¹ I am here concerned with the meanings of in thought, as well as the meanings associated with sounds in public languages. Thus the scope of the discussion will encompass the normativity of mental content.

² It might be thought that proponents of non-normative theories of truth or meaning would have nothing to worry about even if these phenomena were intrinsically normative. For, it might be thought that normative properties can be reduced to non-normative properties. However, there is no plausible conceptual (i.e. definitional) reduction of normative to non-normative notions; and a weaker (merely constitutive) reductionism could not help those who advocate non-normative accounts of our concepts of truth and meaning. Moreover, there is absolutely no reason to think that any of the plausible non-normative non-conceptual accounts of truth or meaning (such as the use theory) could possibly result from combining a normative theory of truth or meaning and a reductive analysis of normativity.

³ The redundancy theory says that ‘The proposition that p is true’ means the same as simply ‘p’, whereas certain more recent forms of deflationism about truth — including the ‘minimalism’ defended in my Truth, 2nd edition, (Oxford University Press, 1998Google Scholar) - propose to define the truth predicate by means of our commitment to the material biconditional, ‘The proposition that p is true ↔ p’.

⁴ Dummett, M., ‘Truth’, Proceedings of the Aristotelian Society n.s. 59, pp.141–162 (1959)CrossRefGoogle Scholar.

⁵ Something like this position has been re-iterated by Wright, Crispin in his book Truth and Objectivity (Cambridge, Mass.: Harvard University Press, 1992)Google Scholar. He maintains that deflationism is wrong on the grounds that truth is a goal, hence a genuine property, not merely a device of generalisation. See the Postscript of my Truth (2nd edition) for further discussion of Wright's argument. Bernard Williams has also expressed sympathy for the view that redundancy-style accounts of truth cannot do justice to its value. See his ‘Truth in Ethics’, Ratio 8, 227–42 (special issue, Truth in Ethics, ed. B. Hooke). For similar ideas see Hilary Putnam's ‘Does The Disquotational Theory of Truth Solve All Philosophical Problems?’ and ‘On Truth’, both reprinted in hisWords and Life, edited by Conant, J., (Cambridge, Mass.: Harvard University Press, 1995)Google Scholar.

⁶ Kripke, S., Wittgenstein: On Rules and Private Language (Oxford: Blackwell, 1982). See especially pages 11, 21, 24 and 87.Google Scholar

⁷ Some other philosophers influenced by Kripke and sympathetic to the intrinsic normativity of meaning are McDowell, John (‘Wittgenstein on Following a Rule’, Synthese, 1984)Google Scholar, Gibbard, Allan (‘Meaning and Normativity’, Philosophical Issues 5 - Truth and Rationality, edited by Villanueva, E., (Atascadero CA: Ridgeview Publishing Company, 1994, 95–115)Google Scholar, Brandom, Robert (Making It Explicit Cambridge, Mass.: Harvard University Press, 1994)Google Scholar, Boghossian, Paul (‘The Rule Following Considerations’ Mind 98,1989, 507–50)CrossRefGoogle Scholar, and Blackburn, Simon (‘The Individual Strikes Back’, Synthese 10, 1984, 281–301)CrossRefGoogle Scholar.

⁸ For references, see footnotes [4] and [6].

⁹ Here I am simplifying the issue in a couple of respects. First: I am focusing on why it is desirable, if one has a certain belief, for that belief to be true rather than false; and I am not explicitly considering the desirabil

¹⁰ Accepting a sentence u is, to a very first approximation, a matter of uttering it to oneself. More accurately it involve relying on that sentence in theoretical and practical reasoning. For a little further discussion see footnote 16.

¹¹ For a more detailed discussion of truth as a device of generalisation, see my Truth op. cit.

¹² For a more elaborate version of this explanation, see Truth, chapter 3, (op. cit.) and Barry Loewer's ‘The Value of Truth’ in Villanueva, E., (ed.,) Philosophical Issues 4, (Atascadero, Cal.: Ridgeview Publishing Company, 1993).Google Scholar

¹³ In this essay I am concerned to show that the norms (T), (B), and (M), provide no basis for the claim that truth or meaning are intrinsically normative. Granted, there are other language-related norms in light of which such a claim might be made:

¹⁴ For an elegant and more sophisticated account of moral virtue along realist lines, see Thomson's, Judith Jarvis ‘The Right and the Good’, Journal of Philosophy 94, 1997, 273–98CrossRefGoogle Scholar.

¹⁵ The emotivist-expressivist position is traditionally taken to include the view that there are no normative facts and that normative pronouncements do not have a truth-value. And in that case it would be obvious that there can be no explaining why truth is morally valuable. But deflationism about truth suggests that what is distinctive and defensible in the emotivist-expressivist position is merely the identification of normative commitment with a kind of desire or pro-attitude, and that the existence of normative truth is not thereby precluded. If this is right, then we can equally recognise normative facts (in a deflationary sense of ‘fact’). However, these facts will not be part of the causal order and will not be susceptible of naturalistic explanation. For further discussion see Truth (op. cit.) and my ‘Gibbard's Theory of Norms’, Philosophy and Public Affairs 22, 1993,67–78.Google Scholar

¹⁶ The present strategy does of course leave us with an unexplained norm, and it might be alleged (a) that this norm can hold only if belief* (i.e. sentence acceptance) and desire* are intrinsically normative; and (b) that meaning must be explained in terms of those notions; therefore, (c) that meaning is intrinsically normative. I agree with (b). Therefore, in order to meet this challenge, I would elaborate the following three pronged criticism of (a):

¹⁷ One might wonder, if we ultimately need to rely on the emotivistexpressivist view, what the point was of any of the preceding argument. For couldn't we have invoked that view at the very beginning of this essay in order to dispose of any suspicion that the explanation of (T) and (M) requires that truth and meaning be intrinsically normative? In fact, the emotivist-expressivist view of (T) and (M) does not undermine the point of my pragmatic explanation of their adoption, because that explanation provides support for the emotivist-expressivist view. For it shows that these normative claims do not depend on our having in mind some analysis of ‘ought to be done’; whereas, from the realist-cognitivist perspective, the belief that something ought to be done would have to be justified in light of some account of what ought-to-be-done-ness amounts to.

¹⁸ The present paper is also published in What Is Truth?, edited by Richard Schantz (Gruyter: Berlin, New York). I would like to thank him and Anthony O'Hear for allowing the paper to appear in both volumes. Many of the ideas in it derive from Chapter 8 of my Meaning; but I have revised and extended that discussion in order to address a variety of problems. I am grateful to Paul Boghossian, Wolfgang Kuenne, Michael Martin, David Owens, Barry Smith, Helen Steward, and Albert Visser for bringing some of these problems to my attention.