Classics in the History of Psychology

An internet resource developed by
Christopher D. Green
York University, Toronto, Ontario

ISSN 1492-3173

(Return to index)


By William H.R. Rivers (1920)

Posted March 2000



It is not long since it was regarded as a sufficient definition of instinct that it is the mode of mental activity proper to animals as distinguished from the intelligence which was believed to be the chief, or even the only, factor of any importance in regulating Man's behaviour. All recent work in psychology 'has shown this distinction to be of little value. On the one hand, it has been found that the behaviour of animals, even such animals as the insects which are regarded as pre-eminent patterns of the instinctive, shows many features, such as adaptability to unusual conditions, which can only be explained by qualities of the same order as those belonging to intelligence.[1] Exact observation on animals has shown that their reactions to their surroundings have not the rigid and mechanical character which was once ascribed to them. Not only do failures occur in the adjustment of action to circumstance, but when these failures occur, or when the conditions are such as would lead to failure if the reactions took their ordinary form, animal behaviour has been found to be capable of modification. On the other hand, we have learnt that the behaviour of man is far less subject to reason and intelligence than was once supposed, and that his reactions to circumstance are often with difficulty to be distinguished from the behaviour of the unreasoning brutes. This absence or deficiency of reason is especially pronounced in those social reactions in which individual differences dictated by reason sink into insignificance before the mass-reactions of the crowd. We are learning, that the behaviour [p. 41] animals does not differ from that of Man in kind, but rather in the relative degree and importance of the different modes of reaction of which the behaviour consists.

A second way of distinguishing between instinct and intelligence is psychologically even less valid than the last. In the higher vertebrates, i.e., in those which have developed a cerebral cortex or neo-pallium as part of their central nervous system, instinct is regarded as the product of sub-cortical activity, while intelligence is held to depend on the activity of the cortex or neo-palium. It is an instructive commentary on the difficulties presented by current definitions of instinct that In the last resort even so psychological a writer as Lloyd Morgan is repeatedly driven to employ this anatomical distinction in his work on instinct and intelligence, thus virtually giving up the attempt to make a psychological distinction between the two.[2]

A third and most important distinction which has been made between instinct and intelligence is that the former is innate and the latter acquired. If an animal or man behaves in a way which is quite independent of any experience it can have acquired in its individual existence, the behaviour is regarded as purely instinctive. if, on the other hand, it were possible to say that the behaviour of an animal or man was wholly determined by the experience of the individual, we should regard the behaviour as an example of pure intelligence. Since, however, it is impossible to exclude innate factors, all that we can do is to recognise as intelligent those components of behaviour which can be ascribed to individual experience.

This difference between instinct and intelligence is one of great value and probably furnishes the best theoretical distinction between the two kinds of behaviour, but when we endeavour to use the theoretical difference as a guide in practice and research, we are met by several difficulties. The distinction is one which is difficult to utilise in practice, for, as soon as an animal has acquired experience of any kind, it becomes a matter of the greatest difficulty to distinguish between the innate and the acquired conditions, while, as already pointed out, in all [p. 42] examples of intelligent behaviour, it is impossible to exclude innate factors. Often, as in the case of insects and other animals which carry out actions of a most complicated kind, wholly independent of individual experience, the distinction is valid and useful. Thus, the butterfly which lays its eggs on a special kind of plant in the absence of any experience derived from the observation of this action by others of its species may be regarded as a typical example of instinct. An especially striking and often quoted example of this kind is that of the yucca-moth (Pronuba yucadella) which, preparatory to laying its eggs in the ovary of the yucca plant, cuts open the pistil and stuffs into it the pollen from another plant, so that at one stroke it both fertilises, and ensures the persistence of, the plant which is essential for the future welfare of its progeny. A still better example is given by the behaviour of the grub of the Capricorn beetle (Cerambyx miles). After a larval life spent wholly in the channels within a tree-trunk which it itself manufactures, this creature, little more than a piece of crawling intestine, as Fabre says, makes elaborate preparations to ensure that after the pupal state it shall escape from the woody prison in which it has itself been for all its life immured.[3] I call this insect a better example than the butterfly or the moth because it is quite impossible that the behaviour of the grub can have been in any way influenced by the imitation of its kind.

The exclusion of individual experience which is possible or even easy in the insect is beset with the greatest difficulties in the case of the higher animals. These difficulties become especially great in those animals, of which Man is the best example, which are born in a state of great-immaturity. The years spent by the child in acquiring experience, which it is impossible to record with any degree of accuracy, make it peculiarly difficult to analyse human behaviour into its innate and acquired components.

One other point about this mode of distinguishing instinct and intelligence may be mentioned. The distinction belongs to the field of biology rather than of psychology. If we were able [p. 43] to analyse every case of behaviour, whether human or animal, into its innate and acquired elements, we should still be little, if at all, nearer the solution of the psychological as opposed to the biological problem. We should not yet have begun to understand the place of consciousness in relation to behaviour, which, whatever may be our interest in the unconscious, must still remain the special task of psychology.

So long as we are considering the subject biologically we may be content with distinctions which depend on whether behaviour is exhibited by Man or animal, whether it is dependent on, or independent of, acquired experience, and on the locality of the physiological processes with which the behaviour is correlated. These modes of distinction, however, will not, or should not, satisfy the psychologist who requires something in the nature of the behaviour itself by means of which he may distinguish the instinctive from the intelligent. Nevertheless in the present state of the subject I believe we shall do best to take as the distinguishing mark of instinct its innate character, even though this character be biological rather than psychological. We shall do best if we devote our inquiries to the attempt to distinguish different kinds of instinct according to their psychological character. It should be our task to analyse the general group of instincts into its component parts just as it has been the main task of psychologists hitherto to analyse the different forms of intelligent behaviour.

In seeking for a criterion by which to distinguish different varieties of instinct, I propose to turn away for a time from the behaviour of insects or other invertebrate animals which are usually taken as our patterns of the instinctive. These animals differ so enormously from ourselves that it is too great an adventure into the unknown to base any distinction on differences between their behaviour and ours. Let us look rather to the behaviour of Man as compared with the animals to which he is more nearly related, and to the behaviour of adult man as compared with the infant, for our clue to the nature of the differences which will enable us to distinguish different classes of instinctive behaviour. [p. 44]

I will begin with a difference taken from the comparison of the human adult with the infant and the animal. An animal or child exposed to danger, which is so recognised as danger that it produces a reaction, tends to give itself to the reaction fully. If it runs away, it tends to run with every particle of the energy which it is capable of putting forth; if it cries, screams, or utters other sound, it tends to do so with all the vigour at its command. In these cases there is no discrimination of the degree of danger. The reaction by flight or cry is the same whether the danger be great or small. In the case of the animal the movement of a shadow thrown by a falling leaf may produce as strong a reaction as the full sight of its deadliest enemy. The child may scream as vigorously after some trivial touch as it does with the pain of a cut or burn. With no discrimination of the degee of danger, there may be complete absence of graduation of the reaction to the nature of the stimulus which occurs even in the animal in its more intelligent behaviour, and is characteristic of the behaviour of the adult man when danger threatens. If the danger is sufficiently great, or if certain lines of behaviour by which the danger would normally be met are frustrated, even the adult man will fail to discriminate the nature of the danger and to graduate his movements accordingly. He will devote every particle of his energy to flight or other form of primitive or instinctive behaviour. Thus, if he becomes angry and assumes an aggressive attitude, his anger and aggression will go far beyond those called for by the needs of the situation. If he flees, his flight may continue long after it has removed him to a safe distance from the source of danger.

In what I have just said I have spoken of the child as tending to scream and of the animal as tending to run away with all the force at their command, because I wish to make clear that the child or animal does not always behave in this thoroughgoing manner. All I wish to imply is that when these reactions take place in their most characteristic manner, they show a complete absence of proportionality between the behaviour and the conditions which call it forth. I assume that when the child and animal are so behaving, they are acting [p. 45] in a manner ill which they would act if their instinctive behaviour had not been modified by experience.

In the last chapter I have adopted the current view that such emotions as fear or anger, with the reactions characteristic of them, are expressions of instinct. When they occur in Man, these reactions are prominent, even the most prominent, elements in that part of his behaviour which can be ascribed to instinct. We have now seen that these reactions, when occurring in their most characteristic form, have the special feature that there is an absence of graduation according to the nature of the conditions by which the behaviour is produced. If they take place at all, they tend to occur in their full strength. This form of reaction is known in physiology as the "all-or-none" reaction,[4] and I propose to adopt this term for the special kind of behaviour I am now supposing to be characteristic of certain forms of instinct.

It may help us to understand this reaction if I give a brief account of its nature in physiology. For this purpose I will begin with the instance in which the principle was discovered by Keith Lucas and Adrian. I will not describe the somewhat complex experimental procedures which were needed to demonstrate the principle, and will give only the essential facts. When a weak electrical stimulus is applied to isolated nerve-fibre, and the impulse which in consequence travels along the nerve is measured, it is found that if the stimulus is weak there is no impulse at all, or more correctly, the electrical behaviour of the nerve gives no evidence of any impulse. If strength of the electrical stimulus is gradually increased, a point is reached when the nerve gives the response normally associated with an impulse passing along its length. If now strength of the stimulus is increased, there is no corresponding increase in the response, and this remains so, however great the increase of the stimulus. If the isolated nerve-fibre is set action at all, it reads with its full strength and produces all effect of which it is capable. [p. 46]

