1 Philosophical Review 99 (1990) 157-84

2 Hancock, Roger briefly discusses a similar problem in Twentieth Century Ethics (New York: Columbia University Press 1974)Google Scholar.

3 This is a first approximation. The theory needs to accommodate less straightforward cases, such as quantification and indirect discourse. For a worked out deflationary theory of truth see Grover, Dorothy Camp, Joe and Belnap, Nuel ‘A Prosentential Theory of Truth’ Philosophical Studies 27 (1975) 73–125CrossRefGoogle Scholar. Ayer subscribed to the outline of a theory provided by Ramsey, F. in ‘Facts and Propositions’ Proceedings of Aristotelian Society Supplementary Vol. 7 (1927)Google Scholar.

4 In Logic and Knowledge, Marsh, R.C. ed. (London: George Allen and Unwin 1956)Google Scholar.

5 Of course this is not true for Russell himself, since he thought that ‘Elizabeth’ should be treated as a definite description.

6 The locus classicus of the suggestion that Ayer's logical analysis faces difficulties in this direction is Geach, P. ‘Assertion’ Philosophical Review 74 (1965)CrossRefGoogle Scholar. Recent attempts to offer a similar analysis that would overcome Geach's objections includes Blackburn, Simon Spreading the Word (New York: Oxford University Press 1984)Google Scholar; and Gibbard, Allan Wise Choices and Apt Feelings (Cambridge, MA: Harvard University Press 1990)Google Scholar.

7 Ayer, A. Language, Truth and Logic (New York: Dover 1952), 102Google Scholar

8 ‘Deflationary theory of truth’ is not Ayer's terminology.

9 Boghossian does not raise this possibility, but even our short discussion of Russell suggests that the grounds might also be considerations pertaining to the logical (as opposed to the surface) grammar of the sentence. More on this issue later.

10 For instance, in chapter VI of Language, Truth and Logic the questions of whether ethical judgments are capable of being true or false and whether they are statements of (empirical) fact are asked interchangeably. This is not a mere accident of Ayer's formulation. A deflationist is committed to identifying facts with true claims. For if there is anything for a fact to be other than a true proposition, the door is open for a robust theory of truth, in which to be robustly true is to describe a fact correctly.

11 ‘Express the same content’ is a place holder for one's favorite notion of semantic equivalence in this context.

12 ‘Values and Secondary Qualities’ in Essays on Moral Realism, Sayre-McCord, G. ed. (Cornell University Press 1988), 167nGoogle Scholar. McDowell is responding in this passage to Blackburn's version of irrealism (or quasi-realism, as Blackburn would prefer to call it).

13 ‘Traditionalist’ here should not mark a political persuasion but rather an attitude towards our conception of ourselves as moral agents.

14 This expression is taken from Mackie, J.L. Ethics: Inventing Right and Wrong (New York: Penguin Books 1977)Google Scholar.

15 One should distinguish here between a project of, so to speak, anatomy of morals, and the project of ethics. The first aims to uncover the psychological principles of our ethical behavior. None of my arguments would be compelling if Mackie's positive morality were part of such a project. However, this project is not the project of ethics, the project that involves figuring out what one ought to do and what kind of person one ought to be. To the extent that Mackie is not merely observing our moral practices but is actually assessing their legitimacy he cannot take his project to be the former.

16 The Phenomenology of Spirit, Miller, A.V. trans. (New York: Oxford University Press 1977)Google Scholar

17 For instance, according to Blackburn, once his preferred anti-realist position (i.e., projectivism) is fully developed ‘it becomes hard to see what else a realist can want what point he can be making in opposition to projectivism’ (‘Reply: Rule-Following and Moral Realism’ in Holtzman, S. and Leich, C. Wittgenstein: to Follow a Rule [Boston: Routledge & Kegan Paul 1981], 164)Google Scholar

