In the theory of psycho-analysis we have no hesitation in assuming that the course taken by mental events is automatically regulated by the pleasure principle. We believe, that is to say, that the course of those events is invariably set in motion by an unpleasurable tension, and that it takes a direction such that its final outcome coincides with a lowering of that tension-that is, with an avoidance of unpleasure or a production of pleasure. In taking that course into account in our consideration of the mental processes which are the subject of our study, we are introducing an ‘economic’ point of view into our work; and if, in describing those processes, we try to estimate this ‘economic’ factor in addition to the ‘topographical’ and ‘dynamic’ ones, we shall, I think, be giving the most complete description of them of which we can at present conceive, and one which deserves to be distinguished by the term ‘metapsychological’.’

It is of no concern to us in this connection to enquire how far, with this hypothesis of the pleasure principle, we have approached or adopted any particular, historically established, philosophical system. We have arrived at these speculative assumptions in an attempt to describe and to account for the facts of daily observation in our field of study. Priority and originality are not among the aims that psycho-analytic work sets itself; and the impressions that underlie the hypothesis of the pleasure principle are so obvious that they can scarcely be overlooked. On the other hand we would readily express our gratitude to any philosophical or psychological theory which was able to inform us of the meaning of the feelings of pleasure and unpleasure which act so imperatively upon us. But on this point we are, alas, offered nothing to our purpose. This is the most obscure and inaccessible region of the mind, and, since we cannot avoid contact with it, the least rigid hypothesis, it seems to me, will be the best. We have decided to relate
${}^1$${}^1$^(1){ }^{1} [See Section IV of ‘The Unconscious’ (1915e).]
pleasure and unpleasure to the quantity of excitation that is present in the mind but is not in any way ‘bound’; ${}^1$${}^1$^(1){ }^{1} and to relate them in such a manner that unpleasure corresponds to an increase in the quantity of excitation and pleasure to a diminution. What we are implying by this is not a simple relation between the strength of the feelings of pleasure and unpleasure and the corresponding modifications in the quantity of excitation; least of all-in view of all we have been taught by psycho-physiology-are we suggesting any directly proportional ratio: the factor that determines the feeling is probably the amount of increase or diminution in the quantity of excitation in a given period of time. Experiment might possibly play a part here; but it is not advisable for us analysts to go into the problem further so long as our way is not pointed by quite definite observations. ${}^2$${}^2$^(2){ }^{2}

We cannot, however, remain indifferent to the discovery that an investigator of such penetration as G. T. Fechner held a view on the subject of pleasure and unpleasure which coincides in all essentials with the one that has been forced upon us by psycho-analytic work. Fechner’s statement is to be found contained in a small work, Einige Ideen zur Schöpfungs- und Entwicklungsgeschichte der Organismen, 1873 (Part XI, Supplement, 94), and reads as follows: 'In so far as conscious impulses always have some relation to pleasure or unpleasure, pleasure and unpleasure too can be regarded as having a psycho-physical relation to conditions of stability and instability. This provides a basis for a hypothesis into which I propose to enter in greater detail elsewhere. According to this hypothesis, every psychophysical motion rising above the threshold of consciousness is attended by pleasure in proportion as, beyond a certain limit, it approximates to complete stability, and is attended by unpleasure in proportion as, beyond a certain limit, it deviates from complete stability; while between the two limits, which may be described as qualitative thresholds of pleasure and

unpleasure, there is a certain margin of aesthetic indifference. . . .’ ${}^1$${}^1$^(1){ }^{1}

The facts which have caused us to believe in the dominance of the pleasure principle in mental life also find expression in the hypothesis that the mental apparatus endeavours to keep the quantity of excitation present in it as low as possible or at least to keep it constant. This latter hypothesis is only another way of stating the pleasure principle; for if the work of the mental apparatus is directed towards keeping the quantity of excitation low, then anything that is calculated to increase that quantity is bound to be felt as adverse to the functioning of the apparatus, that is as unpleasurable. The pleasure principle follows from the principle of constancy: actually the latter principle was inferred from the facts which forced us to adopt the pleasure principle. ${}^2$${}^2$^(2){ }^{2} Moreover, a more detailed discussion will show that the tendency which we thus attribute to the mental apparatus is subsumed as a special case under Fechner’s principle of the ‘tendency towards stability’, to which he has brought the feelings of pleasure and unpleasure into relation.

