Winnicott, D.W. (1949). Hate in the Counter-Transference. Int. J. Psycho-Anal., 30:69-74.
(1949). International Journal of Psycho-Analysis, 30:69-74
Hate in the Counter-Transference
D. W. Winnicott
In this paper I wish to examine one aspect of the whole subject of ambivalency, namely,
hate in the counter-transference. I believe that the task of the analyst (call him a research analyst) who undertakes the analysis of a psychotic is seriously weighted by this
phenomenon, and that analysis of psychotics becomes impossible unless the analyst's
own hate is extremely well sorted-out and conscious. This is tantamount to saying that an
analyst needs to be himself analysed, but it also asserts that the analysis of a psychotic is
irksome as compared with that of a neurotic, and inherently so.
Apart from psycho-analytic treatment, the management of a psychotic is bound to be irksome. From time to time23 I have made acutely critical remarks about the modern trends in psychiatry, with the too easy electric shocks and the too drastic leucotomies.
Because of these criticisms that I have expressed I would like to be foremost in recognition of the extreme difficulty inherent in the task of the psychiatrist, and of the mental nurse in particular. Insane patients must always be a heavy emotional burden on
those who care for them. One can forgive those who do this work if they do awful things. This does not mean, however, that we have to accept whatever is done by psychiatrists and neuro-surgeons as sound according to principles of science.
Therefore although what follows is about psycho-analysis, it really has value to the
psychiatrist, even to one whose work does not in any way take him into the analytic type of relationship to patients.
To help the general psychiatrist the psycho-analyst must not only study for him the
primitive stages of the emotional development of the ill individual, but also must study
the nature of the emotional burden which the psychiatrist bears in doing his work. What
we as analysts call the counter-transference needs to be understood by the psychiatrist too. However much he loves his patients he cannot avoid hating them, and fearing them,
and the better he knows this the less will hate and fear be the motive determining what he does to his patients.
STATEMENT OF THEME
One could classify counter-transference phenomena thus:
1. Abnormality in counter-transference feelings, and set relationships and identifications that are under repression in the analyst. The comment on this is that the analyst needs more analysis, and we believe this is less of an issue among psycho-analysts than among
psychotherapists in general.
2. The identifications and tendencies belonging to an analyst's personal experiences and
personal development which provide the positive setting for his analytic work and make
his work different in quality from that of any other analyst.
3. From these two I distinguish the truly
- 69 -
objective counter-transference, or if this is difficult, the analyst's love and hate in reaction
to the actual personality and behaviour of the patient, based on objective observation.
I suggest that if an analyst is to analyse psychotics or anti-socials he must be able to be so
thoroughly aware of the counter-transference that he can sort out and study his objective
reactions to the patient. These will include hate. Counter-transference phenomena will at
times be the important things in the analysis.
The Motive imputed to the Analyst by the Patient
I wish to suggest that the patient can only appreciate in the analyst what he himself is
capable of feeling. In the matter of motive; the obsessional will tend to be thinking of the
analyst as doing his work in a futile obsessional way. A hypo-manic patient who is
incapable of being depressed, except in a severe mood swing, and in whose emotional
development the depressive position has not been securely won, who cannot feel guilt in
a deep way, or a sense of concern or responsibility, is unable to see the analyst's work as an attempt on the part of the analyst to make reparation in respect of his own (the
analyst's) guilt feelings. A neurotic patient tends to see the analyst as ambivalent towards the patient, and to expect the analyst to show a splitting of love and hate; this patient,
when in luck, gets the love, because someone else is getting the analyst's hate. Would it
not follow that if a psychotic is in a 'coincident love-hate' state of feeling he experiences a deep conviction that the analyst is also only capable of the same crude and dangerous
state of coincident love-hate relationship? Should the analyst show love he will surely at
the same moment kill the patient.
This coincidence of love and hate is something that characteristically recurs in the analysis of psychotics, giving rise to problems of management which can easily take the
analyst beyond his resources. This coincidence of love and hate to which I am referring is something which is distinct from the aggressive component complicating the primitive
love impulse and implies that in the history of the patient there was an environmental
failure at the time of the first object-finding instinctual impulses.