18 Cf. Blackburn, ‘Reply,’ especially 164Google Scholar.

19 Or, under Horwich's deflationism, it is not a ‘complex or naturalistic property.’ See note 25 below.

20 Although the correspondence theory of truth will be my, so to speak, default robust theory of truth, it is important to notice that robust theories of truth other than the correspondence theory of truth might also have consequences that would not permit applying full-blown truth and objectivity to ethical discourse. So an epistemic conception of truth, which identifies truth with certain criteria of verifiability, might question whether ethical discourse satisfies the relevant criteria. Also, a coherence theory of truth, depending on the notion of coherence employed, might require that we unify all areas of discourse that may include true statements, in a way that might put in question whether both theoretical physics and ethical discourse could turn out to contain truths (Boghossian suggests, on similar grounds, that a coherence theory of truth might motivate attempts to reduce one region of discourse to another. See ‘The Status of Content,’ 165n.). If a deflationary theory of truth is adequate and delivers us from these metaphysical burdens, it should be preferred to these forms of robust theories of truth too.

21 Boghossian also takes deflationary theories of truth to imply content skepticism, which could be seen also as a criticism of these theories. I will not address this claim in this paper.

22 The notion of ‘logical form’ is being left intentionally vague, since I do not purport to have an account of it. However, I believe that nothing I want to say depends on a particular view about logical form or deep grammar. The considerations that lead one to a deflationary theory of truth such as the prosentential theory of truth are logical or grammatical under any account of logical form or deep grammar.

23 The prosentential theory of truth is presented in Camp, Grover and Belnap, ‘A Prosentential Theory of Truth.’ For a further defense and revision of this theory see Brandom, Robert ‘Pragmatism, Phenomenalism and Truth Talk’ in French, P. Uehling, T. and Wettstein, H. eds. Midwest Studies in Philosophy XII (Minneapolis: University of Minnesota Press 1988)Google Scholar.

24 This is, of course, a grossly oversimplified presentation of the theory. Just to give an example of an intentional oversight, I am ignoring complications raised by sentences such as ‘Newton's first law is true.’

25 Though I take the prosentential theory of truth to provide the clearest way out of Boghossian's criticisms, it is not the only deflationary theory of truth that can avoid them. A theory of truth along the lines of the one presented by Paul Horwich will work as well. According to Horwich's account ‘is true’ is a predicate. Horwich accepts that one can then say (trivially) about truth that it refers (in a deflationary sense) to a property. But he also claims that a deflationary theory of truth also exhibits that the kind of property that the truth predicate refers to is not, in his words, a ‘complex or naturalistic property.’ See his Truth (Oxford: Basil Blackwell 1990), 38-9.

26 Brandom, ‘Pragmatism, Phenomenalism and Truth-Talk,’ 78Google Scholar; emphasis added

27 Notice that prosententialists themselves do not think they can account for all uses of the word ‘true.’ For instance, sentences such as ‘The true and the beautiful are one’ or ‘it is not easy to find true love’ cannot be accounted for by such a theory. It seems, however, that such uses of the word ‘true’ are irrelevant to the issues of this paper and might merit separate treatment.

28 So Field's definition of a disquotational conception of truth as a conception according to which a correspondence notion of truth ‘serves no useful purpose at all’ makes disquotationalism a philosophical view and not a grammatical analysis of the English word ‘true.’ See Field, Hartry ‘The Deflationary Conception of Truth’ in MacDonald, G. and Wright, C. Fact, Science and Morality (New York: Basil Blackwell 1986), 86Google Scholar.

29 I do not wish to imply that by calling something ‘a technical term in philosophy’ one has thereby showed its futility. I am merely saying that ‘true₁’ should be evaluated according to its contribution to philosophical discussion.

