CHAPTER IV

SENSATION AND BEHAVIOUR

1. SENSATION AND THE ORGANISM

4.1. In the preceding chapters the apparatus for the classification of impulses has been represented as if it were a neutral, self-contained, and completely centralized system which passively registered the simultaneous occurrence of impulses set up by external stimuli and thus came to reflect the significance which the stimuli possessed in the environment of this system. Such a passive apparatus of registration is conceivable, and to consider it served to bring out the general principle of our theory. But it would be something very different from the sort of apparatus which the nervous system constitutes. While it would register the significance of the stimuli in the environment, it would not indicate the special significance which they possess for the living organism of which that apparatus forms a part.

4.2. The exclusive concentration on the order that might be created by the establishment of the connexions between sensory impulses only has been adopted quite deliberately (3.18). It was intended to emphasize one aspect of the more complete picture which, under the influence of behaviourism, has been somewhat neglected during the last generation. The emphasis which was placed during that period on the observable peripheral responses brought it about that the rôle played by the higher nervous centres has been largely disregarded, and that the whole relation between stimulus and response has often been treated as if the higher centres did not exist. In the preceding chapters we have gone to the other extreme and practically disregarded everything except the central effects of any sensory impulse. This temporary disregard of the fact that the nervous system operates within a living and acting organism, which in some measure is capable of adaptive and regulative behaviour apart from control by the higher nervous centres, must now be corrected by an explicit consideration of these facts.

4.3. In the present chapter, therefore, we shall have to examine not only the effects of the sensory on the motor processes, but also have to give much greater attention than we have yet done to the sensory impulses set up by the various processes in the body, that is, to the registration of stimuli which originate in what has appropropriately been called the milieu intérieur, the internal environment, within which the central nervous system functions. Of the latter we shall have to conceive as a sort of apparatus of control superimposed upon a living whole rather than as a self-contained and fully centralized structure of its own.

4.4. In turning to these problems we are entering a field in which the very attitudes which during the past generation have led to a comparative neglect of our main problem have led to great progress and the accumulation of a wealth of new knowledge. We have nothing to add to this and cannot even hope to give the barest outline of all the relevant facts which a more systematic survey of the field would have to take into account. The sole purpose of this chapter is to show how our theory of the determination of sensory qualities fits into the picture of the ‘integrated action of the nervous system’ which is gradually emerging.

4.5. At the same time it should be pointed out, however, that in one respect in which the task which we are undertaking is most in need of a solid foundation, theoretical biology is only just beginning to provide the needed theoretical tools and concepts. An adequate account of the highly purposive character of the action of the central nervous system would require as its foundation a more generally accepted biological theory of the nature of adaptive and purposive processes than is yet available.

4.6. The consideration of the inter-relations between sensory and motor processes will also make it necessary to take more explicit notice of the hierarchical order of the central nervous system. We shall see that the organization of all connexions between sensory and motor processes on many superimposed levels, and the corresponding existence of a hierarchy of centres of increasing comprehensiveness, is of the greatest importance for the understanding of the sensory order.

4.7. In some measure connected with this hierarchical order of the nervous system is the distinction between the phylogenetic and the ontogenetic aspects of the processes in question, or between those connexions which are inherited and those which are acquired by the individual. There is, however, not a great deal which, in the present state of our knowledge, can be said on this question; we shall on the whole have to continue to disregard this distinction and to represent the process of the building up of the sensory order as if it took place in the course of the life of the individual.

4.8. The relation between sensory and motor processes which we shall have to consider is a double one: we shall have to consider both how various complexes of sensory impulses will influence behaviour, and how in turn the motor responses will influence sensory discrimination. The latter question will make necessary some consideration of the interoceptive and proprioceptive impulses, i.e., those impulses which record not external stimuli but various states of different parts of the organism.

4.9. A more systematic discussion of the connexion between the sensory and the motor apparatus would also have to include an examination of the manner in which the efferent (or motor) impulses are themselves ordered so as to produce certain co-ordinated patterns of movement, and of how the bundles of efferent impulses interact with the proprioceptive afferent impulses by which the resulting movements are recorded at the centres. In this respect we can, however, attempt no more than the barest sketch which must serve as an indication of the sort of problems which a fuller elaboration of our theory would have to answer.

4.10. In examining the significance of the proprioceptive impulses we shall have briefly to consider not only the effects which the impulses recording postures and movements accompanying perceptions have on sensory discrimination, and the rôle played in this connexion by the back-reports of responses which are produced by the stimuli at various subcortical levels; but in particular also to examine the significance of the various ‘biogenic’ impulses which are caused by the vegetative processes of the organism and are closely connected with the various ‘urges’, ‘drives’, or ‘wants’. The latter are, of course, essential for any explanation of purposive behaviour.

