² On Kant's view, to say that I have a duty of right not to do X just means that I can justifably be forced to refrain from doing X; the two expressions are interchangeable. As he puts it: ‘Right and authorization to use coercion … mean one and the same thing’ (MS 6: 232). References to Kant's works follow the usual system; unless otherwise indicated, all translations are from Kant, , Practical Philosophy, ed. and trans. Gregor, MJ (Cambridge: Cambridge University Press, 1996).Google Scholar

³ For Kant's view of external freedom, see MS 6: 230 and 382. For the idea that freedom can be restricted only for the sake of freedom itself, see MS 6: 231. The reading of the idea of external freedom I adopt here follows closely that put forward by Ripstein, Arthur in his ‘Authority and coercion’, Philosophy and Public Affairs, 32: 2–35 (2004), pp. 8–11CrossRefGoogle Scholar, and in his Force and Freedom: Kant's Legal and Political Philosophy (Cambridge, MA: Harvard University Press, 2009), chapter 2.

⁴ Throughout this discussion, I use the term ‘freedom’ without qualifcation to refer specifcally to external freedom, since it is exclusively with that kind of freedom that political philosophy is concerned on Kant's view. I make no claims here about the kind of internal freedom that is the concern of Kant's moral philosophy. In particular, I do not claim that the state is necessary for internal freedom; given Kant's famous tenet that ‘a free will and a will under moral laws are one and the same’ (G 4: 447), that would amount to saying that the state is necessary for moral action – a thesis that Kant did not hold, and that I see no reason to endorse.

⁵ Property and contract rights are provisional in the state of nature, because no agent has the authority to enforce them in a way that would be fully justifed. By contrast, the innate right to bodily integrity is enforceable in a state of nature, since each individual has a right of self-defence; but it is imperfectly realized, since no one has the authority to punish transgressions after the fact (once someone's innate right has been violated, nothing can be done about it).

⁶ Similar arguments could be made in terms of other types of rights; indeed, on Kant's view, this has to be the case for any type of right that can legitimately be enforced by the state.

⁷ I take on the issue in my ‘Kant on the right to freedom: a defense’, forthcoming in Ethics.

⁸ I shall not attempt to link the points I make here to the growing literature on Kant's conception of property, as that would unnecessarily weigh down the argument, but I do want to mention at the outset some discussions from which I have benefted. These include Williams, Howard, Kant's Political Philosophy (New York: St. Martin's Press, 1983)Google Scholar, chapter 4, and Kant's Critique of Hobbes (Cardiff: University of Wales Press, 2003), chapters 4–6; Buck, Wayne F., ‘Kant's justifcation of private property’, in den Ouden, B. and Moen, M., eds, New Essays on Kant (New York: Peter Lang, 1987), pp. 227–44Google Scholar; Baynes, Kenneth, ‘Kant on property and the social contract’, The Monist, 72: 433–53 (1989)Google Scholar; Korsgaard, Christine M., ‘Taking the law into our own hands: Kant on the right to revolution’, in Reath, A., Herman, B. and Korsgaard, C. M., eds, Reclaiming the History of Ethics: Essays for John Rawls (Cambridge: Cambridge University Press, 1997), pp. 297–328Google Scholar [reprinted in Korsgaard, , The Constitution of Agency (Oxford: Oxford University Press, 2008), pp. 233–62]Google Scholar; Wood, Allen W., Kant's Ethical Thought (Cambridge: Cambridge University Press, 1999), esp. pp. 245–9Google Scholar; Flikschuh, Katrin, Kant and Modern Political Philosophy (Cambridge: Cambridge University Press, 2000)Google Scholar, chapter 4; Westphal, Kenneth, ‘A Kantian justifcation of possession’, in Timmons, M., ed., Kant's ‘Metaphysics of Morals’: Interpretative Essays (Oxford: Oxford University Press, 2002), pp. 89–109Google Scholar; Bernd Ludwig, ‘Whence public right? The role of theoretical and practical reasoning in Kant's Doctrine of Right’, ibid., pp. 159–83; and Ripstein, Force and Freedom, chapters 4 and 6. I should also acknowledge that my understanding of the Doctrine of Right owes much to Bernd Ludwig's illuminating corrections to the traditional ordering of the text, as put forth in Kant, , Metaphysische Anfangsgründe der Rechtslehre, ed. Ludwig, Bernd (Hamburg: Meiner, 2nd edn, 1998).Google Scholar Finally, let me note that Jeremy Waldron argues for a link between freedom and property broadly construed in his ‘Homelessness and the issue of freedom’, in Liberal Rights: Collected Papers 1981–1991 (Cambridge: Cambridge University Press, 1993), pp. 309–38. The argument I attribute to Kant differs from Waldron's in its detail, but the general spirit is close.

