Chapter 5: Whitehead and Modern Science by C. H. Waddington
C. H. Waddington taught and did research at the Department of Genetics in the University of Edinburgh. He died in 1975.
The invitation to take part in this symposium came at a time which was rather awkward for me. I had for a year or two been thinking that it might be useful to write a book about Whitehead and the relevance of his thought to some of our present controversies, particularly in relation to biology, and the ‘anti-science’ movement. I had in fact decided to do so, when I had cleared off my plate a number of things which were already on it. But I had not fully thought out just what I would have to say, and I confess that I still have not yet done so. So my paper now must have a rather interim, half-prepared character; it will be another year or so before I can hope to have properly sorted out whatever I may find I have to contribute. I shall therefore present here only a rather expanded version of the notes I used when I gave the talk in the Villa Serbelloni, rather than a definitive essay.
The points I want to make come under four headings:
I. Whitehead and Reductionism
The controversy between reductionism and anti-reductionism (which in my young days in the ‘thirties we used to refer to as between material materialism and organicism) is usually stated in something like the following terms. We (and science in general) start by accepting the real existence of certain scientific objects -- atoms, electrons, gravitation, light, etc. The question at issue is whether we can account for everything -- e.g., all biological processes, including behaviour (and some people would include, others exclude, mind and/or conscious self-awareness) -- in terms of those entities, as reductionists and mechanists claimed, or do we have to invoke something else, which might be organizing relations’ or ‘system properties,’ as anti-reductionists and organicists argued.
Whitehead stated that we start somewhere else; not with objects, but with ‘events’ which are four-dimensional happenings, i.e., processes. All knowledge, and all talk, is derived from experiences of events. Scientific objects -- atoms, etc -- are not basic, but are derivative, intellectual constructs invented to assist us to understand events.
This view removes the whole heat from the controversy. Reductionism simply does what it says it is going to do, namely it reduces, from the experienced event to an intellectually constructed object, which is useful in making sense out of the particular aspect of the event we are interested in at the time. If we get interested in some other aspect, e.g., if we change from considering the blood circulation of an animal to its nest-building, we have a perfect right to invent new appropriate objects. But, in the interests of consistency, we shall only invent new objects when we cannot escape from doing so; usually, it is sufficient slightly to modify the description of our previous objects, e.g., when we modify the definition of the atom to accommodate it to new observations about radioactivity or atom-smashing, or new types of chemical combination. The point is, the definition which is given of a scientific object at any given point in history is not inviolable, so that the only change possible would be to add something (‘organizing relations’) to it; what happens is that the definition itself is changed.
II. Whitehead and the Complexity of Events
For Whitehead real existents were events. Each event has a definite character, but this results from the ‘concrescence’ of an infinite number of objects (which are essentially relations with other events, ‘prehensions’) into a unity. In doing science we have, on the one hand, to try to formulate simple objects which express the most important causal relations between events, but at the same time we have to ensure that these objects include (as sub-objects) as many as possible of all those involved in the event. The thrust of Whitehead’s thought is not to simplify unduly; every time you ‘reduce’ you leave something out, and scientific ideas are richer and nearer to nature the less that has had to be omitted in order to reach them.
I applied these ideas in embryology as follows. When I began working, the standard concepts of experimental embryology were such things as ‘potencies.’ A potency to develop into neural tissue was a simple concept, but it was in fact totally vacuous since it could not be analyzed into anything else. Reinforced by Whiteheadian principles, I was not afraid to substitute for this the concept that development depended on the activities of very large numbers of genes. It had to be supposed that these activities were brought together to result in some relatively unified type of action, e.g., to form a definitive nerve cell, rather than a definite muscle cell. This ‘concrescence-like’ process would have to involve a great many control circuits, cybernetic-type interactions and so on. Further, this unity would itself be a process. That implies that the unifying action is not a homeostasis, i.e., a set of interactions which ensure that a certain state of the system is stable; instead it is a homeothesis, which brings it about that a certain process of change is stable. When the developing system is disturbed it returns not to the state it was at when the disturbance occurred, but to some later part of the stabilized pathway of change. The stabilized pathway of change is named a ‘chreod,’ and the whole system of chreods in a complex developing system such as an egg gives rise to an ‘epigenetic landscape.’
Note that this is quite a different conception from that which a few geneticists were beginning, at that time or shortly thereafter, to formulate from a reductionist point of view. They asked the question, What controls the activity of a single gene, turning it on or off? We are by now beginning to get quite good answers to this question as regards bacteria, and are even approaching it for higher organisms. But it is not actually the crucial question, which is, What controls the activity of the complex interacting set of genes, which produces a nerve cell or a muscle cell? We know that in general this control of complex development processes takes place earlier than the control of single genes, but I doubt whether we know any more about it now than we did when I first formulated it, around 1940.