It must be pointed out, however, that strictly speaking it is incorrect to talk of the dominance of the pleasure principle over the course of mental processes. If such a dominance existed, the immense majority of our mental processes would have to be accompanied by pleasure or to lead to pleasure, whereas universal experience completely contradicts any such conclusion. The most that can be said, therefore, is that there exists in the mind a strong tendency towards the pleasure principle, but that that tendency is opposed by certain other forces or circumstances, so that the final outcome cannot always be in harmony
${}^1$${}^1$^(1){ }^{1} [Cf. ‘Project’, end of Section 8 of Part I.-‘Aesthetic’ is here used in the old sense of ‘relating to sensation or perception’.]
${}^2$${}^2$^(2){ }^{2} [The 'principle ‘of constancy’ dates back to the very beginning of Freud’s psychological studies. The first published discussion of it of any length was by Breuer (in semi-physiological terms) towards the end of Section 2(A) of his theoretical part of the Studies on Hysteria (Breuer and Freud, 1895). He there defines it as ‘the tendency to keep intracerebral excitation constant’. In the same passage he attributes this principle to Freud and there in fact exist one or two earlier very brief references to it by Freud himself, though these were not published until after his death. (See Freud, 1941a [1892] and Breucr and Freud, 1940 [1892].) The subject is also discussed at length at the beginning of Freud’s ‘Project’, under the name of ‘neuronic inertia’.]
with the tendency towards pleasure. We may compare what Fechner ( 1873,90 ) remarks on a similar point: 'Since however a tendency towards an aim does not imply that the aim is attained, and since in general the aim is attainable only by approximations. . . ’

If we turn now to the question of what circumstances are able to prevent the pleasure principle from being carried into effect, we find ourselves once more on secure and well-trodden ground and, in framing our answer, we have at our disposal a rich fund of analytic experience.

The first example of the pleasure principle being inhibited in this way is a familiar one which occurs with regularity. We know that the pleasure principle is proper to a primary method of working on the part of the mental apparatus, but that, from the point of view of the self-preservation of the organism among the difficulties of the external world, it is from the very outset inefficient and even highly dangerous. Under the influence of the ego’s instincts of self-preservation, the pleasure principle is replaced by the reality principle. ${}^1$${}^1$^(1){ }^{1} This latter principle does not abandon the intention of ultimately obtaining pleasure, but it nevertheless demands and carries into effect the postponement of satisfaction, the abandonment of a number of possibilities of gaining satisfaction and the temporary toleration of unpleasure as a step on the long indirect road to pleasure. The pleasure principle long persists, however, as the method of working employed by the sexual instincts, which are so hard to ‘educate’, and, starting from those instincts, or in the ego itself, it often succeeds in overcoming the reality principle, to the detriment of the organism as a whole.

There can be no doubt, however, that the replacement of the pleasure principle by the reality principle can only be made responsible for a small number, and by no means the most intense, of unpleasurable experiences. Another occasion of the release of unpleasure, which occurs with no less regularity, is to be found in the conflicts and dissensions that take place in the mental apparatus while the ego is passing through its development into more highly composite organizations. Almost all the energy with which the apparatus is filled arises from its innate instinctual impulses. But these are not all allowed to reach

the same phases of development. In the course of things it happens again and again that individual instincts or parts of instincts turn out to be incompatible in their aims or demands with the remaining ones, which are able to combine into the inclusive unity of the ego. The former are then split off from this unity by the process of repression, held back at lower levels of psychical development and cut off, to begin with, from the possibility of satisfaction. If they succeed subsequently, as can so easily happen with repressed sexual instincts, in struggling through, by roundabout paths, to a direct or to a substitutive satisfaction, that event, which would in other cases have been an opportunity for pleasure, is felt by the ego as unpleasure. As a consequence of the old conflict which ended in repression, a new breach has occurred in the pleasure principle at the very time when certain instincts were endeavouring, in accordance with the principle, to obtain fresh pleasure. The details of the process by which repression turns a possibility of pleasure into a source of unpleasure are not yet clearly understood or cannot be clearly represented; but there is no doubt that all neurotic unpleasure is of that kind-pleasure that cannot be felt as such. ${}^1$${}^1$^(1){ }^{1}

The two sources of unpleasure which I have just indicated are very far from covering the majority of our unpleasurable experiences. But as regards the remainder it can be asserted with some show of justification that their presence does not contradict the dominance of the pleasure principle. Most of the unpleasure that we experience is perceptual unpleasure. It may be perception of pressure by unsatisfied instincts; or it may be external perception which is either distressing in itself or which excites unpleasurable expectations in the mental apparatus-that is, which is recognized by it as a ‘danger’. The reaction to these instinctual demands and threats of danger, a reaction which constitutes the proper activity of the mental apparatus, can then be directed in a correct manner by the pleasure principle or the reality principle by which the former is modified. This does not seem to necessitate any far-reaching limitation of the pleasure principle. Nevertheless the investigation of the mental reaction to external danger is precisely in a position to produce new material and raise fresh questions bearing upon our present problem.
${}^1$${}^1$^(1){ }^{1} [Footrote added 1925:] No doubt the essential point is that pleasure and unpleasure, being conscious feelings, are attached to the ego.

