1.

After many years of neglect, Kant’s political thought has returned as a major point of reference and contention in political philosophy.Footnote ¹ He is now regularly cited in debates about justice, property, political legitimacy, war, immigration, and so forth. I shall not speculate about the reasons for the long period of neglect. As for the reasons for Kant’s revival, the critique of utilitarianism in the Anglo-American world by Rawls and others has clearly been a major factor, but equally important has been the growing interest in cosmopolitan ideas of justice and legitimacy. Whereas Kant’s position on the internal constitution of the state may look dated and conservative, on questions of international order he can appear progressive, indeed visionary, to contemporary eyes.

Of course the exact nature of Kant’s cosmopolitanism is open to debate. His interpreters have located him at different positions on the spectrum between statism and strong cosmopolitanism, much depending on how we understand the requirement that he places on states to enter a league to settle their disagreements without recourse to war. There has also been a fair amount of controversy about his remarks on hospitality to visitors, and what they might entail in relation to immigrants and refugees. This is an issue to be addressed later in the article. But first it is noteworthy that both sides in the debate between statism and cosmopolitanism agree in characterising Kant as an opponent of any form of nationalism. To many students of Kant, this has seemed self-evident. When Isaiah Berlin lectured on ‘Kant as an Unfamiliar Source of Nationalism’, he began by presenting the theme he was going to develop as wholly counterintuitive:

Despite this, Berlin goes on to argue, we can trace a connection between Kant’s idea of the rational will as the source of moral imperatives and the much later idea of collective self-determination, which in the hands of his successors, Herder, Fichte, and others, takes the form of romantic nationalism. But Berlin is clear that this involves a series of intellectual leaps that would have horrified Kant himself. Kant, he says, ‘would have repudiated this misbegotten by-product of his deeply rational and cosmopolitan philosophy’ (Berlin Reference Berlin and Hardy1996: 244). And in this he is followed by all recent commentators. Particularly for those nearer to the statist end of the interpretive spectrum, one of the great attractions of Kant is that he appears to offer an account of state legitimacy that makes no reference at all to national identity.

2.

My aim is to throw some doubt on this picture of Kant as deeply opposed to any form of nationalism. Of course in one respect the question ‘was Kant a nationalist?’ is anachronistic, since nationalism as a doctrine only emerged in the nineteenth century, even if partially anticipated in the eighteenth in the work of Herder, and in some respects Rousseau. We also need to keep clearly in mind that at the time when Kant wrote, the land that would eventually become Germany was divided into a mosaic of small states, most of them still nominally subject to the authority of the Habsburg empire. The powerful kingdom of Prussia, which included Königsberg where Kant lived throughout his life, was the main exception in this respect. In consequence, the idea of creating a unified German state was on no-one’s political agenda in the eighteenth century – even Herder’s proto-nationalism was largely apolitical. In that sense, Kant could never have been a German nationalist even if he had wished it. An all-embracing German nation-state was not on the horizon of possibilities until well after his death.Footnote ² And Kant was aware of this himself. Observing in his lectures on anthropology that ‘the Germans have no national pride’, he proposed that ‘the cause perhaps lies in the fact that they are divided into many small states and do not constitute a whole nation’ (LA, 25: 1409).Footnote ³

So to ask ‘was Kant a nationalist?’ is asking the wrong question. I want to propose the more nuanced question ‘did it matter to Kant that states should be nation-states, and if so what implications does this have for other parts of his political philosophy, including what he might have said about immigration?’.Footnote ⁴ I begin by observing that the German terms ‘Volk’ and ‘Nation’ appear frequently in Kant’s writings, usually rendered into English as ‘people’ and ‘nation’. Kant is quite clear that a nation is not the same as a state and draws attention to the confusion that may arise if they are used as equivalents, commenting for instance that the ‘right of nations’, or what we would now call international law, should really be designated the ‘right of states’, since it is the actions of states that it is meant to govern (MM, 6: 343). It is unfortunate therefore that commentators on Kant sometimes use ‘nation’ in what we might call its ‘United Nations’ sense, when it is actually states and their behaviour that is being discussed.Footnote ⁵ Furthermore, a nation or people is not merely a collective term for those who live under the jurisdiction of the same state. One indication of this is that Kant also applies the term to people he describes as ‘savages’, living in lawless freedom outside of the state. Then, when thinking of people who do live under the auspices of a state, Kant characterises them in a way that sees them as united by more than just a bond of political allegiance. He says that ‘those who constitute a nation can be looked upon analogously to descendants of the same ancestors (congeniti) even though they are not’ (MM, 6: 343). His meaning is not entirely clear, but a plausible reading is that here Kant is recognising that nations think of themselves as communities of descent, even though this is not literally true for every member, since some members might be immigrants or the children of immigrants.