2. EVOLUTION AND THE HIERARCHICAL ORDER OF THE CENTRAL NERVOUS SYSTEM

4.11. The continued existence of those complex structures which we call organisms is made possible by their capacity of responding to certain external influences by such changes in their structure or activity as are required to maintain or restore the balance necessary for their persistence. This involves, even in the most primitive organisms, some capacity of discriminating responses to different physical stimuli, and perhaps even some capacity of ‘learning’,1 although we know very little about the nature of such individual learning (as distinguished from the process of hereditary selection of such individuals as show appropriate adjustments).

4.12. The fact that an organism will respond differently to different external forces acting upon it is, of course, not peculiar to organisms. It would be merely an instance of different causes producing different effects. The peculiar problems presented by organisms appear only where they respond to particular stimuli in the manner which will secure their continued existence, and in so far as they develop specific organs which enable them not only to discriminate between different stimuli, but to react differently to the same stimuli if they appear in different combinations with other stimuli or when the organism itself is in different states.

4.13. It is perhaps worth stressing that the problem of purposive adjustment of organisms to changes arises long before the problem of its purposive behaviour with regard to external objects. The question of what determines (or what is meant by) purposiveness is in the last instance really the same question as that of what ensures the continued existence of the organism. It arises as much in connexion with the normal functioning and growth of the organism, the processes of metabolism and the replacement of damaged parts, as in connexion with those movements of the organism which we commonly describe as behaviour.

4.14. It has already been suggested that in a certain sense any attempt to explain the highly complex kind of purposive action made possible by a developed central nervous system may be premature so long as we do not possess a fully adequate biological theory of the comparatively simpler kind of purposive functioning. Many of the problems often regarded as peculiar to mental phenomena in fact arise already at a much earlier stage, where there can be no question yet of that complex order that is shown in the response to external stimuli which we have described as mind. It cannot be our task here to restate the present position of biological theory with regard to these problems, and we must content ourselves to refer in this connexion to W. B. Cannon’s concept of homeostasis and its development by other authors, and especially to the more recent and most promising work of L. von Bertalanffy. His theory of ‘open systems’ in a steady state (Fliessgleichgewicht) in which ‘equifinality’ prevails because the equilibrium that will be reached will be in some measure be independent of the initial conditions, seems to provide the most helpful contribution to this problem.2 Any further comments we shall have to make with particular reference to the problem of purposiveness will be reserved to the next chapter (5.63–5.76).

4.15. Here we are not directly concerned with the regulative functions of the organism other than those which are effected through the central nervous system. It is merely necessary to remain aware throughout that this system functions within an organism which independently of the former is capable of some adaptive and purposive responses to external causes, responses which are brought about through a system of neuro-chemical regulation. Our task begins essentially where the somatic nervous system makes possible discriminatory responses to a great variety of combinations of stimuli, and particularly where learning becomes the dominant factor.

4.16. The mere fact that the organisms in the course of their evolution develop specific receptor organs which are sensitive only to fairly narrow ranges of stimuli must not be confused with the development of a sensory order and of distinct sensory qualities. It is to be assumed that this development goes hand in hand with the acquisition of different motor responses to the different stimuli. But even an organism which had developed distinct receptors sensitive to all those various stimuli which in a highly developed central nervous system produce different sensory qualities, for that reason could not yet be said to discriminate in terms of a system of sensory qualities similar to that familiar to us.

4.17. While such an organism would perform the simplest kind of classification we have discussed (2.35–2.38), it would still be unable to perform the multiple classifications which alone give rise to the system of sensory qualities. The different stimuli which evoke the different sensory qualities would, if occurring in isolation, all produce different effects, but these different effects would not differ from each other in the specific manner in which the sensory qualities differ from each other.

4.18. The essential characteristic of the order of sensory qualities is that, within that order, each stimulus or group of stimuli does not possess a unique significance represented by the particular response, but that they are given different significance if they occur in combination with, or are evaluated in the light of, an infinite variety of other stimuli which may originate from the external world or from the organism itself.

4.19. Such an order implies that any impulse recording a particular stimulus is connected not merely with one particular motor response, but that some apparatus exists by which the effects of any impulse are adjusted to, and integrated with, the effects of other impulses proceeding within the central nervous system at the same time. In other words, the various sensory impulses, whose effects are thus to be adjusted to each other, must in some manner be brought together before the effect of their joint action is decided.

