Oxford, PHIB Philosophical 0031-8051 2002 43 41000 Blackwell UKScience Publishers Books Ltd Ltd 2002 RECENT WORK TRUTH JULIAN DODD The University of Manchester Substantial Theories and Deflationary Theories The concept of truth has been regarded by many philosophers as both deep and profound. The presumption of depth is demonstrated by a widespread commitment to the project of uncovering what truth consists in: the property F which all and only the truths share, and which is such that truths are true because they are F. 1 Indeed, the familiar story told to undergraduates largely concerns the merits of the rival accounts of the nature of F: correspondence to a fact, membership of some favoured coherent set of propositions, and proving a useful belief to have: the favoured candidates of correspondence theorists, coherence theorists, and pragmatists, respectively. The presumption of profundity, meanwhile, is evidenced by the (apparent) fact that an understanding of the concept of truth provides the key to explaining other concepts and phenomena. Indeed, a list of the supposed explanatory rôles of the concept of truth makes impressive reading. Truth, so a compelling story goes, is the concept by means of which we articulate the relation between mind and world. The correspondence theorist, it has been claimed, is a kind of transcendental realist, her talk of true propositions matching or fitting the facts committing her to a conception of reality and thought as constituting two-self subsistent realms. 2 Coherence theorists and certain identity theorists, 3 reacting against a picture in which thought 1. This way of putting the desideratum of a substantial theory of truth is due to Marian David. See his Correspondence and Disquotation (Oxford University Press, 1995), pp. 65 66. 2. Three philosophers who take this view of correspondence theories of truth are Richard Rorty, Hilary Putnam and John McDowell. See, for example, Rorty s Philosophy and the Mirror of Nature (Princeton University Press, 1979), Ch. 6; Putnam s Reason, Truth and History (Cambridge University Press, 1981), Ch. 3; and McDowell s Mind and World (Harvard University Press, 1994), Ch. 2. 3. For the claim that McDowell offers a kind of identity theory of truth in his Mind and World, see Julian Dodd, McDowell and Identity Theories of Truth, Analysis, 55 (1995), pp. 160 165, and his An Identity Theory of Truth (Palgrave, 2000), Ch. 7; and also Jennifer Hornsby, Truth: The Identity Theory, Proceedings of the Aristotelian Society, 97 (1997), pp. 1 24, reprinted in Michael P. Lynch (ed.), The Nature of Truth (MIT Press, 2001). A helpful brief introduction to identity theories is provided by Stewart Candlish in his Truth, Identity Theory Of, Stanford Encyclopedia of Philosophy (http://plato.stanford.edu/ archives/fall2001/entries/truth-identity). 279, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA
provides, at best, an indirect kind of access to the world, respond by dragging reality into the realm of thought. The concept of truth is thus seen as the place at which the dispute about the nature of mind/world relations is ultimately prosecuted. The supposed explanatory rôle of truth does not end here, however. Truth, it has been claimed, is the concept in terms of which we must explain the concept of meaning: to understand a sentence is to grasp its truth conditions. Truth, according to some, is a substantial norm: what we aim for in belief and judgement. And it has also been maintained that it is the truth of scientific theories that explains their success. I could go on, but the point is surely clear enough by now: a concept of such depth, and with such numerous explanatory connections with other concepts, would seem to be a substantial concept par excellence. Of course, some philosophers have doubted whether truth really has such a profound, yet concealed essence, 4 but the publication of Paul Horwich s Truth in 1990 5 saw this scepticism expressed so pithily and effectively that seekers after a substantial theory felt challenged to defend their general approach to the concept. So it is perhaps fair to say that, subsequent to the publication of Horwich s book, the dispute between those seeking to construct a substantial theory of truth and those who follow Horwich in taking a deflationist line has come to overshadow all others pertaining to that concept. However, before going any further, it would be as well to set out just what it is for a theory of truth to be deflationary. The benchmark claim made by a deflationist is that there is no elusive property F: no property, shared by all and only the truths, which explains their truth. But saying this does not yet distinguish the deflationist attitude from the doctrine which we may call alethic primitivism. For there are two possible reasons why the search for F could be misguided. According to the alethic primitivist, truth s very profundity entails that it cannot be explained in terms of supposedly more basic concepts. Truth is substantial all right; in fact, it is so substantial, and so central to our conceptual scheme, that we cannot explain what truth consists in. As Donald Davidson puts it, the concepts philosophers single out for attention, like truth, knowledge, belief, action, cause, the good and the right, are the most elementary concepts we have, concepts without which... we would have no concepts at all. Why then should we expect to be able to reduce these concepts definitionally to concepts that are simpler, clearer, and more basic? 