A condrrion has long been known and described which occurs after severe mechanical concussions, railway disasters and other accidents involving a risk to life; it has been given the name of ‘traumatic neurosis’. The terrible war which has just ended gave rise to a great number of illnesses of this kind, but it at least put an end to the temptation to attribute the cause of the disorder to organic lesions of the nervous system brought about by mechanical force. ${}^1$${}^1$^(1){ }^{1} The symptomatic picture presented by traumatic neurosis approaches that of hysteria in the wealth of its similar motor symptoms, but surpasses it as a rule in its strongly marked signs of subjective ailment (in which it resembles hypochondria or melancholia) as well as in the evidence it gives of a far more comprehensive general enfeeblement and disturbance of the mental capacities. No complete explanation has yet been reached either of war neuroses or of the traumatic neuroses of peace. In the case of the war neuroses, the fact that the same symptoms sometimes came about without the intervention of any gross mechanical force seemed at once enlightening and bewildering. In the case of the ordinary traumatic neuroses two characteristics emerge prominently: first, that the chief weight in their causation seoms to rest upon the factor of surprise, of fright; and secondly, that a wound or injury inflicted simultaneously works as a rule against the development of a neurosis. ‘Fright’, ‘fear’ and ‘anxiety’ ${}^2$${}^2$^(2){ }^{2} are improperly used as synonymous expressions; they are in fact capable of clear distinction in their relation to danger. ‘Anxiety’ describes a particular state of expecting the danger or preparing for it, even though it may be an unknown one. ‘Fear’ requires a definite object of which to be afraid. ‘Fright’, however, is the name we give to the state a person gets into when he has run into danger without being prepared for it; it emphasizes the factor of surprise. I do not believe anxiety can produce a traumatic neuro-

sis. There is something about anxiety that protects its subject against fright and so against fright-neuroses. We shall return to this point later [p. 25 f.$]{}^1$$]{}^1$]^(1)]{ }^{1}

The study of dreams may be considered the most trustworthy method of investigating deep mental processes. Now dreams occurring in traumatic neuroses have the characteristic of repeatedly bringing the patient back into the situation of his accident, a situation from which he wakes up in another fright. This astonishes people far too little. They think the fact that the traumatic experience is constantly forcing itself upon the patient even in his sleep is a proof of the strength of that experience: the patient is, as one might say, fixated to his trauma. Fixations to the experience which started the illness have long been familiar to us in hysteria. Breuer and Freud declared in $1893{}^2$$1893{}^2$1893^(2)1893{ }^{2} that ‘hysterics suffer mainly from reminiscences’. In the war neuroses, too, observers like Ferenczi and Simmel have been able to explain certain motor symptoms by fixation to the moment at which the trauma occurred.

I am not aware, however, that patients suffering from traumatic neurosis are much occupied in their waking lives with memories of their accident. Perhaps they are more concerned with not thinking of it. Anyone who accepts it as something selfevident that their dreams should put them back at night into the situation that caused them to fall ill has misunderstood the nature of dreams. It would be more in harmony with their nature if they showed the patient pictures from his healthy past or of the cure for which he hopes. If we are not to be shaken in our belief in the wish-fulfilling tenor of dreams by the dreams of traumatic neurotics, we still have one resource open to us: we may argue that the function of dreaming, like so much else, is upset in this condition and diverted from its purposes, or we

may be driven to reflect on the mysterious masochistic trends of the ego.

At this point I propose to leave the dark and dismal subject of the traumatic neurosis and pass on to examine the method of working employed by the mental apparatus in one of its earliest normal activities-I mean in children’s play.

The different theories of children’s play have only recently been summarized and discussed from the psycho-analytic point of view by Pfeifer (1919), to whose paper I would refer my readers. These theories attempt to discover the motives which lead children to play, but they fail to bring into the foreground the economic motive, the consideration of the yield of pleasure involved. Without wishing to include the whole field covered by these phenomena, I have been able, through a chance opportunity which presented itself, to throw some light upon the first game played by a little boy of one and a half and invented by himself. It was more than a mere fleeting observation, for I lived under the same roof as the child and his parents for some weeks, and it was some time before I discovered the meaning of the puzzling activity which he constantly repeated.

