Authors: Athanassios Raftopoulos
Abstract: Any discussion concerning representations in cognitive psychology and science assumes that mental representational contents are carried by brain representational vehicles. Recently Sedivy attacked the view that mental contents are carried by vehicles. Sedivy relies on Dennett’s work to argue that mental contents cannot be carried by vehicles because the former cannot be independently individuated whereas the latter are, by definition, independently individuated by means of their semantic properties. In this paper, I claim that mental contents are carried by neural vehicles and that Sedivy misunderstands Dennett’s work on the relation between the mental and the neural by confounding the algorithmic and the computational level of description. I analyze connectionist representations that Dennett favors and which are the most amenable to Sedivy’s concerning the context dependency of contents on which Sedivy builds her arguments, and argue that in connectionism, there is a clear cut distinction between contents and vehicles that carry these contents.
Keywords: Representational contents; representational vehicles; dynamical systems; physicalism; neural networks