Is novelty of thought a mystery? Old schemata for new problems

Contextual flexibility has been the focus of considerable research in cognitive science. The pessimistic view expressed by Fodor on this issue has been challenged either by modular approaches or by proposals based on common codes/spaces where information can be integrated. I analyse these views and the middle-ground approach explored by SHANAHAN and BAARS (2005), and propose a different non-modular account. The general idea is that flexible integration of information is essentially ensured thanks to the hierarchical and schematic organization of memory, and to the bottom-up/top-down dynamic of its associative activation. In this perspective, context is not a crucial part of the problem, it is instead key to the solution: the different inputs in a context activates schemata that compete and integrate with each other, so that the schemata that are the most coherent with the context will be the most activated as well. I also consider the role that might be played by consciousness in this process, especially with regard to cases of extreme flexibility, that is, cases in which creative thoughts are formed.