Show Summary Details
Page of

(p. 333) Aggression in Relation to Emotional Development 

(p. 333) Aggression in Relation to Emotional Development
(p. 333) Aggression in Relation to Emotional Development

Donald W. Winnicott

Page of

date: 14 December 2017

Originally published in Collected papers: Through paediatrics to psycho-analysis (pp. 204–218). London: Tavistock, 1958.
The first part of this paper was a contribution to a Symposium with Anna Freud, given to the Royal Society of Medicine, Psychiatry Section, 16 January 1950. The second section was given to a private group in January 1955, and the third section to a private group in November 1954. An early draft of this paper, dated December 1947 and probably written for a symposium at the Royal Society of Medicine (see letter to Anna Freud, [CW 3:3:5]), included several additional paragraphs, which have been reproduced in the footnotes.

I. Contribution to Symposium

The main idea behind this study of aggression is that if society is in danger, it is not because of man’s aggressiveness but because of the repression of personal aggressiveness in individuals.

In a study of the psychology of aggression a severe strain is imposed on the student, for the following reason. In a total psychology, being-stolen-from is the same as stealing, and is equally aggressive. Being weak is as aggressive as the attack of the strong on the weak. Murder and suicide are fundamentally the same thing. Perhaps most difficult of all, possession is as aggressive as is greedy acquisition; indeed acquisition and possession form a psychological unit, either is incomplete without the other. This is not saying that acquiring and possessing are good or bad.

These considerations are painful, because they draw attention to dissociations that are hidden in current social acceptance; they cannot be left out of a (p. 334) study of aggression. Also, the basis for a study of actual aggression must be a study of the roots of aggressive intention.

Prior to integration of the personality there is aggression.1 A baby kicks in the womb; it cannot be assumed that he is trying to kick his way out. A baby of a few weeks thrashes away with his arms; it cannot be assumed that he means to hit. A baby chews the nipple with his gums; it cannot be assumed that he is meaning to destroy or to hurt. At origin aggressiveness is almost synonymous with activity; it is a matter of part-function.

It is these part-functions that are organized by the child gradually, as he becomes a person, into aggression. In illness a patient may display activities and aggressiveness not fully meant. Integration of a personality does not arrive at a certain time on a certain day. It comes and goes, and even when well attained it can be lost through unfortunate environmental chance. Nevertheless, purposive behaviour is eventually arrived at if there is health. In so far as behaviour is purposive, aggression is meant. Here immediately comes the main source of aggression, instinctual experience. Aggression is part of the primitive expression of love. A description of this in oral terms is appropriate since I am studying the first love impulses.

Oral erotism gathers to itself aggressive components, and in health it is oral love that carries the basis of the greater part of actual aggressiveness that is, aggression intended by the individual and felt as such by the people around.

All experience is both physical and non-physical. Ideas accompany and enrich bodily function, and bodily functioning accompanies and realizes2 ideation. Also, of the sum of ideas and of memories it must be said that these gradually separate out into that which is available to consciousness, that which is available to consciousness only in certain circumstances, and that which is in the repressed unconscious, unavailable because of intolerable affect.

I am aware that I am mixing the theme of actual aggressiveness with that of aggressive impulse. I do feel, however, that the one cannot be studied without the other. No one act of aggression can be fully understood as an isolated phenomenon; and in fact the study of any one act of a child involves consideration of the following:

  • The child in his environment, with adults caring for him.

  • The child mature according to his chronological and emotional age.

  • The child who, although mature according to his age, contains within himself all degrees of immaturity reaching right back to the primary state.

  • The child as an ill person, having fixations at immature levels.

  • The child in a relatively unorganized emotional state, still liable with more or less ease to regression and to spontaneous recovery from regression.

(p. 335) Aggression at Various Stages

It would be helpful if we could start at the beginning of the individual’s life, but here there is much that is not known with certainty. A complete study would trace aggressiveness as it appears at the various stages of ego development:


  • Pre-integration

  • Purpose without concern


  • Integration

  • Purpose with concern

  • Guilt

Total personal

  • Inter-personal relationships

  • Triangular situations, etc.

