Transfer of Learning

It is a recognized fact that practice in one activity may facilitate performance of a related activity. The evidence for transfer is particularly good in the phenomenon of cross-education, where practice of one bodily part in performing a skill increases the ability of a bilaterally symmetrical part to do the same act. It is not necessary to review these experiments in any detail, since our primary concern is with the mechanism responsible for transfer.

Two general theories have been proposed to explain the phenomenon, the theory of identical elements and the theory of spread. One activity alters another to the degree to which both utilize the same or identical neural elements. On the other hand, there may be some actual training of an apparently unpracticed part, due to the spread of nervous impulses from the centers first acting to those involved in the subsequent act. According to this theory, it would not be necessary for acts to have identical neural elements in order for transfer to occur. It is difficult to test either of these hypotheses experimentally. Very little is yet known of the phenomenon of spread, and the usual "proof" of the identity theory is circular; that is, two tasks are said to have identical elements when practice in one facilitates the other. In predominantly motor skills, such as mirror drawing, one can see that common head, eye, or trunk movements might be involved in practice with either hand. But can we define identity in more specific physiological terms? If not, perhaps the theory is merely an expression of accompanying conditions not essential to transfer.

Abundant evidence of the close relation of the various bodily parts has been presented in our chapters on integrative action. So far, however, there is little to show that modifications of one sensorimotor pathway will be duplicated in the contralateral path. In fact, it will be recalled that Lashley's studies are definitely negative. This suggests that the locus of transfer is central rather than peripheral; and that if there are common elements, they exist as ratios of excitation in two patterns of integration rather than as identical neurones.