Learning and brain lesions: 1

Abstract

Learning can be discussed in a variety of languages, and since the language that is used can reflect preconceptions about the likely mechanisms, it is as well to be clear about the sense in which any particular author is using words that mean different things to different people. I shall assume that any long-term (hours, days or months) change in behaviour is associated with structural changes in the nervous system. There are plenty of good precedents for this. In Octopus, as in other animals, learned changes in behaviour survive a variety of treatments, ranging from anaesthesic to electroconvulsive shock (provided that it is given several hours after training — see Table 11.1) that might be expected to disrupt any mechanism dependent upon on-going electrical activity. ‘Structural change’, in this context, could be at any level, from alterations in the external shapes or connexions of cells, to changes in their biochemistry. The structural changes that result in a long-term learned change in response to a specific stimulus constitute a ‘memory store’, ‘trace’ or ‘representation’ of that event. A memory store is at present a concept; nobody has ever observed one in any animal, only the consequences of its existence. These consequences are behavioural, ‘memory’ is demonstrable, the animal shows by its actions that it ‘remembers’. A learned change in behaviour implies the continued existence of a structural alteration. In contrast a failure to remember is not evidence that the memory store has disappeared. It may have done, but all one can say for certain is that it no longer seems to be determining what the animal does.