6 4. Notably F.P. Ramsey in his Facts and Propositions, Proceedings of the Aristotelian Society, supp. vol. 7 (1927), pp. 153 70; reprinted in Simon Blackburn and Keith Simmons (eds.), Truth (Oxford University Press, 1999). 5. Paul Horwich, Truth, 1st edn. (Blackwell, 1990). A second edition of the book contains a few amendments and a useful postscript (Oxford University Press, 1998). 6. Donald Davidson, The Folly of Trying to Define Truth, Journal of Philosophy, 93 (1996); reprinted in Michael P. Lynch (ed.), The Nature of Truth (MIT Press, 2001), and in Simon Blackburn and Keith Simmons (eds.), Truth. The quotation is from p. 309 of the latter. 280
Davidson, however, is no deflationist. 7 The deflationist, like the primitivist, denies that we may explain what truth consists in, but for very different reasons. If the deflationist is correct, truth is insusceptible of explanation, not because it is too basic a concept, but precisely because it lacks depth and profundity. Truth, on a deflationist reading, is just not philosophically interesting, and hence cannot have the explanatory rôles commonly ascribed to it. Coupled with this thesis concerning truth itself are deflationary theses concerning the word true and our concept of truth respectively. I shall discuss them in turn. It is a point acknowledged by deflationists and substantial theorists of truth alike that true is a device for making indirect or compendious assertions. If one wishes to endorse a proposition without specifying it, or if one wishes to endorse a great many propositions, one uses true in environments such as and (1) What Susan just said is true (2) Whatever Susan says is true respectively. True, we can say, has an expressive function: it enables us to say things which we would not be able to say otherwise. Now, the deflationist s distinctive claim about true is that it exists solely to fulfil this function. 8 In particular, the term does not express an explanatory property which can be analysed in terms of metaphysical or epistemological concepts. The deflationary claim concerning the concept of truth pertains to what it is to grasp the concept. For if true is an expressive device and no more, and if truth is too uninteresting to admit of an account of what it consists in, it would seem to follow that understanding the concept requires very little of us. Specifically, grasping the concept of truth does not require us to bring to bear any putatively more basic, explanatory concepts. Understanding the word true does not demand of us that we employ concepts such as correspondence or coherence. What, then, for a deflationist, is legitimate philosophical work pertaining to truth? Just this: explaining the expressive function of true ; explaining how true comes to have this rôle; setting out a suitably austere account of our concept of truth; and, lastly, attempting to loosen the grip that substantial theories, and, in particular, correspondence theories continue to have on philosophers. The Varieties of Deflationism Two kinds of deflationary theory have been pre-eminent in the last decade or so. These theories differ according to the explanation given of how true is 7. As he himself explains, ibid., p. 310. 8. See Horwich s Truth, 2nd edn., p. 2; and Dorothy Grover, Two Deflationary Truth Theories, in her A Prosentential Theory of Truth (Princeton University Press, 1992), p. 215. 281
able to play its sole, expressive function. According to minimalist theories, true is a metalinguistic predicate, 9 and its performance of its expressive function in (1) and (2) is the result of the fact that, leaving aside paradoxical replacements for p, the following schema holds: (E) <p> is true if and only if p. 10 For example, given that what Susan just said is <Julian is unreliable>, it follows from (E) that asserting (1) is just another way of saying that Julian is unreliable. By contrast with minimalism, the prosentential theory has it that true functions at the level of the object language, as the syncategorematic part of the prosentences it-is-true and that-is-true : anaphoric devices which are to sentences what pronouns are to nouns. 