1 Billington, Ray, Living Philosophy: An Introduction to Moral Thought (London: Routledge, 1988), p. 88. Original emphasis.CrossRefGoogle Scholar

2 Richards, Janet R., The Sceptical Feminist (London: Routledge, 1980), p. 64.Google Scholar

3 Cottingham, John, “Neo-Naturalism and its Pitfalls,” Philosophy, 58 (1983): 455.CrossRefGoogle Scholar

4 Ibid., p. 456. Original emphasis.

5 Hume, David, A Treatise of Human Nature, edited by Selby-Bigge, L. A. (Oxford: Clarendon, 1896).Google Scholar

6 Ibid., p. 469.

7 Ibid.

8 The term “Hume's Law” comes from Hare, R. M., The Language of Morals (Oxford: Oxford University Press, 1952), p. 29.Google Scholar The plausibility of this claim depends on the making of three provisos. First, its concern is not with ought statements that express merely hypothetical imperatives. Second, by ‘ought statement’ we mean a statement that contains the word ‘ought’ or a term which is synonymous with it. Third, the application of the law should be restricted to cases in which the premises and conclusion are not related via linguistic rules. (As John Searle has noted, it is part of the meaning of ‘promise’ that someone who says “I promise toφ”ought φ when it is appropriate to do so; but such a case does not detract from Hume's core doctrine.) All of these points are made by Mackie, J. L. in Hume's Moral Theory (London: Routledge, 1980), p. 62.CrossRefGoogle Scholar

9 The standard interpretation has most recently been disputed by Morris, Michael, The Good and the True (Oxford: Oxford University Press, 1992), pp. 159–62.CrossRefGoogle Scholar However, Morris's interpretation is quite compatible with the claim that the is/ought gap and the fact/value distinction are distinct; one of the aims of his book is to make just this point.

10 Saul Kripke's argument that there cannot be a dispositional analysis of meaning and understanding is an instance of the claim that an ought cannot be derived from an is. See his Wittgenstein on Rules and Private Language (Oxford: Blackwell, 1982)Google Scholar, chap. 2. According to Kripke, it is wrong to think that if one means plus by‘+’, one will answer ‘125’ when asked to give the sum of 68 and 57; obviously, one could mean plus by ‘+’ and yet make a mistake in this (and other) cases. To allow for the possibility of computational error, we must say that if one means plus, one ought to answer ‘125’. Computational error occurs when one fails to answer as one ought. Kripke says that the conclusion to be drawn from this is that “[t]he relation of meaning and intention to future action isnormative, not descriptive” (ibid., p. 37). This way of putting it, however, may prompt the erroneous thought that it is the fact/value distinction, rather than the is/cought gap, that has application to the theory of meaning and understanding.

11 Hume, Treatise, p. 469.

12 Ibid., p. 470. Emphasis added.

13 Ibid., p. 458.

14 Wright, Crispin, Truth and Objectivity (Cambridge, MA: Harvard University Press, 1992), pp. 1–2.Google Scholar

15 This is David Wiggins's way of explaining the putative distinction. See his “Truth, Invention and the Meaning of Life,” reprinted in his Needs, Values, Truth (Oxford: Blackwell, 1987), p. 95.Google Scholar

16 If facts are simply true thoughts, then they are found out rather than found.

17 See Plato, Protagoras, 322b9-d5.

18 Obviously, this view has its origin with Hume, but has been developed more recently by Ayer, A. J., Language, Truth and Logic (London: Gollancz, 1936)Google Scholar, chap. 6; Mackie, J. L., Ethics (London: Penguin, 1978);Google Scholar and Blackburn, Simon, Spreading the Word (Oxford: Clarendon, 1984)Google Scholar, chap. 6. We do not wish to underplay the significant differences between these representatives of the “expressivist” tradition. Ayer and Mackie think that our moral discourse, inasmuch as it embodies the assumption that it aims at truth, is in error and hence stands in need of revision. Blackburn and (Blackburn argues) Hume think that this conclusion may be avoided. Blackburn terms the project of explaining why our moral discourse is the way it is, given that expressivism is true, the enterprise of “quasi-realism” (Blackburn, Spreading the Word, p. 171).

19 Hume, Treatise, p. 470.

20 This problem wasfirstraised by Geach, P. T. in “Ascriptivism,” reprinted in his Logic Matters (Oxford: Black well, 1972)Google Scholar, but has been discussed more recently by, among others, Wright, Crispin in “Realism, Antirealism, Irrealism, Quasi-Realism,” in Midwest Studies in Philosophy, Vol.12, Realism and Antirealism, edited by French, P., Uehling, T. and Wettstein, H. (Minneapolis: University of Minnesota Press, 1988)Google Scholar, and Nick Zangwill, in “Moral Modus Ponens,” Ratio, N.S. 5, 2 (1992): 177-93.

21 Simon Blackburn suggests that the expressivist may solve the puzzle by construing the conditional premise as itself an evaluation. According to Blackburn, the conditional expresses moral approval towards combining a disapproval of stealing with a disapproval of encouraging one's students to steal (Blackburn, Spreading the Word, pp. 192-93). But as Crispin Wright has noted, this makes the person who accepts the argument's premises but not its conclusion guilty of a moral offence, that of failing to have a combination of attitudes which she approves (Wright, Truth and Objectivity, pp. 32-33). As such, Blackburn fails to do justice to the evident fact that the offence in question is a logical one, that of being inconsistent.

22 Ibid.

23 Wright, while being a representative of this strategy, denies that the thinnest possible notion of truth is, strictly speaking, the deflationary notion (ibid., chap. 1).