  • Conflict, conscious and unconscious

What I attempt here is mainly a development of the second of these three themes, the intermediate.3


It is necessary to describe a theoretical stage of unconcern or ruthlessness in which the child can be said to exist as a person and to have purpose, yet to be unconcerned as to results. He does not yet appreciate the fact that what he destroys when excited is the same as that which he values in quiet intervals between excitements. His excited love includes an imaginative attack on the mother’s body. Here is aggression as a part of love.4,i

One can see some degree of this appearing as a dissociation between quiet and excited aspects of the personality, so that children who are ordinarily nice and lovable will ‘act out of character’ and do aggressive things to people they love, not feeling fully responsible for their actions.

If aggression is lost at this stage of emotional development there is also some degree of loss of capacity to love, that is to say, to make relationships with objects.

Stage of Concern

Now comes the stage described by Melanie Klein as the ‘depressive position’ in emotional development. For my purpose I will call this the Stage of (p. 336) Concern. The individual’s ego integration is sufficient for him to appreciate the personality of the mother figure, and this has the tremendously important result that he is concerned as to the results of his instinctual experience, physical and ideational.

The stage of concern brings with it the capacity to feel guilty. Henceforth some of the aggression appears clinically as grief or a feeling of guilt or some physical equivalent, such as vomiting. The guilt refers to the damage which is felt to be done to the loved person in the excited relationship. In health the infant can hold the guilt, and so with the help of a personal and live mother (who embodies a time factor) is able to discover his own personal urge to give and to construct and to mend. In this way much of the aggression is transformed into the social functions, and appears as such. In times of helplessness (as when no person can be found to accept a gift or to acknowledge effort to repair) this transformation breaks down, and aggression reappears. Social activity cannot be satisfactory except it be based on a feeling of personal guilt in respect of aggression.


In my description there now comes a place for anger at frustration. Frustration, which is inevitable in some degree in all experience, encourages the dichotomy: 1. innocent aggressive impulses towards frustrating objects, and 2. guilt-productive aggressive impulses towards good objects. Frustration acts as a seduction away from guilt and fosters a defence mechanism, namely, the direction of love and hate along separate lines. If this splitting of objects into good and bad5 takes place, there is an easing of guilt feeling; but in payment the love loses some of its valuable aggressive component, and the hate becomes the more disruptive.ii

Growth of Inner World

The psychology of the infant from now on becomes more complicated. The individual child becomes concerned not only with the effect on his mother of his impulses, but he also notes the results of his experiences in his own self. Instinctual satisfactions make him feel good, and he perceives intake and output in a psychological as well as in a physical sense. He becomes filled with what he feels to be good, and this initiates and maintains his confidence in himself and in what he feels he may expect from life. At the same time he has (p. 337) to reckon with his angry attacks, as a result of which he feels he becomes filled with what is bad or malign or persecuting. These evil things or forces, being inside him, as he feels, form a threat from within to his own person, and to the good which forms the basis of his trust in life.

He now starts a life-long task of management of his inner world, a task which, however, cannot be started until he is well lodged in his body and able to differentiate between what is inside himself and what is external, and between what is actual and what is his own fantasy. His management of the external world depends on his management of his inner world.

An extremely complex series of defence mechanisms develops, which should be examined in any attempt to understand aggression in a child who has reached this stage of emotional development. It will be impossible here to do more than enumerate some of the ways in which this part of human psychology is relevant to the present theme.

First I will describe the return from introversion, since this is an important and common source of actual aggression.

In health the child’s interest is directed both towards external reality and towards the inner world, and he has bridges between the one world and the other (dreams, play, etc.). In ill-health the child may re-arrange his relationships so that the good is concentrated within and the bad is projected. He now lives in his inner world. He may be said to have become introverted (or pathologically introverted).

