2. Provisional Right and the State: The Standard View

To begin, recall that Kant’s political philosophy is organized around the Universal Principle of Right (UPR). The UPR states that it is both necessary and sufficient for the rightness of an action or a state of affairs (Zustand) that it be consistent with the freedom of others under universal law (MM, 6: 230–1). Remaining in the state of nature (Naturzustand) is wrong ‘in the highest degree’ (im höchsten Grade) (6: 307–8) insofar as it violates this principle. Therefore, we have an obligation to leave the state of nature and enter into a civil condition.Footnote ⁶ Indeed, Kant’s Postulate of Public Right states precisely that, when interaction with others is unavoidable, ‘you ought to leave the state of nature and proceed with them into a rightful condition’ (6: 307).

For Kant, the civil condition (der bürgerliche Zustand) is that in which everyone is under ‘a general external (i.e., public) lawgiving accompanied with power’ (MM, 6: 256), and in which ‘the will of all is actually united for giving law’ (6: 264). On the standard picture, law enters to limit each person’s freedom to her rightful, or compossible, freedom. Whereas exercises of freedom in the state of nature are unilateral, resulting in disagreement with respect to the distribution of rights and a lack of assurance that they will be respected, law resolves disagreement and provides assurance, instantiating an omnilateral will (compare Ripstein Reference Ripstein2009 and Waldron Reference Waldron1996). By the postulate of public right, we must enter such a condition.

But we cannot discharge this requirement alone. Realizing a civil condition requires, among other things, coordination with numerous other persons and institutions. We must be subjected to common laws issued by a legislature, adjudicate disputes by a common arbiter (a judiciary) and ensure that its judgments are enforced by a sufficiently powerful executive. In unfortunate circumstances where we do not receive adequate cooperation from external factors, Kant views our failure to discharge the postulate of public right – our failure to enter a civil condition – as faultless. In such conditions, our acquired rights are provisional, rather than peremptory or non-existent:

In this passage Kant makes clear, first, that possession becomes peremptory only in an ‘actual civil condition’. Given what has come before, this means in a condition in which everyone’s will is in fact unified for lawgiving under an adequate power. Call this the Omnilaterality Requirement. Second, in advance of securing coordination for entering a civil condition, unilateral claims to possess external objects are provisionally rightful so long as they ‘accord with the possibility’ of a civil condition, a condition in which the omnilaterality requirement is satisfied.Footnote ⁷ The introduction of the distinction between provisional and peremptory right in this context motivates the standard view that the category of provisional right loses practical relevance once states come to be. For if (1) right is peremptory (not provisional) wherever there is an actual civil condition, and (2) states are civil conditions, then (3) where there are states, there are sites of peremptory rights. But then PT is false.

This is too quick. Kant defines a civil condition in terms of its satisfaction of the omnilaterality requirement. Whether it is correct, therefore, to identify a state with a civil condition requires better understanding that requirement. We get some purchase on the issue by focusing on the contrasting case of the particular or unilateral will. To approach a more positive characterization, we might additionally notice that Kant uses this term interchangeably with three others: a general (allgemein) will (e.g. MM, 6: 294), a common (gemeinsam) will (e.g. 6: 257) and the unified (vereinigt) will of all (e.g. 6: 302, 314).

The first naturally recalls Rousseau’s Social Contract, where the general will plays a central, but somewhat ambiguous, normative role. Interpreters have distinguished between at least three ways of understanding this central notion. According to the first, the content of the general will is specified in a purely proceduralist way (see e.g. Rousseau Reference Rousseau1997: II.vi.5, II.vi.7). Once a people chooses a decision rule (e.g. majoritarian or dictatorial rule) to govern their interactions, the outcome of that rule with respect to a particular issue just is identical with the general will. The second option is a pure substantivist position, according to which the general will corresponds to an independent conception of the common good (see e.g. Rousseau Reference Rousseau1997: II.iv.7, II.vi.2). The third option, a mixed view, becomes plausible insofar as there is textual evidence for both of the pure options. On such a mixed view, an objective common good imposes constraints on either the outcome or the inputs of the community’s decision procedures to ensure that their outcomes remain within certain bounds (see Sreenivasan Reference Sreenivasan2000). But procedures (or outcomes) within these permissible bounds are authoritative and express the general will. Whereas it is uncertain which of these views Rousseau accepts, Kant’s own characterization of the omnilateral will seems to imply that he adopts either the second or the third understanding (compare Flikschuh Reference Flikschuh2012: 34, 38–41).Footnote ⁸

