Pure Informational Semantics and the Narrow/Broad Dichotomy

The influence of historical-causal theories of reference developed in the late sixties and early seventies by Donnellan, Kripke, Putnam and Devitt has been so strong that any semantic theory that has the consequence of assigning disjunctive representational content to the mental states of twins (e.g. [H2O or XYZ]) has been thereby taken to refute itself. Similarly, despite the strength of pre-theoretical intuitions that exact physical replicas like Davidson's Swampman have representational mental states, people have routinely denied that they have any intentional/representational states. I want to focus on a particular brand of causal theory that is not historical, the so-called pure informational or nomic covariance theories, and examine how they propose to handle twin cases and replicas like Swampman. In particular, I will take up Fodor’s version of the theory, since it is the best worked out specimen in this genre. I will argue that such (non-historical/non-teleological) theories as Fodor’s are bound to assign disjunctive content to twins and representational content to replicas. I will also argue that this consequence should perhaps be welcome. I will end by sketching a picture according to which a pure informational semantics can accommodate both the internalist and the externalist intuitions.