¹ Are there prescriptions whose action-guiding force for agents depends upon no interest at all? There could be if a substantive theory of rationality were correct, for then there would be commands of reason that depend upon no interest or motivational state of the agent. However, part of the circumstances that constitute the setting for the problem discussed in this paper is a sense that no substantive theory of rationality can be made to work.

² See Moore, G.E., Principia Ethica (Cambridge: Cambridge University Press, 1903), esp. pp. 15ff.Google Scholar Strictly speaking, Moore's argument was applied not to conceptions of a person's good, but to good itself. For reasons we need not go into here, Moore believed treating “good for so-and-so” as a basic category to be inherently confused; see Principia, pp. 98ff.

³ For an important early discussion, see Frankena, W.K., “The Naturalistic Fallacy,” ed. Philippa, Foot, Theories of Ethics (New York: Oxford University Press, 1967).Google Scholar

⁴ I say ‘obviously successful’ because it remains possible that once some existing analysis has been fully understood and assimilated, and once all accidental connotations of the notion of ‘good’ have been stripped away (assuming such a thing to be possible), the analysis could emerge as incontestable.

⁵ An alternative approach is to accept the descriptive grammer of discourse about good at face value, but then to say that since nothing answering to its descriptions – nothing with the required connection between descriptive character and prescriptive force – could exist, statements about a person's good are always false. For an example of such an “error theory,” applied to the moral case, see Mackie, J.L., Ethics: Inventing Right and Wrong (New York: Penguin, 1977).Google Scholar

⁶ For two historically important examples of expressivist accounts of value discourse, see Ayer, A.J., Language, Truth, and Logic, 2d ed. (London: Gollantz, 1946)Google Scholar, ch. 6, and Stevenson, C.L., Ethics and Language (New Haven: Yale University Press, 1944).Google Scholar

⁷ For an impressive example of the explanatory power of expressivism, see Gibbard, Allan, Wise Choices, Apt Feelings (Harvard University PressGoogle Scholar: forthcoming).

⁸ I certainly do not mean to have given a definition of ‘natural property’ or ‘naturalism’ in this way. More will be said about naturalism in subsequent sections. For now, let me note only that what seems to me to be of interest in naturalism is not a metaphysical doctrine about what sorts of things are part of – or not part of – “Nature,” but rather a methodological stance reflecting our experience of the ways in which useful predictive and explanatory theories have been achieved. Anyone who would be a naturalist in this methodological sense must regard it as a matter for inquiry, not definition, what sorts of properties can figure in developed, explanatory empirical theories – that is, what sorts of properties are “natural.”

⁹ I use the expression ‘happiness’ rather than ‘pleasure'; in characterizing hedonism because of the unfortunately narrow connotations of the latter. I do intend ‘happiness’ to pick out a class of experiential states.

¹⁰ For a view I find more plausible, see Railton, Peter, “Facts and Values,” Philosophical Topics, vol. 14 (Fall 1986), pp. 5–31.CrossRefGoogle Scholar

¹¹ I will not try to say with any rigor what it is for the value realized in a life to be good for the person living it. It seems uncontroversial in present-day philosophy that, at a minimum, in order for something to be good for a person, that person must exist at the time of realization. By this criterion, it could be part of my good that I have certain experiences or engage in certain activities, but it could not be part of my good that people think well of me after my death. Aristotle believed that one's eudaimonia could be affected – at least for a certain length of time – by events occurring after one's death (Nicomachean Ethics, book I, chapter 10), but that may be a reason for thinking that we should not without qualification translate eudaimonia as “a person's own good” or “happiness.”

¹² According to an important tradition in moral philosophy – utilitarianism – intrinsic nonmoral value is not only independent of moral value, but prior to it, part of its very foundation. This is a view I find compelling, and one which makes our discussion especially relevant to the foundations of ethics. However, I will not attempt to argue for such a view here. Our question about the possibility of naturalism lies within the theory of nonmoral value itself, and does not presuppose any particular moral theory.

¹³ Kant, Immanuel, Critique of Practical Reason, trans. Beck, L.W. (Indianapolis: Bobbs-Merrill, 1956), pp. 128ff.Google Scholar, and Religion within the Limits of Reason Alone, trans. Greene, T.M. and Hudson, H.H. (New York: Harper and Row, 1960), p. 5n.Google Scholar

¹⁴ For discussion of the use of reforming definitions, see Brandt, Richard, A Theory of the Good and the Right (Oxford: Clarendon, 1979), esp. pp. 3–16, 23.Google Scholar

¹⁵ Ramsey, , The Foundations of Mathematics and other Logical Essays, ed. Braithwaite, R.B. (London: Kegan Paul, 1931), pp. 288–89.Google Scholar

¹⁶ Nagel, Thomas, The View from Nowhere (New York: Oxford University Press, 1987), p. 144.Google Scholar

¹⁷ For examples of nonreductionist, naturalistic accounts of moral discourse, see: Miller, Richard W., “Ways of Moral Learning,” Philosophical Review, vol. 94 (October 1985), pp. 502–50CrossRefGoogle Scholar and Sturgeon, Nicholas, “Moral Explanations,” ed. Copp, D. and Zimmerman, D., Morality, Reason, and Truth (Totowa: Rowman and Allanheld, 1986).Google Scholar

¹⁸ I borrow this term from David Copp.

¹⁹ The useful term ‘track’ I have borrowed from Nozick's, Robert discussion of epistemology in Philosophical Explanations (Cambridge: Harvard University Press, 1981).Google Scholar

²⁰ In one famous case, a freighter deemed to be jinxed in virtue of its record of bizarre collisions was found to have an accidental cross–connection in its steering, which caused aberrations only under rather special conditions. Obviously, many of the ships deemed “unlucky” were innocent of special defect in design or construction, but suffered from poor seamanship or mere chance.

²¹ For a seminal discussion of an experience machine, see Nozick, Robert, Anarchy, State, and Utopia (New York: Basic Books, 1974).Google Scholar

²² I follow a standard practice in calling this sort of instrumental conception of rationality ‘Humean’. I do not mean to claim that this conception is, in fact, the whole of Hume's.

²³ At this point, it might be argued that in adopting a broadly Humean account of reasons and abandoning stronger notions of “rational command,” the hedonist has already adopted a form of skepticism about intrinsic good, skepticism of the kind that philosophers have often attributed to Hume himself. According to this argument, the idea of “rational command,” though esoteric, was in fact discovered by philosophers to be a logical presupposition of ordinary discourse about good. When the hedonist gives it up, he in effect engages in a quite extensive revisionism, so much so that he changes the subject.

The hedonist, for his part, can reply that he is unconvinced that the philosopher's idea of “rational command” is at the core of ordinary discourse about a person's good, and that, in any event, the fact that the doctrine of “rational command” leads to so powerful and revisionist a conclusion as skepticism about intrinsic good should suggest that this doctrine stands in need of much greater support than it has thus far received. The need for support will be larger still if the hedonist can make good his attempt to give a naturalistic account of prescriptivity.

²⁴ For an example, see Railton, Peter, “Moral Realism,” The Philosophical Review, vol. 95 (April 1986), pp. 163–207CrossRefGoogle Scholar, and “Facts and Values.”