So Kant uses the idea of a people or nation as a group with a shared identity formed by (real or imagined) common ancestry, and he is obviously very alive to, and interested in, national differences. In his lectures on anthropology, he spends some time commenting on the diverse national characters of the leading European nations – the French, English, Spanish, and so forth. He presents these differences as cultural in nature, and he denies that they can be explained by contrasting forms of government. As he puts it, ‘to claim that the kind of character a people will have depends entirely on its form of government is an ungrounded assertion that explains nothing: for from where does the government itself get its particular character?’ (APPV, 7: 313). He also rejects the reductionist view that national character merely reflects the physical influence of climate and soil, claiming that when peoples migrate en masse, their distinctive character persists, merely adapting to new circumstances. Finally, he notes that national characters only find their proper expression under conditions of freedom, citing the modern Greeks, whose present ‘fickle and groveling’ character is to be explained by the harsh oppression of their Turkish rulers (APPV, 7: 320).

At this point, supporters of the mainstream interpretation of Kant may say that these (purportedly) empirical observations may testify to his curiosity about all things human, but they are strictly irrelevant to his political philosophy. According to that philosophy, reason compels us to enter into a political relationship with those with whom we unavoidably interact, regardless of any cultural similarities or differences. Jeremy Waldron puts it more strongly still: ‘this conception of political community rests upon an assumption of mutual antipathy or least diffidence; it is more like a Hobbesian conception. The idea is that we form political community with those with whom [we] are likely to fight, rather than with those whom we already like and who are, in their identity, already like us’.Footnote ⁶ Moreover, the form of the state can be deduced by asking what is necessary for us to live together on terms of equal freedom. It might then seem, as a corollary, that where the boundaries of the state are placed, in terms of which territory and which people fall under its jurisdiction, is an irrelevance, so long as that form is maintained. Normatively speaking, Kant should remain silent on the state’s national character – whether its citizens belong to a single nation, or to several, or to none at all – so long as its laws and institutions meet the conditions that he lays down under the heading of Public Right.

That, however, is not the position that Kant actually adopts. He implicitly attributes to peoples a right of self-determination, in the form of a right not to be forced against their will into association with another people. This is shown, for example, in his remarks on colonialism in The Metaphysics of Morals, where he argues that it is a major injustice for colonisers to force indigenous peoples into civil association with them as a way of legitimising their unconsented settlement (MM, 6: 353).Footnote ⁷ It is also shown in the passage in Perpetual Peace in which he criticises the European practice whereby the rulers of one state could acquire another ‘by inheritance, exchange, purchase or gift’. A state, he says,

Why does Kant use the organic metaphor of the rooted tree when describing the state here? The question is not about the formal conditions for something being a legitimate state. Rather Kant is trying to explain the nature of the wrong that is committed when a state that had previously been independent is, at the whim of its ruler, transformed into a subsidiary branch of another state. To do that he needs to invoke a fuller conception of the state as not just a legal institution but a social formation that has developed over time, with its own language of public life, culture, practices, and so forth. These are the roots that would be lost if it were annexed to another state whose history and culture were quite different. Kant follows this by saying that the transfer would ‘terminate its existence as a moral personality’. His meaning is not entirely clear, but a plausible interpretation is that the state through its laws has made concrete a particular conception of justice, which is then destroyed in the process of transfer and replaced by an alien legal regime that does not represent the combined will of the citizens. (Kant says that such a transfer ‘contradicts the idea of the original contract, without which the rights of a people are unthinkable’.) If these readings are correct, Kant must be committed to the idea of the self-determination of peoples, at least in the form of a prohibition against their being absorbed by another state against the people’s will.

