Institutionalising Kant’s political philosophy: Foregrounding cosmopolitan right

Cosmopolitan right

While ‘internal freedom’ is the concern of Kant’s moral philosophy, his political philosophy instead concerns ‘external freedom’. ³ Internal freedom refers to persons’ autonomous willing of maxims of action in accordance with principles of pure practical reason, and independent of subjective, arbitrary desire. Unlike internal freedom, external freedom is an inherently relational idea. For Kant, each person, ‘by virtue of his humanity’, has ‘one innate right’, namely ‘freedom (independence from being constrained by another’s choice), insofar as it can coexist with the freedom of every other in accordance with a universal law’ (MM 6:237). When is one’s independence constrained by another’s choice in a way that would infringe upon this innate right? The answer here is not a matter of the content of the maxim on which the other acts, but rather of the private, ‘unilateral’ character of the other’s act. Property acquisition in a pre-political state of nature exemplifies such an act: where one claims property in (say) a piece of land, one thereby unilaterally claims a right to exclude others from possessing that land, and to that extent restricts the innate freedom of those others.

The personal acquisition of property is, Kant thinks, necessary in order for us to pursue our chosen life projects. It is then important to understand how such acquisition can occur without infringing upon the innate right to freedom of others. Kant’s answer is that ‘[i]t is possible to have something external as one’s own only in a rightful condition, under an authority giving laws publicly, that is, in a civil condition’ (MM 6:255). Specifically, the answer is the state. Following Arthur Ripstein (2009), we can say that the state solves three interrelated problems: the problem of unilateralism; the problem of indeterminacy; and the problem of assurance. First, the state, in representing the general and public will of its citizens, can overcome the problem of the unilateral acquisition of property by way of its public, ‘omnilateral’ determination of property rights – it therefore provides the context in which private ownership is possible without infringing upon the independence of others. Second, the state can authoritatively settle potential problems of indeterminacy (where exactly, for example, does the boundary between my land and yours lie?) that could otherwise only be settled by the respective reassertion of unilateral wills. And third, the state provides assurance: for Kant, I am not ‘under obligation to leave external objects belonging to others untouched unless everyone else provides me with assurance that he will behave in accordance with the same principle with regard to what is mine’ (MM 6:256). Kant believes such assurance can only be provided by a coercive, sovereign authority. These three roles for the state – i.e. the solving of the problems of unilateralism, indeterminacy and assurance – are distributed respectively to the legislature, the judiciary and the executive, delivering the Kantian case for a republican state, specifically.

But this deals only with ‘domestic right’ (i.e. conditions of external freedom between individuals). And while there is little interpretative controversy about the institutional solution to domestic right, we have already seen that the same cannot be said about ‘international right’ (i.e. conditions of external freedom between states). The present article, additionally, brings ‘cosmopolitan right’ firmly into view. As it will be understood here, cosmopolitan right refers to conditions of external freedom between political communities (be they states, or non-state ‘peoples’) and those private persons who arrive at the territories of such communities seeking to interact with them (whom I will refer to as ‘visitors’).

In its content, cosmopolitan right is ‘limited to the conditions of universal hospitality’ (PP 8:357). As I will understand this here, such conditions of hospitality entail that (i) a foreign visitor has the right to visit other political communities, for the purposes of striking up interaction (e.g. trade, migration and intellectual exchange) without that visit being treated as a hostile act, but that (ii) the receiving community has the right to refuse such interaction and turn the visitor away, provided that (iii) this ‘can be done without destroying him’ (PP 8:358).

Cosmopolitan right completes Kant’s concern with external freedom by overcoming ‘the exclusion of foreigners from the fold of moral respect, while at the same time securing a space for nations and groups to pursue distinct ways of life’ (Muthu, 2000: 24). As substantiated by the conditions of hospitality, I here take cosmopolitan right to have at least the following two implications. First, it protects political communities against unwelcome approaches of outsiders not covered by domestic or international right: importantly, it entails that the colonial acquisition of foreign lands is contrary to right. ⁴ Second, and more controversially, because it precludes communities turning away persons where doing so would lead to their ‘destruction’, I will take it to amount to a modest right of asylum. As Pauline Kleingeld (2012: 77) has put it, Kant ‘anticipates many of the refugee rights, including the principle of non-refoulment, that were established in the twentieth century’ (see also, for example, Benhabib, 2004: 35; Brown, 2010: 317). The principle of non-refoulment is one that can be found in various documents in contemporary international law. Article 33(1) of the 1951 United Nations Convention Relating to the Status of Refugees, for example, states that:

No Contracting State shall expel or return ‘refouler’ a refugee in any manner whatsoever to the frontiers of territories where his life or freedom would be threatened on account of his race, religion, nationality, membership of a particular social group or political opinion. (United Nations, 1951)