4.20. This does not mean that individual afferent impulses, or groups of such impulses, might not also be uniquely connected at low levels with certain efferent impulses so that, as soon as the former occur, a particular movement is produced. Such a relation would correspond to the ideal simple reflex arc of traditional theory. In the present context such reflex responses are important mainly because of the proprioceptive impulses by which they in turn will be recorded in the higher centres. The original exteroceptive impulse which sets up such a reflex will in consequence arrive at the centres to which it is conducted already accompanied (or rapidly followed) by a report of the spontaneous response of the organism to the external stimulus. The impulse recording the external stimulus is thus already ‘marked’ as meaning (involving) a certain kind of response.

4.21. This sort of relationship we must suppose to recur in relays, or at many successive stages: the initial sensory impulse at the first stage, the spinal cord, setting up both a motor response and a further afferent impulse proceeding on to the higher centres. At the next level it will arrive together with the report of the motor responses it has produced at the lower level, and with other sensory impulses recording other peripheral stimuli, which may be similarly accompanied by the reports of the reflex responses which they have set up at lower levels. At this stage this particular combination of signals may again produce a distinct motor response, so that at the next higher level the bundle of impulses arriving there will include reports of responses which take already a wider range of exteroceptive impulses into account. And as we ascend to higher and higher levels, both the comprehensiveness of the range of external stimuli which are taken into account in any response, and the number of responses effected at lower levels and reported back to these higher levels will constantly increase.

4.22. It is not difficult to see how such an arrangement for the mutual adjustment of the responses to different simultaneously occurring stimuli is made necessary by the development of specific receptors for different kinds of stimuli. So long as the whole organism was merely susceptible to irritation by a wide range of stimuli, and was capable only of a few simple responses, such as contraction and expansion, no special apparatus for co-ordinating the responses to different stimuli was required. But as soon as specific responses became attached to particular classes of stimuli, the mutual adjustment of these responses according to the significance of the particular combination of stimuli became necessary.

4.23. Between the case where specific responses are uniquely attached to particular stimuli, and the case where all responses are decided in view of all the stimuli, there is, of course, an enormous range of intermediate possibilities. Nor need in a given organism only either the one or the other type of arrangement exist. Some other stimuli will be more likely to affect the appropriateness of a given response to one particular stimulus than will be true of others, and there will be more need—or it may be easier for the organism to provide—for the mutual adjustment between those than between others. We must probably assume that, in the course of evolution, the original direct connexions between particular stimuli and particular responses are being preserved, but that control mechanisms are being superimposed capable of inhibiting or modifying these direct responses when they are inappropriate in view of other simultaneously acting stimuli.

4.24. Parallel with this progressively more complex evaluation of the stimuli in the light of an ever more comprehensive collection of other stimuli, a similar organization will operate on the motor side: instead of simple movements of particular muscles, more and more complex patterns of behaviour will be evoked as a whole; and the groups of impulses which evoke this pattern of movements are probably evoked as groups by a few central impulses which ‘stand for’ the whole pattern (4.48–4.51).

4.25. The sketchy manner in which these questions must be treated here ought not to give the impression that these relations are simple. It is neither to be assumed that comparatively simple patterns of stimulation will normally produce comparatively simple responses, and that the integrative action of the higher centres will operate only when more complicated stimulation patterns are involved; nor that the motor responses are built up in a simple additive manner from the effects of individual impulses producing the movement of individual muscles. It is very likely that, just as more than one afferent impulse will generally be required to produce anything like a ‘simple’ (or sharply discriminated) sensation, so a single efferent impulse will as a rule produce somewhat diffuse movements and only the overlapping of many such impulses will produce a clearly differentiated movement of the separate muscles.3

4.26. The increasing differentiation of the different stimuli from each other, and the accompanying increased variability and complexity of the responses to any subgroups of stimuli, involves, as we have seen, that impulses representative of these stimuli are brought together so that they can act upon each other in a manner which reproduces their significant relations. The more comprehensive this adjustment is, the more elaborate must be the centres set aside from, and capable of overruling the effects of, the more direct connexions between stimulus and response.

3. FROM SPECIFIC REFLEX TO GENERALIZED EVALUATION

4.27. It is doubtful whether the ideal simple reflex arc, where the impulse from a single afferent fibre is transmitted to a single efferent fibre, is of any importance, and even whether it occurs at all. But between it as the one extreme ideal type and the other extreme of the ‘voluntary’ or ‘conscious’ responses there exists probably a continuous range of connexions between stimuli and responsies of intermediate types in which processes of classification take place which are more or less analogous to those which determine the system of sensory qualities. The simplest of these intermediate cases which is of interest is that in which a particular motor response becomes attached to every one of a group of sensory impulses, so that any one of the latter will be transmitted to the motor fibre and produce the response in question. This represents the simplest possible instance where we can speak of a classification of the stimuli.