11 Prosentences, so it is said, sometimes function as proforms of laziness, as in (3) Eleanor: Dad has baked us a cake. Susan: If that-is-true, the kitchen is bound to be in a terrible mess. And, crucially, the prosentences that-is-true and it-is-true are also claimed to function as genuinely sentential variables, thus enabling authentic quantification into sentence position. So, for the prosententialist, (2), for example, can be glossed as (4) For each proposition, if Susan says that it-is-true, then it-is-true. The prosententialist agrees with the traditional redundancy theorist that true is not a genuine predicate. 12 Unlike the redundancy theorist, however, she denies that we can say everything without the truth term which we can say with it. 13 We need it precisely to play the expressive rôle outlined above. 9. In what follows I presume that, if true is a predicate at all, it is a predicate primarily of propositions. In this I am in agreement with Horwich (Truth, 2nd edn., pp. 16 17) and Scott Soames (Understanding Truth (Oxford University Press, 1999), Ch. 1). However, we should not define minimalism in such a way that this account of the vehicles of truth becomes essential to the view. Quine s disquotational theory, according to which sentences are truth bearers (Philosophy of Logic (Prentice Hall, 1970), Ch. 1), counts as minimalist since it too is a version of deflationism which holds that true is a genuine predicate. 10. The expression <p> is short for the proposition that p. See Horwich, Truth, 2nd edn., p. 10. 11. For a defence of the prosentential theory, see the papers in Grover s A Prosentential Theory of Truth and Robert Brandom s Pragmatism, Phenomenalism and Truth Talk, in P. French, T. Uehling and H. Wettstein (eds.), Midwest Studies in Philosophy, 12 (University of Minnesota Press, 1988), pp. 75 93. The use of hyphens stresses the point that true in it is true and that is true is supposed to be no more a separable semantic unit than is at in scatter. 12. As Grover, Camp and Belnap put it, [t]ruth, to coin a phrase, isn t a real predicate (A Prosentential Theory of Truth, p. 94). 13. Grover, A Prosentential Theory of Truth, p. 94. The distinctive claim made by a redundancy theorist is that an occurrence of true may always be removed from its environment without semantic loss. For a defence of the redundancy theory, see Ramsey s Facts and Propositions. 282
This is not the place to give a detailed assessment of the relative merits of minimalism and the prosentential theory. It suffices to say that the prosententialist s proposal faces several difficulties, not least of which is the counter-intuitiveness of the claim that it is true and that is true are really prosentences. At the sharp end the problem is the fact that sentences of the form that is G standardly function in the way we expect: is G is used to describe an entity referred to by the demonstrative that. Here is one such example. If Susan says that Julian is late and Eleanor responds by saying that is surprising, it seems obvious that Eleanor uses that to refer to <Julian is late> and is surprising to describe that proposition. But if this is so, then the prosententialist s case has been considerably weakened: her account of that is true would seem to be out of kilter with these familiar, everyday cases, and she must explain away this disanalogy. Naturally, there has been a good deal of discussion of examples such as these, but I am left with the suspicion that minimalism faces fewer obstacles than prosententialism. 14 Correspondence and Truthmaking Truth, according to a deflationist, is nothing more than that whose expression in a language yields a device for making indirect or compendious assertions. There is nothing that truth consists in. A philosopher who suspects that such a position deflates truth beyond recognition will seek to reinflate the concept by putting her finger on something about truth in general which the deflationist has overlooked. What could this something be? One feature of recent writing on truth has been the tendency to think that deflationism s only serious competitor is some correspondence theory or other. 15 So perhaps the deflationist has ignored the (putative) fact that truth is, in Crispin Wright s phrase, seriously dyadic : 16 that truth is a two-term relation between a proposition and a non-propositional item (such as a state of affairs or trope) 17 in a mind-independent world. The correspondence theorist s intuition here is expressed in the so-called truthmaker principle: namely, that every true proposition is made true by something. And the truthmaker principle, at bottom, demands that a truth must have an entity whose existence guarantees its truth. This, in turn, is taken to mean, in D.M. Armstrong s 14. For a lively and (to my mind) telling critique of prosententialism, see Graeme Forbes, Truth, Correspondence and Redundancy, in G. MacDonald and C. Wright (eds.), Fact, Science and Morality (Blackwell, 1986). Richard Kirkham nicely sets out the issues in his Theories of Truth (MIT Press, 1992), Ch. 10. 15. This attitude is represented by Marian David s suggestion that the success of his attack on deflationism leads us towards an acceptance of a correspondence theory (Correspondence and Disquotation, p. 188). 16. See Wright s Truth and Objectivity (Harvard University Press, 1992), p. 83. 17. D.M. Armstrong takes states of affairs to be truthmakers. See his A World of States of Affairs (Cambridge University Press, 1997). The claim that tropes are truthmakers has been made by Kevin Mulligan, Peter Simons and Barry Smith in their Truth-Makers, Philosophy and Phenomenological Research, 44 (1983), pp. 287 321. 283
