How to Think about the Debate over the Reality of Beliefs

Abstract The aim of this paper is to propose a new conceptualization of the distinction between realism and anti-realism about beliefs that is based on the division between natural and non-natural properties, as defined by Lewis. It will be argued that although the traditional form of anti-realism about beliefs, namely eliminative materialism, has failed (as it led to unacceptable consequences), there is a possibility to reformulate the division in question. The background assumption of the proposal is the framework of deflationism about truth and existence: it will be assumed that beliefs can be said to exist and their attributions can said to be true. The aim is to show that even when we buy into such assumptions we can meaningfully distinguish between the realist and anti-realist approach to belief. According to the proposal, the paradigmatic anti-realist view on beliefs should be seen as a conjunction of three claims: that belief attributions do not track objective similarities, that beliefs are not causally active, and that there is no viable way of naturalizing content. It will be shown that seeing the debate in the proposed way has important advantages as it allows the issue of belief realism to be made non-trivial and tractable, and it introduces theoretical unity into contemporary metaphysics of beliefs.