The problem to which we must direct our attention is that of the process by which lines of specific conduction, or response control, are laid down through the cerebral cortex. This problem is substantially identical with that of the physiological nature of learning. We have already reviewed some current theories regarding this process in our discussion of doctrines of animal motivation. Our present task is to consider these views more critically and to endeavor to develop a satisfactory theory for the purposes of our own discussion.
The reader will recall that experiments upon animals, supplemented by observations upon human beings, have suggested the following explanations of learning.
In the first place, all authorities agree that if nervous conduction occurs through a path, this path is thereby reduced in resistance and will conduct more freely in the future. This is the principle of use, exercise, or practice. Furthermore, it is pretty generally agreed that primitive learning requires a principle of random movement or "trial and error" response. The organism must be able to "experiment" with its reactions in order that it should succeed in the adoption of satisfactory behavior in any given situation. The essential problems in connection with learning of this sort appear to be, first, that of the physiological basis of "random movement," and, second, that of the nature of the process of "adoption," "stamping in," or "learning by experience."
We have seen that the majority of psychologists have attributed this fixation process to some feature of the satisfaction of instincts, appetites or desires. Thorndike has formulated the effect in terms of pleasure-pain, comfort-discomfort, or satisfyingness-annoyingness.