⁹ Having a thing as one's property and having it in one's physical possession are thus different things. I use the term ‘possession’ in what follows to refer to physical or empirical possession. Kant also speaks of intelligible possession (see MS 6: 249–50), and suggests that the right to property should be understood in terms of that idea, but I leave that point aside for present purposes.

¹⁰ See MS 6: 248 and 250.

¹¹ Hence Kant's claim that ‘something external would be mine only if I may assume that I could be wronged by another's use of a thing even though I am not in possession of it’ (MS 6: 245).

¹² MS 6: 246 (Gregor translation amended). On the notion of a postulate, with specifc attention to the postulates presented in the Doctrine of Right, see Guyer, Paul, ‘Kant's deductions of the principles of right’, in Timmons, M., ed., Kant's ‘Metaphysics of Morals’: Interpretative Essays, pp. 23–64Google Scholar; see also the reply by Flikschuh, Katrin, ‘Kant's indemonstrable postulate of right: a response to Paul Guyer’, Kantian Review, 12: 1–39 (2007).Google Scholar Kant's most familiar use of the notion of postulate is in the Critique of Practical Reason, where he presents the ‘postulates of pure practical reason’ concerning the immortality of the soul, freedom and the existence of God (see KpV 5: 122–34). There, he defnes a postulate as ‘a theoretical proposition, though one not demonstrable as such, insofar as it is attached inseparably to an a priori unconditionally valid practical law’ (KpV 5:122). That is what Kant seems to have in mind when he says that ‘it is a duty of right to act towards others so that what is external (usable) could also become someone's despite the fact that there is ‘no way of proving of itself [theoretically] the possibility of nonphysical possession or of having any insight into it’ (MS 6:252).

¹³ MS 6: 246.

¹⁴ Let me stress that the problem I am calling attention to here is not that, absent full-fedged property rights, I cannot be certain that the means I need for a given project will be at my disposal. That kind of uncertainty does not as such undermine my ability to set and pursue ends for myself. I can set for myself the end of being the frst human being to walk on Mars, even if there is no telling in advance exactly what means will be required to achieve that end, and no way to be certain that I shall obtain them. No system of property could be expected to solve that problem; more importantly, no system of property need do so, since my freedom does not depend on my ability to pursue any project in particular. The point I am making in the text is different: it is that a system of mere empirical possession undermines freedom because it arbitrarily restricts the means that I can have at my disposal.

¹⁵ For an illuminating discussion of this point, see Arthur Ripstein, ‘Authority and Coercion’, pp. 8–11, and Force and Freedom, chapter 2. Not everyone agrees with this interpretation of Kant's position. Flikschuh, who follows Brandt on this point, writes the following: ‘Under conditions of unavoidable empirical constraint (i.e. the earth's spherical surface) any exercise of choice by one compromises the freedom of everyone else by removing from availability to them external objects of their possible choice’ (Kant and Modern Political Philosophy, p. 134; see also her article ‘On Kant's Rechtslehre’, European Journal of Philosophy, 5: 50–73 [1997], p. 64). That is precisely the conception of external freedom against which Ripstein forcefully argues. I find Ripstein's argument compelling, although I should note that the point I go on to make in the rest of the paragraph holds on either conception of external freedom.

¹⁶ Korsgaard makes a similar point in ‘Taking the law into our own hands’, pp. 325–6, n. 8. Kant himself appears at times sceptical about the legitimacy of communal property, as when he writes, concerning the idea of an original common possession of the land by all men, that ‘the choice of one is unavoidably opposed by nature to that of another’ (MS 6: 267), so that the possibility of ‘particular possession’ is necessary if the land is to be usable. But note that this does not mean that communal property is incompatible with external freedom; only that it requires explicit rules that make it possible for the choices of all to coexist. What Kant rules out is the possibility of spontaneously sharing objects. Note also that the point I am making is compatible with Kant's claim that communal property cannot be invoked to justify private property, since the two stand equally in need of justifcation (see MS 6: 251; cf. Ripstein's discussion in Force and Freedom, chapter 4). My point is that both forms of property are justifed, because both solve the problem posed by a system of mere empirical possession.

