which a legal system, grounded on a priori principles, could exist.

As Kant puts it in his reply to criticism from Benjamin Constant, politics is a problem, the problem of how to arrange society in accordance with freedom and equality.¹¹¹ It is a problem finite human beings face, perhaps perpetually, and that requires judgment, not only law and principles, to manage successfully. It is this thought of Kant’s, though not necessarily ‘Kantian’ thought, that I have tried to understand in this paper, a task which legalistic readings have not undertaken. I don’t claim to have decisively refuted such readings, only suggested that another kind of interpretation is possible, one embraces the active ambiguity in Kant’s conception of sovereignty as being the result of the practical and political perspective from which he wrote and judged.