If one approaches the problems of evolution with a similar readiness to accept that the process may essentially involve very numerous components, one again comes out with a set of questions which are characteristically Whiteheadian rather than present-day orthodox. For instance, one admits that in much of evolution (probably all above the bacteria), evolutionary changes involve enormous numbers of genes, rather than a selection of one or two particular genes (although that occurs in a few instances, possibly, for instance, in industrial melanism). Of course this is a point which Dobzhansky and Sewall Wright have emphasized, both from the practical and theoretical points of view. It reduces to very small proportions, almost negligible, in fact, the importance of the element of chance mutation, on which R. A. Fisher on the one hand and Jacques Monod on the other have reared such super-structures of rather emotional philosophizing.
Moreover, from the Whiteheadian point of view one has to recognize that the evolving events -- actual animals and plants as we meet them in real life -- are influenced by environmental factors as well as genetic. Further, all living things above a very low level of evolution play some role, active or passive, in deciding what environmental influences will act selectively on their populations. All this produces a much more interactive theory of evolution than the conventional ‘chance and necessity.’ The organism draws its genes from an enormously variegated gene pool; it develops under the influence of them and also under those of a probably pretty heterogeneous environment; and, at any given stage of its life, the way its genes and its previous environment have acted up to that point may have considerable effect on the nature of the environment to which it will next be subjected -- if the animal does not like it here it may migrate someplace else, and so on. We are dealing in fact with a Whiteheadian type of interacting network, rather than a straightforward linear sequence of cause and effect of the classical materialist kind.
III. Whitehead and the Nature of Organizing Relations
Absolutely central to Whitehead’s thought is the idea that a unified event, which has a definite characteristic identity or, otherwise expressed, is an organized unit, is built up by the bringing together in some way of its interactions with other entities in the world (‘the concrescence of prehensions’). How is this done? Since Whitehead wrote in the ‘twenties, the standard scientific ideas about the nature of effective interactions between entities have undergone a number of changes. I think they are just about now beginning to catch up with the first phase of Whitehead’s thought on the subject, that is, his thought up to and including Science and the Modern World. They have not yet caught up with his thinking in Process and Reality. I think science will proceed in that general direction, but I am not quite convinced that it will come to exactly the same conclusions as Whitehead does -- partly because I find it very difficult to make up my mind what those conclusions are. However, the first part of the story is considerably simpler.
The ‘classical’, idea of effective interactions was that of simple material causality, with one or two causes producing one or two effects in a simple linear manner. The next stage was to accept that in many entities we have to consider a large number of interacting components, and processes of cause-effect which are not simply linear, but may interest either by being linked into networks, or by various types of feed-back interaction, positive or negative and so on. These are the ideas which were first brought forward in biology by thinkers such as Needham and Woodger (with myself acquiescing on the side-lines) under the name ‘Organizing Relations,’ and by Bertalanffy, who conducted a vigorous propaganda campaign on their behalf, under the name of ‘Systems Theory, and finally by Norbert Wiener, with equal fervour, as ‘Cybernetics.’ These all essentially involved interaction -- though not simple linear interactions -- between material things.
A new step was taken with the introduction of ‘Information Theory.’ It became common to come across sentences such as the following (taken more or less by chance from the manuscript of a book entitled, Information, Explanation and Meaning, sent to me in advance of publication by its author E. H. Nutton). "The most general model of a natural process on which scientific explanation may be based is no longer the movement of a particle under the action of a force, but the storage (or organization) and the transmission of information within a system." ‘Information’ is, of course, not a material entity. This is the essential change made from the classical notion of effective interactions or the systems theory development of it.
However, Information Theory was developed by Shannon and Weaver, originally in relation to the transmission of messages along communication channels, such as the telephone wire. It is mainly a theory of inactive specificity, which does nothing and brings nothing about. This is a profound limitation on its powers of implication or explanation. It is, therefore, rapidly being displaced from the centre of scientific thought by theories which are concerned with active rather than inactive specificity. These are theories of Instructions (Automata Theory) or Programming. These have been, of course, specially developed in relation to computers, but they deal with interactions in terms of something which resembles information in that it is not material, but differs from it in that it does not merely describe a state, as does information, but both describes a process and, further, instructs that this process should be done. It is in terms of vectors, not of scalars.
