The Cost of Correspondence

Abstract

It's hard to know what realism is any more. In 'Concepts and Correspondence',' Scott Shalkowski construes it as a matter of a non-linguistic, non-psychological correspondence between language and the world.' Although he maintains that Goodman's arguments do not undermine realism, what he shows is that they do not refute it a weaker thesis that is readily conceded. 'Any statement can be held true come what may if we make drastic enough adjustments elsewhere in the system.'3 Shalkowski's defense of realism involves costly compensatory adjustments in our theories of logic, language, and knowledge. He thus inadvertently suggests that the rewards of realism may fail to compensate for the sacrifices they demand. Shalkowski takes correspondence to be a metaphysical relation between truths and the world. Contributions of language and psychology, if any, are negligible, as 'it is the world that renders the sentence true' (p. 463). Moreover, a single sort of correspondence is supposed to connect each truth to the world. But in the absence of a plausible candidate, it is doubtful that the world stands in the same relation to phenomenalist and physicalist truths, to truths of psychology and truths of physiology, to truths about what is and truths about what is not. This is what tempts realists to reductionism to the view that truths couched in a single vocabulary are secured by correspondence, and truths expressed in other terms are reducible to or supervenient on the favored truths. The difficulty, as Shalkowski notes, is that reductionism is unvia-