In previous chapters I have cited the work of Head as giving good examples of the process of suppression, and protopathic sensibility as a characteristic example of the content of the unconscious. It will greatly strengthen my argument and help to show that I am dealing with a real character of certain forms of instinct if protopathic sensibility should be subject to the "all-or-none" principle. As a matter of fact this is practically, though not completely, the case. When a region of the skin which is endowed only with protopathic sensibility is stimulated with cold, the intensity of the cold sensation is roughly the same whether the temperature of the stimulating surface is zero or 20C., i.e., whether it is the temperature of ice or about the temperature of a summer day. The sensation due to the colder stimulus radiates over a larger area, which makes it difficult to be absolutely confident that there is no difference in the intensity of the sensation of cold, but we can be confident that when protopathic sensibility reacts to cold, it does so with appropriately or altogether the same strength so far as this can be tested by sensory experience,

The principle also holds good of certain forms of reflex action. Thus, the nature of the reflex known as the "extensor thrust" led Sherrington to think that the strength of the stimulus had no influence upon the amount of the response, and that the reflex occurred either not at all or fully.[5] The mass-reflex recently observed by Head and Riddoch, of which I gave an account in Chapter IV, also obeys the "all-or-none" principle fairly completely. Still more significant is the fact that the heart-muscle responds to stimulation either not at all or fully, this mode of reaction being of especial importance owing to the close relation between the heart and those affective disturbances which are closely connected with instinct.

Thus, the isolated nerve-fibre, the heart, certain forms of reflex action, and the protopathic sensibility of the skin all agree in having characters which only appear in the more complex behaviour of man or animal under conditions which bring instinctive processes into activity. [p. 47]

The "all-or-none" principle may be regarded as only a special case of a wider law holding good of the relation between stimulus and sensation, or between stimulus and reaction. Except at the limits of the range of intensities the normal sensibility of the skin or other sense-organs shows definite proportionality between stimulus and reaction, of which the most exact expression is given by Fechner's formula that the sensation is proportional to the logarithm of the stimulus. Any such exact relation is wholly absent in the case of protopathic sensibility, in the reactions of the "extensor-thrust" or the mass-reflex, and similarly, there is no such exact relation between the conditions setting an instinctive or emotional reaction into being and the strength of the reaction, at any rate in the child, or in the adult human being whose emotions have not been brought well under control by long training and practice.

The Fechner formula has been supposed to hold good of one affective state. It has been pointed out that the amount of pleasure derived from an accession of fortune stands in a definite relation to the fortune we already possess. A gift of half-a-crown will have a very different effect on a beggar and on a millionaire, and it has been supposed that this relation is subject to logarithmic expression; that equal increment of good fortune produce steadily decreasing increments of pleasure. Even if this law could be shown to apply with any degree of exactness, it conceals a highly-developed aspect of the affective life, one in which the crude emotional basis has been elaborated by the addition of highly complex intellectual factors. The states of pleasure and displeasure, at any rate in their more customary forms, are definitely graded. From the point of view here put forward, they must be regarded as states in which the crude emotional basis has undergone great development under the influence of individual experience. The nature of pleasure and displeasure, as well as the relation between what have been called physical and moral fortune, show a certain amount of definiteness of relation between stimulus and affect. This definite and even quantitative relation is not true of the cruder [p. 48] passions which I connect with the instinctive behaviour of the man, and still less is it true of the passions of the child. Here there is not even an approach to any exact proportionality between the fear or other emotion and the condition or conditions by which the emotion has been produced.

I have chosen the "all-or-none " principle and the absence of the relation expressed by Fechner's formula as my examples of the kinds of character by which we may distinguish different forms of instinctive behaviour, because they furnish differences which are capable of exactness of expression and even of measurement. Another character, common to emotive reactions, to protopathic sensibility and to the forms of reflex action I have considered, is their immediate, and as it were unreflective, character. It is characteristic of emotion that it flares up at once and leads immediately to the behaviour characteristic of it. When, on the other hand, the crude affective tendencies which I associate with instinct have been brought under control, and even brief reflexion becomes possible, the emotion will only come into being if the conditions tending to produce it have such force as to sweep before them with their flood the obstacles interposed by intelligence. Similarly, it is characteristic of the reactions of protopathic sensibility that they tend immediately to result in movements approaching in nature those of reflex action, and are quite beyond the control which we normally exert over our more reasoned movements. One of the first signs of the return of the later epicritic sensibility is that this urgency goes, so that stimulation is followed by movements which are adapted to the nature of the stimulus.

