Learning as a process of selection and amplification.

Abstract

These are only a few examples of how learning occurs during development by a process of sculpturing, selecting among the possible variations, combinations and sequences of actions. This is how we build up what may be called programs in the brain (Young 1978). The physiologist will say that these are only, as it were, metaphorical statements, not detailed descriptions of known brain actions. ‘The programmes of elaborate sequences of movement, built up over the course of many years, are not yet on the agenda of practical neurophysiology’ (Phillips & Porter 1978, p 333). Perhaps this statement may prove to be over-pessimistic, as I am sure the authors would hope. Studies of learning, of which I have tried to give a small sample, do begin to show us how programs are assembled by selecting from the vast number of possibilities that are available and then increasing the reliability of the chosen actions. It is particularly pleasing to find that we can begin to understand the various principles that are involved. In Aplysia we see something of them due to the cellular changes. In Octopus we see how learning involves different pieces of neuronal machinery. One serves to distribute signals to allow access to the full repertoire of possibilities. Another part distributes and balances rewards and penalties. In a third the actual learning change occurs and the many minute cells suggest that we may look for the origin of learning among reciprocal reflexes.In the vertebrate brain we see the multiplication of neuronal patterns and centres greatly increased. We get glimpses of how various parts of the brain cooperate in building and executing the programs. Learning involves many parts of the brain, each contributing in a different way. The cerebellum perhaps provides the repertoire of small variations in muscular contraction out of which those are selected that perform the acts that prove necessary. Perhaps the motivations to do these correctly come partly from the aminergic pathways such as those of the locus ceruleus, while the hippocampus provides the power to seek to make new ones. And of course the cerebral cortex and its attendant input and output centres provides the biggest repertoire of all, selection among whose modules has allowed us to build and use a program that studies itself. Surely John Hughlings Jackson would have rejoiced that there has been some advance in understanding of the ‘Relations of Different Divisions of the Central Nervous System to One Another and to Parts of the Body’, which was the title of the first Hughlings Jackson lecture, which he himself gave in 1897.