4.28. The operation of this simple kind of classification is familiar from the experiments with conditioned reflexes and from the phenomenon known as generalization. It has been found tha after a conditioned response has been developed to one stimulus, other ‘similar’ stimuli may also elicit the same response.4 In these instances the grouping of certain impulses as similar has the effect that further impulses which become connected to some of these impulses become also attached to the other impulses forming the group.

4.29. A higher degree of selection or classification is reached when several responses are alternatively connected with each of a given group of sensory impulses, so that which of these responses will be elicited by a particular stimulus depends on which of a number of other sensory impulses occur at the same time with the former. Which response will be produced in this case by any sensory stimulus will depend on what other stimuli accompany it, and any particular sensory impulse may in some context produce results which are similar to those produced by others, and in other contexts produce results which are different.

4.30. On the lower levels on which connexions of this still relatively simple or quasi-reflex type prevail, there will thus already exist some sort of qualitative ordering or discrimination. But it will be greatly limited in two respects: it will involve selective responses to the stimulation of particular sensory receptors for only a very few kinds of responses: and it will select between this limited number of possible responses only with regard to a limited number of simultaneously occurring other stimuli.

4.31. While on these lower levels the discrimination may thus be fairly detailed in so far as it refers to particular responses or functions, it will be specific in the sense that it will be effective only with regard to a particular group of responses, and take into account only a comparatively small range of stimuli. In the famous case of the decapitated frog which is still capable of wiping a drop of acid from its back, the signal evoked by the drop of acid will be sufficiently discriminated as regards location to guide the movement of the leg. But the localization of the stimulus which this proves will probably be specific in the sense of being effective only with respect to this particular response.

4.32. Such limited classification may be effected in subcentres which offer the opportunity for connexions among a limited number of sensory and motor fibres. As the impulse is transmitted to higher and more comprehensive centres, there will arise opportunities for more extensive connexions, and with them will appear the possibility of a more complex discrimination both with regard to the range of different responses and to the variety of stimuli which will contribute to the decision which of the potential responses will take place.

4.33. The increasing opportunities for connexions between fibres carrying sensory impulses from different parts of the body, and the correspondingly increased comprehensiveness of the net of connexions which can be formed on the higher levels, does mean neither that at these higher centres the individual stimuli must always be represented by individual impulses as they are on the lower levels, nor that the lower level connexions are confined to impulses belonging to the same sense modality.5 It means merely that at the lower levels only such other sensory impulses will generally be able to modify the response to a particular impulse as are most immediately relevant to the interpretation of (or most frequently associated with) the particular stimulus; while at the higher levels a wider range of less immediately significant other factors will have an opportunity of modifying the result. Similarly, the increasing comprehensiveness of the connexions possible at the higher levels need not mean the possibility of connexions with a larger number of individual impulses, but may mean merely the possibility of connexions with impulses representing a greater variety of stimuli.

4.34. The responses to any given stimulus thus become at the higher levels more and more liable to be modified by the influence of impulses from other sources. The continuous range of connexions between the simple reflex and conscious action thus becomes one in which an ever-increasing number of different stimuli contribute jointly in determining the response. Even though we are directly familiar only with the classification of stimuli which lead to behaviour that is conscious, and in which this comprehensiveness of the stimuli taken into account has presumably reached the highest degree, at least a great part of observable behaviour is probably guided by processes which are intermediate between this and reflex action.

4. PROPRIOCEPTION OF LOW LEVEL RESPONSES

4.35. The fact that the sensory impulses may evoke responses on many successive levels has great influence on the manner in which they will be discriminated at the higher levels. In so far as sensory impulses evoke such responses at lower levels, they will arrive at the higher levels accompanied by the proprioceptive impulses recording those responses. The higher centres will in consequence at any one time receive reports not only of given external stimuli but also of the body’s spontaneous reaction to those stimuli. The effect of a bright light will not be only a visual impulse but also an impulse reporting the contraction of the pupil, etc. So far as the higher centres are concerned, the self-moving organism must indeed be regarded as part of the environment in which they live.

4.36. Since as a result of the excessive stress placed by the behaviourists on the peripheral responses, certain misconceptions about their significance are still prevalent, it will be necessary to consider with some care the rôle which such peripheral movements can play in the structure of nervous action. The first point which requires emphasis is that peripheral events, in order to influence further central nervous processes, must be reported back to the centres in which these processes take place. It will therefore be neither the resulting movements as such, nor the efferent motor impulses, but the proprioceptive impulses recording these movements which affect the further neural processes. (The theoretical possibility that part of every efferent neural impulse may, as it were, be branched off before it leaves the centre from which it originates, so as to represent there the resulting movement, can be disregarded because there appears to exist no evidence for this.)