Kant rules out a pure proceduralist notion when he writes that the omnilateral will is ‘united not contingently but a priori and therefore necessarily’ (MM, 6: 263) and contrasts it with other kinds of willing: unilateral, bilateral and multilateral. Any actual procedure will typically be contingently multilateral (as in majority rule). Even in the best case, procedures exhibit contingent unanimity. By contrast, Kant holds that the united will’s necessary a priori unity makes it ‘the only will that is lawgiving’ (ibid.). This characterization implies that omnilateral will is given antecedent to any positive assertion of an external right or any positive institution (even if it must in the end be represented by one). More specifically, he treats it as an idea – a regulative concept that cannot be met with anywhere in experience (L-NR, 27: 1393, Drafts of DR, 23: 220, MM, 6: 258–74). From the perspective of Kant’s omnilateral will, even overwhelming majorities (‘the will of everyone but one’) are merely unilateral with respect to the one(s) they exclude. But because a unilateral will cannot serve as a law to an autonomous agent, and because the acquisition of external rights in the state of nature must appear to give law unilaterally, the ‘rational title’ of these rights lies in the idea of their being united or compatible with the dictates of an omnilateral will. Such a will can be united a priori because it is based on principles of pure practical reason (e.g. the UPR), principles willed by each rational agent merely insofar as she is rational. The civil condition represents an external lawgiving that actually realizes these principles, and it is in such a condition that possession becomes peremptory.

On the standard view, a state satisfies the omnilaterality requirement just in case (a) it allows that acquisition be permissible (does not violate the postulate of practical reason with regard to rights) and (b) it is governed by institutions that satisfy certain functional criteria (does not violate the postulate of public right).Footnote ⁹ On this view, we leave the state of nature insofar as we find ourselves under a government that specifies one law to govern all within a geographically bounded territory, adjudicates disputes solely on its basis and enforces decisions with irresistible force.Footnote ¹⁰ The standard view holds that there are no further substantive constraints that the state must meet in order to satisfy the omnilaterality requirement. As Japa Pallikkathayil puts it, the solution to the problems in the state of nature:

In sum, ‘[p]rovisional rights can be made conclusive only through the establishment [of] an institution that defines the scope of our rights and enforces those rights, in other words, the state’ (ibid.). Ripstein appears to concur. For him, we enter a civil condition ‘simply by being subject to laws’ (Reference Ripstein2009: 198; cf. 225). Whereas in the state of nature what is externally right is determined by each person’s power, in the civil condition what is externally right is determined by public institutions. The transition from a situation of provisional to a situation of peremptory right occurs just as soon as the relevant institutions are realized. The standard view is plausible to the degree that states as we know them satisfy the omnilaterality requirement. So far as they do, peremptory right is widely actualized in our world. Perhaps this is why major accounts accord to the idea of provisional right little more than passing attention in their remarks on the transition between the state of nature and the civil condition.Footnote ¹¹

For all its intuitive plausibility, two considerations militate against this view. The first is that just international and cosmopolitan institutions are external conditions on a domestic state’s capacity to render rights peremptory. Existing international bodies are inadequate and human nature and our distribution over great distances makes it impossible to reform them in a fully adequate way. Now, one might grant this, and think that it need not pose significant problems for the standard view. For the standard view can allow that any given state’s territorial claims are provisional with respect to outsiders, while maintaining that, nevertheless, states render the claims of insiders peremptory. Unfortunately, the second consideration shows that this escape route is blocked. For there are internal (domestic) conditions of peremptory right that are not satisfied by existing states and that are in principle impossible to satisfy given the frailty of human nature. That Kant places these two conditions on peremptory right entails accepting PT and rejecting the mainstream view. The next two sections provide textual evidence – all of it drawn from Kant’s mature political philosophy – showing that each of these requirements is a necessary condition of peremptory right.

3. The Right of Nations and Cosmopolitan Right

On the standard view, the existence of the domestic state suffices to make the rights of its citizens peremptory. The first strain of argument against this view begins by recalling that rights in the state of nature are said to have provisional status owing to their unilaterality, indeterminacy and lack of assurance. It then notes that these same conditions apply to states with respect to their territory. That is: (1) a state’s acquisition of territory is necessarily unilateral (at best multilateral) from an outside perspective;Footnote ¹² (2) it is unclear how much territory a state can acquire and there are no rules to govern disputes between states; and (3) states are threatened by outer violence, such that they lack assurance that their claims to territory will be respected. In sum, even if actual states unite the will of all citizens, Kant does not restrict the principles of right to citizens. With respect to non-citizens, any one state’s activity in isolation looks to fall short of the omnilaterality requirement.

To see that Kant takes universalism seriously, note that he locates the title of anyone’s acquiring an external right to a corporeal thing in the common ownership of the earth by all human beings. In Kant’s words, appropriators base their acquisitions ‘on an innate possession in common of the surface of the earth and on a general will corresponding a priori to it’; only this ‘permits private possession’ (MM, 6: 250).Footnote ¹³ Kant is quite clear that it is human beings, not co-nationals, that own the earth in common. Therefore, satisfying the omnilaterality requirement demands that our claims be consistent with the like claims of outsiders, not just the claims of compatriots. Kant himself notes that a nation’s claims to territory cannot be peremptory until relations of right extend to the ‘entire human race’ (6: 266). If rightful relations do not extend this far, then it remains open that current claims to ownership, to governance, etc. are incompatible with a perfectly general and common will. Indeed, appropriation is prima facie inconsistent with the omnilaterality requirement because all are originally common owners of the earth, and so appear to have standing to reject any act of acquisition to which common owners do not unanimously consent. But since the state’s territory is composed of the union of its citizens’ property (plus its public property which derives from theirs), the stateless and those in other states can make claims against nations based on this title of common possession.

Now perhaps the standard view would wish to assert that the omnilaterality requirement is discharged as long as the international order allows acquisition (division of the earth) and state formation (Byrd and Hruschka Reference Byrd and Hruschka2006b). Provided this is true, the claim of each state to its territory is compatible with the like claims of other states. But this response is at odds with the way Kant explicitly argues. Consider a passage in which he lays out the structure of public right:

Kant says that if the principle of outer freedom limited by law (the UPR) is lacking on any level (domestic, international, cosmopolitan), then the framework of the others is undermined. Importantly: when the framework is undermined in this way, our circumstances are circumstances of provisional, not peremptory right.

To sum up: individuals and nations are structurally analogous, and face similar problems, to which they require similar solutions. Whereas individuals ought to enter a state, states ought to enter an association of states. Until an association of states of the right kind comes about, any rights that states might assert hold only provisionally (compare Gregor Reference Gregor and Gregor1996: 15–16).

Might the standard view respond that existing international institutions (WTO, UN, EU, etc.) are sufficient for rendering rights peremptory? There is reason to think not. For the kind of association up to the task must be generally universal and hold throughout time. But this condition is impossible to satisfy, given the demands of global governance.

Kant requires of satisfactory international institutions that they be (1) genuinely universal and (2) capable of governing stably across the earth. For his part, Kant doubts our prospects for realizing these conditions simultaneously. Successfully meeting the first requirement means failing to satisfy the second, and vice versa. Our current international institutions fall short, as must any human-governed institutions we can envision. Kant’s scepticism runs so deep that he deems the end of his project, perpetual peace, an unachievable idea. In advance of such institutions – however closely we may approximate them – external rights are merely provisional. The true conditions of perpetual peace – so too of peremptory right – are impossible to realize.