words, that if a certain truthmaker makes a certain truth true, then there is no alternative world where the truthmaker exists but the truth is a false proposition. 18 In other words, a truthmaker s existence guarantees a proposition s truth by logically entailing it, and so we can represent the truthmaker axiom as (TM) If <p> is true, then there must be at least one entity whose existence entails that <p> is true. Much interesting recent work on truth in the last few years has concerned this principle, although such work has been found in books and articles concerned with matters metaphysical and ontological, rather than in writing primarily about truth itself. 19 Nonetheless, a consideration of the truthmaker principle should form a central part of an examination of the concept of truth. For if it turned out that (TM) were misconceived or unmotivated, the heart would thereby have been removed from the body of the correspondence conception of truth. If (TM) falls, correspondence theories fall with it. 20 Those seeking to argue for the correctness of (TM) tend to appeal to what is claimed to be its naturalness, or else suggest that (TM) is an essential feature of an appropriately realist metaphysics. 21 To my mind, however, neither appeal is successful. 22 First of all, it is far from clear that the correspondence theorist has made the case for why the truth of <Fa> need commit us ontologically to a truthmaking entity (such as a state of affairs or trope) in addition to a and to F. <Fa> is true because that object has that property: why must the proposition as a whole have a worldly correlate? Second, the suggestion that (TM) is a necessary condition of realism is far from convincing. If realism is the doctrine that the entities which make up the world are genuinely mind-independent, it is plain that one can be a realist without signing up to (TM). Someone unconvinced by the truthmaker principle will deny that the world contains truthmakers. But such a denial may be coupled with the distinctively realist claim that those entities which are to be found in the world objects, properties and events, presumably have no dependence upon mental states for their existence. This being so, the truthmaker principle awaits motivation, and it would be precipitate to suggest that the deflationist errs in ignoring it. 18. D.M. Armstrong, A World of States of Affairs, p. 115. 19. The two main sources are Armstrong s A World of States of Affairs and John Bigelow s The Reality of Numbers (Oxford University Press, 1988). 20. David Lewis denies this claim, not because he believes that a correspondence theory can survive the fall of (TM), but because he thinks that (TM) has nothing to do with truth. See his Forget about the Correspondence Theory of Truth, Analysis, 61 (2001), pp. 275 279. 21. The former position is taken by Armstrong (Universals: An Opinionated Introduction, (Westview Press, 1989), p. 89), the latter position by John Bigelow (The Reality of Numbers, pp. 121 134). Bigelow s discussion is the deepest attempt to motivate the truthmaker principle that I know of. 22. An extended sceptical discussion of the truthmaker principle is to be found in my Is Truth Supervenient on Being?, Proceedings of the Aristotelian Society, 102 (2002), pp. 69 86. 284
A Norm of Truth? Another candidate for the feature of truth in general which deflationism omits is the supposed normative significance of truth. Truth, it has been suggested, is what we aim at in assertion, and this is a fact which the deflationist cannot allow for. For if true were merely a device of endorsement, it certainly could not express a norm. 23 How would such an anti-deflationist argument be formulated in detail? Presumably, the normativity of truth is supposed to lie in the correctness of the following principle: (6) One should only assert what is true, whose logical form may be represented as (7) x(one should assert x only if x is true). However, the deflationist should reply by pointing out that the normativity with which we are concerned lies in the following schema: (8) One should assert that p only if p, a schema in which true does not feature. The truth predicate in (6), she will argue, is only being used in its familiar expressive rôle: namely, as a device for the facilitation of generalisation with respect to sentence positions. In short, true only appears in (6) in order to translate a schema which has nothing to do with truth viz. (8) into a genuine universally quantified proposition. But if this is right, then the deflationist need not worry. The fact that true occurs in (6) does not indicate that it expresses a genuinely explanatory property. In (6) the truth predicate figures in an explanation, but we should not be misled by this into thinking that it is truth that does the explaining. Needless to say, this deflationist reply is controversial, insisting, as it does, that there is no illuminating general story to be told about what constitutes a successful assertion. 