¹ Howard-Snyder, Frances, ‘The Rejection of Objective Consequentialism’, Utilitas, ix (1997)Google Scholar. Page references in the text are to this paper, unless otherwise stated.

² Howard-Snyder makes the distinction between ‘objective’ and ‘subjective’ consequentialism a little differently from Railton, Peter, ‘Alienation, Consequentialism and Morality’, Consequentialism and its Critics, ed. Scheffler, Samuel, Oxford, 1988, pp. 113–19Google Scholar. Howard-Snyder's definitions of the various forms of consequentialism relate to actions only. Her discussion is thus limited to what Railton calls ‘act–consequentialism’, rather than to ‘rule– or ‘trait–consequentialism’ See Railton, pp. 117 f. I think that the arguments I make could be extended to the cases of rule- and trait-consequentialism, but I restrict myself to Howard-Snyder's specific formulations of consequentialism and one way of reformulating consequentialism.

³ So while I am intellectually able to make each of the moves which would defeat Karpov, I do not know which series of moves this is.

⁴ Like Howard-Snyder, throughout this paper, when I use ‘best consequences’ without any further qualification, I mean ‘consequences at least as good as any others’.

⁵ Howard-Snyder (244 f.) admits that she is rather vague about the sense of ‘can’ which is, for her, relevant to attributions of moral responsibility. It can be argued that, in the case of blame, ‘ought’ does not imply ‘can’, but ‘could have’. I have discussed this issue in Qizilbash, Mozaffar, ‘Obligation, Human Frailty and Utilitarianism’, Utilitas, vii (1995), 146 fGoogle Scholar. See also the recent discussion of James Griffin, in Griffin, James, Value Judgement: Improving Our Ethical Beliefs, Oxford, 1996, pp. 92–4, 157 fGoogle Scholar.

⁶ It can be argued that the criticisms which apply to objective consequentialism as an action–guiding theory also apply to it as a theory of right and wrong. For example, if a moral theory is insensitive to human weakness, it is implausible that it will provide a criterion of right and wrong human actions. Griffin discusses this issue in Value Judgement, p. 115, as well as in Griffin, James, ‘The Distinction Between Criterion and Decision Procedure: A Reply to Madison Powers’, Utilitas, vi (1994)Google Scholar.

⁷ For Howard-Snyder, on the other hand, if I can ø, in the sense of ‘can’ appropriate to ‘ought’ implies ‘can’ (which I have characterized as sense two), then my choosing to ø must make it ‘at least very likely’; (245) that I will ø. I think this understates her case. If it is up to me whether or not I ø, there must be a sense in which my choosing to ø ensures that I will ø.

⁸ I introduce my own terminology and do not use Howard-Snyder's related terminology involving ‘higher–level’; and ‘lower–level tasks’ (246), since she does not elaborate on her use of terms.

⁹ Howard-Snyder makes this point (245 f.).

¹⁰ As Griffin puts it: ‘we do not know where the limits of “can” lie’. See Value Judgement, p. 96.

¹¹ Qizilbash, 154 f.

¹² James Griffin has made this sort of appeal to ‘ought’ implies ‘can’ in several places, in the context of utilitarianism. His argument can, I suggest, be generalized to the case of consequentialism (objective and subjective). See Value Judgement, pp. 92–4; ‘The Distinction’; ‘On the Winding Road from Good to Right’, Value, Welfare and Morality, ed. Frey, R. G. and Morris, C. W., Cambridge, 1993CrossRefGoogle Scholar; ‘The Human Good and the Ambitions of Consequentialism’, The Good Life and the Human Good, ed.Paul, E. F., Miller, F. D. and Paul, J., Cambridge, 1992Google Scholar; ‘Mixing Values’, Proceedings of the Aristotelian Society, lxv (1991)Google Scholar.

¹³ I am very grateful to James Griffin and Alan Hamlin for their thoughtful comments on the first draft of this paper, and to Roger Crisp for helpful advice in the final stages of preparation. Any error is mine.