The child was not at all precocious in his intellectual development. At the age of one and a half he could say only a few comprehensible words; he could also make use of a number of sounds which expressed a meaning intelligible to those around him. He was, however, on good terms with his parents and their one servant-girl, and tributes were paid to his being a ‘good boy’. He did not disturb his parents at night, he conscientiously obeyed orders not to touch certain things or go into certain rooms, and above all he never cried when his mother left him for a few hours. At the same time, he was greatly attached to his mother, who had not only fed him herself but had also looked after him without any outside help. This good little boy, however, had an occasional disturbing habit of taking any small objects he could get hold of and throwing them away from him into a corner, under the bed, and so on, so that hunting for his toys and picking them up was often quite a business. As he did this he gave vent to a loud, long-drawn-out ’ $\mathrm{O}-\mathrm{o}-\mathrm{O}-\mathrm{O}$$\mathrm{O}-\mathrm{o}-\mathrm{O}-\mathrm{O}$O-o-O-O\mathrm{O}-\mathrm{o}-\mathrm{O}-\mathrm{O} ', accompanied by an expression of interest and satisfaction. His mother

and the writer of the present account were agreed in thinking that this was not a mere interjection but represented the German word ‘fort’ [‘gone’]. I eventually realized that it was a game and that the only use he made of any of his toys was to play ‘gone’ with them. One day I made an observation which confirmed my view. The child had a wooden reel with a piece of string tied round it. It never occurred to him to pull it along the floor behind him, for instance, and play at its being a carriage. What he did was to hold the reel by the string and very skilfully throw it over the edge of his curtained cot, so that it disappeared into it, at the same time uttering his expressive ’ $0-0-0-0$$0-0-0-0$0-0-0-00-0-0-0 '. He then pulled the reel out of the cot again by the string and hailed its reappearance with a joyful ’ $da$$da$dad a ’ [‘there’]. This, then, was the complete game-disappearance and return. As a rule one only witnessed its first act, which was repeated untiringly as a game in itself, though there is no doubt that the greater pleasure was attached to the second act. ${}^1$${}^1$^(1){ }^{1}

The interpretation of the game then became obvious. It was related to the child’s great cultural achievement-the instinctual renunciation (that is, the renunciation of instinctual satisfaction) which he had made in allowing his mother to go away without protesting. He compensated himself for this, as it were, by himself staging the disappearance and return of the objects within his reach. It is of course a matter of indifference from the point of view of judging the effective nature of the game whether the child invented it himself or took it over on some outside suggestion. Our interest is directed to another point. The child cannot possibly have felt his mother’s departure as something agreeable or even indifferent. How then does his repetition of this distressing experience as a game fit in with the pleasure principle? It may perhaps be said in reply that her departure had to be enacted as a necessary preliminary to her joyful return, and that it was in the latter that lay the true purpose

of the game. But against this must be counted the observed fact that the first act, that of departure, was staged as a game in itself and far more frequently than the episode in its entirety, with its pleasurable ending.

No certain decision can be reached from the analysis of a single case like this. On an unprejudiced view one gets an impression that the child turned his experience into a game from another motive. At the outset he was in a passive situation-he was overpowered by the experience; but, by repeating it, unpleasurable though it was, as a game, he took on an active part. These efforts might be put down to an instinct for mastery that was acting independently of whether the memory was in itself pleasurable or not. But still another interpretation may be attempted. Throwing away the object so that it was ‘gone’ might satisfy an impulse of the child’s, which was suppressed in his actual life, to revenge himself on his mother for going away from him. In that case it would have a defiant meaning: ‘All right, then, go away! I don’t need you. I’m sending you away myself.’ A year later, the same boy whom I had observed at his first game used to take a toy, if he was angry with it, and throw it on the floor, exclaiming: ‘Go to the fwont!’ He had heard at that time that his absent father was ‘at the front’, and was far from regretting his absence; on the contrary he made it quite clear that he had no desire to be disturbed in his sole possession of his mother. ${}^1$${}^1$^(1){ }^{1} We know of other children who liked to express similar hostile impulses by throwing away objects instead of persons. ${}^2$${}^2$^(2){ }^{2} We are therefore left in doubt as to whether the impulse to work over in the mind some overpowering experience so as to make oneself master of it can find expression as a primary event, and independently of the pleasure principle. For, in the case we have been discussing, the child may, after all, only have been able to repeat his unpleasant experience in play because the repetition carried along with it a yield of pleasure of another sort but none the less a direct one.

