Psychological ecology.

Abstract

A naturalist account of representational content is developed from five related themes. First, an explanation of content should attempt to give those conditions that are relevant and sufficient for the existence of representational content. Second, we should be guided in the resources we appeal to in order to fulfill these conditions by the explanatory opportunity these resources offer. The only serious explanatory opportunity for a generative theory of content is to be found in an account that appeals to a cognitive system's relations to its environment. Third, there are several equally sufficient kinds of environmental relations that appear to generate representational content (causality, indication, resemblance). Each of these can separately, simultaneously act upon the same representation, generating multiple contents for that representation. Different instances of the same content generating environmental relation can also simultaneously act on the same token, again producing multiple contents. This is the core thesis, that the contents of representations are multiple. Fourth, there is no principled way to set aside some of these contents. All leading, contemporary naturalist theories of content fail to do this: Fodor's, Papineau's, Millikan's and Clark's. Fifth, the core problems of any theory of content, the qua, depth, disjunction and misrepresentation problems, are easily solved by a view that accepts multiple contents. The conclusion is that a naturalist, generative theory of content has nothing to fear from multiple contents, but is actually advantaged by these.