¹ A nonindividualist view, then, would relate persons to one another “internally,” as might seem natural (and even obvious) to human beings whose lives are dominated by feudal or other traditional solidarities, and perhaps to communist men and women too, moved by a “general will” in conditions of (relative) abundance. See my The End of the State (London: Verso, 1987).

² Historical capitalisms commonly rely on markets in labor power, means of subsistence, and capital, and also on large-scale factory systems of production. In addition, historical capitalisms are class societies, with a bourgeoisie and a proletariat. Marx did not always use the term “capitalism” in the same way, but for the most part he did regard some (or all) of these features of historical capitalisms as part of capitalism's definition. The definition proposed above is therefore weaker than Marx's definition, though it is suggested by and consonant with historical materialist claims. Peasant societies without relations of feudal bondage would count as capitalist by this definition, so long as peasants own their means of production. Thus, in contrast to standard Marxian usage, the “simple commodity production” sketched in the first volume of Capital or Adam Smith's “early and rude state of society” – classless societies of (relatively) independent producers who own their own means of production – would be capitalist societies. On definitions of “capitalism” and “socialism,” see my Arguing for Socialism: Theoretical Considerations (Boston and London: Routledge & Kegan Paul, 1984; 2nd edition, Verso: London, 1988), pp. 5–11.Google Scholar

³ An absolute right to the income generated by deployment of productive assets is also a limiting case. Even libertarian proponents of “minimal states”, states that do not redistribute (capitalist) market allocations, acknowledge the right of states to tax individuals to obtain the revenues they require to carry out their (legitimate) operations.

⁴ The term, though commonplace, is misleading. Construed literally, “self-ownership” suggests that there is something, the self, that one owns. But “the thesis of self-ownership does not say that all that is owned is a self, where ‘self’ is used to denote some particularly intimate, or essential, part of the person…. The term ‘self’ in the name of the thesis of self-ownership has a purely reflexive significance. It signifies that what owns and what is owned are one and the same, namely, the whole person….” Cohen, G.A., “Self-Ownership, World-Ownership, and Equality: Part I,” Lucash, Frank S., ed., Justice and Equality Here and Now (Ithaca: Cornell University Press, 1986), p. 110.Google Scholar

⁵ Cf. Cohen, “Self-Ownership,” p. 109: “… each person is the morally rightful owner of himself. He possesses over himself, as a matter of moral right, all those rights that a slaveholder has over a complete chattel slave as a matter of legal right, and he is entitled, morally speaking, to dispose over himself in the way such a slaveholder is entitled, legally speaking, to dispose over his slave. Such a slaveholder may not direct his slave to harm other people, but he is not legally obliged to place him at their disposal to the slightest degree: he owes none of his slave's service to anyone else. So, analogously, if I am the moral owner of myself, and therefore of this right arm, then, while others are entitled to prevent it from hitting people, no one is entitled, without my consent, to press it into their own or anybody else's service, even when my failure to lend it voluntarily to others would be morally wrong.”

⁶ I am grateful to Joshua Cohen for pointing out that, despite what is widely supposed, particularly after C.B. Macpherson's influential study of Locke, in The Political Theory of Possessive Individualism (Oxford: Oxford University Press, 1962)Google Scholar, Locke himself was not, strictly, a self-ownership theorist. In Locke's view, God is the source of all entitlements, including our rights to our own bodies and powers. Thus, for Locke, insofar as we own ourselves, we do so contingently. However, Nozickean libertarians and other neo-Lockean political theorists wisely eschew divine foundations for rights ascriptions. Lockeans today take self-ownership as fundamental, and go on to use the idea as a point of departure for the elaboration of other, putatively inviolable rights.

⁷ Cf. Cohen, G.A., “Self-Ownership,” and “Self-Ownership, World-Ownership and Equality: Part II,” Social Philosophy and Policy, vol. 3, no. 2 (Spring 1986), pp. 77–96CrossRefGoogle Scholar, and “Why Marxists Care More About Nozick than Liberals Do,” unpublished manuscript. Cohen defends self-ownership from its detractors on the left by arguing that the idea does not imply the inequality of condition that Nozick maintains. He argues too that self-ownership in Nozick's sense is assumed in standard Marxian accounts of capitalist exploitation (where workers are held to have entitlements to the surplus value capitalists extract) and in Marx's account of communism (where the free development of each person's powers, understood as the full deployment of those assets self-owners own, is finally realized). In short, if Cohen is right, self-ownership is compatible with (some) Marxian egalitarian aspirations, and is a even a tenet (some) received Marxian views share with precapitalist libertarian political philosophy. This claim is, to say the least, ironic. Self-ownership has been widely derogated, not only by Marxists, as a “bourgeois” notion par excellence. In the end, Cohen claims to agree with this assessment, though for reasons that at present remain obscure. His contention, which serves as my own point of departure, is that traditional dismissals of self-ownership are overly facile, and that if the idea is finally rejected, it will have to be for reasons different from those that Marxists and others on the left have so far contrived.

