ABSTRACT
In computer science, the term "evolution" has meanings such as may apply to learning circuits. In these instances, an "evolutionary learning system" incorporates both a state space whose individual states are modes of information processing, and an "evolutionary" search through these states which follows simulated annealing dynamics. In several papers, Kugler and Turvey and their colleagues have called attention to the inadequacies of the computer metaphor as a basis for accounting for complex, self-organizing (especially living) systems. Symbol strings can adequately account for the classical indicational and injunctional aspects of information. Anatomically at the macroscopic level, most neural networks are spatially 2-dimensional in the same sense that modern silicon chips are: the third dimension is used largely for connectivity, rather than for computation per se. The high anatomic dimensionality of the computing elements in the nervous system lies at the fan out at synaptic connections and at the molecular level where receptors "recognize" ligands as transmitters.