A recovery from pathological introversion involves a new turning out into what is for such a child an external world full of persecutors, and at this point in his recovery the child regularly becomes aggressive. This is an important source of aggressive behaviour. If in a child’s recovery from introversion the attack-in-defence is mishandled by those in charge, the child easily slips back into introversion. Apart from illness, some degree of this state of affairs is met daily in the life of any small child, and the concept is by no means a purely theoretical one. An individual is in a sensitive state on coming round after a period of concentration on a personal task.

It must be remembered that in childhood we are watching the human being only gradually becoming able to distinguish between the subjective and objective. A state of what looks like delusional madness easily appears through the child’s projection of inner world experience. Even the healthy child of two or three commonly wakes in the night and feels he is in a world which (from our point of view) is his own inner world, not the external reality that we can share with him. In daytime small children become deluded in their play activities and, in fact, children can be found to be living chiefly in their inner world when apparently to us they are in our world. This need not be unhealthy, but in the management of such a child we cannot expect to meet with logic, which applies only to external or shared reality. A large proportion even of adults never achieve a reliable capacity for objectivity, and (p. 338) those who are most reliably objective are often comparatively out of touch with their own inner world’s richness.

Three other examples will be given of the way in which the child’s management of his own inner world explains aggressive behaviour.

In the child’s fantasy the inner world is localized primarily in the belly or secondarily in the head or some other specific bodily area.

A child who has reached a certain degree of personality organization meets with an experience such that it is beyond his power to deal with it by identification. For instance, his parents quarrel in front of him at a time when he is fully occupied over some other problem. He manages only by taking the whole experience into himself in order to master it. It can then be said that a fixed state of parents quarrelling is living inside him, and a quantity of energy is thenceforth directed towards the control of the internalized bad relationship. Clinically he becomes tired, or depressed, or physically ill. At certain times the internalized bad relationship takes over, and then the child behaves as if ‘possessed’ by the quarrelling parents. We see him as compulsively aggressive, nasty, unreasonable, deluded.6

Alternatively the child with introjected quarrelling parents periodically engineers quarrelling in the people around him, then using the real external badness as a projection of what was ‘bad’ within. In such a case, there may easily be times of madness with true hallucination of quarrelling voices or people.

In the child’s management of his inner world and in the attempt to preserve in it what is felt to be benign, there are moments when he feels that all would be well if a unit of malign influence could be eliminated. (This is equivalent to the scapegoat idea.)

Clinically there appears a dramatization of ejection of badness (kicking, passage of flatus, spitting, etc.). Alternatively the child is accident-prone, or there is a suicide attempt with the aim to destroy the bad within the self; in the total fantasy of the suicide there is to be a survival, with the bad elements destroyed, but survival may not occur.

The management of inner world phenomena, felt by the child to be in the belly (or head, etc.), from time to time presents so great a difficulty that the child puts on a comprehensive control with depressive mood as the clinical result. This leads to a state of inner deadness which is intolerable. The complementary state of mania is liable to occur. In this the inner world liveliness takes over and activates the child, who may clinically be violently aggressive, without obvious external stimulus for anger. These phases of mania are not the same as that which is called the manic defence, in which there is a denial of inner deadness by artificial activity (the so-called manic defence against depression, Klein). The clinical result of the manic defence is not an aggressive outburst, but a state of common anxious restlessness, hypomania, in which there is mild aggression in the form of untidiness, messiness, irritability with lack of constructive perseverance.

(p. 339) In health, the individual can store badness within for use in an attack on external forces that seem to threaten what is felt to be worth preserving. Aggression then has social value.

The value of this (as compared with maniacal or delusional aggression) lies in the fact that objectivity is preserved, and the enemy can be met with economy of effort. The enemy then does not need to be loved in order to be attacked.iii


The foregoing mainly describes the relation of aggression to what I have called the intermediate stage of emotional development. This stage precedes the total personal, with its interpersonal relationships and the triangular situations of the Oedipus complex, and it follows on after the early stages of ruthlessness, and of the era before purpose and before the integration of the personality.

Aggression that belongs to the stage that I have called total personal is already familiar to the present generation through the accepted work of Freud.

(p. 340) Important sources of aggression date from the very early stages of the development of the human being, and some of these will be traced in the next part of this chapter.

