ABSTRACT

When we assign a function to something, we in fact assert two propositions about it. Firstly, that it forms part of a causal network; and secondly, that the results of the causal network, when observed over the range in which they are expressed, exhibit some general property. Another way of expressing the latter point is to say that the causal network is organized. The concept of function is in fact very closely connected with that of organization, and can be regarded as a derivative of that more general notion. Is organization, then, an illegitimate concept? It probably is so in terms of a crudely mechanical materialist picture of the world, i.e. a picture in which we consider that all existing things can without loss be reduced to the movements and interactions of some ultimate constituent particles. But such a picture has never been more than a theoretical aspiration in biology, and is at present out of date even in the physical sciences. There are now, I think, few scientists who would consider it illegitimate to conclude that groups of elementary constituents may, byentering into close relationships with one another, build up complex entities which then enter into further causal interactions with one another as units. It is this fact, of the integration of groups of constituents into complexes, which in certain respects operate as units, which is spoken of as organization. In so far as it occurs, the concept of function is a legitimate one. If we have some complex entity A which acts as a unit, we can regard it as exhibiting organization of its constituent elements. Suppose that within A we can discern certain sub-units, P, Q, R, then the function of P within the organized system A is the contribution which P

makes towards those types of behaviour in which the unitary character of A is exhibited.