The understanding of cosmopolitan right I have set out here might be accused both of being too substantive, and of not being substantive enough. Some deny that cosmopolitan right amounts to a right of asylum, or indeed to anything publicly enforceable at all. For one recent example here, Christopher Meckstroth claims that Kantian hospitality ‘functions not as a freestanding positive claim demanding enforcement but as a way of ruling out specious justifications for war against those the traditional law of nations permitted one to label enemies’ (Meckstroth, 2018: 537). Specifically, Meckstroth understands Kantian hospitality as a rejoinder to the idea, put forward by Vitoria and Grotius, that ‘when non-state peoples reject trade with Europeans, they violate a sacred right of hospitality, committing an injury (injuria) or harm (laesio) that entitles Europeans to vindicate their right by force’ (Meckstroth, 2018: 550). Kant’s principle of hospitality says that although there is a right to visit, the community being visited also has the right to refuse interaction with the visitors; hence, such refusal does not constitute an injury to which conquest and plunder is a just response. Kantian hospitality, on this view, is understood to entail ‘no enforcement at all’ (and hence no institutions); rather, it ‘was framed just so that it could never be invoked to start a new war in the name of pursuing one’s rights’ (Meckstroth, 2018: 539).

At the very outset of part II of the ‘doctrine of right’ within the Metaphysics of Morals, Kant states that the ‘sum of the laws which need to be promulgated generally in order to bring about a rightful condition is public right’ (MM 6:311). Domestic, international and cosmopolitan right are there explicitly presented as three distinct forms of this public right. By contrast, Meckstroth’s interpretation of cosmopolitan right invites us to conceive of it not as a form of justiciable law, but simply as a ‘moral law’ which potential colonisers ought to recognise and heed. To understand cosmopolitan right in this way, however, is to ignore public right’s explicit concern to realise ‘outer [i.e. external] freedom limited by law’, rather than to articulate the moral maxims on which persons ought to act (MM 6:311). The latter is not sufficient for the realisation of external freedom because, inter alia, ‘before a public lawful condition is established individual human beings, peoples and states can never be secure against violence from one another’ (MM 6:312). Even if cosmopolitan right were to be limited to the precluding of colonialism, therefore, it is far from clear that there ought to be no enforcement, and no institutional implications. ⁵

The case for so limiting the scope of cosmopolitan right is anyway debatable. Meckstroth may be right that Kant’s primary intention with respect to cosmopolitan right was to reject contemporary understandings of just war and, in particular, to preclude colonialism (see also Muthu, 2000). But this much can be happily accepted without also needing to accept the stronger claim that cosmopolitan right does not entail ‘a positive claim to welfare for refugees or to unimpeded communication or to anything else’ (Meckstroth, 2018: 551). Admittedly, the case for reading a positive claim to asylum in Kant is not clear-cut. On the one hand, there is no reference to the notion that visitors cannot be turned away if it will lead to their destruction in discussion of cosmopolitan right in the Metaphysics of Morals. Yet elsewhere in Kant’s writing can be found the following:

Inhospitable stretches of the earth’s surface, such as the sea and the deserts that belong to no one, divide the community of human beings, but in such a way that ships in the one case and camels (the ships of the desert) in the other make possible a visit by one people to another. Whoever does this voluntarily can in any case be turned away, but not fought, by the inhabitants, whoever is involuntarily forced into it (a ship that seeks haven in a storm or the crew of a stranded ship) cannot be again chased into driving danger from the coast or the oasis in which he saved himself, still less can he be captured, but he must be able to find shelter until a suitable opportunity for his departure arises. (DPP 23:173)

It might be argued that this example hardly adds up to an anticipation of 20th-century refugee rights. After all, the ‘save haven’ example is particular in form: there is no mention of a duty to provide refuge to visitors suffering political persecution elsewhere, for instance, but only to the literal and immediate offering of harbour to those at a state’s territory who would otherwise perish. Taken strictly, therefore, Kantian cosmopolitan right cannot be invoked to (for example) claim an obligation for states to accept asylum seekers who are currently in refugee camps in another territory. Nevertheless, the logic of Kant’s argument seems extendable beyond the specific example of sailors in peril that he gives. It seems to imply, for instance, that if asylum seekers reach the territory of another state, and having done so have a convincing case that their life would – for whatever reason – be in peril were they to be turned away again, then the receiving state has an obligation to offer them refuge until such time until the threat of ‘destruction’ has passed.

I therefore agree with Kleingeld (2012: 78) that ‘it is reasonable to read [cosmopolitan right] as entailing refugee rights’ – at least some refugee rights, even if a narrower set than those characterising modern-day refugee practice. To consider this to be reasonable is not to claim to have settled the matter definitively – and for those who remain unconvinced, the argument that unfolds in the proceeding sections of this article can be understood conditionally.