For vectors were the basis of Whitehead’s idea of prehensions, as was pointed out particularly by Victor Lowe in his essay on "The Development of Whitehead’s Philosophy" (Schilpp, PA., ed., The Philosophy of Alfred North Whitehead, 1941, 1951). Whitehead always thought of interactions as (a) involving something much more general than the physical forces contemplated in classical dynamics, involving in fact specificities akin to those contemplated in Information Theory, but (b) always as active ingredients of processes. That makes them very similar to the basic components of recent ideas about Automata Theory and Programming.
Whitehead had to go one degree further. He was concerned with the organization of specific character; every event reacts with every other, but not with all aspects of every other. We have one of those systems of circular causation which are characteristic of the whole Whiteheadian way of thought about organic unity. Event A requires its specific character of A-ness only by means of its interactions with all the other events P,Q,R,S,T, etc. But it interacts only with certain aspects of event P, and which those aspects will be depends on the character of A. Thus the character of A both depends on, and decides, the nature of its interactions with P,Q, and the rest. (Note the parallel with the system of ideas about evolution in which I suggest that the character of a living organism is determined by the particular natural selection it is subjected to, and that the natural selection it is subjected to is dependent on what its character leads it to select out of the range of environments available to it.)
Whitehead came to express this notion by saying that the prehension of one event A for another event B was according to the ‘Subjective Aim’ of A, and instead of ‘prehensions’ he began using the word ‘feelings.’ These are obviously very dangerous terms, and a great deal of thought and discussion is necessary to discover just what Whitehead meant by them. Certainly he intended to convey that the character of A, which was decisive in specifying the nature of its interactions with B, was a dynamic character which had some affinity with an intention or an objective, or instruction, rather than a mere static characteristic, such as a chemical composition is usually considered to be. How far he was justified in introducing clear references to human personality, by using words such as ‘subjective’ or ‘feeling,’ is difficult to decide. I am fairly certain he did not intend a simple pan-psychism in which every entity -- every stick and stone--is supposed to have a ‘stream of consciousness’ in any way comparable to our own consciousness. Perhaps he would have ruffled fewer people’s feelings if he had used, like the topologist Rene Thom, a less obviously loaded term, such as ‘the logos’ of an entity, which Thom defines as a figure which describes the totality of the regulatory mechanisms of a system.
I do not want to pursue the discussion of this matter in this section, the main emphasis of which has been to point out that Whitehead already in the ‘twenties was thinking in terms very close to those of the Automata Theorists and Programmers of the present day. I should, however, also remark that the more subtle developments of Whitehead’s thought seem to have been the inspiration for one of the most thorough and impressive discussions of the evolution of human mentality and language in its relation to cognate activities in earlier evolutionary forms, namely Suzanne Langer’s impressive work, Mind: An Essay on Human Feeling, of which two volumes have so far appeared and a third is promised soon. Thorpe stated that he personally had found Whitehead’s thought of little help in relation to his own work on animal behaviour which was largely concerned with birdsong, but I think Suzanne Langer has shown that it may indeed be illuminating to think of problems of animal communication and eventually human language in terms of instructions, subjective aims or feelings, rather than in terms of information and description of states of affairs.
IV. Whiteheadian Science and the Present Anti-Science Movement
This is the aspect of Whitehead’s relevance to the thinking of the present day that I do not feel ready yet to go into very deeply.
In cannot be denied that at the present day there is a widely disseminated feeling, which affects many people quite deeply, that it is an inescapable consequence of the essential nature of science that it devalues, dehumanizes, impoverishes the relation between man and nature; anyone who regards nature scientifically must find himself alienated and estranged from it. Exponents of this view are Marcuse, Gillespie, Roszak and many others.
Whitehead was the Knight in Shining Armour (some people seem to think him only the White Knight) who rode out to do battle against any and every form of what he called the Bifurcation of Nature -- whether it was the Cartesian Dualism or the alienation which the sensitive soul feels when it learns that water may be represented as H2O. As we have seen knowing which occasions, and when in the development of that line of creatures, makes all the difference. We can never know just how ‘the giraffe got his long neck,’ but like everything else in evolution it is the result of subjectivity and purpose. Now, all the effective components in this picture---the instructional interactions and the events with a specific character, into which they are drawn together -- have the character of intentions, so we are already, in our basic view of science, in a world in which conceptions of Value are at least not foreign and maybe are inescapable. If we approach science from the Whiteheadian point of view, the fortress which the anti-scientists will have to attack is not what they think it is, and may be capable of mounting a rather devastating counter-attack.