If the analyst is going to have crude feelings imputed to him he is best forewarned and so
forearmed, for he must tolerate being placed in that position. Above all he must not deny
hate that really exists in himself. Hate that is justified in the present setting has to be sorted out and kept in storage and available for eventual interpretation.
If we are to become able to be the analysts of psychotic patients we must have reached
down to very primitive things in ourselves, and this is but another example of the fact that
the answer to many obscure problems of psycho-analytic practice lies in further analysis
of the analyst. (Psycho-analytic research is perhaps always to some extent an attempt on
the part of an analyst to carry the work of his own analysis further than the point to which his own analyst could get him.)
A main task of the analyst of any patient is to maintain objectivity in regard to all that the
patient brings, and a special case of this is the analyst's need to be able to hate the patient
Are there not many situations in our ordinary analytic work in which the analyst's hate is
justified? A patient of mine, a very bad obsessional, was almost loathsome to me for
some years. I felt bad about this until the analysis turned a corner and the patient became lovable, and then I realized that his unlikeableness had been an active symptom,
unconsciously determined. It was indeed a wonderful day for me (much later on) when I
could actually tell the patient that I and his friends had felt repelled by him, but that he
had been too ill for us to let him know. This was also an important day for him, a
tremendous advance in his adjustment to reality.
In the ordinary analysis the analyst has no difficulty with the management of his own hate. This hate remains latent. The main thing, of course, is that through his own analysis
he has become free from vast reservoirs of unconscious hate belonging to the past and to inner conflicts. There are other reasons why hate remains unexpressed and even unfelt as such:
1. Analysis is my chosen job, the way I feel I will best deal with my own guilt, the way I
can express myself in a constructive way.
2. I get paid, or I am in training to gain a place in society by psycho-analytic work.
3. I am discovering things.
4. I get immediate rewards through identification with the patient, who is making
progress, and I can see still greater
- 70 -
rewards some way ahead, after the end of the treatment.
5. Moreover, as an analyst I have ways of expressing hate. Hate is expressed by the
existence of the end of the 'hour'.
I think this is true even when there is no difficulty whatever, and when the patient is pleased to go. In many analyses these things can be taken for granted, so that they are scarcely mentioned, and the analytic work is done through verbal interpretations of the patient's emerging unconscious transference. The analyst takes over the role of one or
other of the helpful figures of the patient's childhood. He cashes in on the success of
those who did the dirty work when the patient was an infant.
These things are part of the description of ordinary psycho-analytic work, which is
mostly concerned with patients whose symptoms have a neurotic quality.
In the analysis of psychotics, however, quite a different type and degree of strain is take by the analyst, and it is precisely this different strain that I am trying to describe.
Illustration of Counter-Transference Anxiety
Recently for a period of a few days I found I was doing bad work. I made mistakes in respect of each one of my patients. The difficulty was in myself and it was partly personal but chiefly associated with a climax that I had reached in my relation to one
particular psychotic (research) patient. The difficulty cleared up when I had what is sometimes called a 'healing' dream. (Incidentally I would add that during my analysis and
in the years since the end of my analysis I have had a long series of these healing dreams
which, although in many cases unpleasant, have each one of them marked my arrival at a new stage in emotional development.)
On this particular occasion I was aware of the meaning of the dream as I woke or even
before I woke. The dream had two phases. In the first I was in the gods in a theatre and
looking down on the people a long way below in the stalls. I felt severe anxiety as if I
might lose a limb. This was associated with the feeling I have had at the top of the Eiffel Tower that if I put my hand over the edge it would fall off on to the ground below. This would be ordinary castration anxiety.
In the next phase of the dream I was aware that the people in the stalls were watching a play and I was now related to what was going on on the stage through them. A new kind
of anxiety now developed. What I knew was that I had no right side of my body at all.
This was not a castration dream. It was a sense of not having that part of the body.