30 See his Truth and Objectivity (Cambridge, MA: Harvard University Press 1992).

31 Or capable of being ‘true₂,’ in our words.

32 There are at least two other ways suggested by Wright. One way is to question whether ethical discourse answers the best explanation constraint; roughly, whether the states-of-affairs referred to in this discourse enter in the explanation of the beliefs we form about these states-of-affairs. I consider a similar kind of constraint in my discussion of Harman in the next section of this paper. Another way is to question whether ethical discourse is response-dependent (see Wright, chapter 3, appendix) in a way similar to secondary qualities. I think McDowell has successfully argued that no interesting form of anti-realism follows from accepting this form of response-dependence. In a nutshell, McDowell argues that to infer from a conception of secondary qualities as response-dependent the impossibility of being a realist about them confuses two senses of something being objective. In one sense, something is objective just in case it ‘can be adequately understood otherwise than in terms of dispositions to give rise to subjective states’ (‘Values and Secondary Qualities,’ 170). In this sense, secondary qualities and values are not objective. McDowell goes on to argue in ‘Values and Secondary Qualities’ that, given the relation between values and attitudes, to attribute this kind of objectivity to values is a quite straightforward conceptual impossibility. Someone who would try to argue that values are objective in this sense would be falsifying, rather than vindicating, moral experience. In another sense, something is objective just in case ‘it is there to be experienced, as opposed to being a mere figment of the subjective state that purports to be an experience of it’ (170). In this sense, secondary qualities and values can be seen as objective and existing independent of the experience of them. For instance, in the case of ‘red,’ ‘an object's being such as to look red is independent of its actually looking red to anyone on any particular occasion … an experience of something as red can count as a case of being presented with a property that is there anyway independent of the experience itself’ (168-9). If McDowell is right on this point, one should also reject an inference from the fact that discourse about values involves a necessary reference to our sensibilities to the conclusion that failure to be committed to this discourse cannot be understood as a cognitive shortcoming. This is an inference that Wright suggests at a number of points in Truth and Objectivity (see, for instance, 200).

33 This holds on the assumption that the case against anti-realism about beliefs can be barred in the same way I have suggested that it can be barred in the case of ethics. In any case, it is hard to see how the notion of a cognitive shortcoming can get off the ground without a notion of belief or an appropriate substitute for it.

34 McDowell, makes a similar point in his Projection and Truth in Ethics: The Lindley Lecture (Lawrence, KS: University of Kansas 1987)Google Scholar.

35 I say more about how we may see ourselves as gaining knowledge in the ethical realm in the discussion of Harman's views in the last section of this paper.

36 ‘Empiricism and the Philosophy of Mind’ in Minnesota Studies in the Philosophy of Science (Minneapolis: University of Minnesota Press 1956)

37 A similar line of argument is pursued by Horwich in Truth, ch. 4.

38 Realism and Truth (Princeton: Princeton University Press 1984), 93

39 Devitt, 93. I am being deliberately unfair to Devitt. As it stands these considerations do not seem to go very far in the way of showing that we need a notion of ‘truth.’ I am ignoring a number of steps in Devitt's argument to the effect that we need a notion of ‘truth’ to explain such situations. But my aim is merely to show that, insofar as these explanations of the phenomena are cogent, a deflationary notion of truth will do just as well as a robust one.

40 The idea of such a gradient is suggested by Putnam, Hilary in The Many Faces of Realism (LaSalle, IL: Open Court 1987), 27Google Scholar.

41 ‘Pragmatism, Davidson and Truth’ in LePore, E. and MacLaughlin, B. eds. Truth and Interpretation (New York: Basil Blackwell 1986)Google Scholar; emphasis added

42 This is a strand in Davidson's argument against skepticism in ‘A Coherence Theory of Truth and Knowledge’ in LePore and MacLaughlin, Truth and Interpretation.

43 I am not endorsing the idea that this is an appropriate epistemology for any field of inquiry.

44 See The Nature of Morality (New York: Oxford University Press 1977), ch. 1.

45 For this line of criticism of Harman, see Sturgeon, Nicholas ‘Moral Explanations’ in Copp, D. and Zimmerman, M. eds. Morality, Reason and Truth (Totowa, NJ: Rowman and Allanheld 1985)Google Scholar.

46 It is important to notice that this does not amount to conceding that there is no explanation of the observation that could be given at a different level and that would use moral facts.

47 I am indebted to Annette Baier, Robert Brandom, Phil Kremer, Hans Lottenbach, John McDowell, Jennifer Nagel, an editor, and two anonymous referees of this journal for helpful comments and suggestions.