24 What counts as success, it is argued, depends upon what 23. This argument has its origin in the work of Michael Dummett. See his Truth, Proceedings of the Aristotelian Society, 59 (1959), pp. 141 162; reprinted in his Truth and Other Enigmas (Duckworth, 1978). Crispin Wright expands upon the argument, aiming to block the suggestion that the norm of truth is that of warranted assertibility, in Chapter 1 of his Truth and Objectivity. For discussions of Wright s argument, see, e.g., Ian Rumfitt, Truth Wronged, Ratio, 8 (1997), pp. 100 107; Max Kölbel, Wright s Argument From Neutrality, Ratio, 10 (1997), pp. 35 47; Julian Dodd, There Is No Norm of Truth: a Minimalist Response to Wright, Analysis, 59 (1999), pp. 291 299; and Alex Miller, On Wright s Argument Against Deflationism, Philosophical Quarterly, 51 (2001), pp. 527 531. 24. Indeed, Horwich does not take this line, preferring to argue that the fact that true is playing its familiar expressive function in (6) indicates that the deflationist can accommodate the norm of truth. See Horwich s Realism Minus Truth, Philosophy and Phenomenological Research, 56 (1996), pp. 877 881 and his Postscript to the second edition of Truth. For Crispin Wright s 285
takes the place of p in (8): the subject-matter of what is asserted. Suffice it to say that if the case can be made for there being no norm of truth, the deflationist should not be criticised for omitting such a thing from his account. Deflationism, Non-Factualism and Truth-Value Gaps Non-factualism about a sphere of discourse is the doctrine that the declarative sentences of the discourse in question do not have truth conditions, and hence can be neither true nor false. A familiar example of such a non-factualist doctrine is expressivism in ethics. According to an expressivist, in uttering Stealing money is wrong one does not say something which can be true or false; one is merely giving vent to one s negative feelings towards stealing money. 25 Now, it seems that one could not find a more substantial metaethical issue than the question of the correctness (or otherwise) of expressivism. However, according to an anti-deflationist line of thought, the correctness of deflationism would trivialise this issue, since deflationism is incompatible with the possibility of there being declarative sentences which are meaningful yet neither true nor false. Consequently, the fact that the truth of deflationism would so quickly rule out expressivism (and other forms of non-factualism) might be taken to be a reductio of the former. The question which needs to be focused on is this: does deflationism really rule out the possibility of truth-value gaps? 26 Horwich s thought is a simple one. According to the deflationary point of view, ethical declaratives express genuine propositions since they can form that -clauses referring to the objects of the propositional attitudes. And, given that ethical declaratives express propositions, there is no obstacle to such declaratives taking the place of p in (E) <p> is true if and only if p. convincing reply to Horwich, see his Response to Commentators, Philosophy and Phenomenological Research, 56 (1996), pp. 911 941. Wright goes on to criticise (albeit indirectly) the view that there is no norm of truth for the deflationist to accommodate in IV of his immensely helpful and thought-provoking Truth: A Traditional Debate Reviewed, in S. Blackburn and K. Simmons (eds.), Truth, pp. 203 238. 25. For defences of expressivism, see A.J. Ayer, Language, Truth and Logic (Victor Gollancz, 1936), Ch. 6; Simon Blackburn, Spreading the Word (Oxford University Press, 1984), Ch. 6; and Allan Gibbard, Wise Choices, Apt Feelings (Harvard University Press, 1990). 26. This question has recently been the subject of a little industry in analytical philosophy. Philosophers arguing that deflationism is incompatible with there being truth-value gaps include Horwich, Truth, 2nd edn., pp. 85 85; Wright, Truth and Objectivity, pp. 35 36; Paul Boghossian, The Structure of Content, Philosophical Review, 99 (1999), pp. 157 184; and Alex Miller and John Divers, Why Expressivists about Value Should Not Love Minimalism about Truth, Analysis, 54 (1994), pp. 12 19. The compatibilist position has been supported by, amongst others, Michael Smith, Why the Expressivists about Value Should Love Minimalism about Truth, Analysis, 54 (1994), pp. 1 12, and Frank Jackson, Graham Oppy and Michael Smith, Minimalism and Truth Aptness, Mind, 103 (1994), pp. 287 302. Richard Holton s Minimalism and Truth-Value Gaps, Philosophical Studies, 97 (2000), pp. 137 168 provides an excellent disentangling of the issues. 286
