Learning and Neural Structure

Studies of the relation of learning to neural structure generally have one of two problems in mind: (1) the significance of the amount of functional tissue, and (2) the localization of the engrams of specific habits. Since there is almost universal opinion that the cerebral hemispheres are mainly concerned with the patterns of complex action, attention has been focused upon the relation between cortical mass and the physiological limits of acquisition and retention.

It is clear that the modifications in behavior produced through training have to reside somewhere or somehow in the nervous system. We have progressed beyond the notion that individual ideas are stored in separate brain cells; and fixation is now generally assumed to reside in the synapse, habits depending upon specific neural paths and their intercellular junctions. "Are the engrams so definitely localized?" asked Lashley. To answer this question he trained rats in a habit, then removed some of the cerebral tissue, and after recovery from the operation tested them for retention of the habit.

By making systematic ablations in different cerebral areas and in various amounts, he was able to show that the engrams must be localized in areas of rather wide extent. The more complex the performance, the more difficulty was experienced in establishing the locus of the essential neural pattern. Visual form discrimination was abolished by complete destruction of the occipital lobes. Maze habits, however, showed little evidence of localization in any circumscribed area. They survived small injuries but were abolished by large ones, independent of locus.