Wednesday, April 16, 2008

Davidson, Belief

I think there might be something like an inconsistency in Davidson’s theories of meaning and belief. Consider the following set of claims, all of which, I think, Davidson has to accept:

1. Intentional indeterminism: the apparatus of folk-psychology always underdetermines the total set of beliefs and desires that we attribute to an intentional system. The non-psychological facts and the commitments of psychology as such (i.e. the common commitments of any psychological theory in virtue of which the theory is psychological) underdetermine – inductively, deductively, and abductively – what beliefs and desires an intentional system has.
2. General agreement: in general, folk that are equally well-informed about an intentional system in non-psychological terms agree, in the individual cases, both on what beliefs and desires the system has, and how to determine further what beliefs and desires the system has.
3. We must explain (2) in terms of the commonly-held beliefs of the folk.
4. These beliefs are not among the basic commitments of psychology as such. (From (1) and (2).)
5. Facts about commonly-held beliefs that determine the evidential procedures governing the proper application of a word (like “believes) are among the facts that constitute, in part, the meaning of that word. (I take this to be one of the reasons that Davidson says that meaning and belief are so closely related. I also take it to be a consequence of most varieties of semantic holism.)
6. Facts about the meaning of “believes” are among the basic commitments of psychology as such.

But then, from (3), (5), and (6), it seems we can derive the conclusion that the facts that explain (2) are facts about the meaning of “believes,” and so among the basic commitments of psychology as such. This is immediately contrary to (4).

I think Davidson might get out of this by denying (3), but it still seems that the sort of meta-semantics underlying (5) is going to make it difficult to say that the commonality that explains (2) is not, in part, constitutive of the meaning of “believes.”

No comments: