We pass, then, to cognition in the usual acceptation of the term—to what we speak of as knowledge in the proper and narrower sense. My contention is that this is a mode of relatedness which science must endeavour to treat on precisely the same lines as it deals with any other natural kind of relatedness. At some stage of evolution it came into being, whereas in the fire-mist, and for long afterwards, it was not in being. None the less when it did come into being, it was dependent on, and for our interpretation it logically implies, underlying physiological processes, as they in turn imply physico-chemical processes, in each case serially correlated. It is pre-eminently selective. And just as any physiological process, however externally conditioned, is grounded in[72] the constitution of the organism, as such, so too is any cognitive process grounded in the constitution of the organism as one in which this higher type of relatedness has supervened. Again, just as the physiological constitution implies a prolonged racial preparation, describable in terms of that mode of relatedness we name heredity, so, too, does any cognitive process imply, not only this racial preparation of the biological kind, but also an individual preparation of the psychological kind—implies relatedness to what we call, rather loosely, prior experience—which itself implies a concurrent physiological preparation.
Now there can be no doubt that awareness is a characteristic feature of the knowledge of cognition, whether it be present or absent in knowledge in the more extended sense. We must just accept this as what appears to be a fact. In science we do not pretend to say why facts are what they are and as they are. We take them as they are given, and endeavour to trace their connexions and their implications. Accepting, then, awareness as given, we must ask: Awareness of what? It is sometimes said that cognition is aware of itself. I am not sure that I understand what this means. If we are speaking of the cognitive relation, which is an awareness relation, the question seems to be whether a relation of awareness is related to itself. But of course if a field of cognitive relatedness be regarded as a complex whole, any part may be related to the rest, and the rest to any part. That kind of self-awareness—if we must so call it!—is eminently characteristic of cognition in the higher forms of its development. On these terms cognition is aware of itself—though the mode of statement savours of ambiguity.
Let us next ask whether there is awareness of the underlying cortical process. If we are speaking of direct awareness, apparently not. The correlation between the two is only discoverable through a very elaborate and complex[73] application of further cognition in interpretative knowledge. We only know the correlated cortical process by description, as Mr. Bertrand Russell would say,[74] and never by direct acquaintance.
Parenthetic reference must here, I suppose, be made to psycho-physical parallelism. But it shall be very brief. The sooner this cumbrous term with its misleading suggestions is altogether eliminated from the vocabulary of science the better. The locus of the so-called parallelism is, we are told, the cortex of the brain. But the cortical process is only an incident—no doubt a very important one, but still an incident—in a much wider physiological process, the occurrence of which, in what we may speak of as primary cognition, implies events in the external world. It is of these events that there is direct physical, physiological, and cognitive knowledge. Of course there are also inter-cortical relations which underlie the relations of those ideal cognita (Spencer's faint class) that supplement the primary cognita which imply direct stimulation of sensory receptors (Spencer's vivid class). It is questionable whether any form of cognition, properly so called, is possible in their absence. Now I see no objection to labelling the fact (if it be a fact) that the cognitive process implies a physiological process in which, as in a larger whole, the cortex plays its appropriate part, by the use of some such convenient correlation-word as psycho-physical; but only so long as this does not involve a doctrine of parallelism; so long as it merely means that cognition implies, let us say, certain underlying cortical changes. Of course it implies a great deal more than cortical process only; but this may perhaps be taken for granted. My chief objection to the word 'parallelism' is that it suggests two separate orders of being, and not two types of relationship within one order of being for scientific study.[75] We do not speak of parallelism between physiological and physico-chemical processes. We just say that scientific interpretation proceeds on the working hypothesis that there is a correlation of such a kind that physiological process implies a physico-chemical basis. So too, I urge, we should be content to say that scientific interpretation proceeds on the working hypothesis that there is a correlation of such a kind that cognitive process implies a physiological basis.
It may be said that Spencer accepted the so-called identity hypothesis which does not lie open to the objection that it suggests two orders of being. He believed[76] 'that mind and nervous action are the subjective and objective faces of the same thing', though 'we remain utterly incapable of seeing or even imagining how the two are related'. Well, we may call them in one passage the same thing, we may speak, in another passage, of the antithesis between them as never to be transcended, and we may try to save the situation by reference to duality of aspect. But this kind of treatment does not help as much towards a scientific interpretation. It is true that, in yet another passage, speaking of the correlation of the physical and the psychical, Spencer says:[77] 'We can learn nothing more than that here is one of the uniformities in the order of phenomena.' Then why not leave it at that? And if there be a constant and uniform correlation which is 'in a certain indirect way quantitive', it would seem that we do see, as far as science ever professes to see, 'how the two are related.' We see, or conceive, how they are related in much the same way as we do in the case of the connexion between the physiological and the physico-chemical, and in numberless other cases. Both parallelism and identity will have to go by the board in a philosophy of science. They must be replaced by the far more modest hypothesis, which seems to express all that they really mean for science, that cognition always implies certain physiological processes in the organism.
