¹ My discussion makes various simplifying assumptions. I do not discuss various indirect forms of consequentialism that assess something, not in terms of its value, but in terms of conformity with something else (e.g., a rule) that is assessed on the basis of its consequences. Also, I do not discuss satisficing, rather than maximizing, forms of consequentialism. My guess is that everything I want to say could be said mutatis mutandis of these less standard forms of consequentialism.

² As far as I know, this useful language was first introduced in Kagan, Shelly, The Limits of Morality, Oxford, 1989Google Scholar, ch. 1.

³ Bentham, Jeremy, An Introduction to the Principles of Morals and Legislation, London, 1970Google Scholar, ch. 1, §11; and Moore, G. E., Principia Ethica, Cambridge, 1903, §§ 17, 89Google Scholar.

⁴ See Scanlon, T. M., ‘Contractualism and Utilitarianism’, Utilitarianism and Beyond, ed. Sen, A. and Williams, B., New York, 1982, esp. p. 115Google Scholar; and Boyd, Richard, ‘How to Be a Moral Realist’, Essays on Moral Realism, ed. Sayre-McCord, G., Ithaca, 1989, esp. pp. 215 fGoogle Scholar.

⁵ Hume, David, An Enquiry Concerning the Principles of Morals [1751], ed. Nidditch, P. H., Oxford, 1975, § ixGoogle Scholar, part 1.

⁶ Whereas unanimity may be the only decision rule acceptable to all ex post, majority-rule can be acceptable to all ex ante. Cf. Rae, Douglas, ‘Decision-Rules and Individual Values in Constitutional Choice’, American Political Science Review, lxiii (1969)Google Scholar; Taylor, Michael, ‘Proof of a Theorem on Majority Rule’, Behavioral Science, xiv (1969)Google Scholar; and Mueller, Dennis, Public Choice, New York, 1979Google Scholar, ch. 11.

⁷ For more discussion of this claim, see Brink, David O., ‘The Separateness of Persons, Distributive Norms, and Moral Theory’, Value, Welfare, and Morality, ed. Morris, C. and Frey, R., New York, 1993Google Scholar.

⁸ Nozick, Robert, Anarchy, State, and Utopia, New York, 1974, p. 29Google Scholar.

⁹ Ibid., p. 30; cf. John Rawls, A Theory of Justice, Cambridge, MA, 1971, esp. §§ 5–6.

¹⁰ This is similar to the responses to constraints found in Scheffler, Samuel, The Rejection of Consequentialism, Oxford, 1982Google Scholar, ch. 4; and Kagan.

¹¹ See Williams, Bernard, ‘A Critique of Utilitarianism’, in Smart, J. J. C. and Williams, B., Utilitarianism: For and Against, New York, 1973Google Scholar, and ‘Persons, Character, and Morality’, repr. in B. Williams, Moral Luck, New York, 1981.

¹² Scheffler, The Rejection of Consequentialism, esp. chs. 1–3; also see Scheffler, , Human Morality, New York, 1992Google Scholar, esp. chs. 6–7.

¹³ However, there are other interpretations of moral dilemmas that do not threaten voluntarism. See Brink, David O., ‘Moral Conflict and its Structure’, Philosophical Review, ciii (1994), 215–47CrossRefGoogle Scholar.

¹⁴ These thoughts may also suggest a reply to Rawls's argument that utilitarianism violates the strains of commitment within a well-ordered society. (See Rawls, esp. pp. 175–83, 496–502.) If a well-ordered society is one in which citizens are regulated by a sense of justice, informed by a utilitarian conception of impartiality, then utilitarianism may not impose undue strains of commitment.

¹⁵ Aristotle discusses friendship in Nicomachean Ethics, books viii–ix. He recognizes various forms of friendship – complete friendship among the virtuous, friendship for pleasure, friendship for mutual advantage, familial (including parental) friendship, and civic friendship – all of which involve reciprocal good will that is sustained by friends sharing thought and discussion. I discuss these and other aspects of Aristotelian friendship in ‘Eudaimonism, Love and Friendship, and Political Community’, Social Philosophy & Policy, xvi (1999), 252–89Google Scholar.