Nor shall we be helped in our hesitation between these two views by further considering children’s play. It is clear that in

their play children repeat everything that has made a great impression on them in real life, and that in doing so they abreact the strength of the impression and, as one might put it, make themselves master of the situation. But on the other hand it is obvious that all their play is influenced by a wish that dominates them the whole time-the wish to be grown-up and to be able to do what grown-up people do. It can also be observed that the unpleasurable nature of an experience does not always unsuit it for play. If the doctor looks down a child’s throat or carries out some small operation on him, we may be quite sure that these frightening experiences will be the subject of the next game; but we must not in that connection cverlook the fact that there is a yield of pleasure from another source. As the child passes over from the passivity of the experience to the activity of the game, he hands on the disagreeable experience to one of his playmates and in this way revenges himself on a substitute.

Nevertheless, it emerges from this discussion that there is no need to assume the existence of a special imitative instinct in order to provide a motive for play. Finally, a reminder may be added that the artistic play and artistic imitation carried out by adults, which, unlike children’s, are aimed at an audience, do not spare the spectators (for instance, in tragedy) the most painful experiences and can yet be felt by them as highly enjoyable. ${}^1$${}^1$^(1){ }^{1} This is convincing proof that, even under the dominance of the pleasure principle, there are ways and means enough of making what is in itself unpleasurable into a subject to be recollected and worked over in the mind. The consideration of these cases and situations, which have a yield of pleasure as their final outcome, should be undertaken by some system of aesthetics with an economic approach to its subject-matter. They are of no use for our purposes, since they presuppose the existence and dominance of the pleasure principle; they give no evidence of the operation of tendencies beyond the pleasure principle, that is, of tendencies more primitive than it and independent of it.
1 [Freud had made a tentative study of this point in his posthumously published paper on ‘Psychopathic Characters on the Stage’ (1942a) which was probably written in 1905 or 1906.]

Twenty-five years of intense work have had as their result that the immediate aims of psycho-analytic technique are quite other to-day than they were at the outset. At first the analysing physician could do no more than discover the unconscious material that was concealed from the patient, put it together, and, at the right moment, communicate it to him. Psychoanalysis was then first and foremost an art of interpreting. Since this did not solve the therapeutic problem, a further aim quickly came in view: to oblige the patient to confirm the analyst’s construction from his own memory. In that endeavour the chief emphasis lay upon the patient’s resistances: the art consisted now in uncovering these as quickly as possible, in pointing them out to the patient and in inducing him by human influence-this was where suggestion operating as ‘transference’ played its part-to abandon his resistances.

But it became ever clearer that the aim which had been set up-the aim that what was unconscious should become con-scious-is not completely attainable by that method. The patient cannot remember the whole of what is repressed in him, and what he cannot remember may be precisely the essential part of it. Thus he acquires no sense of conviction of the correctness of the construction that has been communicated to him. He is obliged to repeat the repressed material as a contemporary experience instead of, as the physician would prefer to see, remembering it as something belonging to the past. ${}^1$${}^1$^(1){ }^{1} These reproductions, which emerge with such unwished-for exactitude, always have as their subject some portion of infantile sexual life-of the Oedipus complex, that is, and its derivatives; and they are invariably acted cut in the sphere of the transference, of the patient’s relation to the physician. When things have reached this stage, it may be said that the earlier neurosis has now been replaced by a fresh, ‘transference neurosis’. It has

been the physician’s endeavour to keep this transference neurosis within the narrowest limits: to force as much as possible into the channel of memory and to allow as little as possible to emerge as repetition. The ratio between what is remembered and what is reproduced varies from case to case. The physician cannot as a rule spare his patient this phase of the treatment. He must get him to re-experience some portion of his forgotten life, but must see to it, on the other hand, that the patient retains some degree of aloofness, which will enable him, in spite of everything, to recognize that what appears to be reality is in fact only a reflection of a forgotten past. If this can be successfully achieved, the patient’s sense of conviction is won, together with the therapeutic success that is dependent on it.