⁸ However, the forced removal of bodily parts is not always offensive to considered intuitions. Forced haircuts, circumcisions, castrations and sterilizations, and also mandatory inoculations, drug and alcohol testing, and testing for infectious diseases, have defenders. It may be that hair and blood and even prepuces and ovaries somehow fall into a different category from kidneys and other organs. But I don't think so. I would suggest instead that in cases where bodily invasions seem warranted, the inviolability of the body is still a pertinent – though overridden – consideration, swamped perhaps by the requirements of citizenship or by a determination to prevent harm to others or for paternalistic reasons. The appropriateness of these countervailing claims is debatable. But their appeal (to some of us) shows that our intuitions about bodily inviolability, though powerful and general, are not so overwhelming that they carry for everyone in all cases.

⁹ Ronald Dworkin's views on justice also share this feature, though in developing a view of equality that is “ambition-sensitive” by definition – rather than in consequence of an “agreement” to reward (freely undertaken) expenditures of effort differentially and, more generally, to respect the consequences of (free) choice – Dworkin comes perilously close to a Lockean view of desert.

¹⁰ However, in Rawls's view, basic rights and liberties must be distributed equally (and to the greatest extent possible), regardless of efficiency considerations. For Rawls, the principles of justice are lexically ordered. It is only after basic rights and liberties are equally distributed that “the difference principle,” which allows inequalities in the distribution of other primary goods, pertains.

¹¹ It is one thing to use state power to enforce particular conceptions of the good, and something else again to move towards realization of an ideal order through the transformative effects of institutional arrangements. Against the likely rejoinder that political education per se inevitably infringes respect for individuals' conceptions of the good, I would therefore assert – following Mill and a host of other political philosophers, both liberal and nonliberal, and also following Rawls himself – that political institutions necessarily educate, and that a central task of political philosophy is to articulate a conception of the citizen to educate towards. Thus, On Liberty is not only a “constitutional” tract, defending principled limitations on societal and state interferences with individuals' lives and behaviors. It is also an investigation of the institutional arrangements best suited for forming liberal citizens: individuals disposed to tolerance and the advancement of their own and each others' autonomy. Similarly, for proponents of Rawlsian justice, a paramount task is to contrive a strategy for making individuals more just – that is, more disposed to implement the egalitarian vision Rawlsian justice expresses.

¹² See my Arguing for Socialism: Theoretical Considerations, Chapter 2.

¹³ Kant, The Foundations of the Metaphysics of Morals, Part 2. The more familiar version of the categorical imperative would have us “Act only according to that maxim by which you can at the same time will that it should become a universal law.” However, Kant insists that all formulations of the categorical imperative, including the formulation alluded to above (“Act so that you treat humanity, whether in your own person or in that of another, always as an end and never as a means only”) are strictly equivalent.

¹⁴ Cf. Rawls, A Theory of Justice, passim and esp. pp. 251–257; and “Kantian Constructivism in Moral Theory,” The Journal of Philosophy, vol. 77 (Sept. 1980).

¹⁵ Cf. Rawls, A Theory of Justice, pp. 265–73.

¹⁶ For a contrary view of the implications of Rawlsian justice, see my Arguing for Socialism: Theoretical Considerations, Chapter 2.

¹⁷ See, among others, Roemer, John, A General Theory of Exploitation and Class (Cambridge: Harvard University Press, 1982)CrossRefGoogle Scholar, passim and esp. pp. 8–23.

¹⁸ I do not think we need accede to this pessimistic conclusion, though I cannot claim to be able to defend an optimistic prognostication “within the realm of reason alone.” For some elements of an argument supporting the historical feasibility of the egalitarianism implicit in Kantian moral theory, see my The End of the State.

¹⁹ Strictly speaking, a successful assault on Lockean justice is not by itself a reason to support Rawls's theory because there are other plausible alternatives. There is, for example, utilitarian justice: roughly, the view that a distribution is just if and only if it has the consequence of maximizing aggregate or average utility (in comparison with alternative possible distributions). But I have already said enough in support of Rawlsian justice or, at least, its Kantian defense of human dignity to justify regarding it as the serious competitor to the position I impugn. It is worth noting that utilitarianism is often faulted precisely for its inability to defend human dignity adequately – for treating human beings as bearers of utility and therefore only as means and not, as Kant would have it, ends in themselves.

²⁰ John Locke, Second Treatise of Government, paragraph 33.

²¹ Cf. Nozick, Robert, Anarchy, State, Utopia (New York: Basic Books, 1974), pp. 175–82Google Scholar; and Gauthier, David, Morals By Agreement (Oxford: Oxford University Press, 1986)Google Scholar, Chapter 7.

²² What follows draws on an idea of Gauthier's, developed in Morals by Agreement. Gauthier uses the Lockean proviso in the course of an extended argument for the possibility of grounding moral judgments in (contractual) agreements. Since the use I propose for the proviso is different and more circumscribed, many of the subtleties implicit in Gauthier's own elaboration of the idea are omitted here.

²³ Cf. Nozick, Anarchy, pp. 174–178.

²⁴ Technological innovation may ultimately make access to internal assets more nearly person-neutral than has been the case until now – as has already occurred, to some degree, with organ transplantations. It is instructive to reflect on real or imaginable innovations that break down the person-specificity of inalienable assets in order to appreciate the generality, pace Locke, of the Lockean problem of justifying the original appropriation of things.

²⁵ For the present, I assume, following Nozick and Gauthier, that utility adequately represents welfare or well-being.

²⁶ Strictly speaking, only a full-fledged Lockean person, who owns himself and parts of the external world too, requires capitalist property relations. Self-ownership, as noted, can survive capitalism's demise.