To throw further light on the nature of Kant’s objection to state annexations, whether these occur through military invasions or through rulers treating their states as commodities or bargaining chips to be used when forming alliances and so forth, it is helpful to consider Arthur Ripstein’s alternative reading. This focuses on the wrong that is committed when a ‘self-certifying’ legal system is disrupted by an external force that has not been authorised by the system’s subjects:

This understands the wrongness of annexation in terms of a formal break in the continuity of a legal system – the replacement of one rule of recognition by another, to use Hart’s vocabulary (Hart Reference Hart1961: ch. 6). Whether the substantive rules of the system – the laws governing property rights, for example – also change is another matter. We can imagine an annexation that leaves them completely unaltered.Footnote ⁸ We can also conceive of an annexation whereby the legal system is improved, as measured by Kant’s equal freedom standard. Suppose that at the time when the two Germanies were divided, the Federal Republic had been able to annex the Democratic Republic and had then replaced parts of the latter’s legal system with laws that better guaranteed rights of private property and freedom of speech. Although we might prefer that the two Germanies had united voluntarily, to describe the annexation merely as a wrongful act privileges procedure at the expense of substance.

Ripstein supports his reading of Kant by laying stress on the forceful nature of annexations. ‘The legal order has authority over those subject to it because it has solved the problem of right for them; another legal order that seeks to impose itself forcibly interferes with the provision of a rightful condition’ (Ripstein Reference Ripstein2021: 52). He extends this to the case of ‘bloodless invasion’ which, even if it preserves private rights, ‘wrongs the invaded country and its citizens by subjecting them to what must be, from their perspective, arbitrary force’ (Ripstein Reference Ripstein2021: 54). But every legal order will use force to establish itself in the face of those within its jurisdiction who would rather remain independent or join some other state (and who will therefore regard this particular use of force as ‘arbitrary’). So on Ripstein’s view the use of force is permissible as a way of establishing an effective legal system, but once the system is in place no more force may be used for this purpose, even if the aim is to improve a rightful condition or to extend its scope (as in the two Germanies example). This seems like an arbitrary privileging of the legal status quo.

As Stilz points out, the basic problem with Ripstein’s reading of Kant is that it cannot explain the moral basis of the state’s right to independence: why the citizens themselves have a right to resist being incorporated against their will into an alien legal order. It fails to recognise the interest that people have, not just in having their private rights properly protected, but in playing an active role, as citizens, in determining the regime of justice under which they will live. This gives them a claim that the state should act as their representative, a claim that is denied even in the hypothetical case in which America bloodlessly annexes Canada but then grants the former Canadians equal rights in the enlarged USA. However, this still leaves it an open question how we should understand ‘the people’ when a claim of this kind is being made. Are they simply the set of individuals who up to now have been united merely by common subjection to a legal order? Or should they also be understood as a nation in the historical-cultural sense of the term?

3.

As we have already seen, Kant’s language when he objects to rulers treating their states as commodities suggests that he has this richer, national conception of the state in mind. We might learn more if we knew how he would react to a real case in which one hitherto independent state was forcibly annexed by another. In fact, Kant was a witness to just such an event: the dismemberment of Poland by Austria, Prussia, and Russia. It is worth reminding ourselves briefly of the relevant historical facts. Up until 1772, Königsberg formed part of eastern Prussia, an enclave surrounded by Polish territory (though with access to the Baltic). In the First Partition of Poland in that year, Prussia took a slice of territory that connected East Prussia to the rest of Prussia to the west, so it ceased to be an enclave. A further chunk was taken in 1793, and then with the Third Partition in 1795, Poland vanished as a political entity, not to be reconstituted as an independent state until 1918. Had Kant chosen to walk north-eastwards from Königsberg after 1795, the frontier he would have first reached would no longer be with Poland but with Russia.

Kant could hardly have failed to notice these momentous events going on around him. His essay on Perpetual Peace is said to have been inspired by the Peace of Basel, the first Treaty of which involved Prussia withdrawing from its coalition with Austria and Russia to partition Poland and settling its western border with France. What is remarkable is that he never refers explicitly to these developments, either in his published writing or in his correspondence. You might think that the break-up of Poland was a clear case of a state’s moral personality being terminated, and something therefore which Kant should have criticised. In The Metaphysics of Morals, he explicitly prohibits ‘either a war of extermination (bellum internecinum) or of subjugation (bellum subiugatorium) which would be the moral annihilation of a state (the people of which would either become merged in one mass with that of the conqueror or reduced to servitude)’ (MM, 6: 347).Footnote ⁹ How could he have failed to see that his own state, Prussia, was involved in just such a war?

Might Kant have thought that there were such serious deficiencies in the way that Poland was constituted politically that it could not claim the right to independence that every state could otherwise assert? Enlightenment thinkers generally took a negative view of the Polish-Lithuanian Commonwealth, regarding its peculiar constitution dominated by the noble class as oppressive of the common people (Davies Reference Davies1981; vol 1, ch. 10) and Kant appears to have shared this view.Footnote ¹⁰ In his brief comments on Poland in his lectures on anthropology, his main criticism of the constitution is that it lacks an authority able to enforce the law effectively. The Poles, he says, ‘want freedom and law, but no executive power, a demand that is completely absurd’ (LA, 25: 1185). But although he regards this as a significant defect, he nowhere suggests that Poland should not count as a state at all. In general, he did not believe that a state’s right to autonomy depended on it having already adopted the republican form that he regarded as fully realising his principles of public right.

Although Kant’s essay on Perpetual Peace never mentions Poland explicitly, it has been claimed that he had the partition of that country directly in mind while writing it (see Kupś Reference Kupś2022). He first states, in the fifth preliminary article of perpetual peace that ‘no state shall forcibly interfere in the constitution and government of another state’, which might seem to deliver a clear verdict in Poland’s favour and against the partitioning powers. However, Kant then considers a possible counter-argument, perhaps echoing claims currently being made by Prussian and Russian propagandists (Kupś Reference Kupś2022: 16-18):

But he immediately goes on to narrow down the circumstances in which such a defence of intervention could be given:

Might Kant have considered Poland as a case where internal division made external interference permissible? Not plausibly. In 1794, between the Second and Third Partitions, came the national uprising which united Poles in an unsuccessful attempt to expel the occupying Russian and Prussian forces (see Davies Reference Davies1981: vol. 1, ch. 18). The only internal enemies were senior figures accused of collaborating with the invaders. Poland was indeed ‘merely struggling with its internal ills’.

So why did Kant not condemn outright such a clear ‘violation of the rights of an independent people’? It is possible that he did so in private discussion, but he makes no mention either of Poland or of the war in any of his published texts or correspondence from that period.Footnote ¹¹ Was he perhaps worried about censorship (though how would that explain his silence in letters to friends)? In 1794, he was in trouble with Frederick William II for what he had said about religion in Religion within the Boundaries of Mere Reason, and in reply to the king had promised that he would ‘abstain entirely from all public lectures on religious topics. … and not only from lectures but also from publications’ (C, 11: 530). Perhaps he felt that it was prudent tacitly to extend this promise to commentary on Prussia’s foreign policies. In November of that year he wrote that ‘I hope that once peace is established, which seems to be near, the limits of what an author is allowed to write will be defined more precisely, so that one can feel secure about what is permissible’ (C, 11: 531). It has also been suggested that this insecurity may explain the otherwise rather puzzling paragraph that introduces Perpetual Peace (see Kuehn Reference Kuehn2001: 383).

I have found only two places in which Kant deigns to acknowledge the fate that has befallen Poland. As we have seen already, in the various versions of his lectures on anthropology, Kant regularly incorporates a section in which he discusses nations and their characteristics. In early versions, the Poles are included as a distinct nation. In the 1784-5 series, for example, they are the sixth nation to be listed, and described as ‘vivacious, but without much wit and inventiveness’, ‘frivolous and therefore like to choose the French as models’, poor hosts and bad at paying their debts, and so forth (most of Kant’s national pen portraits are somewhat malicious) (LA, 25: 1412). However, in the final version of the lectures, published as a book in 1798, the list of nations to be discussed is shortened to just five (the French, English, Spanish, Italian, and Germans), Russia is described as insufficiently developed to be included, and, Kant cryptically remarks, ‘Poland is no longer at this stage’ (APPV, 7: 319). He seems here to imply that since Poles have ceased to possess a state, it is irrelevant to discuss their national character.

The second place is a short postscript that Kant wrote for a bilingual Lithuanian-German dictionary in 1800. Lithuanian had been the main language of the Grand Duchy of Lithuania which formed part of the Polish-Lithuanian Commonwealth before the Partition, and it was also spoken in parts of Prussian Poland that had now been absorbed into Prussia. Kant is clearly concerned that the language might be in danger of dying out under the new regime, hence his endorsement of the dictionary. He values its preservation for two reasons. One is scientific: he says that ‘the still unmixed language of a very old tribe of people’ can help us understand ‘the ancient history of the migrations of peoples’ (Kant Reference Kant, Kant and Louden2007b; 8: 445). The other is practical: he says that ‘every small people in a country’, or what we would today call an ethnic or national minority, should be taught its language in its purest form in schools and churches. So he wants the Lithuanian speakers who are now incorporated into Prussia to have access to the best version of their national language. He clearly felt that their identity as a national group was under threat now that the protection of the Polish state had disappeared.

I have dwelt on the Polish case for the light it might shed on the nature of Kant’s commitment to state independence. Had Kant felt inclined to comment on the dismemberment of Poland, he would perhaps have revealed whether it was only the interference by one state in the internal affairs of another he cared about, or also the political integrity and self-determination of the Polish people. If he had held a view of the kind attributed to him by Ripstein, he would have condemned the successive annexations on procedural grounds, as disruptions of an established legal order, but once the dismemberment was complete and all the land occupied by Poles had been incorporated into well-functioning states, he would have no further reason to object to the outcome. Conversely, had he attributed to the Polish people as a people the ‘original right to unite itself into a commonwealth’, then he would have regarded their enforced political separation as an ongoing injustice that should be rectified, by force if necessary. It is therefore very unfortunate that, as we have seen, he seems to have been loath to say or publish anything bearing directly on this question that might prove politically controversial.

4.

It is time to take stock. What have we discovered about Kant’s attitude to the nation-state? We know that he distinguished nations from states and was sufficiently interested in nations and their peculiarities to discuss them repeatedly in his lectures on anthropology. We also know that he believed in the integrity of states, and their rights of self-determination, and was strongly opposed to their status being changed either by force or by the machinations of monarchs. But I have not yet been able to forge a direct link between these two themes. I have not shown, in particular, that Kant thinks that all states should aspire to be nation-states, in the sense that their subjects ought to be united by ties of national identity as well as political obligation. Although his remarks about the socially rooted character of the state are suggestive, they fall short of a conclusive demonstration that Kant believed in what we would now call national self-determination.

Can we go further here? I do not want to suggest that Kant can be treated as the prototype for late twentieth/early twenty-first century liberal nationalism, any more than he can be treated as a prototype for nineteenth-century romantic nationalism. The most convincing arguments for liberal nationalism rely on the fact that we live today in democratic welfare states, which can be stable only if grounded in the solidarity that a shared national identity can provide (see Miller Reference Miller1995: ch. 4). Kant’s more austere idea of the Rechtsstaat at least arguably does not need such support. It is not democratic, and it is not committed to social justice as we would understand it, even if it can incorporate measures to avoid citizens becoming dependent through falling into poverty, as Ripstein (Reference Ripstein2009: ch. 9) has argued. So to find Kantian reasons for favouring states that have a national basis, we need to look elsewhere.Footnote ¹²

The place to look is Kant’s argument in favour of a plural world of many states, as against a single world government, which Kant describes as a universal monarchy, thereby implicitly assuming that such a government could not take the form of a republic. Describing such a government as despotic, but liable to collapse into anarchy, he points to national differences of language and religion as nature’s way of ensuring that humanity is guided towards the best outcome, namely a voluntary league of nation-states committed to international law as a way of resolving their disputes (Kant Reference Kant and Reiss1971: 113-4; 8: 367).Footnote ¹³ In contrast, as he puts it in a draft of Perpetual Peace, ‘unity of languages and of religions and of types of government would soon make for migration and a melting together of peoples, hence universal monarchy, which is harmful’ (D, 23: 170). Whereas critics of nationalism are wont to blame nationalism for giving rise to international conflicts, Kant’s much more balanced view is that as nations increasingly adopt republican constitutions, going to war will require the consent of their citizens, who will be very reluctant to bear either the economic or the human costs of doing so. So, Kant says, ‘nature wisely separates the nations’, for this is the prophylactic against the imperial ambitions of any one of them, but it also unites the nations, when they see it is in their mutual interest to create machinery for resolving their disputes. ‘The spirit of commerce sooner or later takes hold of every people and it cannot exist side by side with war’ (Kant Reference Kant and Reiss1971: 114; 8: 368).

