Content without Misrepresentation: A Pragmatic Theory

Abstract

In Chapters 2 and 3 we have seen the arguments for the thesis that no naturalistic theory of mental content can allow for misrepresentation. The attempts to define meaning by a restricted class of uses are unsuccessful, as are attempts to ground a new analytic/synthetic distinction on idealized causal roles, inferential roles, conceptual roles, or adaptational roles. All these attempts were circular or begged the question by assuming semantic properties in their analyses. So, either we consider all uses meaning-fixing or we must reject a use theory altogether. Yet, in Chapter 1 we saw both that non-use theories have various insurmountable problems and also that they do not make room for misrepresentation. So we must have a use theory, and the only non-circular, non-question-begging use theory involves all actual and counterfactual uses, the entire conceptual role of the representation, to define meaning. I will call such a theory of content, built on an unlimited, non-idealized, unmodified conceptual role, Strict Conceptual Role Theory (SCRT). In accepting this theory that has been forced upon us, we also must accept what it entails: There can be no misrepresentation, and thus no error or false belief. I come reluctantly to this conclusion, since it does great damage to our intuitive views of representation and belief. The problem is that we have no theory that succeeds in accommodating our intuitions. So we are forced into a dilemma: either accept nihilism or have a use theory which does not allow misrepresentation. It is not a happy ending to the search for a theory of content, but a conceptual role theory with no misrepresentation is the best we can do. In pursuing this result, we reject that last empiricist dogma of meaning, that meaning requires the possibility of error (and that representation presupposes the possibility of misrepresentation). To help sustain this result, I propose a pragmatic theory of concept change that captures most, but not all, of what we wanted to get from error and misrepresentation. Such a theory will not cure this lack of misrepresentation, but it is recommended a possible treatment. It will allow us to make do with that part of the notion of meaning that we can coherently retain, and which mitigates some of the pain and apparent absurdity of the no-misrepresentation result.