We have thus seen that international institutions must be realized for states’ claims to territory to attain peremptory status. Moreover, if the ‘principle of outer freedom limited by law’ remains unsatisfied at any level (domestic, international or cosmopolitan), ‘the framework of all the others is unavoidably undermined and must finally collapse’ (MM, 3: 611). Read naturally, this means that a failure to realize institutions of the right kind at the cosmopolitan level undermines rights’ peremptory status at the international and domestic levels, and that an analogous failure at the international level undermines rights’ peremptory status at the domestic level. But if so, it would seem that rights everywhere are unavoidably provisional. Minimally, a domestic state is insufficient for rendering claims to outer objects peremptory, pace the received view.

In spite of the strong textual evidence for these conclusions, one might think that there is an easy reply. The reply admits that states’ rights with respect to one another and the stateless must be provisional, but holds that nevertheless individuals’ rights under one state are made peremptory by its giving of law.Footnote ¹⁴ Put differently, a state’s rights with respect to outsiders might be provisional due to inadequate international and cosmopolitan institutions, at the same time that individuals’ rights under it are peremptory. Although this flies in the face of the natural reading of Kant’s claims that inadequacy at one level infects all the rest, perhaps these passages can be given a different reading. I do not wish to deny this possibility. Still, in the next section, I argue that domestic conditions on peremptory right are more demanding than is typically thought. For there are substantive norms of freedom that constrain a state’s lawgiving activity, as commentators have tended to appreciate. What commentators have missed is that failure to satisfy these norms also prevents states from realizing peremptory right. Thus, even if the argument in this section fails to establish PT (because it fails to rule out the possibility that states are internal but not external sites of peremptory right), the next section establishes that domestic states that fall short of the true republic are not themselves sites of peremptory right. Together with the results of this section, this entails that all claims to acquired rights are provisional.

4. The True Republic and the Right of States

This section argues for two conclusions, both of which point in the direction of PT. The first is that, even on the standard view, a state’s capacity to render rights peremptory depends upon its formal lack of private purposes. But any actual state is governed by human beings, and actual state agents use their offices to pursue private purposes with depressing regularity. For Kant, this fact makes the problem of instituting a rightful condition the most difficult to solve. But if a rightful condition is necessary for peremptory right, then this difficulty makes hard walking for the standard view. The second is that, contrary to the standard view, a state’s commitment to pursue only public purposes by means of its coercive apparatus is a mere necessary condition for converting provisional into peremptory rights. In addition, Kant holds that states must make freedom the basis of their deployment of coercion. So far as states coerce subjects for reasons other than freedom, the distribution of rights they recognize is merely provisional.

Recall that, on the received view, states that meet the formal conditions of minimal legality are civil conditions, i.e. they satisfy the omnilaterality requirement. Since meeting this requirement suffices for peremptory right, PT is false. We might note that this is a curious position, given that on the Kantian account the state necessarily comes into being through force. What distinguishes the state from other merely powerful agents? Certainly, if I were to approach everyone in my geographical region with my lackeys, and demand to receive a substantial portion of the existing claimed property for the protection of the rest (and this in the name of all), my actions do not appear to obligate you (compare Huemer Reference Huemer2013). What is distinctive about states?

One plausible answer is that states are essentially public entities (TPP, 8: 352). As Ripstein puts it:

Whereas individuals have private ends, states and public officials constitutively lack private ends. Their constitutive lack of private ends means that states and their agents can act in the name of everyone, even when they are exercising coercion against the express consent of some. Publicity sustains a state’s ‘character as a rightful condition’ and distinguishes it from other powerful but unilateral agents.

It is reasonable to object that the historical record is replete with abuses of power, where public offices are exploited to pursue private purposes. Why should the fact that an ideal state acts publicly permit actual states (that do not so act) to lay down peremptory rights and duties? Kant shares the worry. In his ‘Idea for a Universal History’ (as well as in the later Anthropology), he observes that, if human beings are not subject to a master laying down law for all backed with sanctions (i.e. a sovereign), they tend to abuse and dominate others (IUH, 8: 23, Anth, 7: 327). But all eligible masters are imperfect humans, likely to abuse their power unless constrained by another master. But that master must also be a human being, and so on.Footnote ¹⁵ Though public offices might be theoretically constrained to pursue public purposes, this does not ensure that the individuals occupying those offices will act accordingly.