II. Very Early Roots of Aggression

In its simplest form the question that we ask is: does aggression come ultimately from anger aroused by frustration, or has it a root of its own?

The answer is necessarily highly complex unless a deliberate effort is made to cut through the great mass of clinical fact that goes to make up our daily analytic practice. If we do this, however, we run the risk of being accounted unaware of what we have in fact deliberately ignored.

We can say that in the primitive love impulse we shall always be able to detect reactive aggression, since in practice there is no such thing as a complete id satisfaction. Is it necessary therefore to attempt to dissect down? I think it is necessary because of the confusion that results from failure to do so. This is especially true in view of the fact that the primitive love impulse is operative at a stage when ego growth is only starting, when integration, for instance, is not an established fact. There is a primitive love that is operative when there is not yet a capacity for taking responsibility. In this era there is not even ruthlessness; it is a pre-ruth era, and if destruction be part of the aim in the id impulse, then destruction is only incidental to id satisfaction. Destruction only becomes an ego responsibility when there is ego integration and ego organization sufficient for the existence of anger, and therefore of fear of the talion. However early anger and fear can be detected, there is still room for recognition of those ego developments before which it is not sensible to talk of the individual’s anger.

Hate is relatively sophisticated and cannot be said to exist in these early stages. It is necessary therefore to examine aggression apart altogether from the reactive aggression that inevitably follows the id impulse because of failure of id experience due to the operation of the reality principle.

It is convenient then to say that the primitive love impulse (id) has a destructive quality, though it is not the infant’s aim to destroy since the impulse is experienced in the pre-ruth era.

From this assumption it is possible to go into the matter of the root of the destructive element in the primitive love (id) impulse.

To simplify matters, the variable factor of birth trauma can be left out, and a normal or non-traumatic birth can be taken for granted. By normal here I mean that the birth is felt by the infant to be the result of his own effort. Neither delay nor precipitation interfered with this (see Birth Memories, Birth Trauma, and Anxiety [CW 3:4:8]).

The early id experiences bring into play a new element for the baby, instinctual crises, characterized by a preparatory period, a climax, and a period (p. 341) following some degree of satisfaction. Each of these three phases brings its own problems for the infant.

Our task is to examine the pre-history of the aggressive element (destructive by chance) in the earliest id experience. We have at hand certain elements which date from at least as early as the onset of foetal movements namely motility. No doubt a corresponding element on the sensory side must eventually be added. Can this motility that dates from intra-uterine life, and that persists in infancy (and indeed throughout life), be linked up with the activity inherent in id experience proper? Indeed, is this activity to be classified as an id or an ego element? Or is it better to allow an undifferentiated ego-id phase (Hartmann, 1952) and to leave aside the attempt to classify motility on the ground that it appears before ego-id differentiation?

Each infant must be able to pour as much as possible of primitive motility into the id experiences. Here no doubt comes the truth of the need the infant has for the frustrations of reality since id satisfaction if it could be complete and without hindrance would leave the infant with that which is derived from the motility root unsatisfied (Riviere, 1936).

In the pattern of id experience that belongs to any one infant there is x per cent of primitive motility included in the id experience. There is then (100 − x) per cent left over for use in other ways and here indeed is a reason for the vast difference in the experience of various individuals in regard to their aggressiveness. Here also is the origin of one kind of masochism (see later).

It is profitable to examine the patterns that evolve round this matter of motility (Marty et Fain, 1955).

In one pattern, the environment is constantly discovered and rediscovered because of motility. Here each experience within the framework of primary narcissism emphasizes the fact that it is in the centre that the new individual is developing, and contact with environment is an experience of the individual (in its undifferentiated ego-id state, at first). In the second pattern the environment impinges on the foetus (or baby) and instead of a series of individual experiences there is a series of reactions to impingement. Here then develops a withdrawal to rest which alone allows of individual existence. Motility is then only experienced as a reaction to impingement.