24 Jackson, Frank, “Realism, Truth and Truth Aptness,” Philosophical Books, 35 (1994): 168–69.Google Scholar

25 As Jackson notes, our response to pluralism concerning truth mirrors that given by Quine to the philosopher who takes there to be different types of existence (ibid., p. 168). As we may remember, Quine's claim is that there are different types of thing that exist, rather than different types of existence.

26 Including Wiggins, Needs, Values, Truth, pp. 95-96.

27 Wiggins, David, “Cognitivism, Naturalism and Normativity: A Reply to Peter Railton,” in Reality, Representation and Projection, edited by Haldane, J. and Wright, C. (Oxford: Oxford University Press, 1993), p. 307.Google Scholar

28 Moore, G. E., Principia Ethica (Cambridge: Cambridge University Press, 1903), p. 44.Google Scholar

29 Railton, Peter, “What the Non-Cognitivist Helps Us to See, the Naturalist Must Help Us to Explain,” in Reality, Representation and Projection, edited by Haldane, J. and Wright, C. (Oxford: Oxford University Press, 1993), p. 291.Google Scholar

30 Wiggins, David, “A Neglected Position?”, in Reality, Representation and Projection, edited by Haldane, J. and Wright, C. (Oxford: Oxford University Press, 1993), p. 335–36.Google Scholar

31 Ibid., p. 336.

32 Ibid., p. 331.

33 Davidson, Donald, “Mental Events,” reprinted in his Essays on Actions and Events (Oxford: Oxford University Press, 1980), p. 222.Google Scholar

34 If we were to find attractive an analogy between values and secondary qualities (and we are not committing ourselves to the view that we should), the points of the previous few paragraphs should persuade us that such an analogy can only go so far. Let us suppose that an object's being red consists in its looking red (to normal observers in standard circumstances). Even if this claim is correct, the equivalent claim for an object's having an evaluative property does not hold. Impressed by the open question argument, we should say that an object's possessing a certain value does not consist in its having a power to evoke responses of a given kind, but in its meriting those responses. (This point is well made by McDowell, John, “Values and Secondary Qualities,” in Morality and Objectivity, edited by Honderich, T. [London: Routledge, 1985], p. 118.)Google Scholar

35 This beautiful phrase, and the general tenor of the final remarks of this section, come from John McDowell's hugely thought-provoking John Locke lectures in Oxford in 1991, since published us Mind and World (Cambridge, MA: Harvard University Press, 1994).Google Scholar

36 See, for instance, Socrates' argument against Thrasymachus’ conception of phusis (nature) in Republic I.

37 A notable exception to this tendency is Baldwin, Thomas, G. E. Moore (London: Routledge 1990), pp. 80–86.Google Scholar Although Baldwin's arguments for resisting the conflation of the is/ought gap with the naturalistic fallacy overlap with ours, his considerations on naturalism and anti-naturalism, which follow from his own critical distinction between reductionism and anti-naturalism, are quite different from ours.

38 Moore, Principia Ethica, p. viii and chap. 5 passim. It should not be concluded from Moore's characterization of the “second question” of ethics thai, at this point, he disregards Hume's warning about is and ought. True, Moore maintains that ethical judgments on conduct should be supported by empirical generalizations about the likely effects of the actions they recommend. However, he always took care to add that such generalizations “also involve an ethical judgment proper, i.e., the judgment that certain effects are better, in themselves, than others” (ibid., p. 146). He thereby endorses Hume's view that the major premise of any moral reasoning must itself be evaluative. To that extent Moore's ideal utilitarianism differs from, for example, Bentham's hedonistic variety, which does involve a disregard of Hume's Law.

39 Ibid., p. 2.

40 Ibid., p. 20.

41 Ibid., p. 7.

42 Ibid., p. 10. Original emphasis.

43 See, for example, Geach, P. T., “Good and Evil,” Analysis, 17 (1957): 33–42.CrossRefGoogle Scholar

44 Moore, Principia Ethica, p. 10. Original emphasis.

45 Ibid., chap. 2 passim.

46 Ibid., chap. 4 passim.

47 Ibid., p. 15.

48 Ibid., p. 14.

49 Ibid., p. 114. Original emphasis.

50 Ibid.

51 Ibid., pp. 183-86.

52 Ibid., p. 188 (emphasis added). For the purposes of the present argument we shall ignore the intuitionistic nature of Moore's claims in chap. 6 of Principia Ethica.

53 Ibid., p. 191.

54 Ibid., p. 217.

55 “[T]his small attention wou'd … let us see that the distinction of vice and virtue is not founded merely on the relations of objects” (Hume, Treatise, p. 470).

56 Ruskin, John, Modern Painters (London, 1843)Google Scholar, Part 4, chap. 12.

57 W. K. Wimsatt and Monroe Beardsley have further sanctioned its semantic extension by the introduction, in literary studies, of their famous “intentional fallacy.” See Wimsatt, W. K. and Beardsley, Monroe, “The Intentional Fallacy,” in The Verbal Icon, edited by Wimsatt, W. K. (Lexington: University of Kentucky Press, 1954).Google Scholar

58 In her “G. E. Moore on the Ideal,” Mary Midgley offers a stimulating, though internally contradictory, interpretation of Principia Ethica, which is due to her non-cognitivist reading of the naturalistic fallacy. See Mary Midgley, “G. E. Moore on the Ideal,” in her Heart and Mind: The Varieties of Moral Experience (London: Methuen, 1981).Google Scholar

59 We would like to thank Tracy Bowell, Susan Ford, Axel Stern, Nick Unwin and two anonymous referees for their patient and thoughtful criticism of previous drafts of this paper.