Revising the classical postulates on memory: the non-linear dynamical approach

Abstract

The absence of parsimony and problems related to the testability of models of memory cast in doubt the acceptability of the research program that, since Ebbinghaus and Bartlett, underlies the works in that area of study. This domain of investigation, based on a mechanistic and Platonic conception of organism, is built on two postulates that may be discussed: the postulate of mediationism and the postulate of complexity of memory processes. Simultaneously, the Platonic and mechanistic conception tends to neglect the question of  variability in memory performances. Adapting alternative conceptual tools based on nonlinear dynamics, we try to supplement these perspectives. We defend the proposition of nonmediationism, the idea of emergence of complex behavior from a small number of simple processes, and the notion of “randomness by essence”. Finally, we examine the possibility of rapprochement between the works in nonlinear dynamics and the studies on memory in a behaviouristic perspective.