In a third pattern, which is extreme, this is exaggerated to such a degree that there is not even a resting place for individual experience, and the result is a failure in the primary narcissistic state to evolve an individual. The ‘individual’ then develops as an extension of the shell rather than of the core, and as an extension of the impinging environment. What there is left of a core is hidden away and is difficult to find even in the most far-reaching analysis. The individual then exists by not being found. The true self is hidden, and what we have to deal with clinically is the complex false self whose function is to keep this true self hidden. The false self may be conveniently society-syntonic, but the lack of true self gives an instability (p. 342) which becomes more evident the more society is deceived into thinking that the false self is the true self. The patient’s complaint is of a sense of futility.

The first pattern is what we call healthy. It depends for its formation on good-enough mothering, with love expressed (as at first it can only be expressed) in physical terms. The mother holds the baby (in womb, or in arms) and through love (identification) knows how to adapt to ego needs. Under these conditions, and under these alone, the individual may start to exist, and starting to exist to have id experiences. The stage is set for the maximum of infusion of motility into id experiences. There is a fusion of the x per cent of motility potential with erotic potential (with x quantitatively high). Nevertheless even here there is (100 − x) per cent of motility potential left out of the pattern of fusion, and available for pure motility use.

It must be remembered that the fusion allows of experience apart from the action of opposition (reaction to frustration). That which is fused with the erotic potential is satisfied in instinctual gratification. By contrast, the (100 − x) per cent unfused motility potential needs to find opposition. Crudely, it needs something to push against, unless it is to remain unexperienced and a threat to well-being. In health, however, by definition, the individual can enjoy going around looking for appropriate opposition.

In the second and third patterns it is only through environmental impingement that the motility potential becomes a matter of experience. Here is ill-health. To a lesser or a greater degree, the individual must be opposed, and only if opposed does the individual tap the important motility source. This is satisfactory while environment consistently impinges, but:

  • Environmental impingement must continue.

  • Environmental impingement must have a pattern of its own, else chaos reigns since the individual cannot develop a personal pattern.

  • This means dependence, out of which the individual might not grow.

  • Withdrawal becomes an essential feature in the pattern. (Except in the extreme degree, with true self hidden; then even withdrawal is not available as a primitive defence.)

When the second and third patterns are operative there can be no health, and no treatment is of avail unless it changes the basic pattern in the direction of the pattern I have described first. Patients who have developed according to the second and third patterns do, however, come to analysis, and they may seem at first to be able to make especially good use of the analyst’s work done on the false assumption that the patient really exists.

Here is a special comment on the positive value of the neurotic patient’s resistances. The fact of these resistances, which can be analysed, gives a good prognosis. The absence of resistances leads to a diagnosis of disturbance in the early patterning of the kind I have described.

(p. 343) It would follow from these considerations that it is not possible to bring about a higher degree of fusion of motility and erotic potentials by analysis except in those who are normal by this method of classification. Where the first pattern is not established there cannot be a fusion except in a secondary way, through the ‘erotization’ of aggressive elements. Here is a root for compulsive sadistic trends, which can turn round into masochism. The individual feels real only when destructive and ruthless. He tries to bring about relationships through interplay with another individual by finding an erotic component to fuse with the aggression which is not in itself much more than pure motility. Here the erotic achieves fusion with motility, whereas in health it is more true to say that motility fuses with the erotic.

It is probable that in the perversions two kinds of masochism can be distinguished; one kind comes from a sadism which is an erotization of a crude motility urge, and the other kind is a more direct erotization of the passive of active motility; and it would appear that the development is directed one way or the other according to whether the first partner was masochistic or sadistic. The partnership produces a relationship which is valued the more because relationships were feeble when developed out of the erotic life, owing to relative lack of fusion of motility elements into the erotic life.

The sense of real comes especially from the motility (and corresponding sensory) roots, and erotic experiences with a weak infusion of the motility element do not strengthen the sense of reality or of existing. In fact such erotic experiences may be avoided precisely because they lead the subject to a sense of not existing, that is to say, in individuals whose early pattern is not of the variety that I have placed first in my description.