As Horwich puts it, ethical propositions provide perfectly good and useful instances of the equivalence schema instances which are needed to formulate generalizations (e.g. in logic) that cover such propositions. 27 At this point, the two possible lines of reply would seem to be open to the deflationist. The first response is to follow Horwich in accepting the force of the argument, but denying that it acts as a reductio of deflationism. According to Horwich, the moral of the argument is, not that we should ditch deflationism, but that we should reformulate expressivism so it becomes the claim that the meaning of, for example, x is good, consists in the fact that it is asserted when and only when the speaker wants x. 28 The second response is more bullish: namely to make a case for the compatibility of deflationism and expressivism (the latter standardly formulated). Which strategy is to be preferred? The second, I would argue. The problem with Horwich s reformulation of expressivism is that it is difficult to see how it can be a mere reformulation. For if, pace expressivism, ethical pronouncements do indeed have truth conditions, then the expressivist s distinctive thesis has been denied. But another point needs to be made too. For expressivism was intended to be an anti-realist thesis: a way of giving substance to the thought that the ethical realm is not one in which genuine discoveries can be made. Horwich s reformulated doctrine the claim that the meanings of ethical propositions are determined by the speaker s desires is arguably not an anti-realist thesis at all. For as formulated by Horwich, the expressivist thesis would not seem to entail that there cannot be mind-independent ethical facts, for example. Consequently, Horwich s suggestion fails to do justice to the intuition motivating expressivism in the first place. What, then, of the second form of reply: the thesis that deflationism and expressivism are compatible? Let us call the property of having truth conditions truth-aptness. Pace Horwich, the suggestion made by compatibilists is that a deflationary account of truth may be coupled with a robust account of truth-aptness. Let us see how this enables the deflationist to be an expressivist about value. Consider the minimalist version of deflationism about truth. Roughly, such a minimalist claims that there is no more to a proposition s being true than is given by (E) <p> is true if and only if p. Must a minimalist hold that we may replace the instances of p in (E) with moral declaratives, as Horwich believes? No. 29 A declarative sentence can take the place of p in (E) only if it has truth conditions: only if it is truth-apt. (If the substituend is not truth-apt, the instance of (E) will end up false because the left-hand side of the biconditional will be false, while the right-hand side 27. Horwich, Truth, 2nd edn., p. 84. 28. Ibid., p. 85. 29. In what follows I am indebted to Frank Jackson, Graham Oppy and Michael Smith, Minimalism and Truth Aptness. 287
will be neither true nor false.) And, crucially, a deflationist may consistently hold that moral declaratives are not truth-apt, and for the usual expressivist reason: namely, that moral declaratives, by virtue of being action-guiding, express feelings or attitudes, rather than truth-apt contents. The point can be put another way. A sentence can only take the place of p in (E) if it expresses a proposition. (Propositions are, by definition, truth-apt.) So a deflationist could hold both that (E) tells us all there is to know about the property of truth, and that moral declaratives do not express propositions, and hence are not truth-apt. Indeed, this is exactly the way that Ayer puts it. 30 Horwich s mistake here would seem to be that of supposing that a deflationist about truth must also take a deflationary view of truth-aptness (= what it is to express a proposition). He supposes that a deflationist about truth will hold that truth-aptness is just a matter of syntax: that for a sentence to be truth-apt is merely for it to be declarative. We have just seen that this assumption is groundless. (E) may tell us all there is to know about truth, and yet it be a substantive question which kinds of sentences, by virtue of being apt for truth, can take the place of the instances of p in (E). One question is what makes for truth among the truth-apt; another question, which should not be confused with the first, is what makes for truth-aptness among the complete range of linguistic items. 