Published online by Cambridge University Press: 26 January 2009
Derek Parfit, Philip Pettit, and Michael Smith defend a version of consequentialism that covers everything. I argue that this version of consequentialism is false. Consequentialism, I argue, can only cover things that belong to a combination of things that agents can bring about.
1 Parfit, Derek, Reasons and Persons, Oxford, 1984, p. 25Google Scholar. Shelly Kagan makes a similar claim, writing that ‘that 'absolutely every kind of thing is a potential evaluative focal point’ for consequentialism, such as ‘atoms, the weather, sewer systems, suns’. See his ‘Evaluative Focal Points’ in Morality, Rules and Consequences, ed. Hooker, Brad, Mason, Elinor and Miller, Dale, Edinburgh, 2000, p. 151Google Scholar.
2 Pettit, Philip and Smith, Michael, ‘Global Consequentialism’, in Morality, Rules and Consequences, p. 122Google Scholar.
3 See Pettit, Philip, ‘Consequentialism’, in A Companion to Ethics, ed. Singer, Peter, Oxford, 1991Google Scholar.
4 Throughout this paper, I use ‘good’ as an evaluative term and ‘ought’ and ‘reason’ as deontic terms, and I use ‘ought’ as an overall term and ‘reason’ as a pro tanto term. The differences between these different kinds of terms will become clear from the implications that I attribute to claims that contain these terms.
5 This distinction is sometimes conflated with the distinction between versions of consequentialism on which the criterion of rightness is the same as the decision-making procedure and versions on which it is not, which was first made by Bales, R. E. in his ‘Act-Utilitarianism: Account of Right-Making Characteristic or Decision-Making Procedure?’ American Philosophical Quarterly, viii (1971)Google Scholar. Railton, Peter, in his ‘Alienation, Consequentialism and the Demands of Morality’, Philosophy and Public Affairs, xiii (1984)Google Scholar, calls the latter distinction that between ‘objective’ and ‘subjective’ consequentialism. Pettit, Philip and Brennan, Geoffrey, in their ‘Restrictive Consequentialism’, Australasian Journal of Philosophy, lxiv (1986)Google Scholar, call the latter distinction that between ‘unrestrictive’ and ‘restrictive’ consequentialism.
6 See, for example, Smart, J. J. C., ‘An Outline of A System of Utilitarian Ethics’, in Smart, J. J. C. and Williams, Bernard, Utilitarianism: For and Against, Cambridge, 1973CrossRefGoogle Scholar.
7 See, for example, Adams, Robert Merrihew, ‘Motive Utilitarianism’, Journal of Philosophy, lxxiii (1976)Google Scholar.
8 See, for example, Brandt, Richard B., A Theory of the Good and the Right, Oxford, 1979Google Scholar, and Hooker, Brad, Ideal Code, Real World, Oxford, 2000Google Scholar.
9 Of course, their versions of consequentialism can also insert the theory of the good into a combination of more than one theory of the right, like rule consequentialism does, or into some other theory of the right. And, of course, their versions of consequentialism can also claim that both X and Y are everything, that both X and Z are everything, that both Y and Z are everything, or that X, Y and Z are everything. But neither Parfit nor Pettit and Smith seem to intend their versions of consequentialism to do this, and even if they did, it would not affect my arguments.
10 Pettit and Smith, pp. 122 f. By ‘evaluand’, they mean both what I call a Y and what I call a Z. In making this claim about versions of indirect consequentialism, they are quoting Simon Blackburn, but they do so approvingly.
11 For all Pettit and Smith say here, their version of consequentialism could also adopt the collective rather than the straightforward theory of the right. But they do not seem to intend their version of consequentialism to do this, and even if they did, it would not affect my arguments.
12 Pettit and Smith, p. 122.
13 After all, Pettit, and Smith, claim that global consequentialism tells us, among other things, ‘whether it is right for it to be rainy or cloudy or sunny’ (p. 122)Google Scholar.
14 It would then use ‘right’ as an evaluative term rather than as a deontic term (see n. 4).
15 Pettit and Smith could also claim that global consequentialism uses the term ‘right’ the way other views use this term with regard to things an agent can bring about, and uses the term ‘right’ to mean ‘instrumentally good’ with regard to all other things. In that case, global consequentialism would consist of two parts: first, an addition to consequentialism's theory of the good that is not itself a version of consequentialism, and second, the version of consequentialism that I will call ‘semi-global consequentialism’ below. My argument would then still apply to global consequentialism's first part, which is enough to reach the conclusion I will reach at the end of this section.
16 It would then use ‘right’ as a pro tanto term rather than as an overall term (see n. 4).
17 Parfit, p. 25 (italics added).
18 However, by ‘possible’, Parfit may simply mean ‘physically possible’, rather than ‘possible for agents to bring about’.
19 Parfit, p. 32.
20 Ibid., p. 32.
21 Of course, it may not be true that Clare can benefit the stranger only if she does not love her child and can love her child only if she does not benefit the stranger. Parfit discusses a version of Clare's case in which this assumption is dropped (p. 33), but I am here only concerned with cases in which this assumption is true.
22 Claim (7) and claim (8) together imply claim (9) because ‘right’ is being used as an overall term here (see n. 4).
23 Parfit, p. 33.
24 Ibid.
25 Ibid., p. 36.
26 It would then use ‘right’ as a pro tanto term rather than as an overall term (see n. 4).
27 Claim (12) and claim (13) do not together imply claim (14) because ‘reason’ is a pro tanto term (see n. 4).
28 I assume here that the term ‘act’ is defined in a way that includes any way an agent can bring something about. Those who disagree with this definition can simply replace ‘acting’ with ‘bringing something about’ in what follows.
29 Of course, strengthening or weakening a motive of love will normally take more than just two acts, but this does not affect my argument.
30 In this respect, combined act consequentialism is similar to the biographical consequentialism defended by Crisp, Roger in his ‘Utilitarianism and the Life of Virtue’, Philosophical Quarterly, xlii (1992)Google Scholar, and to the broad consequentialist criterion for the evaluation of acts suggested (but rejected in favour of a non-consequentialist view) by Dancy, Jonathan in his Moral Reasons, Oxford, 1993, pp. 234–52Google Scholar, and his ‘Parfit and Indirectly Self-Defeating Theories’, in Reading Parfit, ed. Dancy, Jonathan, Oxford, 1997Google Scholar. According to Crisp, agents ought to live in such a way that the total amount of value in the history of the world is brought as close as possible to the maximum. According to Dancy, when evaluating acts, consequentialism should not just take into account the effects of an act itself, but also the effects of all conditions that have to be satisfied for the act to be performed. Crisp and Dancy argue for these views on consequentialist grounds, rather than by combining these consequentialist grounds with an appeal to the claim that ‘ought’ implies ‘can’. If their arguments are sound, they give consequentialists an additional reason to accept combined act consequentialism.
31 For very helpful comments on earlier drafts of this paper, I would like to thank Jonathan Dancy, Brad Hooker, an anonymous referee for this journal, and an audience at the conference of the International Society for Utilitarian Studies in March 2000.