We are left with the conclusion that a great deal happens prior to the first feed, even if ego organization is immature. The summation of motility experiences contributes to the individual’s ability to start to exist, and out of primary identification to repudiate the shell and to become the core. The good-enough environment makes this development possible. Only if the early environment is good enough does it make sense for us to discuss the early psychology of the human infant, since, unless the environment has been good enough, the human being has not become differentiated, and has not come up as a subject for discussion in terms of normal psychology. Where the individual does exist, however, we may say that one main way in which the ego and id, now differentiated, maintain a relationship, and keep a relationship in spite of the difficulties that belong to the operation of the reality principle, is through the fusion of a high proportion of primary motility potential in with erotic potential.

From these there follow other ideas that concern the problem of the external nature of objects. This subject is discussed in the third part of this chapter.

(p. 344) III. The External Nature of Objects

In psycho-analytic practice, when an analysis has gone a long way, the analyst gets a privileged view of the early phenomena of emotional growth.

I have recently been struck by the following idea, derived from clinical work, that when a patient is engaged in discovering the aggressive root the analyst is more exhausted by the process, one way or another, than when the patient is discovering the erotic root of the instinctual life.

Immediately it will be observed that the material that concerns me here is that which is associated in our minds with the word ‘de-fusion’. We assume a fusion of aggressive and erotic components in health, but we do not always give proper significance to the pre-fusion era, and to the task of fusion. We may easily take fusion too much for granted, and in this way we get into futile arguments as soon as we leave the consideration of an actual case.

It must be conceded that the task of fusion is a severe one, that even in health it is an uncompleted task, and that it is very common to find large quantities of unfused aggression complicating the psychopathology of an individual who is being analysed.

In analysis, if this be true, we have to deal with separate expressions of the aggressive and erotic components, and to hold each separately for the patient who, in the transference, cannot achieve a fusion of the two. In severe disorders that involve failure at the point of fusion, we find the patient’s relationship to the analyst aggressive and erotic in turn. And it is here that I am claiming that the analyst is more likely to be tired by the former than by the latter type of partial relationship.

The immediate conclusion to be drawn from this observation is that in the early stages, when the Me and the Not-Me are being established, it is the aggressive component that more surely drives the individual to a need for a Not-Me or an object that is felt to be external. The erotic experiences can be completed while the object is subjectively conceived or personally created, or while the individual is near to the narcissistic state of primary identification of earlier date.

The erotic experiences can be completed by anything that brings relief to the erotic instinctual drive, and that allows of forepleasure, rising tension of general and local excitement, climax and detumescence or its equivalent, followed by a period of lack of desire (which may itself produce anxiety because of the temporary annihilation of the subjective object created through desire). On the other hand, the aggressive impulses do not give any satisfactory experience unless there is opposition. The opposition must come from the environment, from the Not-Me which gradually comes to be distinguished from the Me. Erotic experience can be said to exist in the muscles and other tissues involved in effort, but this erotism is of a different order from that of the instinctual erotism associated with specific erotogenic zones.

(p. 345) Patients let us know that the aggressive experiences (more or less de-fused) feel real, much more real than do the erotic experiences (also de-fused). Both are real, but the former carry a feeling of real, which is greatly valued. The fusion of the aggression along with the erotic component of an experience enhances the feeling of the reality of the experience.

It is true that to some extent aggressive impulses can find their opposition without external opposition; this is displayed normally in the fish-movements of the spinal column that date from prenatal life, and abnormally in the to-and-fro (futile) movements of ill infants (rocking, or tension denoting a magical internal and invisible to-and-fro movement). In spite of those considerations can one not say that in normal development opposition from outside brings along the development of the aggressive impulse?

In normal birth the opposition encountered provides a type of experience which gives effort a head-first quality. Although birth is often not normal, so that it becomes a vast complication, and although birth may take place by breech instead of by head, there seems to be a general validity in the association between pure effort and a head-first relationship to opposition. This could be tested out by observation on infants who are making an effort to feed according to my theory they can be helped by a degree of opposition to the top of the head.