As I woke I was aware of having understood at a very deep level what was my difficulty
at that particular time. The first part of the dream represented the ordinary anxieties that
might develop in respect of unconscious fantasies of my neurotic patients. I would be in
danger of losing my hand or my fingers if these patients should become interested in
them. With this kind of anxiety I was familiar, and it was comparatively tolerable.
The second part of the dream, however, referred to my relation to the psychotic patient.
This patient was requiring of me that I should have no relation to her body at all, not even
an imaginative one; there was no body that she recognized as hers and if she existed at all she could only feel herself to be a mind. Any reference to her body produced paranoid
anxieties because to claim that she had a body was to persecute her. What she needed of
me was that I should have only a mind speaking to her mind. At the culmination of my difficulties on the evening before the dream I had become irritated and had said that what
she was needing of me was little better than hair-splitting. This had had a disastrous
effect and it took many weeks for the analysis to recover from my lapse. The essential
thing, however, was that I should understand my own anxiety and this was represented in
the dream by the absence of the right side of my body when I tried to get into relation to
the play that the people in the stalls were watching. This right side of my body was the
side related to this particular patient and was therefore affected by her need to deny absolutely even an imaginative relationship of our bodies. This denial was producing in
me this psychotic type of anxiety, much less tolerable than ordinary castration anxiety.
Whatever other interpretations might be made in respect of this dream the result of my
having dreamed it and remembered it was that I was able to take up this analysis again
and even to heal the harm done to it by my irritability which had its origin in a reactive anxiety of a quality that was appropriate to my contact with a patient with no body.
- 71 -
Postponement of Interpretation
The analyst must be prepared to bear strain without expecting the patient to know anything about what he is doing, perhaps over a long period of time. To do this he must be easily aware of his own fear and hate. He is in the position of the mother of an infant
unborn or newly born. Eventually, he ought to be able to tell his patient what he has been through on the patient's behalf, but an analysis may never get as far as this. There may be
too little good experience in the patient's past to work on. What if there be no satisfactory relationship of early infancy for the analyst to exploit in the transference?
There is a vast difference between those patients who have had satisfactory early experiences which can be discovered in the transference, and those whose very early
experiences have been so deficient or distorted that the analyst has to be the first in the patient's life to supply certain environmental essentials. In the treatment of the patient of the latter kind all sorts of things in analytic technique become vitally important that can
be taken for granted in the treatment of patients of the former type.
I asked an analyst who confines his attention to neurotics whether he does analysis in the
dark, and he said, 'Why, no! Surely our job is to provide an ordinary environment, and
the dark would be extraordinary.' He was surprised at my question. He was orientated towards analysis of neurotics. But this provision and maintenance of an ordinary environment can be in itself a vitally important thing in the analysis of a psychotic, in fact
it can be, at times, even more important than the verbal interpretations which also have to be given. For the neurotic the couch and warmth and comfort can be symbolical of the
mother's love; for the psychotic it would be more true to say that these things are the
analyst's physical expression of love. The couch is the analyst's lap or womb, and the
warmth is the live warmth of the analyst's body. And so on.
Objective Hate Under Test
There is, I hope, a progression in my statement of my subject. The analyst's hate is
ordinarily latent and is easily kept latent. In analysis of psychotics the analyst is under
greater strain to keep his hate latent, and he can only do this by being thoroughly aware
of it. Now I want to add that in certain stages of certain analyses the analyst's hate is
actually sought by the patient, and what is then needed is hate that is objective. If the
patient seeks objective or justified hate he must be able to reach it, else he cannot feel he
can reach objective love.
It is perhaps relevant here to cite the case of the child of the broken home, or the child
without parents. Such a child spends his time unconsciously looking for his parents. It is notoriously inadequate to take such a child into one's home and to love him. What
happens is that after a while a child so adopted gains hope, and then he starts to test out the environment he has found, and to seek proof of his guardians' ability to hate
objectively. It seems that he can believe in being loved only after reaching being hated.
