Models and Analogues in Biology

By J. E. HARRIS Department of Zoology, University of Bristol

It will be abundantly clear to anyone who reads this symposium, as it was to all who took part in it, that its subject is the scientific method in biology. Not only is this illustrated by the definitions of 'model' and analogue which have been essayed by several contributors as well as by the wide scope of the topics reviewed; it is still further emphasized by the presence of two papers by R. Brown and R. C. Rainey which could equally have been presented at any ordinary meeting of the Society without the use of the terms model or analogue at all. This is as it should be. We are in fact discussing the way in which theoretical and practical research in biological science is done--not on the organism itself but on a mental substitute for it.

It would be difficult and largely pointless to try to present a detailed summary of the symposium; the present essay is intended more as a sort of personal commentary by the writer, whose limited knowledge in many of the fields will be all too obvious. The responsibility for errors in this account is his alone; the credit for any merit it may have belongs not only to those whose papers appear in the present volume but equally to the unnamed contributors to an always stimulating discussion.

It would doubtless be convenient to begin with a definition of the terms model and analogue. But a relatively small amount of time was spent on semantic distinctions at the meeting; it was as if the subject under discussion was so clearly understood that what we chose to call it was immaterial. Williams has rightly emphasized the impossibility of teaching the scientific method, but in a group of practising scientists, the naming of a widely used tool is naturally less important than the manner of using it.

Edwards and Kacser have each provided in their introductions a generally acceptable statement of the position; Williams has attempted a more rigorous formulation of it. We are of course concerned not with simple geometrical reproductions of static systems but with dynamic, functioning features of the living organism and with their representation. In any such application of biological science, it is suggested that the initial stage is the formulation of a schema--the word is used here to denote a limited representation of the system. This representation is intended to fulfil the rigorous conditions laid down by Williams in his definition of a 'model'--that all propositions true of it must be true of the original. This shadowy schema must in practice be clothed with substance by expressing it in some language--verbal, mathe

-250-