4.37. This means not only that, even where distinct motor responses to the individual stimulus take place, it will still be the (proprioceptive) sensory impulses and not the motor impulses themselves which are important for our purposes, but also that, once a certain peripheral response has become the regular effect of any group of stimuli, it will no longer need actually to occur, since the reports of its occurrence will be associatively evoked by the original stimulus. The emphasis placed by the behaviourists on actual movements, and their efforts to discover at least traces of such movements in the form of ‘implicit speech’ and the like, were thus misplaced. They are not required and the establishment of their existence would not help to answer the problem of what, for instance, constitutes thought.6

4.38. Nevertheless, it is true that the sensory order with which we are concerned is both a result and a cause of the motor activities of the body. Behaviour has to be seen in a double rôle: it is both input and output of the activities of the higher nervous centres. The actions which take place independently of the higher centres help to create the order of the sensory impulses arriving at that centre, while the actions directed from that centre are determined by that order.

4.39. The evaluations of sensory impulses arriving at the highest centres may be compared to the appreciation of the events on the road observed by a person who is being driven in a car, or to the judgements of the pilot of an aeroplane which is being steered by an automatic pilot. In these instances different observed events will lead the passenger of the car, or the pilot of the plane, to expect certain responses of the car or the plane, and those events will come to ‘mean’ for the person particular kinds of responses of the vehicle, just as certain kinds of stimuli mean certain spontaneous responses of the body. The sight of an oncoming car will come to mean the sensation of the car in which the person rides drawing to the right, and the sight of a red traffic light will mean the feeling of the car slowing down. Very soon what will actually be noticed will no longer be that normal response, but only its absence if it fails to occur.

4.40. The position of the highest centres in this respect is somewhat like that of the commander of an army (or of the head of any other hierarchical organization), who knows that his subordinates will respond to various events in a particular manner, and who will often recognize the character of what has happened as much from the response of his subordinates as from direct observation. It will also be similar in the sense that, so long as the decision taken by his subordinates in the light of their limited but perhaps more detailed observation seems appropriate in view of his more comprehensive knowledge, he will not need to interfere; and that only if something known only to him but not to his subordinates makes those normal responses inappropriate will he have to overrule their decisions by issuing special orders.

4.41. In the same manner as, for instance, the captain of a battleship may sometimes recognize the nature of an observed object less from his direct perception of it than from the responses of his ship, so the brain often may get a direct report merely about the action of one of a large class of stimuli and yet be able to recognize its character from the almost simultaneous reports of the responses of the body directed by lower levels. At the same time these responses of the lower centres to particular stimuli, of which the higher centres have no reports, may be governed by general ‘directives’ issued by the higher centres. (We shall see presently that this ‘set’ of the whole organism which determines what the response to a particular stimulus shall be, may in turn be determined either by processes in the highest centre or be the result of subcortical regulation).

4.42. So far as the higher centres are concerned, a given combination of external stimuli will not merely mean that such and such other external events are to be expected, but also that certain adjustments of the organism are taking place. The significance of the effect of cold on the skin will not only be that certain action is indicated, but also that certain responses of the body will automatically come about—not merely a report of a single external stimulus but at the same time also of a change in the state of a great part of the body.

4.43. While it is on the whole more likely that responses via the lowest centres will be innate for the individual, that is, acquired by the race in the course of evolution, while the responses effected by the higher centres will be largely based on individual experience, this cannot be regarded as a universal rule. Probably some inherited responses are effected on fairly high levels, while some learned responses may, after sufficient repetition, become almost completely automatic and be effected at low levels.

4.44. It should also be noted that the degree of modifiability of the response to a particular stimulus by other simultaneous stimuli need not vary in strict correspondence with the extent to which these responses can be altered by individual experience: an acquired response to a given stimulus may be uniquely determined by that one stimulus, while an inherited response may be capable of considerable variation according to the accompanying circumstance.

5. POSTURES AND MOVEMENTS CONNECTED WITH PERCEPTION

4.45. The first group of motor responses to sensory stimuli which we must consider further are those which assist perception directly and which might almost be described as part of the act of perception. We have already mentioned the classical instance of the kinesthetic sensations connected with the focusing of the eye. The familiar effects of displacing the eyeball or of crossing the fingers on the localization of the sensations affected belong to the same category. It is becoming increasingly clear that these are merely special instances of a very general phenomenon, and that the proprioceptive reports of the body postures and movements designed to help perception serve always as a sort of indispensable background for the proper evaluation of the stimulus.