These considerations suggest that, concerning actual states’ capacity to satisfy the constraint of publicity, Kant rejects the mainstream view that most minimally decent states satisfy it as a matter of course. Structural problems (e.g. the fact that we can only ever imperfectly constrain public officials to comply with their duty to pursue by means of their coercive power only public purposes) make it unlikely that most state actors act purely in pursuit of public goods, even when they are constrained by a constitution, and even when they are subject to sanctions for noncompliance. If publicity is required to secure peremptory rights, and publicity is no necessary part of statehood, then the standard view is false.

A defender of the standard view might argue that this misses the point. The point is not that state agents invariably act publicly in the common interest, but that to have a state is in the common interest (e.g. Flikschuh Reference Flikschuh2012 passim). We can flesh out the idea as follows. For Kant, the UPR is a law of reason and generates the postulate of public right. These principles issue infallibly from each individual’s capacity for practical reason (alternatively, rational will). Since all human beings have a will that legislates the UPR, and since the UPR requires a state (by way of the postulate), any actual state can be identified as the object of omnilateral agreement. There may be obligations to bring the laws closer to what justice requires, but that is no argument against thinking that states make rights peremptory. So understood, the view is not that the state is bound always to act publicly in accordance with an omnilateral will (though this an ideal against which states can be evaluated), but that it is always in line with the omnilateral will to have a state. Moreover, part of what it is to have a state is to treat the distributions of rights under it as peremptory (compare Ripstein Reference Ripstein and Ellis2012: 65, MM, 6: 371–2).

It is certainly true that Kant thinks that something normatively important happens insofar as power is consolidated in a domestic state. But the claim in question here is that what happens in these cases is that provisional rights become peremptory. Alas, there is strong textual evidence against this claim. Consider the following passage, which directly follows Kant’s claim that changes to the form of the existing constitution need to be brought about not through revolution, but rather through reform by the sovereign (MM, 6: 340):

Here, Kant makes clear that the idea of the original contract constrains ‘the constituting authority’ to gradually bring about the kind of government required by reason (compare TPP, 8: 349). For Kant, the original contract is the idea of a people unifying itself for the giving of law (MM, 6: 315). Following the contract tradition that he inherits, the original contract serves as the criterion of state legitimacy, as well as the criterion for identifying improvements for existing states (ibid.). But, unlike his predecessors, Kant does not think that this contract is an actual historical event, depending for its shape upon what actual parties agree to; his account is, as Helga Varden convincingly argues, ‘non-voluntarist’ in that way (Varden Reference Varden2008). Instead, history gives us good reason to believe that governments as we know them originated in force ‘to bring about the submission of the people’. Governments begin when power is concentrated in a sufficiently high degree to force coordination on one set of solutions to the problems that exist in the state of nature. So far as it is true that all existing solutions are defective, the idea of the original contract makes a stringent demand: such governments, having achieved this submission, must make ‘freedom the principle and indeed the condition for any exercise of coercion, as is required by a rightful constitution of a state in the strict sense of the word’. Only a rightful constitution, that is, one that limits its use of coercion to what is necessary to secure compossible freedom is literally a state. Until such a state is realized, ‘no absolutely rightful condition of civil society can be acknowledged, but only provisional right within it’.

On its face, this conflicts with the standard view, which would have it that the mere exit from a condition of complete external lawlessness is sufficient for entering a rightful condition. Though it may be true that satisfying the omnilaterality requirement demands a state, it also demands a particular kind of state, namely a pure republic, one that conditions coercion on extending freedom and protecting rights (innate and acquired).

Attention to Kant’s definition of the pure republic reinforces just how demanding this condition of peremptory right is. The republic is one of four ‘conditions’ or ‘states’ that Kant identifies in the Anthropology (see also Ellis Reference Ellis2005: 114, Ripstein Reference Ripstein2009: 338, Wood Reference Wood2014: 98), which was published for the first time the year after his Doctrine of Right:

1. A. Law and freedom without force (anarchy)

2. B. Law and force without freedom (despotism)

3. C. Force without freedom and law (barbarism)

4. D. Force with freedom and law (republic) (Anth, 7: 330–1)

Having introduced these categories, Kant repeats that ‘only the last combination deserves to be called a true civil constitution’ (Anth, 7: 331). This makes sense of his claim in the Doctrine of Right that a pure republic is a necessary condition of making rights peremptory. Only a pure republic – not despotic and imperfect states wherever they might be – counts as a true civil condition, and Kant is quite clear in the Doctrine of Right that ‘only in a civil constitution can something be acquired peremptorily’ (MM, 6: 264).

