The fuzzy paradigm for knowledge representation in cerebral dynamics

Abstract The fuzzy paradigm is suggested to be more successful than its competitors in representing our present-day knowledge of the neural functions. A neurophysiological basis for this proposal comes from the neuronic level (fluctuations of excitability, unknow neural codes,…) and, mainly, from the striking findings of Lashley, Luria and Gonzalo concerning the high residual function of the cortical tissue after traumatic and surgical lesions. A conceptual model based on modular co-operation and functional multiplicity of the same anatomical areas along with the gradual distribution of cortical functions is presented. Some suggestions for experimentally obtaining sensorial functions memberships, μ k ( x ) , are also included. Finally, we propose a set of fuzzy theoretical constructs (quantity of function, Mk(R); modal weight, Fk; lesion, ƒ(x, L) ; residual function, Rk(L); excitability, E(x,t); permeability, αij) that enables us to bring neurophysiological data and fuzzy models into closer agreement. Digital simulation of the membership construction process as well as of these fuzzy constructs for different cortical lesions are also included.