¹⁶ I think that this makes my view about associative duties reductionist in Scheffler's technical sense. See Scheffler, Samuel, ‘Relationships and Responsibilities’, Philosophy & Public Affairs, xxvi (1997), esp. 195Google Scholar. But notice that this sort of reductionism contains no commitment to the voluntarist claim that all obligations are voluntarily undertaken.

¹⁷ For communitarian conceptions that tie the content, as well as the ground, of associative duties to the terms of past association, see F. H. Bradley, ‘My Station and its Duties’, repr. in F. H. Bradley, Ethical Studies, 2nd edn., Oxford, 1927; Maclntyre, Aiasdair, After Virtue, London, 1981Google Scholar, esp. ch. 15; Sandel, Michael, Liberalism and the Limits of Justice, New York, 1982Google Scholar; Walzer, Michael, Spheres of Justice, New York, 1983Google Scholar; and Taylor, Charles, Sources of the Self, Cambridge, MA, 1989Google Scholar. The fact that my conception does not tie the content of associative duties to past association allows me to avoid Simmons's worries about grounding special obligations in morally imperfect associations. See Simmons, A. John, ‘Associative Political Obligations’, Ethics, cvi (1996), esp. 266Google Scholar.

¹⁸ Broad, C. D., ‘Self and Others’, Broad's Critical Essays in Moral Philosophy, ed. Cheney, D., London, 1971, esp. pp. 279 fGoogle Scholar.

¹⁹ Ibid., p. 280. As I (would like to) understand him, Broad should say that the altruism that common sense approves is of variable weight, rather than limited scope.

²⁰ Sidgwick, Henry, The Methods of Ethics, 7th edn., London, 1907, p. 427Google Scholar.

²¹ Ibid., p. 439.

²² Ibid., pp. 431–9.

²³ However, the intrinsic normative significance of special relations cannot be captured by recognizing the intrinsic value of associative relations within a consequentialist view. For instance, the consequentialist can assign special intrinsic value to friendship. But this won't allow the consequentialist to claim that an agent has reasons to give priority to his own friend when he could provide comparable or greater benefits to the friend of someone who is a perfect stranger to him.

²⁴ Cf. Nagel, Thomas, The View from Nowhere, New York, 1986, pp. 152 fGoogle Scholar.

²⁵ Nozick, pp. 153–5. Nozick contrasts these two conceptions of justice in holdings, whereas I want to contrast two different conceptions of associative obligation.

²⁶ This voluntarist principle, which insists that all obligations must be voluntarily undertaken by the agent, should not be confused with the voluntarist principle, discussed earlier (§ 3), according to which moral demands must be ones that agents are able (psychologically) to fulfil.

²⁷ See Scheffler, Samuel, ‘Families, Nations, and Strangers’, Lindley Lecture Series, Lawrence, University of Kansas, 1995Google Scholar, and ‘Relationships and Responsibilities’. In the latter work Scheffler offers his own rationale for associative duties. But I don't under-stand that rationale and won't try to assess it here.

²⁸ I have argued against analytical naturalism elsewhere; see Brink, David O., Moral Realism and the Foundations of Ethics, New York, 1989, ch. 6CrossRefGoogle Scholar and ‘Realism, Naturalism, and Moral Semantics’, Social Philosophy & Policy, xviii (2001)Google Scholar. I have also argued against Sidgwick's philosophical intuitionist account of utilitarianism; see Brink, David O., ‘Common Sense and First Principles in Sidgwick's Methods’, Social Philosophy & Policy, xi (1994)Google Scholar. Both Nagel and Hare think that if we model (i) the interpersonal combinatorial problem on (ii) the problem of how an individual would combine the interests of different individuals if she imagined living these lives seriatim then we are committed by temporal-neutrality to utilitarianism's person-neutrality. See Nagel, Thomas, The Possibility of Altruism, Oxford, 1970, pp. 138 f.Google Scholar; and Hare, R. M., Moral Thinking, Oxford, 1981, pp. 110 fCrossRefGoogle Scholar. But the model does not require the utilitarian conclusion. It would do so only if (iii) the individual treated these distinct lives (to be lived seriatim) as if they were parts or stages of one single super-life. But the question is precisely whether it is reasonable to understand (ii) as (iii).