During the second world war a boy of nine came to a hostel for evacuated children, sent from London not because of bombs but because of truancy. I hoped to give him some treatment during his stay in the hostel, but his symptom won and he ran away as he had
always done from everywhere since the age of six when he first ran away from home. However, I had established contact with him in one interview in which I could see and interpret through a drawing of his that in running away he was unconsciously saving the inside of his home and preserving his mother from assault, as well as trying to get away
from his own inner world which was full of persecutors.
I was not very surprised when he turned up in the police station very near my home. This
was one of the few police stations that did not know him intimately. My wife very generously took him in and kept him for three months, three months of hell. He was the most lovable and most maddening of children, often stark staring mad. But fortunately
we knew what to expect. We dealt with the first phase by giving him complete freedom
and a shilling whenever he went out. He had only to ring up and we fetched him from
whatever police station had taken charge of him.
Soon the expected change-over occurred, the truancy symptom turned round, and the boy
started dramatizing the assault on the inside. It was really a whole-time job for the two of
us together, and when I was out the worst episodes took place.
Interpretation had to be made at any minute of day or night, and often the only solution in a crisis was to make the correct interpretation,
- 72 -
as if the boy were in analysis. It was the correct interpretation that he valued above
The important thing for the purpose of this paper is the way in which the evolution of the
boy's personality engendered hate in me, and what I did about it.
Did I hit him? The answer is no, I never hit. But I should have had to have done so if I had not known all about my hate and if I had not let him know about it too. At crises I would take him by bodily strength, and without anger or blame, and put him outside the front door, whatever the weather or the time of day or night. There was a special bell he could ring, and he knew that if he rang it he would be readmitted and no word said about the past. He used this bell as soon as he had recovered from his maniacal attack.
The important thing is that each time, just as I put him outside the door, I told him something; I said that what had happened had made me hate him. This was easy because
it was so true.
I think these words were important from the point of view of his progress, but they were mainly important in enabling me to tolerate the situation without letting out, without
losing my temper and every now and again murdering him.
This boy's full story cannot be told here. He went to an Approved School. His deeply rooted relation to us has remained one of the few stable things in his life. This episode from ordinary life can be used to illustrate the general topic of hate justified in the
present; this is to be distinguished from hate that is only justified in another setting but
which is tapped by some action of a patient (child).
A Mother's Love and Hate
Out of all the complexity of the problem of hate and its roots I want to rescue one thing,
because I believe it has an importance for the analyst of psychotic patients. I suggest that the mother hates the baby before the baby hates the mother, and before the baby can
know his mother hates him.
Before developing this theme I want to refer to Freud's remarks. In Instincts and their
Vicissitudes (1915) (where he says so much that is original and illuminating about hate),
Freud says: 'we might at a pinch say of an instinct that it "loves" the objects after which it strives for purposes of satisfaction, but to say that it "hates" an object strikes us as odd, so we become aware that the attitudes of love and hate cannot be said to characterize the
relation of instincts to their objects, but are reserved for the relations of the ego as a whole to objects. …' This I feel is true and important. Does this not mean that the personality must be integrated before an infant can be said to hate? However early
integration may be achieved—perhaps integration occurs earliest at the height of
excitement or rage—there is a theoretical earlier stage in which whatever the infant does
that hurts is not done in hate. I have used the word 'ruthless love' in describing this stage.
Is this acceptable? As the infant becomes able to feel a whole person, so does the word hate develop meaning as a description of a certain group of his feelings.
The mother, however, hates her infant from the word go. I believe Freud thought it possible that a mother may under certain circumstances have only love for her boy baby;
but we may doubt this. We know about a mother's love and we appreciate its reality and
power. Let me give some of the reasons why a mother hates her baby, even a boy.
A. The baby is not her own (mental) conception.
B. The baby is not the one of childhood play, father's child, brother's child, etc.
C. The baby is not magically produced.
D. The baby is a danger to her body in pregnancy and at birth.
E. The baby is an interference with her private life, a challenge to preoccupation.
F. To a greater or lesser extent a mother feels that her own mother demands a baby, so
that her baby is produced to placate her mother.
