The debate on truth

As my previous posts indicate, I've been thinking a lot about truth. Let me make a big picture comment. I find it striking just how much the 20th century debate is driven by naturalism, even physicalism. A notable episode in the history, for instance, is Putnam's departure from metaphysical realism on the grounds that if metaphysical realism is accepted, it is going to be a scientific realism, and hence will lead to a physicalism incompatible with the language-world connections that metaphysical realism needs, as well as with ethics, so that we should reject metaphysical realism. It is widely accepted that if there is going to be a genuine language-world or mind-world truth connection of the sort that non-minimal correspondence theories claim, it will be causal and hence natural.

This makes certain portions of the debate not very interesting to me. It's obvious to me that physicalistic causal theories of reference have no hope of working, because we will all over the place need normative concepts (standard conditions, normal observers, typical cases, etc.), and attempts to spell these out in non-normative ways (evolutionary or statistical) only provide opportunities for the ever fun parlor game (a game for playing which can get tenure, which fact I am grateful for) of "Pin the Counterexample on the Theory".

At the same time, I think one can stand back from the truth debate as a whole, and say that there is something we can learn from the debate: it is hard to be a realist of a non-minimal sort (i.e., one with a positive theory of truth that goes beyond Schema (T)) while yet being a physicalist. There is an intellectual strain.

Nonetheless, I do not think the debate is worthless for those of us who are not naturalists. For some of the arguments in the debate probably work just fine even if one does not assume naturalism. Non-naturalism isn't itself a theory of truth! The task of finding a theory of truth is a really hard one for the non-naturalist, too. However, the non-naturalist labors under the advantage that some of the difficulties in the literature in coming up with a theory of truth are artifacts of the assumed naturalism.

It is worth noting that the traditional theist has a tool available for the theory of truth, a tool that should help progress to be made, namely a commitment to the existence of an extensionally correct and logically non-trivial characterization of truth:

1. A sentence is true if and only if God believes what it says.

I do not propose (1) as a definition of truth. Nor is it a reduction of "true" to a non-semantic concept, since it presupposes the semantically loaded relation of x's believing what sentence s says. Nonetheless, it is just as much a "definition" of truth as Tarski's recursive one is: it gives an extensionally correct and non-circular characterization of which sentences are true. (One might worry about circularity if God is defined as omniscient. But then let's not define God as omniscient. Just define him as the perfect being, or as pure actuality, or as the necessarily existing first cause.)