31 Truth, Realism/Anti-Realism Disputes, and Alethic Pluralism A common assumption is that truth is the concept by means of which we must prosecute realism/anti-realism disputes. So, for example, we see Michael Dummett suggesting that such disputes concern whether the truth conditions of the discourse s sentences are verification-transcendent: whether the sentence s sentences may be true or false beyond our ken. 32 Richard Rorty, meanwhile, seems to think that the correctness of his kind of deflationism brings with it an appreciation that disputes as to realism simply cannot get going, predicated, as he thinks they must be, upon truth s being a more robust property than it really is. 33 A deflationist sensitive to the seriousness of realism/anti-realism disputes will, of course, deny that they concern the nature of truth. As Horwich explains, 34 to a deflationist of this kind, an on-going project is that of showing that truth is irrelevant to such disputes. We have seen already one such 30. Ayer, Language, Truth and Logic, pp. 87 ff. and 107 respectively. Naturally, it follows from Ayer s position that, contra Horwich, that -clauses completed with ethical indicatives do not refer to propositions. 31. Here I am paraphrasing Jackson, Oppy and Smith, Minimalism and Truth Aptness, p. 290. 32. See, for example, Dummett s Realism and Anti Realism, in his The Seas of Language (Oxford University Press, 1993), pp. 462 478, for a helpful introduction to his views. 33. This kind of attitude is to be found in Rorty s Is Truth a Goal of Enquiry? Donald Davidson versus Crispin Wright, Philosophical Quarterly, 45 (1995), reprinted in Michael P. Lynch (ed.), The Nature of Truth. 34. Horwich, Truth, 2nd edn., pp. 7 8. 288
example of how this can be done: a deflationist can construe the dispute between the evaluative realist and her expressivist opponent as concerning truth-aptness rather than truth itself. Crispin Wright, however, has an alternative approach to the question of what is at stake in realism/anti-realism disputes, an approach which is the product of three theses. First, Wright claims that deflationism about truth is a non-starter because it fails to capture truth s normative significance. 35 Second, Wright argues for a minimalist account of truth-aptness, what we might call disciplined syntacticism: 36 namely, that a class of sentences are truth-apt just in case they are genuinely declarative, and there exist standards by means of which the sentences use are judged to be correct and incorrect. 37 Third, Wright argues that any predicate counts as a genuine truth-predicate just in case it satisfies a set of basic platitudes about truth: for example, that to assert a statement is to present it as true; that <p> is true if and only if p; that truth is distinct from justification; and so on. 38 What follows from these claims? Well, if deflationism fails, we are under no obligation to relocate realism/antirealism disputes away from the concept of truth. And if Wright s version of minimalism about truth-aptness is correct, then the disputed discourses all count as truth-apt: moral and modal discourse, for example, contain disciplined declaratives. Consequently, truth-aptness cannot be the crux of a realism/anti-realism debate. So it looks as though Wright will be on the look out for a way of prosecuting disputes as to realism which, after all, makes use of the notion of truth rather than truth-aptness. This is where his third thesis comes in. For if a predicate counts as a truth predicate merely by virtue of satisfying a set of platitudes, this leaves the door ajar for alethic pluralism: the proposal that there may be many truth-predicates, all satisfying the basic platitudes, but expressing properties which differ in their metaphysical strength. Disputes as to realism will then turn on the kind of truth-predicate applicable in the target discourse. 39 Anti-realists with respect to a given discourse will take the discourse s truth predicate to express a property which merely satisfies the basic platitudes; realists with respect to the relevant discourse will argue that its truth predicate expresses a property which has additional substance. In the course of his illuminating study, Wright suggests four respects in which the property expressed by a discourse s truth-predicate could have 35. Wright, Truth and Objectivity, Ch. 1. 36. This way of putting it is due to Jackson, Oppy and Smith, Minimalism and Truth Aptness, p. 293. 37. Wright, Truth and Objectivity, p. 74. 38. See, for example, Wright s Précis of Truth and Objectivity, Philosophy and Phenomenological Research, 56 (1996), p. 864. 39. See Wright s Truth and Objectivity, pp. 37 38, 141 142. Wright s book is the subject of an excellent symposium in Philosophy and Phenomenological Research, 56 (1996). The contributions by Horwich ( Realism Minus Truth, pp. 877 881) and Mark Sainsbury ( Crispin Wright: Truth and Objectivity, pp. 899 904), together with Wright s replies ( Response to Commentators, pp. 911 939), involve a detailed and lengthy discussion of the merits of Wright s pluralism with respect to truth. Another philosopher advocating alethic pluralism is Michael P. Lynch. See his Truth in Context (MIT Press, 1998), pp. 129 36. 289
more depth than an anti-realist would allow. First, in a reformulation of the familiar Dummettian semantic issue, there could be judgements which are true and yet not superassertible. 40 Second, there could be judgements which are superassertible because they are true, as opposed to being true because they are superassertible. Third, the discourse may exhibit cognitive command: its being a priori that, if two speakers differ in their judgements, then at least one is in cognitive error. 41 Fourth, the facts stated in the discourse may have a wide cosmological rôle: that is to say, these facts may be cited in order to explain things other than, or other than via, our being in attitudinal states which have such facts as their objects. 42 I cannot do justice to the subtlety of Wright s discussion here. 43 It should be noted, though, that his pluralist approach has been the subject of a good deal of sceptical comment. There would seem to be two major worries. To begin with, there is the problem of explaining the validity of mixed arguments : arguments containing propositions from discourses subject to truth predicates of different strengths. 44 Presumably, the validity of such arguments would be explicable in terms of the weakest truth predicate applicable to the argument s premises. However, this manoeuvre can only be made at some considerable cost: it would follow that a proposition could be evaluable by means of a strong truth predicate when it occurs with propositions from the same discourse, but a weaker one when in mixed company. Such a consequence would take the pluralism to a new level of implausibility. The second concern with Wright s pluralism focuses on its motivation. As Frank Jackson has argued, 45 it would be a mistake to be look at the multifarious things that exist (chairs, electrons, societies, works of music, numbers, and the like) and then conclude that this showed that there were different kinds of existence. The right thing to say, surely, is that existence is monistic; there are just different kinds of thing which exist. But if this is correct, then Wright comes under immediate pressure from an analogous source. Why should we not regard truth as monistic, any differences between the truths expressed in different discourses being differences pertaining to what is true: the judgements respective subject matters? 46 Indeed, it is noticeable that Wright himself elucidates the features of cognitive command and wide cosmological rôle without making use of the word true. Given that we distinguish differences in the contents of sentences anyway, theoretical economy tells against alethic 40. A judgement is superassertible if some actually accessible state of information justifies its assertion, and then will continue to do so no matter how enlarged upon or improved (Wright, Précis of Truth and Objectivity, p. 865). 41. See Truth and Objectivity, Chs. 3 and 4. 42. Ibid., Ch. 5. 43. See Jim Edwards, Debates About Realism Transposed to a New Key, Mind, 103 (1994), pp. 59 72, for an extensive critical appraisal of Wright s project. 44. Here I rely heavily on Mark Sainsbury s discussion. See his Crispin Wright: Truth and Objectivity, pp. 900 901. 45. Jackson, Realism, Truth and Truth Aptness, Philosophical Books, 35 (1994), pp. 168 169. 46. This question has also been pressed by Horwich, Realism Minus Truth, pp. 880 881, and by Sainsbury, ibid., pp. 900 901. 290
pluralism. Disputes as to realism, I venture to suggest, have nothing to do with the nature of truth. Concluding Remarks I hope I have given the reader a flavour of some of the most significant issues that have arisen in the past decade concerning the concept of truth. Inevitably, shortage of space has prevented me from considering, for example: the nest of issues surrounding the liar paradox; the truth theoretic approach to meaning; the interpretation of the work of Tarski; and the various attempts to defend anti-realist and internal realist views of truth. Academics and students in search of an accessible and rigorous introduction to these issues, and to those which I have addressed, should make Richard L. Kirkham s Theories of Truth their first port of call. 47 When it comes to the questions I have discussed in this essay, my sympathies should be obvious. However, the to-ing and fro-ing over the question of the viability of deflationism shows no sign of abating as we progress through the first decade of the twenty-first century. This is a sign of depth rather than stagnation. 47. Richard L. Kirkham, Theories of Truth (MIT Press, 1992). 291