G. The baby hurts her nipples even by suckling, which is at first a chewing activity.
H. He is ruthless, treats her as scum, an unpaid servant, a slave.
I. She has to love him, excretions and all, at any rate at the beginning, till he has doubts about himself.
J. He tries to hurt her, periodically bites her, all in love.
K. He shows disillusionment about her.
L. His excited love is cupboard love, so that having got what he wants he throws her away like orange peel.
M. The baby at first must dominate, he must be protected from coincidences, life must
unfold at the baby's rate and all this needs his mother's continuous and detailed study. For
- 73 -
instance, she must not be anxious when holding him, etc.
N. At first he does not know at all what she does or what she sacrifices for him. Especially he cannot allow for her hate.
O. He is suspicious, refuses her good food, and makes her doubt herself, but eats well with his aunt.
P. After an awful morning with him she goes out, and he smiles at a stranger, who says:
'Isn't he sweet!'
Q. If she fails him at the start she knows he will pay her out for ever.
R. He excites her but frustrates—she mustn't eat him or trade in sex with him.
I think that in the analysis of psychotics, and in the ultimate stages of the analysis, even
of a normal person, the analyst must find himself in a position comparable to that of the
mother of a new-born baby. When deeply regressed the patient cannot identify with the analyst or appreciate his point of view any more than the foetus or newly born infant can
sympathize with the mother.
A mother has to be able to tolerate hating her baby without doing anything about it. She cannot express it to him. If, for fear of what she may do, she cannot hate appropriately
when hurt by her child she must fall back on masochism, and I think it is this that gives rise to the false theory of a natural masochism in women. The most remarkable thing
about a mother is her ability to be hurt so much by her baby and to hate so much without
paying the child out, and her ability to wait for rewards that may or may not come at a later date. Perhaps she is helped by some of the nursery rhymes she sings, which her baby enjoys but fortunately does not understand?
'Rockabye Baby, on the tree top,
When the wind blows the cradle will rock,
When the bough breaks the cradle will fall,
Down will come baby, cradle and all.'
I think of a mother (or father) playing with a small infant; the infant enjoying the play and not knowing that the parent is expressing hate in the words, perhaps in birth
symbolism. This is not a sentimental rhyme. Sentimentality is useless for parents, as it
contains a denial of hate, and sentimentality in a mother is no good at all from the infant's
point of view.
It seems to me doubtful whether a human child as he develops is capable of tolerating the full extent of his own hate in a sentimental environment. He needs hate to hate.
If this is true, a psychotic patient in analysis cannot be expected to tolerate his hate of the
analyst unless the analyst can hate him.
Practical Problem of Interpretation
If all this is accepted there remains for discussion the question of the interpretation of the
analyst's hate to the patient. This is obviously a matter fraught with danger, and it needs the most careful timing. But I believe an analysis is incomplete if even towards the end it has not been possible for the analyst to tell the patient what he, the analyst, did unbeknown for the patient whilst he was ill, in the early stages. Until the interpretation is
made the patient is kept to some extent in the position of infant, one who cannot
understand what he owes to his mother.
An analyst has to display all the patience and tolerance and reliability of a mother
devoted to her infant, has to recognize the patient's wishes as needs, has to put aside other
interests in order to be available and to be punctual, and objective, and has to seem to want to give what is really only given because of the patient's needs.
There may be a long initial period in which the analyst's point of view cannot be (even
unconsciously) appreciated by the patient. Acknowledgment cannot be expected because
at the primitive root of the patient that is being looked for there is no capacity for
identification with the analyst, and certainly the patient cannot see that the analyst's hate
is often engendered by the very things the patient does in his crude way of loving.
In the analysis (research analysis) or in ordinary management of the more psychotic type of patient, a great strain is put on the analyst (psychiatrist, mental nurse) and it is
important to study the ways in which anxiety of psychotic quality and also hate are
produced in those who work with severely ill psychiatric patients. Only in this way can
there be any hope of the avoidance of therapy that is adapted to the needs of the therapist
rather than to the needs of the patient.
- 74 -
Article Citation [Who Cited This?]
Winnicott, D.W. (1949). Hate in the Counter-Transference. Int. J. Psycho-Anal., 30:69-