A criticism of my understanding of cosmopolitan right may also come from an opposite direction, however. It might be claimed that cosmopolitan right isn’t constrained to relations between political communities and individual visitors, but also encompasses relations between individuals across borders (Flikschuh, 2000: 151; Niesen, 2007: 91). Some theorists also want to extract from (or put into) cosmopolitan right a wider range of rights, such as those contained within contemporary human rights discourse (Eleftheriadis, 2003). My line of argument in this article does not depend upon these more ambitious interpretations of cosmopolitan right; but nor does it depend upon their rejection. I therefore remain agnostic about such interpretations. For my argument to go through, I require only that cosmopolitan right includes the limited right to asylum I have suggested here.

The argument from states’ moral personality

For some, the institutional implications of international right follow by simple analogy with domestic right. Just as individual persons must leave an anarchic state of nature and enter a republican state, so states themselves must leave the international state of nature and enter under the coercive authority of a federal world government. Support for this simple analogy can indeed be found in places across Kant’s writings, such as when it is claimed that it is through a union ‘analogous to the union through which a people becomes a state’ that states can realise conditions of public right between themselves (MM 6:350). Across this section and the next, I consider what I take to be the two most prominent normative rejections of the idea that international right can entail a world government, which call into question differing aspects of the simple analogy. Each of the two forms of argument appeals, in a different way, to the notion of the state as a particular kind of person.

The first such argument is one that appeals to states’ distinctive moral personality, and is articulated most extensively by Katrin Flikschuh (2010). ⁶ This argument runs as follows. Like individual persons, states possess moral personality: they have a will, from which it follows that they can be held morally responsible, and which therefore allows them to be ‘appraised as individuals, who in their natural condition (that is, in their independence from external laws) already wrong one another’ by remaining in such a condition (PP 8:354). However, while individuals and states are both moral agents, with resultant moral obligations to leave their respective states of nature, ‘they are not tokens of the same type. The crucial difference between them is that individuals’ wills are juridically non-sovereign, whereas states’ wills are juridically sovereign’ (Flikschuh, 2010: 480). States’ wills, because they are sovereign and general, provide the institutional solution to the matter of domestic right. But precisely because states’ wills have this sovereign character, they cannot be compelled to enter into an equivalent sovereign body at the global level, since to compel them to do so ‘would amount to a denial of their distinctive moral status as belonging to that type of moral agent whose will is juridically sovereign’ (Flikschuh, 2010: 480). Indeed, not only can states not be forcibly compelled into a global sovereign, they ought not voluntarily to enter into one either, since this would still amount to the alienation of states’ distinctive moral personality.

States’ juridical sovereignty does not mean they can avoid obligations of international right, however – indeed, the opposite is the case:

insofar as we accept that as a predicate of their distinctive moral personality states ought not to (be made to) give up their sovereignty, states’ failure in turn to acknowledge the obligations that attach to their moral personality would be equally unacceptable. A moral agent who fails to acknowledge the obligations that attach to their moral status fails to treat himself as moral agent. (Flikschuh, 2010: 480)

This kind of argument is bolstered by the claim that states have, or can progressively develop, the capacity reliably to self-legislate their duties of international right. As Patrick Capps and Julian Rivers (2010: 245) put it, ‘Unlike human nature, which is flawed, the nature of states is, for Kant, reformable, and it is this point that holds the key to the viability of a confederation’. In this vein, cosmopolitan thinkers sympathetic to Kant have appealed variously in their work to the prospect or possibility of ‘responsible cosmopolitan states’ (Brown, 2011), ‘statist cosmopolitanism’ (Ypi, 2011), cosmopolitan ‘democratic iterations’ within states (Benhabib, 2006) and cosmopolitan ‘learning processes’ (Habermas, 2006). Such notions support the idea that states are able to come to autonomously recognise and act upon the demands of international right, absent sovereign power above them (see also Perreau-Saussine, 2010).

Cosmopolitan right, however, has self-confessedly not been the focus of the argument from states’ moral personality. While Flikschuh is keenly aware of the systematic and interlocking nature of Kant’s doctrine of right, she is nevertheless explicit that her ‘focus will be on the transition from domestic to international right’ (Flikschuh, 2010: 476). Let us assume that the argument from states’ moral personality succeeds as a reason for rejecting a world government when the focus is so constrained. What happens when we expand the focus to include cosmopolitan right?

Recall that in the domain of cosmopolitan right, the relevant interactions are not between state and state, but rather between state (or other political community) and visitors to that state. ⁸ This third, cosmopolitan context of interaction amounts to a third state of nature: it is a third context in which agents possess the capacity to infringe upon each other’s innate right to freedom by way of their unilateral wills (domestic and international right do not themselves overcome the lawless nature of interaction between states and visitors). Now, it is a premise of the argument from states’ moral personality that states possess something – i.e. juridical sovereignty – that other kinds of moral agent do not possess. Therefore, rather than a context in which all parties are juridical sovereigns (as in the international context), we now, at the cosmopolitan level, enter a context in which it may be the case that one party (i.e. a state) is a juridical sovereign and the other party (i.e. the individual visitor) is not. Since the individual visitor is not juridically sovereign, it seems that the argument from states’ moral personality must say – by analogy with individuals in the domestic case – that the visitor must enter into the jurisdiction of a global sovereign that can represent the global public will, resolve indeterminacy and provide assurance regarding matters of cosmopolitan right, while also maintaining that states must not enter into any such global sovereign.

Let me illustrate the challenge this presents with reference to the asylum aspect of cosmopolitan right in particular. A visitor arrives in the territory of a state and requests asylum – per the principle of hospitality, the receiving state ought only to refuse asylum if doing so will not lead to the destruction of the asylum seeker. But this in itself is only an abstract moral principle, not a legal determination – and it is the latter that right requires. Suppose the receiving state believes that it can indeed reject the asylum seeker without it leading to their destruction (this is precisely the assumption of ‘safe country of origin’ lists drawn up by various European states today, for instance). ⁹ In contrast, the asylum seeker reaffirms that they believe they will face their destruction if they are refused; that, they say, is why they are seeking asylum in the first place. Both parties here are interpreting the general principle of hospitality in the way that seems right to them; in other words, they both express their private, unilateral will. But what is required is a general, public will.

Can the confederal league of nations help here? I do not believe so. A confederal league is insufficient, first of all, because the visitor, as an individual, private person rather than a formal representative of another state, is not represented within the league at all – they are thus excluded from any purported public will which, with respect to cosmopolitan right, is therefore in fact merely the unilateral, private will of states considered as a collective. One might argue in response that such individuals can be considered indirect co-authors of a global public will on account of their respective state citizenships. Kleingeld takes this view, for example, stating that:

in a republic, those who determine the laws that are to enact cosmopolitan right are representatives who are elected by and accountable to their constituents. Thus, individual citizens can at the same be conceived as world citizens who co-legislate indirectly. (Kleingeld, 2012: 90)

One might claim that in the ideal league, no persons would be rendered stateless. But this cannot be a satisfactory response. It will always remain possible that disasters, human or natural, befall states such that all or some of their citizens become stateless and in need of asylum. The possibility of such events cannot be credibly idealised away, and since it cannot, all individuals, in a free league, remain potentially dependent upon a league of states within which their own will is not represented. ¹⁰ Individuals abstracted from states, as the agents for whom matters of asylum are particularly pertinent, are not appropriately represented as part of a global public will from within a confederal league – and since they are not, their external freedom remains to be finally secured.

Perhaps though, the idea of a confederal league can be tinkered with in a way that both respects the juridical sovereignty of states and recognises the need for individuals to be represented within a cosmopolitan public will. Such might be thought to be the aspiration of modern theorists who have offered various models of ‘cosmopolitan democracy’. In some of these models, a global parliament comprising two chambers is hypothesised: the existing UN assembly, suitably reformed, acts as the first chamber in which states are represented; and a second chamber is added in which individuals worldwide are represented (see, for example, Held, 1995). Such a global parliament might produce laws regarding asylum and other aspects of cosmopolitan right that a global judiciary and other agencies could interpret in specific cases (determining, for example, whether a particular country was a ‘safe country of origin’ for a specific asylum seeker in a specific instance). Where it was further stipulated that representation in such a parliament, and subjection to the jurisdiction of such a court, should be legally voluntary in the case of states but mandatory in the case of individuals, would this not sufficiently respond to the differentiated moral personalities at play?

Such an amendment to the idea of a league of nations cannot, in my view, succeed as a Kantian amendment, because so long as states’ membership remains voluntary, there fails properly to be realised the reciprocity which is inherent in the idea of external freedom. As Kant puts it, the idea of the innate right to freedom ‘already involves… innate equality, that is, independence from being bound by others to more than one can in turn bind them’ (MM 6:237). The institutional form presently under consideration would deny individual visitors this independence: states would be able to participate in definitively binding individuals to law, while retaining the possibility of avoiding being bound in turn. Consider what this means in practice: where asylum seekers appeal to a global court against the rejection of their asylum application by a receiving state, citing their likely destruction were they to be expelled, and win a judgement in their favour, the receiving state in question would in practice be legally entitled to withdraw from the league, ignore the judgement and expel the applicant anyway.

Of course, the argument from states’ moral personality says that states cannot morally do this – instead they ought, as a corollary of their appeal to their unique moral personality, to recognise their obligations in the domain of cosmopolitan right. Yet, practically, they would be entitled to do as they please. And there is no equivalent legal entitlement for individuals: cosmopolitan right institutionalised in this way affords asylum applicants no legal right to withdraw themselves from jurisdiction of the league and to continue to do what seems right to them instead by, for instance, simply entering the territory of the receiving state surreptitiously and evading domestic authorities. This lack of legal reciprocity between subjects of cosmopolitan right undermines the idea that an amended free league could represent the institutional conditions of external freedom, since individuals are not assured of ‘the independence of being bound by others to no more than one can in turn bind them’ (MM 6:237).

