¹ The hegemonist conception of moral theory lies beneath a wide array of objections made to various moral theories. The best known is the complaint that we cannot use consequentialist theories as guides, since agents too often do not know enough about future consequences of acts. Another common theme is the communitarian idea that various liberal moral codes tend to be self-undermining. If we adhere to these standards in a strict way, the argument goes, values which even the liberal endorses will erode. Finally, many writers assume that moral codes can be too demanding, not in the sense that they make unfair demands, or that they place undeserved burdens upon us, but in the different sense that the theory exceeds our motivational capacities. However interesting these objections are, I shall make no attempt to discuss them here.

² Railton, Peter, ‘Alienation, Consequentialism and the Demands of Morality’, Philosophy and Public Affairs, xiii (1984)Google Scholar .

³ See Kappel, Klemens, ‘Has Dancy Shown a Problem in Consequentialism?’, Theoria, lxv (1999)Google Scholar . This paper was prepared in response to Dancy, Jonathan, ‘Parfit and Indirectly Self-defeating Theories’, Reading Parfit, ed. , Dancy, Oxford, 1997Google Scholar .

⁴ Consequentialism, i.e. the view that the only thing that matters is that one brings about as much good as possible, is not a moral theory in this sense, among other reasons because consequentialism as such remains neutral on what the good is. Rather, the term ‘consequentialism’ denotes a broad class of moral theories (or would-be theories) with certain features in common. Neither is Kantianism, broadly conceived, a moral theory, but a denominator of a class of moral theories (or attempts to specify moral theories). Thus, given a sufficiently precise interpretation of, say, Kant's Categorical Imperative, it would become a moral theory, just as a precise version of a consequentialist theory supplemented with an axiology would be a moral theory. Virtue theory is also really a class of theories, rather than one theory. At least some versions of virtue theory fits the above characterization of a moral theory, since virtue theory provides its distinctive specification of right- and wrong-making characteristics of actions. Thus, one version of virtue theory holds that what makes a particular action right is the fact that the virtuous person would choose to perform that action. Another version holds that what makes a particular action right is the fact that it possesses properties such as being temperate or being benevolent, properties that it takes a virtuous person to detect. Either way, virtue theory contains a specification of right- and wrong-making characteristics of actions.

⁵ It is not necessary that truth is here understood realistically.

⁶ Non-cognitivists might want to state this notion without making use of the concept of truth. This can be done, I believe, but we need not do so here.

⁷ There is also a difference between desiring to perform some action that X requires, and desiring to perform the action under the description of being required in the circumstances by X. Similar remarks apply to the notions of wanting some action and deciding to perform the actions. We can ignore these issues here, however.

⁸ Consider again a moral theory based on the Categorical Imperative. The Categorical Imperative might be interpreted as specifying right actions, and as not directly prescribing what dispositions and motivations agents ought to have. On this reading of the Categorical Imperative, it is a secondary and partly empirical question what dispositions and motivations agents ought to have. It might be that Kant and many Kantians would not endorse a theory like that. Perhaps they would prefer moral theories that directly prescribe motives and dispositions. Be that as it may, the important point is that moral theories other than utilitarianism and other forms of consequentialism make room for motivational indirection. Analogously, even virtue theory could be framed so as to permit motivational indirection. What is takes is a virtue theory which specifies right action (say, in terms of the action that the virtuous person would chose in the circumstances), but refrains from specifying directly which dispositions and motives agents ought to possess.

⁹ Note here that even for a fixed basic theory T, and a fixed world, ‘M’ does not refer to one unique set of norms. For each agent (or even for each agent at various stages in his or her life) there is an ideal set motivations and dispositions. Fortunately, however, we can ignore this and other complications.

¹⁰ Notice that though one is not compelled to do so, one can accept both non-hegemonism and indirection. That is, one could hold that a correct moral theory need not figure in the motivational life of any agent, but if it does, there is room for motivational indirection.

¹¹ For an early statement of this distinction, see Bales, Eugene, ‘Act-Utilitarianism: Account of Right-Making Characteristics or Decision-Making Procedure?’, American Philosophical Quarterly, viii (1971)Google Scholar .

¹² Stocker, Michael, ‘The Schizoprenia of Modern Ethical Theories’, Journal of Philosophy, xciii (1976)Google Scholar .

¹³ Stocker, 463.

¹⁴ Stacker, 463.

¹⁵ Stocker, 453 f.

¹⁶ Stocker, 454 f.

¹⁷ See , Parfit's discussion of this issue in his Reasons and Persons, Oxford, 1984Google Scholar , pt. 1.

¹⁸ Williams, Bernard, ‘The Structure of Hare's Theory’, Hare and Critics, ed. Seanor, N. and Fotion, D., Oxford, 1988, p. 190Google Scholar .

¹⁹ , Dancy, ‘Parfit and Indirectly Self-defeating Theories’, 12Google Scholar . See also Dancy, Jonathan, Moral Reasons, Oxford, 1993Google Scholar, esp. ch. 10. I discuss Dancy's views in Kappel, ‘Has Dancy Shown a Problem in Consequentialism?’

²⁰ See also my extended discussion of Dancy's views on this matter in Kappel, ‘Has Dancy Shown a Problem in Consequentialism?’

²¹ , Dancy, ‘Parflt and Indirectly Self-defeating Theories’, 18Google Scholar .

²² , Dancy, ‘Parfit and Indirectly Self-defeating Theories’, 17Google Scholar .

²³ A referee for this journal made this objection.

²⁴ Similar remarks apply to my discussion of objections to the non-hegemonist conception of moral theory in part III. Although these objections are raised in discussions of consequentialism, the objections to externalism, as far as I can see, are general.

²⁵ , Dancy, ‘Parfit and Indirectly Self-defeating Theories’, 12Google Scholar .

²⁶ See the remarks in n. 8.

²⁷ Langenfuss, William ‘Consequentialism in Search of a Conscience’, American Philosophical Quarterly, xxvii (1990), 135 fGoogle Scholar .

²⁸ As a referee for this journal did.

²⁹ Referee's report, p. 2.

³⁰ Referee's report, p. 2.

³¹ Griffin, James, ‘The Ambitions of Consequentialism’, The Good Life and the Human Good, ed. Paul, E., Miller, F., Paul, J., Cambridge, 1992, p. 131Google Scholar .

³² Rawls, John, A Theory of Justice, rev. edn., Cambridge, Mass., 1999, p. 115Google Scholar .

³³ Ibid.

³⁴ Compare to this remark that Stacker makes (Stocker, 463): ‘once we begin to believe that there is something beyond such activities as love which is necessary to justify them, it is only by something akin to self-deception that we are able to continue them’.

³⁵ My thanks to Roger Crisp, Peter Vallentyne and Torbjörn Tannsjo for helpful comments and encouragement, and to an anonymous referee for Utilitas who provided valuable critical remarks on an earlier version of this paper. Thanks also to participants in a workshop at Roskilde University held in May 2000, where an earlier version of this material was presented.