4.46. Recent investigations of the relation between sensation and movement show that this connexion is even closer than had been commonly supposed and that practically all sensory impulses are evaluated in the light of, or corrected for, simultaneous muscular activities. V. von Weizsaecker, to whom we owe a great deal of knowledge on this question, speaks with considerable justification of a complete ‘entwinement’ (Verflechtung) of sensation and movement.7 This seems to apply as much to the evaluation of external stimuli in the light of the simultaneous proprioceptive impulses as reciprocally to the evaluation of the latter in the light of the usually accompanying exteroceptive impulses. Stretching my leg downwards means for me that I expect to feel the ground, and stretching my whole body means that I expect it to cool more rapidly than in a crouched position, etc., etc. The proprioceptive impulses thus receive their significance as much from the exteroceptive ones which are associated with them, as the reverse is true.

4.47. Every sensory situation thus means, among other things, that various movements will have such and such effects, and the totality of the simultaneous exteroceptive and proprioceptive impulses forms the background, as it were, against which the individual impulse is evaluated. It might even be said that every single sensory impulse is probably multivalent, capable of producing various different sensations, and that which sensation it will produce will depend on what other impulses occur at the same time.

6. PATTERNS OF MOTOR RESPONSES

4.48. The manner in which the separate motor impulses are co-ordinated so as to produce complex patterns of behaviour consisting of many simultaneous and successive movements can be considered here only very briefly. We must probably assume that these patterns can be elicited as wholes by a few signals sent out from the higher centres, and that we have thus on the motor side to deal with a phenomenon of ‘bundling’ which in some respects is the converse of the process of classification on the sensory side. As in the latter case, different complexes of sensory impulses will be represented at the higher centres by a few ‘representative’ impulses, so that a few central impulses may suffice to evoke bundles of motor impulses producing complex patterns of behaviour. The particular manner in which this behaviour is executed may then be determined by the interplay of motor and sensory impulses at lower levels.

4.49. These behaviour patterns, however, must not be conceived as fixed but as highly variable. Just as at the higher centres it will not be only one particular sensory impulse, but any one of a class of many different combinations of impulses, which will give rise to a particular response, so the motor signal sent out from the higher centres will be for the execution not of one particular pattern of co-ordinated movements but for any one of a class of such patterns. Such a class of patterns will consist of those different combinations of movements which under different conditions will produce a particular result. Which of these patterns will be put into effect will be decided in the light of the whole sensory position.

4.50. At the higher centres the connexions will thus increasingly exist, not between particular stimuli and particular responses, but between classes of stimuli and classes of responses, and between classes of classes of stimuli and classes of classes of responses, etc. The order given by the highest centre in response to a particular situation may thus be of the kind which we have called a general ‘directive’ for an action of a certain class, and it may be only at lower levels that the appropriate response is selected from the class of behaviour patterns which in different situations may produce the desired result.

4.51. The extent to which behaviour patterns can be adjusted to the sensory situation probably varies with the level which is in control. There is reason to believe that some highly stereotyped or ‘mechanical’ patterns, such as those of the movements of flying and running, are co-ordinated on a fairly low level and that even at these low levels the execution is constantly controlled and modified by sensory signals from the kinesthetic receptors and the semicircular canals. At higher levels the pattern of movement will be variable to a higher degree.

4.52. We can again not concern ourselves here with the question to what extent behaviour patterns are innate to the individual and how the innate and the learned behaviour pattern interact.8 There can be little doubt that even fairly complex behaviour patterns, or rather classes of behaviour patterns from which a selection will be made in the light of the whole sensory situation, are innate and can be elicited by fairly simple stimuli. A well-known instance is the evocation of the maternal behaviour in the rat by definite chemical stimuli.9

4.53. The selection of the particular behaviour pattern from the class of such patterns appropriate to the result aimed at, must not be conceived as taking place in one act. The choice of a kind of behaviour pattern and its continued control, modification, and adjustment while it takes place, will be a process in which the various factors act successively to produce the final outcome. It is not as if the whole behaviour pattern were determined upon before any movement takes place, but rather that during the process of execution further adjustments are constantly made to secure the result.

4.54. In connexion with these continuous adjustments, made while the movement proceeds, the interaction between the exteroceptive and the proprioceptive impulses and the operation of the ‘feed-back’ principle10 become of special significance. In the first instance, the sensory representation of the environment, and of the possible goal to be achieved in that environment, will evoke a movement pattern generally aimed at the achievement of the goal. But at first the pattern of movement initiated will not be fully successful. The current sensory reports about what is happening will be checked against expectations, and the difference between the two will act as a further stimulus indicating the required corrections. The result of every step in the course of the actions will, as it were, be evaluated against the expected results, and any difference will serve as an indicator of the corrections required.