When the focus is on international right, the argument from states’ moral personality might be able to explain why states cannot be subjected to a global sovereign. But this is not sufficient to settle the institutional question, because it leaves unexplained how it can rightly be the case that states cannot be subjected to a global sovereign and yet individual visitors, as juridical non-sovereigns, must be. The appeal to states’ moral personality offers no reason for individuals to accept this legal inequality, even if it offers a reason for states mutually to recognise each other’s juridical sovereignty in the international context. We are seemingly at an impasse.

Flikschuh presents the argument from states’ moral personality as part of a ‘systemic solution’ to a ‘sovereignty dilemma’ for Kant (Flikschuh, 2010: 490). The dilemma, which arises in the international context, is that Kant is committed both to the inherently enforceable nature of right, and to the juridical sovereignty of states. The proposed solution, as we have seen, is to emphasise that as a corollary of states’ distinctive moral personality, states are obliged to self-enforce their international obligations as determined by the free league. This solution is ‘systemic’ because the institutional implications and possibilities for international right are considered in light of the institutional demands of domestic right, rather than merely by way of analogy with the latter. But what I have argued here, in effect, is that the argument from states’ moral personality isn’t systemic enough: it proposes an institutional solution for international right in light of domestic right, but it doesn’t look the other way, and similarly consider the prospects for cosmopolitan right in light of this supposed solution. When we do so, we seem to be led to replace one dilemma with another: instead of the clash between the inherently enforceable nature of right and states’ juridical sovereignty, we have the clash between states’ juridical sovereignty and the idea of equality inherent in the innate right to freedom.

The systemic solution offered by the argument from states’ moral personality is thus incomplete. What might a fuller systemic solution to the institutional question look like, one that took account not just of the interaction between domestic and international right, but also of cosmopolitan right? I cannot pursue this question at any length here, but one possibility is that it is actually world government that is best placed to provide such a solution. A sovereign, federal world government within which all subjects of cosmopolitan right (i.e. states and individuals) were represented could preserve equality between them, consistent with their innate right to freedom. A federal world government that administered only cosmopolitan right (and not international right, which could still be administered confederally between states) would impinge upon no more of states’ juridical sovereignty than was necessary in order to account for cosmopolitan right. Given that this solution would involve a partial restriction of states’ juridical sovereignty, it might be thought to be only a partial solution. A ‘partial solution’, however, is also how Flikschuh (2010: 472) describes the argument from states’ moral personality itself as a response to the original sovereignty dilemma. Such partial solutions are perhaps intrinsic to a systemic approach to the institutionalisation of public right.

States’ physical personhood: The argument from restricted competence

Let us now consider the second argument for a league of nations as the answer to the institutional question, again developed with particular focus on international right, and see if it can be sustained when cosmopolitan right is brought into view.

We can call this second argument the argument from restricted competence, and it runs as follows. Unlike domestic right, international right does not require institutionalisation in the form of a (world) republic because international right must achieve comparatively fewer things. To recall, at the level of domestic right, there are three discrete problems at hand: the problem of unilateralism; the problem of indeterminacy; and the problem of assurance. By contrast, according to the argument from restricted competence, at the level of international right there exists only one of these problems, namely the problem of indeterminacy. This claim relies on our accepting that ‘Kant’s discussion of conflict between states turns entirely on the right to engage in defensive war’ (Ripstein, 2009: 227). Assuming international right is confined to this issue, the problem of indeterminacy persists because there can be good faith disagreements between states about who is acting aggressively and who is acting in self-defence which need to be impartially resolved. But there are no problems of unilateralism or assurance because (i) states are not the sorts of entities that acquire property, unilaterally or otherwise and (ii) since they don’t acquire anything, there is no problem of assurance of acquisition to be settled. Therefore, international right has no need of a legislature, nor of an executive, since those agencies are institutional responses to the respective (and in this context non-existent) problems of unilateralism and assurance. All that is required is a judiciary, as a response to the indeterminacy problem.

The notion that states don’t acquire property might strike the reader as strange, given that states claim and defend territories, which we might intuitively think of as equivalent to property claims. But in the argument from restricted competence, it is supposed that:

[t]he state does not acquire its territory; its territory is just the spatial manifestation of the state … The state is always necessarily in possession of its territory, just as a person is always in possession of his or her body. (Ripstein, 2009: 228)

A further qualification is required, because it might be thought that even if a global judiciary is all that international right requires, such a judiciary ought still to be sovereign in its decision-making, meaning that it would still be possible to speak of the need for an ultra-minimal world government, rather than a free league. To avoid this conclusion, a further claim is made: the judiciary must not be sovereign, because for it to be sovereign would be to open up the possibility that a state be forced to war in order to defend another state, whereas ‘no other state has an enforceable right that others put themselves in danger to defend it’ (Ripstein, 2009: 230).

I take there to be two problems with the argument from restricted competence so described. The first problem is the central claim of restricted competence itself: a consideration of cosmopolitan right can render this problem clear, and I turn to this task imminently. The second problem is that there is no reason to suppose that a sovereign judiciary need lead to states being obligated to go to war on behalf of others. A global judiciary can make sovereign judgements without those judgements needing to involve the ordering of state X to defend state Y against state Z; it is no conceptually necessary part of resolving a case of indeterminacy regarding two opposing claims of self-defence that the court must compel a third party to enforce its judgement. Now, it might be claimed that without a method of enforcement, a judiciary evidences no practical sovereignty worthy of the name. But even if that is true, there is no good reason why third-party states must be that method, rather than a global executive – indeed, the idea that states would be obligated to go to war in defence of others seems to arise only because the initial claim of restricted competence itself renders a global executive unavailable.

What then of that claim? I want to grant that the argument from restricted competence might succeed when considering international right in abstraction, but argue that it cannot sustain when bringing cosmopolitan right into view. Moreover, when cosmopolitan right is indeed brought into view, the argument from restricted competence points instead towards a world government as the answer to the institutional question. ¹¹

The initial argument from restricted competence rests upon the notion that states’ territories are morally analogous to persons’ bodies. If such an equivalence pertains, then it follows that ‘Anyone who enters its [i.e. a state’s] territory without its authorization enters the state itself; should such a person overstay his welcome, he commits a wrong analogous to battery, rather than one analogous to theft’ (Ripstein, 2009: 228). In infringing upon the bodily integrity of the state, one infringes upon the state’s innate right to freedom, by analogy with persons’ innate right to bodily freedom in the individual case. Bodies are entitled to occupy space; that another may wish to occupy that same space, and yet now cannot, is not an undue restriction of the other’s freedom, but is instead only an instance of the way in which, by occupying space, one agent can legitimately affect the context of the freedom of choice of another. Occupying space is not a restriction of another’s freedom, whereas interfering with a body that already occupies that space is. Therefore, ‘If I invade the space you occupy, you can push me away’, as a ‘hindrance to a hindrance to freedom’ (Ripstein, 2009: 378).

By analogy, it should follow that if someone, without authorisation, enters the space a state occupies, that state can ‘push away’ the ‘invader’ too. Yet the question arises as to how any such understanding of states’ territories can be rendered compatible with the asylum aspect of cosmopolitan right. This aspect, in setting out the circumstances in which states cannot turn visitors away from their territories, in effect (if we accept the analogy with persons’ bodies) delineates the circumstances in which a state, which may wish to expel the visitor, must nevertheless accept ‘battery’. But to say this would be to strike at the very heart of the idea of an innate right to freedom, which is itself the foundational notion of Kant’s doctrine of right. If states’ territories really were morally akin to persons’ bodies for Kant, then he would have to have said that a state, in repelling asylum seekers, never infringes upon the latter’s freedom, but merely (adversely) changes those asylum seekers’ context of freedom of choice, by occupying space that would otherwise be available as refuge.

That Kant does not say this, and also cannot credibly be understood as legislating for battery, should lead us to conclude that states’ territories in fact differ from persons’ bodies in a morally relevant way. Yet it is also true that for Kant, property represents external means by which agents pursue private ends, and since states, as public entities, ought not to have private ends, it is not at all obvious that states’ territories should be understood, from a Kantian perspective, as property (Mikalsen, 2011: 312). The way forward here, I believe, is to emphasise states’ distinctive kind of embodiment. If states’ territories are analogous to persons’ bodies in one respect (namely that they can each be described as the ‘spatial manifestation’ of their respective subjects), they are also disanaologous in at least the following two ways. First, states’ territories, unlike persons’ bodies, are contingent. It is not necessary that a state’s territory takes the specific form that it does. A person’s body, by contrast, is importantly physically determinate (on account, for example, of one’s genetic makeup). We could not be anything other than the bodies that we are; ¹² and without certain of our body parts, we will not be at all. By contrast, while some area of territory is necessary to the very existence of a state, there is no particular patch of territory that for any one state must be part of its physical manifestation. ¹³

Second, states’ bodies, unlike persons’ bodies, are artificial. They are at root an imaginative act, just as is, indeed, ‘the state’ itself. It is possible to quit this imaginative act, and to conceive of the land that makes up a state’s body as, simply, land. To put it another way, ‘state body’ and ‘patch of land’ are differing senses of the same referent (Frege, 1948). We switch between these senses depending on the context: someone discussing international relations may find the former sense pertinent; an archaeologist, by contrast, will be unlikely to. For the latter, the idea of this patch of land as the body of the state is not useful or important; that sense of the referent, therefore, drops away entirely in such a context. It is of course equally possible to conceive of a person’s body in a different sense: as physical matter, for instance. For a surgeon performing a complex operation, this may well be the most useful sense. But the surgeon never lets the sense that this is also someone’s body drop from view entirely, since that sense remains clearly important. Nor, I suggest, does (or ought) anyone else who may at times have cause to think of someone’s body primarily in terms of physical matter.

The relevance of states’ contingent and artificial embodiment to the institutional question becomes clear once we move beyond international right to focus on cosmopolitan right. The essential contingency of states’ embodiment makes it conceptually and practically meaningful to ask whether there are reasons to resist states’ exclusionary claim to their territories. In the asylum aspect of cosmopolitan right Kant anticipates such a reason, by way of two claims: all human beings having ‘a right to be wherever nature or chance (apart from their will) has placed them’ (MM 6:262), and their having original ‘possession in common of the earth’s surface’ (PP 8:358). The relation of the former (the ‘right to be somewhere’) to the latter (‘common possession’ of the earth) admits of differing interpretations: for some, Kant proceeds from the right to be somewhere to the idea of common possession (e.g. Flikschuh, 2000); for others, it is the notion of common possession of the earth which grounds the right to be somewhere (e.g. Pinheiro Walla, 2016). Whichever it may be, that there does for Kant exist a right to be somewhere is of significance for our present purposes: it means that when an asylum seeker presents themselves at the border of state, the state’s contingent embodiment runs up against the necessary embodiment of the asylum seeker, who makes a claim to entry by way of appeal to this right.

Where such a claim is determined to have merit under cosmopolitan right, what has happened? On the side of the asylum applicant there has been a successful possessive claim, a claim of the right to be within the state’s territory. But what about on the part of the state? Here the artificial nature of the state’s embodiment is important; in my view, the only way to rationalise what has happened is to let the sense ‘territory as state body’ drop away entirely, and instead take up a more individualistic sense, namely ‘territory as claimed possession of the state’s existing citizens’. From this perspective, the exclusionary possessive claim to the state’s territory made by the existing citizens – which we now can understand as a property claim, since individual persons do make property claims over land – is successfully challenged by the rival possessive claim (i.e. the right to be somewhere within the state’s territory) of the visitor. ¹⁴

Refusing to switch to this individualistic sense of state territory with respect to cosmopolitan right, and instead continuing to conceive of the state as a singular, embodied person, necessitates absurd conceptual gymnastics. One such feat would be to say that when an asylum seeker makes a successful possessive claim to enter the state, a part of that body then ceases to be part of the state’s embodiment for the length of time the asylum seeker remains in the state. But which piece of land? The piece that the visitor occupies at any one time? This implies, bizarrely, that the body of the state is constantly morphing from one shape to another as refugees move about. Moreover, the pieces of land that refugees occupy do not cease to be under the jurisdiction of the state – and presumably, however we conceive of the body (territory) of the state, we will want to understand it as co-extensive with its legal jurisdiction.

An alternative conceptual contortion is to attempt to conceive of the state’s territory both as its embodiment and simply as land that the refugee claims on grounds of the right to be somewhere. I said previously that a surgeon, when performing an operation, may have two senses of the same referent in view: ‘person’s body’ and ‘physical matter’. These two senses, however, are not in explicit tension with each other: a person’s body is physical matter. By contrast, the two senses ‘body’ and ‘land over which a rightful possessive claim can be made by someone else’ are, from a Kantian perspective, in necessary tension with each other: one’s control over one’s own body is at the heart of the innate right to freedom, and yet the latter sense implies the infringement of such control. Both senses of the land that makes up states’ territories cannot be kept in view at the same time without contradiction.

I suggest then that the imaginative fiction of the state having a body cannot hold at the level of cosmopolitan right. We must disassemble the artificial person in order to be able coherently to represent cosmopolitan political activity. When international right is considered in abstraction, we might be able to conceive of states’ territories as (contingent, artificial) bodies, and we therefore might be able to accept the claim of international right’s restricted competence. But this is not enough to resolve the institutional question in favour of a league of nations. If cosmopolitan right necessitates ceasing to conceive of states’ territories as states’ bodies, and conceiving of them instead as the property exclusively claimed by the states’ citizens (claims potentially challenged by a particular class of visitor), then, as with domestic right (even if not with international right), the ‘problem of unilateralism’ with respect to the acquisition of property is again a live issue; if a state’s citizens want to vindicate the acquisition of their state’s territory, they must be prepared to enter into a global-level law-making body that can render such property claims appropriately ‘omnilateral’ (cf. Ypi, 2014).

Since the problem of unilateralism is a live issue, then the problem of assurance is too: in the particular cosmopolitan context we have been interested in here, we might say that if a potential asylum seeker cannot be assured that a legitimate claim to the right to be somewhere will be respected by the state they arrive at, then they are themselves under no obligation to respect that state’s claim to its territory and may, for example, seek to enter the state anyway. The solution to the problem of assurance, to recall, is sovereign executive power. A problem of indeterminacy was never denied by the argument from restricted competence even at the level of international right, but clearly there will recur a problem of indeterminacy with respect to cosmopolitan right too, since the general principle of hospitality will always require determination in specific instances according to the particularities of any one case. The solution here is a global court competent to resolve such cases.