In order to make it easier to understand this ‘compulsion to repeat’, which emerges during the psycho-analytic treatment of neurotics, we must above all get rid of the mistaken notion that what we are dealing with in our struggle against resistances is resistance on the part of the unconscious. The unconsciousthat is to say, the ‘repressed’-offers no resistance whatever to the efforts of the treatment. Indeed, it itself has no other endeavour than to break through the pressure weighing down on it and force its way either to consciousness or to a discharge through some real action. Resistance during treatment arises from the same higher strata and systems of the mind which originally carried out repression. But the fact that, as we know from experience, the motives of the resistances, and indeed the resistances themselves, are unconscious at first during the treatment, is a hint to us that we should corfect a shortcoming in our terminology. We shall avoid a lack of clarity if we make our contrast not between the conscious and the unconscious but between the coherent ego ${}^1$${}^1$^(1){ }^{1} and the repressed. It is certain that much of the ego is itself unconscious, and notably what we may describe as its nucleus; ${}^2$${}^2$^(2){ }^{2} only a small part of it is covered by the term ‘preconscious’. ${}^3$${}^3$^(3){ }^{3} Having replaced a purely descriptive

terminology by one which is systematic or dynamic, we can say that the patient’s resistance arises from his ego, ${}^1$${}^1$^(1){ }^{1} and we then at once perceive that the compulsion to repeat must be ascribed to the unconscious repressed. It seems probable that the compulsion can only express itself after the work of treatment has gone half-way to meet it and has loosened the repression. ${}^2$${}^2$^(2){ }^{2}

There is no doubt that the resistance of the conscious and unconscious ego operates under the sway of the pleasure principle: it seeks to avoid the unpleasure which would be produced by the liberation of the repressed. Our efforts, on the other hand, are directed towards procuring the toleration of that unpleasure by an appeal to the reality principle. But how is the compulsion to repeat-the manifestation of the power of the repressed related to the pleasure principle? It is clear that the greater part of what is re-experienced under the compulsion to repeat must cause the ego unpleasure, since it brings to light activities of repressed instinctual impulses. That, however, is unpleasure of a kind we have already considered and does not contradict the pleasure principle: unpleasure for one system and simultaneously satisfaction for the other. ${}^3$${}^3$^(3){ }^{3} But we come now to a new and remarkable fact, namely that the compulsion to repeat also recalls from the past experiences which include no possibility of pleasure, and which can never, even long ago, have brought satisfaction even to instinctual impulses which have since been repressed.

The early efflorescence of infantile sexual life is doomed to extinction because its wishes are incompatible with reality and with the inadequate stage of development which the child has reached. That efflorescence comes to an end in the most distressing circumstances and to the accompaniment of the most painful feelings. Loss of love and failure leave behind them a permanent injury to self-regard in the form of a narcissistic scar, which in my opinion, as well as in Marcinowski’s (1918), contributes more than anything to the ‘sense of inferiority’

which is so common in neurotics. The child’s sexual researches, on which limits are imposed by his physical development, lead to no satisfactory conclusion; hence such later complaints as ‘I can’t accomplish anything; I can’t succeed in anything’. The tie of affection, which binds the child as a rule to the parent of the opposite sex, succumbs to disappointment, to a vain expectation of satisfaction or to jealousy over the birth of a new baby-unmistakable proof of the infidelity of the object of the child’s affections. His own attempt to make a baby himself, carried out with tragic seriousness, fails shamefully. The lessening amount of affection he receives, the increasing demands of education, hard words and an occasional punishment-these show him at last the full extent to which he has been scorned. These are a few typical and constantly recurring instances of the ways in which the love characteristic of the age of childhood is brought to a conclusion.

Patients repeat all of these unwanted situations and painful emotions in the transference and revive them with the greatest ingenuity. They seek to bring about the interruption of the treatment while it is still incomplete; they contrive once more to feel themselves scorned, to oblige the physician to speak severely to them and treat them coldly; they discover appropriate objects for their jealousy; instead of the passionately desired baby of their childhood, they produce a plan or a promise of some grand present-which turns out as a rule to be no less unreal. None of these things can have produced pleasure in the past, and it might be supposed that they would cause less unpleasure to-day if they emerged as memories or dreams instead of taking the form of fresh experiences. They are of course the activities of instincts intended to lead to satisfaction; but no lesson has been learnt from the old experience of these activities having led instead only to unpleasure. ${}^1$${}^1$^(1){ }^{1} In spite of that, they are repeated, under pressure of a compulsion.