This idea is usually expressed in the following terms: ‘An infant does not thrive on perfect adaptation to need. A mother who fits in with a baby’s desires too well is not a good mother. Frustration produces anger and this helps the infant to gain enhanced experience’. This is true and not true. In so far as it is untrue, it neglects two factors one is that the infant does need perfect adaptation at the theoretical start, and then needs a carefully graduated failure of adaptation; the other is that this statement leaves out of consideration the lack of fusion of the aggressive and the erotic roots of experience, whereas in theory at least, the de-fused state (or the fore-fusion state) must be studied.

Those who make the statement more or less as quoted here only too easily assume that aggression is a reaction to frustration, that is to say, to frustration during erotic experience, during a phase of excitement with instinctual tension rising. That there is anger at frustration in such phases is only too obvious, but in our theory of the earliest feelings and states we need to be prepared for aggression that precedes the ego integration that makes anger at instinctual frustration possible, and that makes the erotic experience an experience.

It can be said that each baby has a potential of zonal erotic instinct, that this is biological, and that the potential is more or less the same for each baby. By contrast the aggression component must be extremely variable; by the time that we observe a baby’s anger at frustration at a feeding delay a great deal has happened that has made the baby’s aggressive potential great or little. To get to something in terms of aggression corresponding to the erotic potential (p. 346) it would be necessary to go back to the impulses of the foetus, to that which makes for movement rather than for stillness, to the aliveness of tissues and to the first evidence of muscular erotism. We need a term here such as life force.

No doubt the life-force potential of each individual foetus is more or less the same, just as is the erotic potential of each baby. The complication is that the amount of aggressive potential an infant carries depends on the amount of opposition that has been met with. In other words, opposition affects the conversion of life force into aggression potential. Moreover, excess of opposition introduces complications that make it impossible for the existence of an individual who, having aggressive potential, could achieve its fusion with the erotic.

It is not possible to go further with this argument without considering in detail the fate of the life force of the (prenatal) infant.

In health the foetal impulses bring about a discovery of environment, this latter being the opposition that is met through movement, and sensed during movement. The result here is an early recognition of a Not-Me world, and an early establishment of the Me. (It will be understood that in practice these things develop gradually, and repeatedly come and go, and are achieved and lost.)

In ill-health at this very early stage it is the environment that impinges, and the life force is taken up in reactions to impingement the result being the opposite to the early firm establishment of the Me. In the extreme there is very little experience of impulses except as reactions, and the Me is not established. Instead we find a development based on the experience of reaction to impingement, and there comes into existence an individual that we call false because the personal impulsiveness is missing. In this case there is no fusion of the aggressive and erotic components, since the Me is not established when erotic experiences occur. The infant does indeed live, because of being seduced into erotic experience; but separately from the erotic life, which never feels real, is a purely aggressive reactive life, dependent on the experience of opposition.

It has been necessary, in this description, to discuss two extremes in the attempt to lead the way to a description of the common state in which some degree of lack of fusion has been a feature. The personality comprises three parts: a true self, with Me and Not-Me clearly established, and with some fusion of the aggressive and erotic elements; a self that is easily seduced along lines of erotic experience, but with the result of a loss of sense of real; a self that is entirely and ruthlessly given over to aggression. This aggression is not even organized to destruction, but it has value to the individual because it brings a sense of real and a sense of relating, but it is only brought into being by active opposition, or (later) persecution. It has no root in personal impulse, motivated in ego spontaneity.

(p. 347) The individual may achieve a false fusion of the aggressive and erotic by converting this pure de-fused aggression into masochism, but for this to occur there must be a reliable persecutor, and the reliable persecutor is a sadistic lover. In this way masochism can be primary to sadism. However, in following the development of an emotionally healthy human being we see sadism as primary to masochism. In health sadism implies successful fusion, that which is absent in the conditions in which masochism develops straight out of the pattern of reactive aggression, unfused.

The main conclusion to be made out of these considerations is that confusion exists through our using the term aggression sometimes when we mean spontaneity. The impulsive gesture reaches out and becomes aggressive when opposition is reached. There is reality in this experience, and it very easily fuses into the erotic experiences that await the new-born infant. I am suggesting: it is this impulsiveness, and the aggression that develops out of it, that makes the infant need an external object, and not merely a satisfying object.

