The Adaptive Self-Organization of Serial Order in Behavior: Speech, Language, And Motor Control

Abstract

This chapter illustrates how a small number of network principles and mechanisms can be used to discuss many topics related to the adaptive self-organization of serial order in behavior. The most important unifying concepts that arise in this framework are those of adaptive resonance, adaptive context-mediated avalanche, adaptively invariant short-term memory (STM) order information in an item field, and an adaptively tuned self-similar masking field. All of these concepts suggest that the functional units of network activity are inherently nonlinear and nonlocal patterns that coherently bind a network's local computations into a context-sensitive whole. The program of classifying the adaptive resonances that control different types of planned serial behavior promises to antiquate the homunculi that burden some contemporary theories of intelligent behavior, and to end Neisser's nightmare of “processing and still more processing” with a synthetic moment of resonant recognition.