Published online by Cambridge University Press: 13 January 2009
Does morality override self-interest? Or does self-interest override morality? These questions become important in situations where there is conflict between the overall verdicts of morality and self-interest, situations where morality on balance requires an action that is contrary to our self-interest, or where considerations of self-interest on balance call for an action that is forbidden by morality. In situations of this kind, we want to know what we ought simpliciter to do. If one of these standpoints over-rides the other, then there is a straightforward answer. We ought simpliciter to act on the verdict of the overriding standpoint.
For purposes of this essay, I assume that there are possible cases in which the overall verdicts of morality and self-interest conflict. I will call cases of this kind “conflict cases.” The verdict of morality in a conflict case would be a proposition as to what we ought morally to do, or as to what we have the most moral reason to do; the verdict of self-interest would be a proposition as to what we ought to do in our self-interest, or as to what action is best supported by reasons or considerations of self-interest. These propositions are action-guiding or normative in a familiar sense. The conflict between morality and self-interest in conflict cases is there-fore a normative conflict; it is a conflict between the overall verdicts of different normative standpoints. I take it that the question of whether morality overrides self-interest is the question of whether the verdicts of morality are normatively more important than the verdicts of self-interest. In due course, I will explain the idea of normative importance as well as the ideas of a normative proposition and of a reason.
1 I do not assume that the overall verdicts of morality and self-interest are always that some particular action is required. For all that I say, quite different verdicts are possible, including the verdict that a situation is a moral dilemma. For simplicity, I limit attention to cases in which morality and self-interest require a particular action.
2 Nothing of importance to my argument turns on the questions that divide realists from antirealists or cognitivists from noncognitivists in ethics.
3 The overall verdict of “Reason” could be complex. For example, it could be that several options are equally acceptable overall.
4 See Williams, Bernard, “Moral Luck,” in his Moral Luck (Cambridge: Cambridge University Press, 1981), pp. 22–24, 36–39.CrossRefGoogle Scholar See also Slote, Michael, Goods and Virtues (Oxford: Oxford University Press, 1983), pp. 77–107;Google Scholar and Wolf, Susan, “Moral Saints,” Journal of Philosophy, vol. 79 (1982), pp. 419–39.CrossRefGoogle Scholar
5 Plato, Republic, 359d-360b.
6 This point is made by Scheffler, Samuel in his Human Morality (New York: Oxford University Press, 1992).Google Scholar
7 A view of this kind was suggested by Thomas Hill, Jr., in discussion.
8 Strictly speaking, I may not need to assume the possibility of cases of conflict. For even if cases of conflict are not possible, there would still be the two kinds of verdict, and we could ask whether one kind overrides the other.
9 Richard Kraut discusses the idea of the good for a person in his essay “Desire and the Human Good,” Proceedings and Addresses of the American Philosophical Association, vol. 68, no. 2 (November 1994), pp. 39–54.
10 See my Morality, Normativity, and Society (New York: Oxford University Press, 1995).
11 This is an account of “normative reasons.” For a similar view, see Smith, Michael, The Moral Problem (Oxford: Basil Blackwell, 1994), p. 95.Google Scholar
12 Elsewhere, I have argued that moral standards are relevantly authoritative just when, roughly speaking, society needs their currency among its members in order to flourish. That is, their currency would be in the interest of the society as a whole. See my Morality, Normativity, and Society.
13 See my Morality, Normativity, and Society, ch. 9. The standard calling on me to satisfy my basic needs is not subjective in the sense I explained before. I cannot take the space to explain this here.
14 Samuel Scheffler says that the “claim of overridingness” is “the claim that it can never be rational knowingly to do what morality forbids” (Human Morality, p. 52).
15 Michael Smith defends a similar view in The Moral Problem, pp. 85–91, 130–202, esp. pp. 182–84.
16 Williams, “Moral Luck,” pp. 22–24, 36–39. Williams's discussion is difficult to follow, so I cannot claim to have the only or the best interpretation.
17 Michael Slote discusses “admirable immorality” in his Goods and Virtues, pp. 77–107.
18 See Wolf, “Moral Saints.”
19 The idea of the good for a person is discussed by Richard Kraut in “Desire and the Human Good.” Thomas Hurka has developed an account of personal excellences in Perfectionism (New York: Oxford University Press, 1994).
20 Philip Clark has stressed this point, in discussion.
21 Let me briefly discuss two possible objections to the conclusion that R must be normatively more important than S. (1) Perhaps R is exactly as important normatively as S. But then, what would constitute them as exactly equally important? It could not be the assessment of R by S and vice versa. This is simply to argue in a larger circle than we argue if we suppose that R is identical to S. There would then have to be some additional standard T that specifies criteria such that R and S are equally important. Now there is the problem of the relative status of R, S, and T, which raises the very issues that were supposed to be escaped by supposing that R and S are equally important. (2) Perhaps R and S are normatively “incomparable.” But then all we have is that a standard of indeterminate significance ranks S as superior to all other standpoints. This is not sufficient to constitute S as normatively superior simpliciter; at most it shows that S is superior from the standpoint of R. Moreover, there might be some other standpoint T that is superior to R and that assesses S as normatively inferior to some other standpoint. To eliminate this possibility, we would have to suppose that no standpoint is superior to either R or S. But then there must be some other standard that specifies criteria of evaluation according to which no standard is superior to R or to S. This begins the regress anew. I see no way to avoid a regress problem.
22 Hence, in most situations of deliberation it is false that there is something we could choose that would be best, period. There may be some situations in which this thought is true, however; for if all the special reasons speak in favor of the same option, then any plausible candidate for the standard of Reason would select that option as best, period. Compare the problem of social choice that was explored by Kenneth Arrow in his Social Choice and Individual Values, 2d ed. (New Haven, CT: Yale University Press, 1963). Arrow argued that there is no function that takes individual preference rankings over social options and yields a social preference ranking over those options–no function that meets certain theoretically desirable conditions. I am arguing that there is no standard that takes the rankings over sets of options given by the various special standards for choice and yields a supreme ranking over these sets of options–no standard that meets certain theoretically desirable conditions, including especially that it be comprehensive and supreme. In this note, I am pointing out that in certain special circumstances only one ranking could plausibly be accepted. In particular, if all the special standards rank the options the same way, then that is the only way that the supreme standard or Reason could rank the options.