4.55. In this process the intervention of the highest centres is probably needed only to give the general directions, while the execution and current adjustment is left to the guidance of the lower centres. Once the ‘course is set’, the deviations will be automatically corrected by the differences between the expected and the effective stimuli acting as the signs which produce the correction. Such responses to a difference between expectations and outcome are merely a special case, on the one hand, of the general principle that a response to any new stimulus is determined by the pre-existing sensory state, and, on the other, of the capacity of the nervous system to respond in a particular manner to certain kinds of relations between impulses rather than merely to particular impulses. Both these characteristics of the higher centres, the predominant importance of the pre-existing excitatory state, and the tendency to respond to differences between expected and realized impulses, will have to be considered further in the next chapter.

7. BIOGENIC NEEDS AND DRIVES

4.56. We still have not yet noticed the prime sources of activity of the organism, namely those changes in its constitution or balance which occur periodically as a result of its normal vegetative processes and which make action by the organism necessary if it is to survive. In consequence of our stress on the sensory organization we have so far treated the whole problem as if it were mainly one of adaptation of the organism to changes originating in the environment. But even more important than the question why the organism will behave differently in different environments is the question why it will at different times behave differently in the same environment.

4.57. Even more than before our discussion must here presuppose a great deal which belongs to theoretical biology and physiology. As has already been pointed out (4.13), there is really no other difference between the problem of the purposive internal functioning of the organism and that of its purposive behaviour towards its environment, than that the latter raises the problem of a comprehensive order of the various external stimuli which determines how in different combinations they will modify each other’s effects. This includes the problem of why internal stimuli may bring it about that a given organism will at different times respond differently to the same set of external stimuli.

4.58. What is significant here for our purposes is not so much the precise nature of the physiological processes which determine such states as hunger, thirst, and the like, but what the ‘attitudes’, ‘dispositions’, or ‘sets’ corresponding to these physiological states mean for the responses of the organism towards its environment. Since these various ‘needs’ or ‘drives’ may be produced by visceral, glandular or general metabolic processes, it is convenient to refer to them by the generic term ‘biogenic needs’.11

4.59. It may be mentioned at once that these ‘needs’ resulting from the spontaneous vegetative processes of the body are, of course, closely related to, and sometimes practically indistinguishable from, another kind of attitudes or sets such as fear or rage, which, though usually caused by some sensory perception, also consist of a disposition for a certain range or type of actions. The problems which these ‘emotions’ or ‘feelings’ raise are thus very similar to those raised by the needs in the narrower sense. It would be difficult to decide whether the sexual urge provoked by a sensory impression is in this sense a ‘need’ or an ‘emotion’. Similarly, appetite may be stimulated by the smell of some delectable food without hunger being present, or a sense of fear caused by bodily processes without any sensory (i.e., exteroceptive) experience inspiring the fear.

4.60. While we shall in the first instance consider the significance of needs and wants proper, and leave to the next sections any more specific comments on emotions, most of what is to be said about needs applies equally to emotions. Both involve not only a disposition of the organism towards a certain class of actions, but also a special receptivity for certain classes of stimuli. As a result of a peculiar state of balance the whole organism comes to ‘like’ or ‘dislike’ particular kinds of stimuli. We shall later in connexion with ‘attention’, (6.26–6.27) have to consider more fully the nature of this state of excitatory preparedness.

4.61. It is perhaps useful to distinguish between the term ‘set’12 as the name for the preparedness of the organism for certain kinds of actions, and the term ‘expectancy’ for the increased receptivity for certain kinds of stimuli which will elicit the corresponding responses. But such a distinction between the sensory and the motor aspect of what is essentially a relation between a class of stimuli and a class of responses must not lead us to treat them as if they were really separate. The important point is their close connexion, the fact that the organism will be disposed to respond in a particular manner to any one of a class of stimuli.

4.62. At this stage of the exposition it is yet too early to try to show even in outline how such a state of need, which at first may produce merely an aimless increase of motor activity, may become directed towards the purposive search for certain kinds of stimuli, so that the animal searches for food or for a sex partner, which, when found, will produce the consuming activity. This will have to be attempted in the next chapter.

4.63. Our present purpose was merely to show that in addition to the ‘objective’ significance which the different stimuli will acquire for the organism as the result of their regular association with other stimuli, they will also acquire a special ‘subjective’ or ‘pragmatic’ significance through their capacity of satisfying certain needs. The connexions which will give them this significance will operate not only through certain stimuli producing certain actions if the need is present, but also through the need of making the organism search for stimuli of the appropriate kind. This evaluation of stimuli with respect to goals which are determined by the momentary needs will have to be considered further when we examine the general problem of how a representation of the environment enables the organism to act ‘purposively’ (5.64 ff.).