It is worth adding here that although Kant himself refers to language and religion as the cultural features that divide nations from one another, the anti-imperialist, pro-pluralist argument that he makes does not require that national identities should take any specific form. What matters is that people should regard their political community as sufficiently distinct from others that they find great value in preserving its independence and resisting dismemberment or annexation. In principle, this could be achieved by what has come to be known as ‘constitutional patriotism’ where the focus of people’s loyalty is the state’s formal constitution and the principles it embodies.Footnote ¹⁴ Whether a narrowly political identity of this kind could really serve the purpose assigned to it in Kant’s argument is an open question, but it is important to say, looking in the other direction, that there is nothing in Kant that speaks in favour of a closed, ethnically based form of national identity. How culturally ‘thick’ or ‘thin’ a people’s specific identity needs to be to unite them in favour of independence is an empirical question that he does not address.

So, to sum up, for Kant it is valuable for the state to have a national basis, not for reasons of domestic justice, but in order that it should form part of an international order that can over time become law-governed and peaceful, but without becoming unitary and despotic. Having such a basis makes states indigestible by larger states that might otherwise wish to absorb them. Knowing that they will encounter resistance by people who wish to preserve their language and religion and the institutions that embody them, states that seek to turn themselves into empires by conquest will hold back. So although the threat of war will remain present until an inclusive league of nations is formed, this is still better than ‘an amalgamation of the separate nations under a single power which has overruled the rest and created a universal monarchy’, which Kant describes as a ‘soulless despotism’ (Kant Reference Kant and Reiss1971: 113; 8: 367).

5.

If for the reasons just given Kant thinks it desirable for states to have a national basis, and to preserve their distinctiveness against any pressure towards cultural assimilation, what does this imply for his (hypothetical) view about immigration – hypothetical because this is something about which Kant had virtually nothing to say? (His mention of migration in the passage cited above as one of the consequences of cultural homogenisation is quite exceptional in this respect.) This should not surprise us. Although migration as a phenomenon clearly existed in Kant’s world, and states sometimes attempted either to encourage or to deter people from moving, they did not have migration policies as we would understand them today – policies setting quotas for different categories of immigrants, policies to encourage immigrants to integrate and become citizens, and so forth. It was therefore not a topic on which a political philosopher would expect to pass judgement.

Despite this, Kant’s name has regularly been invoked in recent philosophical work on immigration and especially on work concerning the state’s duties to refugees. Several authors have attempted to use Kant’s political philosophy to ground extensive duties of this kind, taking as their starting point Kant’s claim, in his short discussion of ‘Cosmopolitan Right’ in Perpetual Peace, that a visitor who arrives on a state’s territory cannot be turned away if this would lead to his destruction.Footnote ¹⁵ In the same place, however, Kant also says that although visitors must be received without hostility, they can only stay permanently with the consent of the host community, which for reasons of self-protection that community may refuse to give (he endorses the very restrictive policies that China and Japan had adopted towards Western traders). It is also noteworthy that when Kant gives examples of the people who must be allowed to stay, he cites ‘a ship that seeks haven in a storm or the crew of a stranded ship’ and assumes that their need for refuge is only temporary. The beleaguered traveller ‘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’.Footnote ¹⁶ All of this suggests that the duty Kant is supporting is a much more limited duty to grant temporary refuge to those who arrive involuntarily by sea or land (or who having arrived voluntarily are now unable to leave safely).

The most prominent recent defence of the expansive reading of Kant on immigration has come from the German scholar Karoline Reinhardt, with whom I have had an exchange in the pages of Studi Kantiani, so here I will just briefly rehearse our disagreement (see Reinhardt Reference Reinhardt2022; Miller Reference Miller2022). According to Reinhardt, ‘[Kant’s] account allows for a high degree of cross-border freedom of movement and provides the argumentative resources for unconditional refugee protection’ (Reinhardt Reference Reinhardt2022: 167). In particular, she aims to show that Kant’s position on cosmopolitan right implicitly commits him to a demanding version of the duty to accept refugees, as defined by the 1951 Geneva Convention.Footnote ¹⁷ The state is obliged to admit everyone who would otherwise face a threat of serious harm in their home countries, whether to body or to mind, and regardless of how many are in that position. To reach this conclusion, Reinhardt interprets the destruction (Untergang) to which a person must not be exposed so as to include the attack on moral personality that occurs when a state persecutes one of its citizens. The argument, then, is that when a refugee is denied her claim to asylum and forced to return to the state where she is subject to persecution in one form or another, she is in danger of being ‘destroyed’, not necessarily in the physical sense, but in the sense that she is no longer able to act on her own understanding of who she is – that is, she is liable to suffer from some kind of mental breakdown.