All three problems that lead Kant to offer a sovereign state as the solution to domestic right, then, reappear in cosmopolitan right. The argument from restricted competence itself therefore points in the direction of a world government when the full scope of the doctrine or right is brought into view, since we come to realise that the competences required at the cosmopolitan level are not restricted.

The compatibility of cosmopolitan right and federal world government

I have argued that neither the argument from states’ moral personality, nor the depiction of states’ physical personhood entailed by the argument from restricted competence – both seemingly plausible as Kantian arguments for a league of nations when considering international right in particular – can be sustained when the concerns of cosmopolitan right are brought into view. I have also claimed – tentatively in the first case, and more firmly in the second case – that the two arguments themselves point in the direction of a world government when cosmopolitan concerns are brought to the fore. In closing, I want briefly to consider the counter-claim that world government is not conceptually compatible with international and/or cosmopolitan right. This has certainly been taken to be Kant’s own view by some of his interpreters. Nevertheless, world government is, in my view, compatible with the elements of Kant’s political philosophy.

Consider first the following claim to the incompatibility of cosmopolitan right with world government:

If Kant had adopted the possibility of a world government, the problems of citizens and strangers [i.e. cosmopolitan right] would not have occurred. All human beings would have been considered as members. It is therefore obvious that this cosmopolitan right could arise in Kant’s writings only after he had dropped the possibility of such a world government. (Mertens, 1996: 332)

It might be thought, however, that this only kicks the conceptual problem back to the level of international right. For while – as was made clear in the first section of this article – it was not in fact Kant’s view that political communities needed to be organised as states in order to be relevant subjects of cosmopolitan right, the subjects of international right are indeed taken to be sovereign states specifically. But where there exists a sovereign world government – however minimal – then states (defined partly by their own sovereignty) would no longer exist, since they could no longer themselves be fully sovereign under a world government. Interpreters of Kant often take his statement that a ‘state of nations… would be a contradiction’ (PP 8:354) to be making this kind of point when considering a world government in respect of international right. Moreover, Kant explicitly states that it is only by a voluntary league, and ‘not a federation (like that of the American states)’, that ‘the idea of a public right of nations [can] be realised’ (MM 6:351).

When Kant claims in Toward Perpetual Peace that a ‘state of nations’ would be a contradiction, he says more specifically that it would ‘contradict the supposition’ of international right. This is true enough in one sense: if there did not exist a world of separate states, then there would be no problem of international right to be solved. But to suppose that it follows from this that the institutional solution to this problem must entail the continued existence of states is to confuse the initial subjects of right with those subjects as they would be transformed in the corresponding rightful condition. The world as it faced Kant was a world containing separate sovereign states. As such, there necessarily exist problems of international and cosmopolitan right, for which states are relevant subjects. But to say that states are relevant subjects of right is not to say they must remain states when the relevant rightful condition is finally institutionalised. Compare with domestic right: there the relevant subjects are individuals in the state of nature. In order to institute domestic right, however, they must cease being individuals of that kind, and instead become state citizens. A similar thing can be said about states: while the initial subjects of (for example) cosmopolitan right are, today, states and visitors, that does not mean they will remain such once the rightful condition between them has been institutionalised. The view offered here is that they may instead become, respectively, largely autonomous federal units, and visitors who, even if they may find themselves ‘stateless’, now possess a world citizenship that allows them to pursue their external freedom at the cosmopolitan level.

Kant himself may in any case not have meant his statement about a contradiction in the idea of a ‘state of nations’ to be a claim about conceptual incoherence. Kleingeld’s (2004) argument that for Kant the ‘contradiction’, such as it is, relates to the premature ‘fusing together’ of separate states, and that such a view is compatible with the idea that a sovereign world government remains the institutional ideal, seems to me plausible. After all, Kant does say in Toward Perpetual Peace – the same text in which he makes the ‘contradiction’ claim – that a world government is ‘correct in thesi’ (PP 8:357), which would be a strange thing to say if he really did think it was conceptually incompatible with his system of right.

What, finally, of the substance of the claim in the Metaphysics of Morals that a voluntary league is the only form in which international right can be realised? This too turns out not to be a conceptual claim, but rather an empirical one. Here again the ideal of a world government seems to be affirmed in principle: were it not practically impossible, in Kant’s view, to govern a world government effectively in practice, then a voluntary league would not be the only form in which international right could be realised. Indeed, Kant seems rather to equate the ideal of perpetual peace with a world government: precisely because, for him, a world government is not practicable, the idea of perpetual peace ultimately remains an ‘unachievable idea’ (MM 6:350). In its stead, we must take the voluntary league as our best available approximation.