Many infants, however, have a massive aggressive potential that belongs to reaction to impingement, that becomes activated by persecution: in so far as this is true the infant welcomes persecution, and feels real in reacting to it. But this represents a false mode of development since the infant needs continued persecution. The quantity of this reactive potential is not dependent on biological factors (which determine motility and erotism) but is dependent on the chance of early environmental impingement, and therefore, often, on the mother’s psychiatric abnormalities, and the state of the mother’s emotional environment.

In adult and mature sexual intercourse, it is perhaps true that it is not the purely erotic satisfactions that need a specific object. It is the aggressive or destructive element in the fused impulse that fixes the object and determines the need that is felt for the partner’s actual presence, satisfaction, and survival.


1. I would now link this idea with that of motility (cf. Marty et Fain, 1955).

2. cf. Sechehaye’s term: ‘symbolic realization’.

3. In Part II of this chapter, I attempt to deal with the theme of aggression relative to the early stages of ego development.

4. This has been called ‘pre-ambivalent’, but this term avoids the issue of the integration of part-object and whole object, breast and mother who holds and cares.

Editorial Note i [1947:] In practice aggression becomes controlled by the child himself. Babies do not regularly bite the breast even when they have teeth, although it is quite certain they have it in them to do so. They have ideas. They can bite on substitute hard objects and conceive of and seek indestructible objects for reassurance. Also they begin to split their objects into good and bad, that is into those that they wish to protect and those they feel should be destroyed.

5. I should now say ‘idealized and bad’ instead or ‘good and bad’ (1957).

Editorial Note ii [1947:] It will be noted that the basis for theft has now been stated, namely the purposive aim of taking the good things from the mother’s body. That is to say, like other forms of aggression, theft is at origin a part of the love impulse. Every thief is unconsciously looking for the mother from whom he can claim the right to “steal” in the act of loving her.

6. This state of affairs is related to that which Anna Freud has termed ‘identification with the aggressor’ (1937). The work of Melanie Klein introduced us to the concept of the omnipotent control of inner phenomena as a defence.

Editorial Note iii [1947:] Aggression in the Relationship Between Mature Persons

Only after all this are we prepared to consider the more mature child whose instinctual drives have come to include the genital organization. But the story is the same, with the accent on genital rather than on oral experience. At the same time the child’s own idea of his personality becomes more capable of description in the familiar terms of behaviour enriched by phantasy, conscious and unconscious, instead of in less familiar terms of management of internal and external reality.

Now inter-personal relationships can be studied along with rivalries, jealousies and death wishes. But these matters are not so difficult to grasp in these times when the once revolutionary theories of Freud are generally accepted, at least as ideas.

It is but seldom that actual aggressive behaviour can be explained on a basis of mature emotional development with genital organization completed. The problem remains, however, would there be wars if all men and women were emotionally mature? Would adolescents, if mature for their age, need to fight to kill, or to possess? Would mature adults require their adolescent children to risk death in the initiation rite? Possibly. In considering this we must be prepared to consider the value of war, not only to study the causes of war in order to be able to end wars.

In practice a human being is only to some extent and in certain circumstances a being dominated by a simple genital organization, the rest of the psyche can never be ignored; and society itself which cannot marry or propagate, is almost entirely moulded on the pregenital organizations of individuals, and on the child’s conception of himself that I have tried to indicate by outline in this paper.

For social health, what is needed is the personal maturity of individuals, from which the rest follows. For bringing about this personal maturing of individuals environmental conditions of love and stability are absolutely essential.

in My theme is not exhausted. Two examples of important themes not dealt with, yet vital to an understanding of actual aggressivity:

  • A powerful background to social unrest is the resentment felt by children and adults when they come to recognize that in the early stages of their personal development they were dependent, doubly dependent because blissfully unaware of the fact, on their mother’s care. It is both men and women who develop this hate of the providers of essentials and the democratic idea is in part an attempt to get behind this memory of dependence.

  • in The hate of external reality itself on account of its disillusioning power.