Full recordShow full item record
AbstractIn this paper we propose a mechanism in the brain for supporting consciousness. We leave open the question of the origin of consciousness itself, although an acausal origin is suggested since it should mesh with the proposed quasi-acausal network dynamics. In particular, we propose simply that fixed-point attractors, such as exemplified by the simple deterministic Hopfield network, correspond to conscious moments. In a sort of dual to Tononi's Integrated Information Theory, we suggest that the "main experience" corresponds to a dominant fixed point that incorporates sub-networks that span the brain and maximizes "relatedness." The dynamics around the dominant fixed point correspond in some parts of the system to associative memory dynamics, and to more binding constraint satisfaction dynamics in other areas. Since the memories that we are familiar with appear to have a conscious origin, it makes sense that a conscious moment itself corresponds in effect to what amounts to memory recollection. Furthermore, since Hopfield-like networks are generative, a conscious moment can in effect be seen as a living, partially predicted memory. Another primary motivation for this approach is that alternative states can be naturally sensed, or contrasted, at the fixed points.