Now I fully endorse the obligation of states today not to return a refugee to a place where she is liable to be persecuted – that is not what is in question here. What is in question first is how we should read Kant himself, and then second what it is reasonable to impute to him by way of entailment in the case of immigration. And there are three respects in which Reinhardt’s interpretation goes well beyond anything that Kant might have said.

First, even setting aside the difference between physical death or injury and damage to moral personality, there is a relevant difference between destroying someone by, say, forcing them back out to sea in an unseaworthy vessel, and exposing them to persecution by another state, where direct responsibility for the harm done to them lies with the actual persecutors. Kant takes a discriminating view about the imputation of consequences when there is an intervening agent, as we know from his famous treatment of the assassin who comes to your door asking whether you are sheltering his intended victim (Kant, Reference Kant and Kant1996b: 612-13; 8: 427). When you answer him truthfully, you are not then held responsible for the murder that follows, whereas if you lie, but the victim is killed anyway, you are: people are held responsible for the consequences of wrongful acts, but not for doing what duty requires. This leaves it an open question how consequences are to be imputed in cases where an action is neither forbidden nor required: in the case we are considering, the state is neither duty-bound to allow the visitor to stay nor duty-bound to require him to leave. It would, though, be implausible to hold that whenever an action of mine has the unintended effect of enabling another to act wrongly, I should be held responsible for the outcome of that wrongful act. Only a simple consequentialist (which Kant as we have just seen is certainly not) could believe that. A garage-owner should not be held responsible for the damage that ensues when he sells a can of petrol to an arsonist pretending to be a stranded motorist. Likewise, the state cannot be held responsible for the persecution that another state may inflict on a person it has turned away. It might have a reason to allow the person to stay in such a case, but it does not have an obligation, given that it will not be the direct cause of the harm that ensues if the person is required to leave – in contrast to the case where expulsion will directly expose the person expelled to physical danger (such as shipwreck).Footnote ¹⁸

Second, the examples that Kant gives show that he has in mind cases in which very small numbers of people – the crew of a ship, for example – are in need of shelter. Reinhardt, in contrast, wants to argue that the duty to admit refugees extends to cases in which millions of people are involved, right up to the point where there is no physical space left in which they can be accommodated (Reinhardt Reference Reinhardt2022: 175). But clearly this makes a big difference in terms of the impact that the arrival of newcomers may make on the integrity of the state – a point I shall return to shortly. A limited duty not to turn away the physically endangered brings no such risk with it. Here, we can look for an analogy with the individual duty of rescue, which can be accepted as a positive duty of justice precisely because the occasions on which we are required to act on it are expected to be few and far between – so it imposes no significant constraint on our personal autonomy (see my discussion in Miller Reference Miller2020). In much the same way, a nation-state’s right of self-determination is not significantly infringed by requiring it to offer sanctuary to a small number of endangered travellers, whereas it would be if it were unconditionally required to host millions of refugees.Footnote ¹⁹

Third, Kant makes it clear that the duty owed to the arriving stranger is quite narrow. As he puts it, ‘the stranger cannot claim the right of a guest to be entertained, for this would require a special friendly agreement whereby he might become a member of the native household for a certain time’ (Kant Reference Kant and Reiss1971: 106; 8: 358). In other words, he will need to look out for himself and hope that someone will behave charitably towards him if he is in need.Footnote ²⁰ Here, there is a fairly stark contrast with the duties that states today owe to the refugees whom they admit. Because the legitimacy of a modern state requires that it should adequately protect the human rights of everyone who falls under its jurisdiction, refugees once accepted are owed protection of a long list of human rights, including rights such as those to housing and health care that may be quite expensive to fulfil. They may also be entitled to stay for a period of years, even in many cases indefinitely. So the cost of meeting asylum claims today is potentially very much greater than the virtually zero cost of Kant’s requirement that a visiting stranger must not be treated ‘with hostility’ when arriving on the state’s territory, where hostility in this context means treating him in a way that leads directly to his destruction – either by attacking him, or by forcibly ejecting him when it is unsafe to do so.Footnote ²¹

6.

I have signalled already that I am not trying to defend Kant’s position vis-a-vis the modern practice of refugee protection, which is clearly far more humane. Instead I am highlighting the gulf between what Kant in fact says, and what modern interpreters have sought to extrapolate from these texts. But suppose Kant had actually been confronted with the prospect of mass migration on the scale that many European societies have had to face in recent years: would he have shifted his position to accommodate the changed circumstances, or would he have had reason to hold firm?