I propose, therefore, to adopt as the distinguishing marks of one class of instincts: firstly, the absence of exactness of discrimination, of appreciation and of graduation of response; secondly, the character of reacting to conditions with all the energy available; and thirdly, the immediate and uncontrolled character of the response. It is interesting to note that Head and Gordon Holmes have found these characters to hold good in large measure of the activity of the optic thalamus, the essential nucleus of which they have shown to be the central [p. 49] representative of the protopathic aspect of peripheral sensibility and the central basis of emotive reactions. As I have already pointed out, it is clear that in this case we have to do with a structure which has come down from an early stage of the development of the nervous system. The optic thalamus is now hidden sway in the interior of the brain, overlaid and buried by the vast development of the cerebral cortex. Just as I have supposed that emotive and instinctive reactions are buried within the unconscious, hidden from consciousness by the vast development of those reactions which are associated with intelligence, so do we find that the organ of the emotions and instinctive reactions has been buried under the overwhelming mass of the nervous structure we know to be pre-eminently associated with consciousness.

It is interesting to note that the line of argument which I have followed has brought us to the view of Lloyd Morgan that instinct is the product of subcortical activity, but with the very important difference that I regard such structures as the thalamus as the organs only of certain forms of instinct, and have attempted to distinguish these forms of instinct by means of definite characters of the mental processes involved, and of the behaviour by which the instinct becomes manifest.

It must be remembered that this attempt to mark off one kind of instinctive behaviour by its psychological character has been based almost entirely on the study of human behaviour. It is now necessary to consider briefly how far these distinctions apply to the behaviour of those animals we have come to regard as our patterns of the instinctive. It is quite clear that the characters which I have taken as the special marks of certain instinctive aspects of human behaviour do not apply to those actions which are universally regarded as characteristic forms of the instinctive behaviour of the insect. It is certain that the "all-or-none" principle does not hold good of the activity of the bee when constructing the cells of the honeycomb, nor even the cruder art of such an animal as the grub of the Capricorn beetle which I have cited as a typical example of innate behaviour. The actions of these animals, certainly those of the [p. 50] bee, require in large measure the fine discrimination and delicacy of adjustment which remind us of epicritic rather than of protopathic sensibility. The way in which an insect will often carry out a set of activities dependent on its inherited tendencies when the external conditions are different from the ordinary, thus depriving these activities of all value, may perhaps be regarded as a sign that the insect is subject in some measure to the working of the "all-or-none" principle, but this is something different from the nature of the reactions themselves.

If we were to take the characters I have considered as marks of certain forms of instinct, it is evident that the behaviour of the insect could not be thus explained, but that some other principle must be in action, giving to its behaviour the power of discrimination and graduation of response. The lines taken by the development which has conferred this power are probably widely different from those which have been followed in the case of the vertebrata. The vast difference between the nervous system of an insect and that of a vertebrate animal would lead us to expect a correspondingly wide difference in the nature of the controlling and graduating mechanisms of the two kinds of animal. If the views here put forward seem worthy of adoption as a working hypothesis by students of insect-behaviour, it will become their business to seek out the nature of the controlling and graduating mechanism by which the originally crude modes of response of the insect have been modified and regulated.

I have in this chapter attempted to show that it is possible to distinguish two kinds of instinctive behaviour according as they do or do not exhibit certain characters. The characters which I have used as a means of distinguishing the instincts which are especially obvious in the innate behaviour of Man resemble in many respects the characters of protopathic sensibility, of a person who is dependent on the activity of the thalamus or of the isolated spinal cord, and in the domain of pure physiology the characters of the isolated nerve and of the heart. The discriminative and graduated activity of the more elaborate instinct of the insect, and also of certain forms of innate behaviour in Man, resemble in its general nature the [p. 51] epicritic sensibility of the skin and the activities of the body generally when fully under the influence of the cerebral cortex. It will be convenient to have terms for these two different kinds of instinct and instinctive behaviour, and I propose that they shall be named after the two kinds of characters which, through the work of Head, can be recognised in cutaneous sensibility. I shall, therefore, in this book, speak of instinctive behaviour as protopathic or epicritic according as it is or is not subject to the "all-or-none" principle, and according as it is not or is capable of graduation in relation to the conditions which call it forth.


[1] See the series of papers on "Instinct and Intelligence," Brit. Joun. Psych., vol. iii. (1910), pp. 209-270.

[2] Lloyd Morgan, Instinct and Experience, London (1912).

[3] J. H. Fabre, The Wonders of Instinct, London (1918), p. 49.

[4] See E. D. Adrian, Journ. of Physiol., vol. xlv. (1912), p. 389; ibid., vol. xlvii. (1914), p. 460; Brain, vol, xli. (2918), p. 26.

[5] The Integrative Action of the Nervous System, London (1906),p. 74.