¹⁷ Indeed, insofar as libertarianism relies on the possibility of full-fedged pre -institutional property rights, it is incompatible with what I go on to say in the rest of the article. For a powerful defense of the libertarian view of rights, see Nozick, Robert, Anarchy, State, and Utopia (New York: Basic Books, 1974).Google Scholar

¹⁸ MS 6: 247. The last clause could be taken to suggest that Kant is endorsing a theory of property based on frst acquisition, but that would be reading too much into the passage. First acquisition may be necessary for a claim of right in the state of nature to be provisionally valid, but it cannot be suffcient for a rightful claim, as we shall see in the next section. On Kant's use of the idea of a permissive law of practical reason, the seminal discussion is Brandt, Reinhard, ‘Das Erlaubnisgesetz, oder: Vernunft und Geschichte in Kants Rechtslehre’, in Brandt, Reinhard, ed., Rechtsphilosophie der Aufklärung (Berlin: de Gruyter, 1982), pp. 233–85.Google Scholar For useful accounts of Brandt's position in English, see Flikschuh, ‘On Kant's Rechtslehre’, pp. 63–5; Kant and Modern Political Philosophy, pp. 134–9; and ‘Kant's indemonstrable postulate of right’, pp. 22–3. On Brandt's view, the importance of the permissive law ‘consists in its mediatory role between prescription and prohibition. Something which is strictly speaking prohibited is provisionally permitted in order to make possible the eventual realization of the universal principle of Right’ (Brandt, ‘Das Erlaubnisgesetz’, p. 244; quoted and translated by Flikschuh in ‘On Kant's Rechtslehre’, p. 65). Hruschka, Joachim has challenged this reading in ‘The permissive law of practical reason in Kant's Metaphysics of Morals’, Law and Philosophy, 23: 45–72 (2004).Google Scholar There, he argues that a permissive law is not one that allows individuals to commit unjust acts, but rather one that makes it possible for merely permissible acts to have consequences for rights.

The reading I adopt here is closer in spirit to Brandt's than to Hruschka's, although I should mention two counts on which I disagree with Brandt. First, his reading strikes me as too historicized. I do not think that the postulate permits one to do something that is strictly speaking wrongful, but that constitutes a necessary step toward a rightful condition. Rather, I think that the postulate allows one to do something that would be wrongful, were it not that freedom itself demands that we be permitted so to act. Actions authorized by the postulate are nevertheless merely provisionally rightful on my reading, not because they are only treated as rightful while we await a rightful condition, but rather because acting on the permission contained in the postulate ultimately amounts to leaving the state of nature (on which see Section V below). Second, I disagree with Brandt about what exactly requires the permission contained in the postulate. On Brandt's reading, as we saw above (see note 15), it is the mere taking possession of an object. Once again, I follow Ripstein in thinking that a person's taking possession of an object or of a piece of land is not in itself a restriction of external freedom in the sense that Kant is concerned with; hence I do not think that it requires special permission. What requires permission on my view, as I go on to explain in the text, is specifcally the use of force to exclude others from a certain object.

¹⁹ MS 6: 312. In the Lectures on Ethics, Kant suggests that the need for a state amounts to the need for three things: ‘[A] universal legislation that establishes right and wrong for everyone, a universal power that protects everyone in his right, and a judicial authority that restores the injured right’ (Kant, , Lectures on Ethics, ed. Heath, P. and Schneewind, J. B. [Cambridge: Cambridge University Press, 1997], 27: 590).CrossRefGoogle Scholar I take what Kant calls public distributive justice in the passage I quote in the text to include both universal legislation and judicial authority. I do not stress the distinction between the legislative and the judiciary here, since I take it that their combined activity is necessary to make the content of law determinate in particular cases. For an interpretation of Kant's position that explicitly distinguishes between the two, see Ripstein, Force and Freedom, chapter 6.