8. EMOTIONS AND THE JAMES-LANGE THEORY

4.64. The second kind of dispositions, emotions, are dispositions for a type of actions which in the first instance are not made necessary by a primary change in the state of the organism, but which are complexes of responses appropriate to a variety of environmental conditions. Fear and anger, sorrow and joy, are attitudes towards the environment, and particularly towards fellow members of the same species, which may become attached to, and then regularly evoked by, a great many different classes of stimuli.

4.65. This means that a great variety of external events, and also some conditions of the organism itself, may evoke one of several patterns of attitudes or dispositions which, while they last, will affect or ‘colour’ the perception of, and the responses to, any external event. In the mental order of events, that is in the influence which external stimuli can exercise on further mental processes and on behaviour, these states will occupy positions which in many ways will be similar to those of the sensory qualities: the occurrence of any one of them will be capable of modifying the result of a given sensory situation in the same kind of way in which the appearance of a new sensory experience could do so.

4.66. Emotions may thus be described as ‘affective qualities’ similar to the sensory qualities and forming part of the same comprehensive order of mental qualities. But they differ, of course, in some respects from the sensory qualities and must be regarded as forming a distinct sub-system of the more comprehensive mental order. The relation between the order of affective qualities and the order of sensory qualities must be conceived as somewhat similar to the relations between the orders of the different sensory modalities which also form sub-systems of the more comprehensive order of all sensory qualities (3.75).

4.67. The most conspicuous difference between the order of the sensory qualities and the order of the affective qualities is that, while the former is organized with spatial relationships as one of its main ordering principles, the affective qualities do not refer to particular points in space. They represent not qualities of particular things but rather a condition of an interval of time as a whole. They will refer not to what is to be expected of an external position but are rather a temporary bias or preference for certain types of responses towards any external situation.

4.68. These important differences between sensory and affective qualities, however, do not alter the fact that the general principle by which they are determined is the same. The similarity of the response to different stimuli will in both instances be determined by the fact that the corresponding different nervous impulses will evoke the same following of other impulses. Similar emotions, as similar sensations, are nervous impulses which evoke the same following and which are therefore functionally equivalent and classified as the same kind of event. The main difference is that within the sensory sub-system of the mental order the classifying connexions will be mainly with other impulses representing sensory stimuli, while in the affective sub-system the classifying connexions will be mainly with impulses representing certain types of behaviour.

4.69. But although the order of affective qualities will constitute a sub-system in the more comprehensive system of qualities (in the sense that the impulses belonging to it will be less closely connected with impulses in other parts of the larger system than they are connected among themselves), this does not prevent this subsystem from contributing to the differences between sensory qualities, and vice versa. By becoming connected with sensory qualities those differently organized qualities can add, as it were, an additional dimension to the order of sensory qualities; and similarly the differences between the different sensory qualities associated to different groups of the latter may assist in enriching the variety of differentiations between the former.

4.70. This account of the determination of the affective qualities of course corresponds very closely to the familiar James-Lange theory of emotions As we said before of the Berkeleyan theory of spatial vision (3.40–3.42), the James-Lange theory also may be regarded as a special case of the theory of mental qualities here outlined. The modifications which are required to make the James-Lange theory fit into our scheme are practically the same as those which we had to make with regard to the rôle which proprioceptive impulses play in determining the perception of space. We shall not regard the actual sensations produced by the various bodily accompaniments of a given stimulus as determining its affective values, but merely the following of physiological impulses which record the states of the body and which, in the same manner in which such a following can determine the peculiar functional significance which we know as sensory qualities, can also determine affective qualities.

4.71. We do, therefore, not propose to say, with William James, that emotions are ‘a set of kinesthetic and organic sensations’. We shall merely contend that the connexions with impulses recording certain connected sets of changes in the general state of the body can give certain central impulses that peculiar position in the whole system of mental events which we know as the different affective qualities.

4.72. The James-Lange theory of emotions (like Berkeley’s theory of spatial vision) would thus appear to be justified in its endeavour to reduce the qualitative attributes of those mental events to relations between different impulses which, if fully evaluated, might evoke certain other sensations. Both theories, however, fall short of a real answer to their problem, and in fact merely shift the problem to a different point, because they attempt to explain the quality of one kind of experience by reference to qualities occurring in another kind of experience, which latter they take as not requiring explanation. In so far as they were concerned with only that one kind of mental quality this procedure was inevitable. But if we consistently follow up and generalize the principle underlying those theories, there remain of course no given mental qualities; we are forced to replace the whole system of qualities by a system of relations between initially undifferentiated elements which can be conceived to be isomorphic with the system of qualities which we have to explain.