That Kant conceives of the republic as achieving the right combination of force, freedom and law explains why he moves freely in the Doctrine of Right between the idea of a pure republic (the rational form of a state) and the idea of a state making freedom ‘the condition for any exercise of coercion’. The true republic limits its exercises of legal coercion to those necessary for ensuring compossible freedom. This is the same notion of the idea of the republic that Kant had endorsed throughout the critical period. As early as in the Critique of Pure Reason, he defines the ideal state as a republic demanding the maximal extension of freedom (A316/B372–3). Somewhat later (in IUH) he writes that ‘the greatest problem for the human species … is the achievement of a civil society universally administering right’ (8: 22). Such a society must realize ‘the greatest freedom’ for all and is the ideal of a ‘perfectly just civil constitution’. The freedom enshrined by the just constitution makes possible ‘a thoroughgoing antagonism of its members’ and yet constrains such antagonism by ensuring ‘the most precise determination and security of the boundaries of this freedom so that the latter can coexist with the freedom of others’ (ibid.).

In the Doctrine of Right, Kant says that the absence of the pure republic implies that ‘no absolutely rightful condition of civil society can be acknowledged, but only provisional right within it’ (MM, 6: 341). Since the absence of a pure republic is (apparently) part of the human condition (it is a mere idea), it follows that so too is provisional right. Thus states (that is, states that are not ‘literally state[s]’ – those that we find in experience) do not suffice for the existence of a condition in which what belongs to each is peremptorily secured. Realizing the pure republic is necessary in addition. Kant was apparently serious when he spoke of crooked timber (IUH, 8: 23).

Might one argue that a state succeeds in making freedom the principle of coercion insofar as its lawgiving coheres with the idea of the original contract? If so, then the standard view has a ready reply. For in his earlier political writings, Kant had argued for a minimalist interpretation of this demand. On this early view, a state gives law in conformity with the idea of the original contract just in case it does not demand anything of its subjects to which they could not possibly consent, and external freedom consists in obeying just those laws that satisfy this condition (OCS, 8: 297, EW TPP, 8: 349n; cf. Kersting Reference Kersting, Ameriks and Höffe2009: 251).Footnote ¹⁶ The demand to make freedom the condition of any exercise of coercion is – on this conception – satisfied by states that allow relatively little freedom, provided, for example, that they do not tax in a disproportionate manner (OCS, 8: 297n.). Attributing this view to Kant secures the desired result that PT is false (it is possible for actually existing states to legislate in this way – indeed, most do). But this strategy purchases the falsity of PT at the expense of depriving the Kantian theory of critical resources for reform. For the idea of the original contract is meant to serve as the criterion of right for ‘every existing social and political confederation’ (Kersting Reference Kersting, Ameriks and Höffe2009: 250). Thus, reading the requirements of the original contract in this minimal way leaves Kant’s theory without resources for evaluating satisfactory regimes as better or worse. Perhaps more significantly for our purposes: by the time he writes the Doctrine of Right, Kant seems to have abandoned the idea that giving law in line with what an entire people could consent to is a sufficient condition of legitimacy. Although he continues to treat the impossibility of consent as a sufficient condition for the illegitimacy of a law (e.g. in his treatment of the hereditary nobility, MM, 6: 329), the idea that the bare possibility of consent to law suffices for legitimacy is nowhere to be found.Footnote ¹⁷ Possible consent becomes, in the mature philosophy, a mere necessary condition of legitimacy. Thus this strategy is multiply unpromising.