²⁰ I am grateful to T. M. Scanlon for pressing me to consider this possibility.

²¹ See Locke, , Two Treatises of Government, ed. Laslett, P. (Cambridge: Cambridge University Press, 1967).Google Scholar For recent discussions of Locke's views on property, see notably John Simmons, A., The Lockean Theory of Rights (Princeton, NJ: Princeton University Press, 1992)Google Scholar, chapter 5; and Sreenivasan, Gopal, The Limits of Lockean Rights in Property (Oxford: Oxford University Press, 1995).Google Scholar An important contribution of these two discussions is that they suggest that Locke's position can be reconciled with an egalitarian outlook; as far as I can see, this does not affect the points I make in the text. There are of course many different interpretations of Locke's theory of property, and I do not claim that the argument I present holds against all of them; I only mean to address one particularly sharp (but perhaps oversimplifed) version of the theory.

²² In support of this idea, one might argue, as Ripstein does in ‘Authority and coercion’, that the indeterminacy problem stems from the general fact that an abstract principle cannot fully determine its own application to particular cases. Thus, even if we all accept the same version of Locke's theory, questions will arise about the exact manner in which objects can be acquired or abandoned, and about the boundaries – both temporal and spatial – of the objects that are acquired. The point is easy to illustrate. Take for instance the question of when an object should count as abandoned. Surely, my land does not cease to be mine just because I walk away from it for fve minutes. But what if I leave it unattended for a year? Or two years? Or ten? When exactly does it become up for grabs? Locke's theory by itself has no defnite answer to offer. Of course, as Ripstein rightly emphasizes, that does not mean that there is no way to settle these questions; it just means that abstract principles of right such as those that make up Locke's theory – or any other similar theory – cannot possibly settle every particular case. Consequently, rights in the state of nature cannot have the crisp contours that they should have. A legal system, and hence a state, is required to remedy this problem (see Ripstein, ‘Authority and coercion’, pp. 26–32). Note that Ripstein does not hold that the indeterminacy problem is the only factor that makes rights provisional in the state of nature; indeed, in the most recent and detailed statement of his view, he argues that there are three problems that explain why rights are provisional in the state of nature, including the enforcement problem. The three problems in turn explain the need for the three branches of republican government (see Ripstein, Force and Freedom, chapter 6). I do not mean to deny that there are problems other than the enforcement problem that affect rights in the state of nature, but I do think that that problem is in an important sense the fundamental one, for reasons I go on to outline in the text.

²³ See Locke, Second Treatise, §§7–9. Note that Locke's view about the justifcation of the use of force in the state of nature is logically independent from his particular theory about how rights are acquired or transferred in the state of nature. One can disagree with him on these latter points and still think that there are rights in the state of nature whose enforcement is justifed regardless of who carries it out.

²⁴ Locke famously fnds three problems with property rights in the state of nature. The one I mention here is the third: that, in the absence of a power able to impose the demands of the law of nature against any given individual, the enforcement of the law of nature will be unreliable, and often dangerous for one undertaking it (see Second Treatise, §126). Locke also mentions that individuals in the state of nature are likely to be biased and ignorant, so that they will often fail to recognize the demands that the law of nature makes on them, despite the fact that its content is ‘plain and intelligible to all rational creatures’ (ibid., §124); and that individuals being judges in their own cases will likely be biased in their application of the law of nature (see ibid., §125). Joining political society is in our interest, Locke argues, because a good state will solve these three problems, and hence make property rights more secure; but the private enforcement of property rights in the state of nature remains entirely legitimate.

²⁵ MS 6: 232, 231 and 257.

²⁶ MS 6: 257. Cf.: ‘Prior to entering [a civil condition], a subject who is ready for it resists with right those who are not willing to submit to it and who want to interfere with his present possession’ (MS 6: 257; emphasis added; see also MS 6: 267). The central importance for Kant's approach of who is entitled to use coercion is stressed notably in Ludwig, ‘Whence public right?’, pp. 172–3.

²⁷ What would constitute jointly suffcient conditions to justify the use of force in the state of nature is a diffcult question, which I must leave to one side here. Note that the question is not merely academic, since the international order as we know it is arguably a state of nature in the relevant sense.