If we do speak of mind and nervous action as two faces of the same thing, it seems pretty certain that the one face is not directly aware of the other. When we speak of awareness in cognition we must therefore, it appears, exclude any direct awareness of concurrent physiological processes. Of what, then, is there awareness? Primarily perhaps of some occurrence in the external world. But the difficulty here is that, in the simplest case of human cognition there is awareness of so many things and in such varying degrees. There may be primary awareness of events in the external world (Spencer's vivid series), awareness of the relations involved in these occurrences as such, of the relations of these to ideal re-presentations of like kind (Spencer's faint series), of the relations of any or all of these to behaviour as actually taking place or as ideally re-presented; and all in different degrees within a relational meshwork of bewildering complexity, which we have not, as yet, adequately unravelled. The essential point to bear in mind is that the cognitive relation always involves relatedness of many terms, and that its discussion involves the analysis of what, in the higher phases of its existence, is probably the most complex natural occurrence in this complex world.
I cannot here follow up further the difficult problem of cognition[78]—save to add one or two supplementary remarks. First: it is, I suppose, fairly obvious that any given field of cognitive relatedness comprises all that is then and there selectively cognized. Just as, in the very extended sense of the word 'knowledge', the earth knows, in gravitative fashion, the whole solar system, as does also any one of the planets, so, in the restricted sense, is knowledge co-extensive with all that is, selectively, in cognitive relationship with the organism or that part of the organism which is the locus of awareness. I speak here of the locus of awareness in just the same sense as I might speak of the earth as a locus of gravitative knowledge of the solar system. The locus of awareness is just a specialized portion of the whole relational web. In other words, the relatedness is of the part-whole kind, where whole means rest of the whole other than the specific part. In any such integrated system the part implies the whole—which, by the way, is quite a different matter from saying that the part includes the whole, or, as I understand the words, is equivalent to the whole. But, whereas gravitative knowledge is reciprocal—the sun knowing the earth in the same fashion as the earth knows the sun—cognitive knowledge is not reciprocal. My cognitive awareness of a spinning-top does not imply that the spinning-top is in like manner aware of me. The part knows the whole in a way that the whole does not know the part. The relationship of the part to the rest of the whole is not reciprocal or symmetrical. This we must just accept as a given feature of cognitive relatedness.[79]
Another very important point is that cognitive relatedness is effective. By this I mean that just as, when the earth is in gravitative relation to the sun and the other planets (the constitution of nature being what it is), changes take place because the parts of the system as a whole are in this field of effective relatedness; so too, when the organism is in cognitive relation to its environment, changes in this system also take place just because a part of the whole system is in cognitive relatedness to the rest of the system. That means that the cognitive relation really counts—that it is not merely an epiphenomenal accompaniment of changes which would be precisely the same if it were absent. The 'sum of energy' presumably remains constant. There is no necessary interference with physical principles. But we know of so many cases in which the direction of change may be changed without any alteration of the 'amount of energy', as the phrase goes, that I see no reason, based on physical science,[80] for denying this kind of effectiveness, within a field of cognitive relatedness, if the facts seem indubitably to point to its existence. To assert that the presence or absence of cognitive relatedness makes absolutely no difference appears to me, I confess, little short of preposterous; to urge that it may be brought under the rubric of physico-chemical relatedness surely involves the ignoring of differentiating features, which science should not ignore. But, on the other hand, to invoke an immaterial psychic entity[81]—unless this merely names the relatedness itself[82] as gravitation names the gravitative relatedness—appears to me quite unwarranted in the scientific universe of discourse.[83]
I must, however, draw to a conclusion. I cannot but think that Spencer failed to bring cognition and the conscious awareness it involves into really close touch with the rest of his philosophy of science. No such double-aspect theory as he accepted affords a satisfactory avenue of scientific approach. But where Spencer failed, who has come within measurable sight of success? We are only just beginning to see our way to stating the problem in such a form as to bring it within the purview of science. What we must insist on, as followers, at a distance, of Herbert Spencer, is the treatment of this type of relatedness on lines similar to our treatment of other types of relatedness within one order of nature.
Surveying his work as a whole, we may confidently assert that Spencer brought to a conclusion a great task, and was himself great in its execution. The present generation can, perhaps, hardly realize how potent his influence was on the thought of the latter half of the last century. Many of his conclusions ran counter to those which were, in his day, widely accepted. If only they seemed to him to be true, however, he held to them with a tenacity which his opponents branded as obstinacy. But as he himself said: