What is at issue between epistemic and traditional accounts of truth?

Abstract

I will discuss those epistemic accounts of truth that say, roughly and at least, that the truth is what all ideally rational people, with maximum evidence, would in the long run come to believe. They have been defended on the grounds that they can solve sceptical problems that traditional accounts cannot surmount, and that they explain the value of truth in ways that traditional (and particularly, minimal) accounts cannot; they have been attacked on the grounds that they collapse into idealism.

I show that all these claims are mistaken. The system of statements accepted by an adherent of an epistemic account who also accepts the equivalence scheme is the same as that accepted by an adherent of a traditional account who also accepts a remarkably strong thesis of epistemic optimism. The singling out of one rather than another claim within this system as defining ‘true’ cannot make as much difference as to imply idealism or refute scepticism.

However, it can make all the difference when it is a matter of explaining the value of truth. For a crucial point in such explanation depends on what can be soundly substituted for what in intensional contexts; above all those governed by such verbs as ‘know’, ‘hope’, ‘believe’, ‘value’. That is, it depends on what expressions are intensionally equivalent. And one point of singling out one formulation as definitional can be to settle just this.

But though some epistemic theorists have deemed ability to explain the value of truth a merit of their account (and lack of this ability a fatal defect of traditional accounts, of minimal accounts in particular), it turns out that minimal accounts of ‘true’ fit a sound account of our valuing of truth in a way that epistemic accounts do not.

In the course of this argument I rebut related positions: e.g. Dummett's, that minimal definitions fail because they cannot account for the point of having a notion of truth, and that an account of the practice of assertion is what would fill this lacuna. I argue to the contrary that if the point of the notion could not be explained on the basis of a traditional definition, it could not be explained at all.

Notes

1 The difference in our understandings of 1—his involves reference to propositions—does not affect the point being made here.

2 An earlier instance of what looks like the same mistake is in Sellars 1963: 224], where he suggests that ‘that p is true’ can be construed as ‘it is assertible that p’ and on that basis says it is derivable that it entails‘p’.

3 I am sure that much this point was made by some philosopher, I think female, in the late 70s or early 80s; but I have not been able to locate the relevant article.

4 Tarski's scruples about using this convention with sentence variables were needless; cf. Fox 1989: 165–6].

5 Or not unless, in the language of the quoted sentence, ‘Water is H₂O’ does not mean that water is H₂O. But this qualification too is uncontroversial. For the rules for translation (with Tarski) or for uniform substitution (with minimal or semiminimal definitions) ensure that the result is not a genuine counterinstance to 1.

6 The late David Lewis liked this argument, but not my gloss on it that many of the strongest arguments merely verbalize incredulous stares.

7 Of course, this would not hold if ‘intensional’ applied to any context that was not extensional, e.g. the context created by the suffix ‘in those very words’.

8 Here I set aside the other, ‘hyperintensional’ idiom, merely remarking that it seems to me that an adequate account of it will surely make it clear that what is not known is partly something about language.

9 I intend this not as a significant bit of information; by ‘intensional equivalence’ I understand equivalence within just such intensional contexts.

10 Ellis heard me give an earlier version of this as a seminar at La Trobe University in 1998, and to his credit he then professed himself persuaded by this reply that on this score at least epistemic accounts do not after all have the advantage he had suggested over traditional ones.

11 The locus classicus for this point is Campbell 1957: chap. 2].

12 Cf. Quine 1969: 344]: ‘the notion of belief is more useful that that of necessity … some obscure notions are, on the grounds of utility, more worth trying to salvage than others’.