²⁸ See MS 6: 255–6.

²⁹ MS 6: 255.

³⁰ MS 6: 255 (Gregor translation modifed).

³¹ MS 6: 237.

³² As he puts it, freedom consists in ‘independence from being constrained by another's choice’ (MS 6: 237).

³³ Let me stress once again that what violates my right to freedom here is not the fact that your choice deprives me of the possibility of accessing the land you claim for yourself. If you exclude me from a certain area by covering it with your body, you do not thereby make me unfree, since I remain free to set and pursue ends for myself everywhere else in the world; the same goes if what prevents me from stepping on a piece of land is the fact that you claim it for yourself (see notes 15 and 18 on this point). The violation I point to in the text occurs only when you use force to prevent me from going where I have chosen to go. By doing so, you force a particular course of action upon me, and you thereby give your choice (to keep me out of the territory) priority over my choice (to go inside the territory); that is what violates my right to freedom.

³⁴ MS 6: 255–6. Kant uses similar language in §42, when he writes: ‘No one is bound to refrain from encroaching on what another possesses if the other gives him no equal assurance that he will observe the same restraint towards him’ (MS 6: 307).

³⁵ The relation between Kant's and Hobbes's views is obviously a complex and multifaceted one. For a detailed treatment of the matter, see Williams, Kant's Critique of Hobbes. For further discussion of Kant's use of the idea of assurance, and how it differs from Hobbes's, see Ripstein, Force and Freedom, chapter 6.

³⁶ Hobbes, , Leviathan, ed. Curley, E. (Indianapolis: Hackett, 1994)Google Scholar, chapter xiv, para. 5.

³⁷ Cf. Kant's claim that ‘he who does not have enough power to protect each one among the people against the others does not have the right to command the people either’ (EF 8: 383).

³⁸ It is perhaps worth stressing here that, even if I commit a wrong against you, I do not thereby lose my right to freedom. I make it the case that some uses of force against me will be consistent with my right to freedom that would otherwise not be (provided that the agent making use of force has the authority to do so). But I certainly do not lose my right to freedom altogether; if I did, then any use of force against me whatsoever would be justifed, which is clearly not the case.

³⁹ Here I am paraphrasing Ripstein's suggestion that ‘[t]he abstract principles that enable us to understand our obligations with respect to the property of others do not tell us how to apply them to particulars’ (‘Authority and coercion’, p. 28).

⁴⁰ MS 6: 256; cf. MS §§11 and 14, and TP 8: 295.

⁴¹ MS 6: 314. The point is obviously indebted to Rousseau, who famously claims that the general will must come from all and apply to all – see Rousseau, , Of the Social Contract, in ‘The Social Contract’ and Other Later Political Writings, ed. and trans. Gourevitch, V. (Cambridge: Cambridge University Press, 1997)Google Scholar, book 2, chapter 4, para. 5. To be exact, two things are required for a will to be general on Kant's view : frst, it must aim exclusively to protect rights, since that is the only purpose required by freedom, and hence the only purpose that can legitimately be imposed on individuals (the only genuinely public purpose); second, it must aim to protect the rights of all individuals whose freedom is codependent – that is, of all who share a given territory (a point about which I say more below).

⁴² Kant's view is thus not that our general will realizes conditions of reciprocity because we actually consent to its actions, but rather that we are rightfully presumed to consent to its actions because they are necessary for our free coexistence (which is the case, in turn, because these actions realize conditions of reciprocity). The former position is ruled out on Kant's view, since it would mean that we depend on one another's actual consent, and hence on one another's choices, for our being in a rightful condition, when the whole point of the rightful condition is precisely to prevent such dependence from arising.

⁴³ MS 6: 307.

⁴⁴ According to Max Weber's well-known defnition, the second condition I mentioned in the previous paragraph is suffcient for something to count as a state. On Kant's view, by contrast, an agent has to fulfl both conditions of reciprocity to have the authority to use force, and hence to count as a genuine state. For Weber's position, see ‘Politics as a vocation’, in The Vocation Lectures: Science as a Vocation, Politics as a Vocation, ed. Owen, D. and Strong, T. B., trans. Livingstone, R. (Indianapolis: Hackett, 2004).Google Scholar