²⁹ Interestingly, whereas Kagan recognizes that any moral conception, including utilitarianism, must meet the demand for a philosophical rationale, he presses this demand only against friends of constraints and options, not against utilitarianism itself See Kagan, pp. 18 f.

³⁰ I discuss these assumptions and their implications at greater length in ‘Self-love and Altruism’, Social Philosophy & Policy, xiv (1997)Google Scholar.

³¹ Locke, John, An Essay Concerning Human Understanding, ed. Nidditch, P. H., Oxford, 1975, bk. II, ch. xxcii, §§ 8, 15, 17–21, 23, 26Google Scholar.

³² (a) Non-responsible agents might usefully be praised or blamed for forward-looking (e.g., deterrent) reasons; but they do not deserve praise or blame, (b) In claiming that ‘person’ is a ‘forensic’ concept, Locke means not only that only persons can be held responsible but also that holding P2 responsible for Pi's actions makes sense only if P2 = P1. I'm here appealing to the former claim. I doubt the latter claim is true; I suspect responsibility presupposes deliberative continuity, rather than identity.

³³ This conception of personhood and responsibility is well represented in both ancient and modem traditions. Cf. Plato, , Republic, 437e–442cGoogle Scholar; Aristotle, , De Anima, ii.2Google Scholar and Nicomachean Ethics, 1102b13–1103a3, 1111b5–1113a14; Cicero, , De Officiis, 1.iiGoogle Scholar; Butler, Bishop, Fifteen Sermons, i.8, ii.12–15Google Scholar; Reid, Thomas, Essays on the Active Powers of the Human Mind, ii.2Google Scholar; Kant, Immanuel, Groundwork, 396, 437, 448Google Scholar; Critique of Pure Reason, A534/B562, A553–54/B581–82, A802/B830; Critique of Practical Reason, 61–2, 87; and T. H. Green, Prolegomena to Ethics, bk. II.

³⁴ The logic of identity will require that personal identity be defined in terms of non-branching psychological continuity. Similar mentalistic views are defended by Parfit, Derek, Reasons and Persons, Oxford, 1984, Part IIIGoogle Scholar; and Shoemaker, Sydney, ‘Personal Identity: A Materialist's Account’ in Shoemaker, S. and Swinburne, R., Personal Identity, Oxford, 1984Google Scholar.

³⁵ See ‘Self-love and Altruism’.

³⁶ Aristotle, , Nicomachean Ethics, i.7Google Scholar.

³⁷ I discuss issues about the scope of such concern at greater length in ‘Self-love and Altruism’, § X.

³⁸ Cf. Singer, Peter, ‘Famine, Affluence, and Morality, Philosophy & Public Affairs, i (1972)Google Scholar; and Murphy, Liam, ‘The Demands of Beneficence’, Philosophy & Public Affairs, xxii (1993)Google Scholar. Murphy's Co-operative Principle appeals to full compliance to set an upper limit on how much sacrifice beneficence can require; no agent can be asked to sacrifice more under partial compliance than she would under full compliance. If the consequentialist can incorporate Murphy's Co-operative Principle in a way that is not ad hoc, she can plausibly argue that consequentialism, though not complacent, is not overly demanding.

³⁹ These claims about accommodation bear some resemblance to claims Nagel and Scheffler make about the reconciliation of personal and impersonal moral demands. See Nagel, Thomas, Equality and Partiality, New York, 1991Google Scholar; and Scheffler, Human Morality, esp. ch. 8. The idea that partial and impartial moral demands might be best accommodated in different ethical spheres also bears some resemblance to Hegelian claims that different aspects of freedom are realized best in the different spheres of the nuclear family, civil society, and the state.

⁴⁰ I have benefited from useful discussion with an audience at the International Society of Utilitarian Studies, March 2000, at Wake Forest University.