At this point, one might press the objection that Kant’s claim that the true civil constitution is an ideal (rather than something that can be instantiated in experience) is beside the point. As long as we can do a sufficiently good job of approximating it, we ought to recognize peremptory rights. But although this is a plausible philosophical worry, it does not impugn the textual evidence presented above for the claim that Kant’s arguments entail PT. My reply to the worry that Kant places implausibly demanding constraints on making rights peremptory is no different. In the next section, I more fully flesh out some of the implications of PT and consider some strong textual evidence against it.Footnote ¹⁸

5. Implications and Some Doubts

I have argued that Kant is committed to PT. Acquired rights must discharge the omnilaterality requirement, which demands that the principles of right are fully satisfied at all levels – domestic, international and cosmopolitan. But it is natural to wonder what precisely this means for Kant’s political philosophy. Although space constraints rule out a full account here, we can get purchase on the options by noting one way of inferring too much from Kant’s acceptance of PT.

After considering much of the textual evidence introduced above and correctly noting that, on Kant’s theory of right, ‘the strict dichotomy between the civil and natural conditions’Footnote ¹⁹ assumed by earlier contract theorists ‘disappears’, scholars Herb and Ludwig draw the following two conclusions (Reference Herb and Ludwig1993: 313, my translation). First, they correctly note that Kant’s view forces us to consider mine and yours and state sovereignty alike as matters of provisional right. Second, they claim that it follows from this that each ‘individual state’ – and (presumably) each historical acquisition of property – ‘represents now merely a transitional stage in a process of the global realization of right’ (314). As merely so many ‘episodes’ in the process of realizing cosmopolitan institutions of right, states are granted sovereignty only to prevent a ‘backslide’ into the state of nature. State sovereignty qua provisional right is a necessary condition of realizing cosmopolitan institutions and moving farther away from the state of nature and closer to a genuine civil condition (315). Still, it is ultimately to give way to cosmopolitan institutions as they become available (cf. Ypi Reference Ypi2014).Footnote ²⁰

While the first conclusion follows from a natural reading of the text (as I have shown above), it is important to see that the second goes beyond PT by presupposing a particular reading of the normative status of provisional right. For taking state sovereignty to be a mere moment in a transition towards cosmopolitan right implies that, once cosmopolitan right is realized, sovereign states as they have formed are to play no governing role. But this does not follow from the claim that rights to sovereignty are provisional. For on an anticipatory account of provisionality (of the sort defended e.g. in Hasan Reference Hasan2018), provisional rights are held in anticipation of a rightful condition. For all that, some provisional arrangements of territory and property might survive the transition (or the closest possible approximation to the transition that we can attain). If so, they will have a permanence that goes beyond what Herb and Ludwig seem willing to countenance (compare Ellis Reference Ellis2005: 89–94). Perhaps it would even be wrong to eradicate them without compensation. Flikschuh, who follows Herb and Ludwig on this point, writes that as provisional gradually transforms into peremptory right, the ‘coming together of peoples … envisages a gradual dismantling of boundaries that hinder the development of relations of Right between subjects’ (Reference Flikschuh2000: 176). This allows us to put my objection in different language: different accounts of provisional right identify different boundaries for dismantling. In assuming that PT entails the dismantling of all but cosmopolitan boundaries, theorists sympathetic to it have moved too quickly.

PT does not, therefore, entail any particular account of provisional normativity, but that is a virtue, not a vice. For scholars continue to disagree sharply amongst themselves concerning just this, and we should avoid begging important questions. On the strongest view, provisional rights impose full constraints, imperfect only insofar as persons lack full assurance of their protection (Byrd and Hruschka Reference Byrd and Hruschka2006a). If this is correct, then any boundaries generated by historical acquisitions become (in the limit) permanent aspects of right, not to be transcended by, but to be encoded in, international and cosmopolitan institutions, revised only as required by movement towards right. On the weakest view, provisional rights impose no genuine constraints at all (Brudner Reference Brudner2011). All historical boundaries are eventually to be transcended, as the actually legislative general will realizes itself. The fact that provisionality can be understood in either of these ways (and in many ways in between) entails that, in terms of its normative implications, PT should be amenable to a wide array of Kantian theorists.

