¹ We restrict our remarks to direct, universalist, act-consequentialism throughout.

² Scheffler, S., The Rejection of Consequentialism, Oxford, 1982Google Scholar; Parfit, D., Reasons and Persons, Oxford, 1984, p. 27Google Scholar; Nagel, T., The View from Nowhere, Oxford, 1986, chs. 8, 9Google Scholar; Darwall, S., ‘Agent-Centred Restrictions from the Inside Out’, Philosophical Studies 1 (1986), 291–319CrossRefGoogle Scholar; Pettit, P., ‘Universalizability without Utilitarianism’, Mind, xcvi (1987), 74–82CrossRefGoogle Scholar; McNaughton, D. and Rawling, P., ‘Agent-Relativity and the Doing-Happening Distinction’, Philosophical Studies, lxiii (1991), 167–85CrossRefGoogle Scholar; McNaughton, D. and Rawling, P., ‘Honouring and Promoting Values’, Ethics, cii (1992), 835–43CrossRefGoogle Scholar; McNaughton, D. and Rawling, P., ‘Deontology and Agency’, The Monist, lxxvi (1993), 81–100CrossRefGoogle Scholar; Dancy, J., Moral Reasons, Oxford, 1993, chs. 10, 11Google Scholar; McNaughton, D. and Rawling, P., ‘Value and Agent- Relative Reasons, Utilitas, vii (1995), 31–47.CrossRefGoogle Scholar

³ For a discussion of these difficulties see our ‘Agent-relativity and the Doing-Happening Distinction’, § II.

⁴ Strictly, we need to add a rider to the effect that the AR rule is not equivalent in ethical content to any AN rule.

⁵ Nagel, , p. 165.Google Scholar

⁶ Terminology varies: Scheffler talks of agent-centred prerogatives (=options) and restrictions (=constraints).

⁷ The term ‘option’ is an unfortunate one, since it characterizes this alleged feature of morality from a consequentialist rather than a deontological perspective – one can only exercise one's option not to maximize the good where one's interests are threatened if one has a duty to maximize the good, which is just what many deontologists deny.

⁸ Our formalization enables us to capture the agent-relativity in all three of these areas. For details see ‘Value and Agent-relative Reasons’.

⁹ Skorupski, J., ‘Agent-Neutrality, Consequentialism, Utilitarianism … A Termino-logical Note’, Utilitas, vii (1995), 49–54.CrossRefGoogle Scholar

¹⁰ For details see ‘Honoring and Promoting Values’, 838. Skorupski also distinguishes the rational from the irrational egoist by claiming that the latter offers a reason which is agent-neutral but not universalizable; that is, he takes the fact that an act is of benefit to him as a reason for everyone to perform that act. ‘[T]he irrational egoist's principle ineliminably contains a rigid designator in a certain position in the reason-predicate’ (p. 50). We agree with Skorupski that properly formulated rules should not contain rigid designators.

¹¹ There is nothing to prevent a deontologist adhering to both rules; he or she need not take the (clannist) view that only his or her children matter.

¹² Skorupski might claim that deontology was only agent-relative ‘as such’ if all the rules in a deontological stystem had to be agent-relative. But this is clearly too strong; all deontologies include rules which are agent-neutral on any account.

¹³ We take it that his discussion of Browne, 's proposal in § 8Google Scholar is not intended as an argument to support his formalization as against ours. We agree with him that the position he calls ‘agent-relative teleology’ is confused, and we have argued as much in ‘Values and Agent-relative Reasons’. But deontology, which does not take the theory of value to be prior to the theory of the right, does not suffer from these confusions.

¹⁴ McNaughton, David makes these points about impartiality in Moral Vision, Oxford, 1988, pp. 167–8.Google Scholar

¹⁵ We are grateful to Brad Hooker for comments on an earlier version of this paper.