Significance of information theory to neurophysiology

Abstract

las re~:entl~y remfided us the concepts pi that Descartes not n~ly had structural mcd.i.fication i)f living tissues to account for . me:nov, but also dearly had before him the concepts of the scanning mechanism, It is as well therefore to begin by considering in what respects OUT ideas care in advance of those of -Descartes. respect I believe are they Pnly 2n one e ssentially in advance of his. We now have as a background to our .thoughts the concept of evolution, and together with it the notion that certain attributes of living matter have survival value. The pht:nompAa comprised by the term '$nind" are included among these attributes, wi.th the resul,t that, recently rema:-&&, we have a as Professor Adrian (I) has "tendency nowadays to regard the relation between mind Cand ma%ter as one which need r1o.t; g5,ve rise to much difficultyt~, The idea, in:~cmceivd.ble t3 nesoartes, now commonplace among biologists that the phenomenon of mind is a byproduct of matter in action dates, as we sot: from the h2storical jntroduction to this symposium by Cherry (4), at least to Julien de la Mettrie (1740). In addrition to this, presentday 'biologists have a krir,d of modesty which would be quite foreign to Descartes, for we realise now that our own picture &' the universe is extremely partisan. This modesty is forcea on us not only by the realisation of the fact that our sense organs have a severely restricted sensitivity "co the various ways in which energy can manifest itself, but also we realise that the types of things we can think about and the ways i.n which we can think about them, are determined by the layout of our nervous s,ystem. For' example, as a member of the primates, we have a disproportior-atel-~. large amount of 012 forebrains devoted to vision, and we are deEied the kind of interpretation oi' nature that would be made if this was replaced by the sense of smell. Nevertheless, in spite of these advances, many of Descartes* difficulties are still with us* The problems of the mechanism of memory and of pattern recognition or 'stimulus equivalence' are today a particular stumbling block. And Information Theory 5s I believe particularly relevant to Neurophysiology because some of our difficulties may exist through an inadequate understanding of *he problems of coding, and of making representations; problems which appear to be the 'bread and butter( of "Information Theory".