The Dual Aspects Theory of Truth

Introduction

Consider the following 'principles':²

(Norm of Belief Schema) Necessarily, a belief of <p> is correct (relative to some scenario) if and only if p (at that scenario) — where 'p' has the aforementioned content <p>.

(Generalized Norm of Belief) Necessarily, for all propositions <p>, a belief of <p> is correct (relative to some scenario) if and only if <p> is true (at that scenario).

Both 'principles' appear to capture the aim(s) of belief. (NBS) particularizes the aims to beliefs of distinct content-types. (GNB) generalizes these aims of beliefs as truth. Properly understood, the instances of (NBS) appear to be at least approximately true as does (GNB). Of course, one might harbor concerns about whether these are mere appearances or whether these 'principles' are properly formulated. Putting such concerns aside, one might naturally wonder how these two 'principles' (or their proper reformulations) are fundamentally related. More specifically, are the instances of (NBS) explanatorily basic if they are true? Or, alternatively, must (NBS) be explained by (GNB)? My interest in the explanatory relationship of these 'principles' is tied to my interest in the debate between the deflationist and inflationist about truth.

For my purposes, deflationism is the view that the nature of truth is metaphysically transparent. Paul Horwich (1990; 1998, ch. 4) makes this idea more precise by suggesting that truth is a pleonastic, quasi-logical property with no underlying substantial nature — unlike the property of being water, this property of truth does not admit of a constitution theory any more than negation does. More generally, deflationism has it that the nature of truth is exhausted by, roughly speaking, the (easily grasped) validity of transitioning back and forth between using a content and ascribing truth to that content (Horwich, 1990; Horwich, 1998; Horwich, 2005; Horwich, 2010, esp. ch. 2; Soames 1999; Field 2001a; Armour-Garb & Beall, 2005). The inflationist rejects deflationism.

It is easy to get the impression that the explanatory relationship between the aforementioned 'principles' is central to this debate (Cf. Wright, 1992; Wright, 2001; Lynch, 2004; Lynch, 2009; Shah & Velleman, 2005, 523-5). In particular, one can get the impression that one's view on this explanatory relationship fixes (even if it does not strictly entail) whether or not one is committed to deflationism. The idea is that those who think that the instances of (NBS) are explanatorily basic must be deflationists whereas those who think that the schema must be explained by the generalization are inflationists. Furthermore, it is assumed that the motivation for rejecting deflationism at least partly stems from the alleged plausibility of explaining the schema by way of the generalization. Indeed, it is sometimes claimed that there is no way that the instances of (NBS) could be true unless there is something along the lines of (GNB) to explain them.

My aim in this paper is to show that this picture of the dialectic between the deflationist and inflationist is a false dichotomy. This picture is accurate in portraying that the deflationist is committed to thinking that the instances of the schema are explanatorily basic. To think that the generalization had explanatory priority would effectively be to think that there is something that having the property of truth can explain, and it is very difficult to see how that could be the case if truth had only a metaphysically transparent nature. Nevertheless, the picture is inaccurate in portraying that the converse is also true: thinking that the instances of the schema are explanatorily basic does not force one to accept deflationism. In the first half of this paper, I defend the claim that the instances of (NBS) as explanatorily basic. In the second half, I suggest that we may, nonetheless, reject deflationism. What emerges from this combination of views is the dual aspects theory of truth.

I Initial Clarifications

For discussion purposes, some clarification of (NBS) and (GNB) is necessary.

First, I note that both 'principles' are to be understood on their de dicto reading rather than their de re reading. On the de re reading, an instance of (NBS) would...