This is where it matters how Kant viewed the state, not just from a formal perspective, but also, as it were, sociologically. If, as I have argued, he saw it as a nation-state, held together not only by legal constraints but also by bonds of language, religion, and common ancestry, then inward migration on a large scale becomes potentially problematic. Depending on who the incoming migrants are, and the content of the prevailing national identity, their arrival may change the state’s character in ways that the existing body of citizens may reasonably be unwilling to accept. Recall his objection to monarchs transferring states to one another as if they were commodities, which ‘contradicts the idea of the original contract, without which the rights of a people are unthinkable’. Although the transformation of a state by mass immigration may seem less dramatic than the state being annexed to some other state, the structural change involved seems equally in violation of the idea of a determinate body of people governing themselves over time through the rule of law.Footnote ²² Kant of course did not believe that everyone living under the state’s jurisdiction must be an active citizen with voting rights. So he might have been open to the suggestion that immigrants should be allowed to come in simply as passive citizens, or as what we would now call denizens. But even that would require the consent of the citizen body, who might want to use cultural or other criteria of selection so as to maintain their state’s integrity as a political community.

There is one place at which Kant does actually refer to immigration, and it might appear that what he says there contradicts the interpretation I have just given.Footnote ²³ He remarks that:

By ‘the lord of the land’ Kant means the person or body that is sovereign, and he assumes here as he does elsewhere that this is not constituted democratically by all of the subjects (otherwise his comment makes no sense: if the subjects are going to look askance at immigration, they will not decide to authorise it). So in general whether or not to encourage immigration remains at the discretion of the sovereign, provided that it does not impinge on the ownership of land, and the question that then remains is whether this discretion extends to cases in which the scale of immigration would make it socially transformative. Here, it is important to keep in mind that all of the sovereign’s decisions must be subject to the test that they could be made by the united will of the people. As Kant puts it ‘what a people (the entire mass of subjects) cannot decide with regard to itself and its fellows, the sovereign can also not decide with regard to it’ (MM, 6: 329). So there must be a contrast between an unpopular decision to admit a particular group of immigrants, which the sovereign is entitled to take, and a decision to admit immigrants on such a scale that the entire character of the society is altered, which on my reading of Kant would fall outside the terms of the implicit contract on which the state is founded. A decision of the latter kind is therefore one that the sovereign has no right to make.

I believe there is a better Kant-inspired response to the issue of large-scale refugee movements than the one proposed by Kant scholars such as Reinhardt and Kleingeld, who defend an open-ended right of admission. As we know, Kant thought states were duty-bound to attempt to create a league or congress that would enable them to resolve international disputes over territory or trade without going to war with one another. Since mass movements of refugees, if uncontrolled, can also be seen as a threat to peaceful co-existence, it would be a natural extension of the aims of such a league to create a mechanism for controlling such movements. This would make the protection of refugees into a collective task shared between a number of states, which would agree among themselves as to how the task was to be divided up, covering both the process of vetting asylum applications and housing those who qualified to receive it. However, membership of any such scheme would also remain voluntary, meaning that any member-state would retain the right to opt out if its citizens felt that it was being asked to carry too heavy a burden. So under this interpretation of Kantian moral cosmopolitanism, the duty of hospitality, in circumstances of mass migration, becomes the duty to enter into an association of states to regulate migration, to negotiate in good faith within the association, and then to carry out what has been decided – say to process and/or resettle a certain number of refugees – subject to the condition that membership in the association remains voluntary, with states able to withdraw if unreasonable demands are placed on them by the league.

7.

Let me quickly summarise the argument I have been attempting to make. I have not tried to squeeze Kant into a nationalist box of some kind, which as I pointed out at the beginning would be thoroughly anachronistic. I have however tried to show that the idea of the nation plays a significant role in his political thought. It conditions the way in which he thinks about the integrity and independence of states, revealed especially in his opposition to involuntary transfers and annexations. I lamented the fact that Kant appears to have sworn himself to silence on the matter of Prussia’s involvement in the break-up of Poland, which might otherwise have revealed more about the significance he attached to national identity. I then went on to argue that the nation-state must form the context for a Kantian view of immigration, and that recent interpreters seeking to bring Kant into line with progressive views today have ignored that context as they perform their stretching exercises. It is fine to argue for a generous policy on refugee admissions, but quite misleading to enlist ‘the sane, rational, liberal internationalism of the great Königsberg philosopher’ in support.