Relativism and Representation

Abstract

Relativism is really a family of doctrines, many of them nebulous, elusive, yet seductive. Most seductive of all is the doctrine that truth is relative, a thesis we might encapsulate with the claim that something can be true with respect to one conceptual framework, false with respect to another. Plato's argument that this view is self-refuting is often regarded as one of the few decisive refutations in philosophy, but it so quickly dispatches its target that we are left wondering why the doctrine of relative truth captivated so many able thinkers in the first place. In True For I sketched an account of relative truth that was as plausible as I could make it without so diluting it that it no longer deserved the label 'relativistic'. The goal was not to defend any sort of relativism, but to locate reasons for the perennial appeal of various relativistic themes like the doctrine of relative truth. Quite apart from any problems involving self-refutation, I was led to the conclusion that the very picture which motivates many versions of relativism and accounts for their appeal is at odds with the doctrine of relative truth. Even after we give the relativist the benefit of every doubt, it remains difficult to find some one thing that could be true with respect to one conceptual framework (as I called the things truths were supposed to be relative to) and false with respect to another. Each candidate leads to a severe tension, requiring that two conceptual frameworks be sufficiently different for something to be true in one and false in the other, while yet being sufficiently alike to contain this same thing. In the end, my search for a bearer of relative truth values came up empty. Foster thinks that I quit too soon, backing his claim with two sorts of putative examples of relative truth and something like a formal criterion for their existence. Foster's formal criterion relies on Goodman's relation of extensional isomorphism. This relation is a bit complicated, but for our purposes we can think of it pretty much as isomorphism between relational structures of the sort familiar in abstract algebra and model theory, the key idea being preservation of structure (an additional constraint on Goodman's