What psycho-analysis reveals in the transference phenomena of neurotics can also be observed in the lives of some normal people. The impression they give is of being pursued by a malignant fate or possessed by some ‘daemonic’ power; but psychoanalysis has always taken the view that their fate is for the most part arranged by themselves and determined by early infantile influences. The compulsion which is here in evidence differs in
${}^1$${}^1$^(1){ }^{1} [This sentence was added in 1921.]
no way from the compulsion to repeat which we have found in neurotics, even though the people we are now considering have never shown any signs of dealing with a neurotic conflict by producing symptoms. Thus we have come across people all of whose human relationships have the same outcome: such as the benefactor who is abandoned in anger after a time by each of his protegés, however much they may otherwise differ from one another, and who thus seems doomed to taste all the bitterness of ingratitude; or the man whose friendships all end in betrayal by his friend; or the man who time after time in the course of his life raises someone else into a position of great private or public authority and then, after a certain interval, himself upsets that authority and replaces him by a new one; or, again, the lover each of whose love affairs with a woman passes through the same phases and reaches the same conclusion. This ‘perpetual recurrence of the same thing’ causes us no astonishment when it relates to active behaviour on the part of the person concerned and when we can discern in him an essential char-acter-trait which always remains the same and which is compelled to find expression in a repetition of the same experiences. We are much more impressed by cases where the subject appears to have a passive experience, over which he has no influence, but in which he meets with a repetition of the same fatality. There is the case, for instance, of the woman who married three successive husbands each of whom fell ill soon afterwards and had to be nursed by her on their death-beds. ${}^1$${}^1$^(1){ }^{1} The most moving poetic picture of a fate such as this is given by Tasso in his romantic epic Gerusalemme Liberata. Its hero, Tancred, unwittingly kills his beloved Clorinda in a duel while she is disguised in the armour of an enemy knight. After her burial he makes his way into a strange magic forest which strikes the Crusaders’ army with terror. He slashes with his sword at a tall tree; but blood streams from the cut and the voice of Clorinda, whose soul is imprisoned in the tree, is heard complaining that he has wounded his beloved once again.

If we take into account observations such as these, based upon behaviour in the transference and upon the life-histories of men and women, we shall find courage to assume that there really does exist in the mind a compulsion to repeat which overrides the pleasure principle. Now too we shall be inclined to ${}^1$${}^1$^(1){ }^{1} Cf. the apt remarks on this subject by C. G. Jung (1909).
relate to this compulsion the dreams which occur in traumatic neuroses and the impulse which leads children to play.

But it is to be noted that only in rare instances can we observe the pure effects of the compulsion to repeat, unsupported by other motives. In the case of children’s play we have already laid stress on the other ways in which the emergence of the compulsion may be interpreted; the compulsion to repeat and instinctual satisfaction which is immediately pleasurable seem to converge here into an intimate partnership. The phenomena of transference are obviously exploited by the resistance which the ego maintains in its pertinacious insistence upon repression; the compulsion to repeat, which the treatment tries to bring into its service is, as it were, drawn over by the ego to its side (clinging as the ego does to the pleasure principle). ${}^1$${}^1$^(1){ }^{1} A great deal of what might be described as the compulsion of destiny seems intelligible on a rational basis; so that we are under no necessity to call in a new and mysterious motive force to explain it.

The least dubious instance [of such a motive force] is perhaps that of traumatic dreams. But on maturer reflection we shall be forced to admit that even in the other instances the whole ground is not covered by the operation of the familiar motive forces. Enough is left unexplained to justify the hypothesis of a compulsion to repeat-something that seems more primitive, more elementary, more instinctual than the pleasure principle which it over-rides. But if a compulsion to repeat does operate in the mind, we should be glad to know something about it, to learn what function it corresponds to, under what conditions it can emerge and what its relation is to the pleasure principle -to which, after all, we have hitherto ascribed dominance over the course of the processes of excitation in mental life.
${}^1$${}^1$^(1){ }^{1} [Before 1923 the last clause read: ‘the compulsion to repeat is as it were called to its help by the ego, clinging as it does to the pleasure principle.’]

What follows is speculation, often far-fetched speculation, which the reader will consider or dismiss according to his individual predilection. It is further an attempt to follow out an idea consistently, out of curiosity to see where it will lead.

Psycho-analytic speculation takes as its point of departure the impression, derived from examining unconscious processes, that consciousness may be, not the most universal attribute of mental processes, but only a particular function of them. Speaking in metapsychological terms, it asserts that consciousness is a function of a particular system which it describes as Cs. ${}^1$${}^1$^(1){ }^{1} What consciousness yields consists essentially of perceptions of excitations coming from the external world and of feelings of pleasure and unpleasure which can only arise from within the mental apparatus; it is therefore possible to assign to the system Pcpt.-Cs. ${}^2$${}^2$^(2){ }^{2} a position in space. It must lie on the borderline between outside and inside; it must be turned towards the external world and must envelop the other psychical systems. It will be seen that there is nothing daringly new in these assumptions; we have merely adopted the views on localization held by cerebral anatomy, which locates the ‘seat’ of consciousness in the cerebral cortex-the outermost, enveloping layer of the central organ. Cerebral anatomy has no need to consider why, speaking anatomically, consciousness should be lodged on the surface of the brain instead of being safely housed somewhere in its inmost interior. Perhaps we shall be more successful in accounting for this situation in the case of our system Pcpt.-Cs.

Consciousness is not the only distinctive character which we ascribe to the processes in that system. On the basis of impressions derived from our psycho-analytic experience, we assume that all excitatory processes that occur in the other systems leave

permanent traces behind in them which form the foundation of memory. Such memory-traces, then, have nothing to do with the fact of becoming conscious; indeed they are often most powerful and most enduring when the process which left them behind was one which never entered consciousness. We find it hard to believe, however, that permanent traces of excitation such as these are also left in the system Pcpt.-Cs. If they remained constantly conscious, they would very soon set limits to the system’s aptitude for receiving fresh excitations. ${}^1$${}^1$^(1){ }^{1} If, on the other hand, they were unconscious, we should be faced with the problem of explaining the existence of unconscious processes in a system whose functioning was otherwise accompanied by the phenomenon of consciousness. We should, so to say, have altered nothing and gained nothing by our hypothesis relegating the process of becoming conscious to a special system. Though this consideration is not absolutely conclusive, it nevertheless leads us to suspect that becoming conscious and leaving behind a memory-trace are processes incompatible with each other within one and the same system. Thus we should be able to say that the excitatory process becomes conscious in the system Cs. but leaves no permanent trace behind there; but that the excitation is transmitted to the systems lying next within and that it is in them that its traces are left. I followed these same lines in the schematic picture which I included in the speculative section of my Interpretation of Dreams. ${}^2$${}^2$^(2){ }^{2} It must be borne in mind that little enough is known from other sources of the origin of consciousness; when, therefore, we lay down the proposition that consciousness arises instead of a memory-trace, the assertion deserves consideration, at all events on the ground of its being framed in fairly precise terms.

If this is so, then, the system Cs. is characterized by the peculiarity that in it (in contrast to what happens in the other psychical systems) excitatory processes do not leave behind any permanent change in its elements but expire, as it were, in the phenomenon of becoming conscious. An exception of this sort
${}^1$${}^1$^(1){ }^{1} What follows is based throughout on Breuer’s views in [the second section of his theoretical contribution to] Studies on Hysteria (Breuer and Freud, 1895). [Freud himself discussed the subject in The Interpretation of Dreams, Standard Ed., 5, 538 and it had previously been fully considered in his ‘Project’ of 1895 (1950a), Part I, Section 3. He returned to the topic later in his paper on the ‘Mystic Writing-Pad’ (1925a).]

2 [Standard Ed. 5. 538.1
to the general rule requires to be explained by some factor that applies exclusively to that one system. Such a factor, which is absent in the other systems, might well be the exposed situation of the system Cs., immediately abutting as it does on the external world.

Let us picture a living organism in its most simplified possible form as an undifferentiated vesicle of a substance that is susceptible to stimulation. Then the surface turned towards the external world will from its very situation be differentiated and will serve as an organ for receiving stimuli. Indeed embryology, in its capacity as a recapitulation of developmental history, actually shows us that the central nervous system originates from the ectoderm; the grey matter of the cortex remains a derivative of the primitive superficial layer of the organism and may have inherited some of its essential properties. It would be easy to suppose, then, that as a result of the ceaseless impact of external stimuli on the surface of the vesicle, its substance to a certain depth may have become permanently modified, so that excitatory processes run a different course in it from what they run in the deeper layers. A crust would thus be formed which would at last have been so thoroughly ‘baked through’ by stimulation that it would present the most favourable possible conditions for the reception of stimuli and become incapable of any further modification. In terms of the system Cs., this would mean that its elements could undergo no further permanent modification from the passage of excitation, because they had already been modified in the respect in question to the greatest possible extent: now, however, they would have become capable of giving rise to consciousness. Various ideas may be formed which cannot at present be verified as to the nature of this modification of the substance and of the excitatory process. It may be supposed that, in passing from one element to another, an excitation has to overcome a resistance, and that the diminution of resistance thus effected is what lays down a permanent trace of the excitation, that is, a facilitation. In the system Cs., then, resistance of this kind to passage from one element to another would no longer exist. ${}^1$${}^1$^(1){ }^{1} This picture can be brought into relation with Breuer’s distinction between quiescent (or bound) and mobile cathectic energy in the elements
${}^1$${}^1$^(1){ }^{1} [This passage is foreshadowed in the later half of Section 3 of Part I of the ‘Project’.]