Still, even if provisional rights need not inevitably be transcended, to say that they are provisional must mean something. And on at least one plausible account, it means that they can be defended with force, but that they generate no correlative duties on the part of anyone to obey (Korsgaard Reference Korsgaard2018: 27–30; cf. MM, 6: 257). But if this is right, then there would appear to be strong evidence against attributing PT to Kant. Consider the following passage.

Here, Kant is explicit: The fact that there is no fully just state, the fact that the idea of the republican constitution cannot be instantiated in experience, the fact that international institutions are bound to be imperfect, these are no grounds for concluding that no rightful condition exists. Instead, if the stately forms we observe in the world of experience are approximately close to the idea of a state, then there is a civil condition. Even though a constitution ‘may be afflicted with great defects and gross faults and be in need eventually of important improvements’ it still holds in general as a ‘rightful constitution in the general sense of the term’. And as long as there is such a rightful constitution, then there is a duty to obey it (MM, 6: 319). And if there is a right to enforce law,Footnote ²¹ as well as a correlative duty to obey, then by hypothesis the right of such a state cannot be a provisional right.

One response is to reject Korsgaard’s analysis of provisionality and argue that, although it may be the correct account of the notion as it operates in the state of nature, a different analysis is required once something approximating a state exists.Footnote ²² For given strong textual evidence for PT, states simply do not make rights peremptory. But then it must be possible for persons to be obliged to obey states and to respect one another’s claims, even though rights remain provisional.Footnote ²³

However, it is also possible to account for these duties of obedience more concessively by noting that acting as if a properly civil condition exists might be thought a necessary condition of bringing it about that a true civil condition exists. Understood in this way, perhaps it is true that provisionally rightful states do not generate duties to obey based on their actual and defective lawgiving activity and (partial) conformity with right. Instead, the duty to obey the ‘power above you’ derives from the duty generated by the postulate of public right to enter a civil condition. If so, the duty we violate in disobeying the state above us is not owed to that particular state. It is rather the perfectly general duty we owe to humanity to enter a rightful condition, so far it is in one’s power, and to do one’s part in ending conditions of war.

I submit that, given Kant’s central normative concerns, a reading of the second kind makes good sense. For recall that what is at issue is what we are duty-bound or obligated to do, and duties have a special character in the Kantian system. Kantian duties are famously necessary (MM, 6: 222; G, 4: 414), universal (6: 225, 4: 433), categorical (6: 221, 4: 414–16) and self-legislated (6: 227, 4: 432). On this account, the notion of adventitious political duties acquired through an act of choice appears paradoxical. For political duties of this kind (e.g. to respect historical claims to property, territory and sovereignty) have an avowed origin in force and another’s exercise of her power of choice to take control (6: 224, 259). They are geographically bounded. They are enforced by pathological sanctions (6: 219). And the laws which ground these imperatives – at least in the case of acquired rights – are legislated by others (compare Hirsch Reference Hirsch2017).

In the limit, the Doctrine of Right is supposed to demonstrate that these apparently contingent and bounded duties, enforced as they are via pathological incentives, made effective as they are through other agents’ willkürliche choice, really must be seen as products of our own lawgiving reason (Schaefer Reference Schaefer, Robinson and Surprenant2017).Footnote ²⁴ This is achieved by showing that acknowledging these very duties is the sine qua non of satisfying a priori laws of pure practical reason (e.g. the Universal Principle of Right and the postulates). But because the principles of right (postulated by practical reason) make this in the first instance possible, the omnilaterality requirement is satisfied only when institutions satisfy these principles and extend across the globe. Until then, we are bound to do what is necessary to bring this condition about. It is plausible that to bring this about requires acknowledging duties to comply with provisional authorities.

On this account, such duties derive from the general duty to enter a rightful condition. They are provisional in the sense that they are held in accord with the possibility, but not yet the actuality, of a rightful condition. Existing states are not, therefore, themselves rightful conditions. Rather, acknowledging a duty to obey leaves open the possibility that they will become rightful conditions. Only in this way can we progress toward peremptory right and perpetual peace in a way that does not depend upon mere chance (MM, 6: 372).