On Transference1

(1956). International Journal of Psycho-Analysis, 37:386-388

On Transference1

D. W. Winnicott

My contribution to this Symposium on Transference deals with one special aspect of the subject. It concerns the influence on analytical practice of the new understanding of infant care which has, in turn, derived from analytical theory.

There has often, in the history of psycho-analysis, been a delay in the direct application of analytical metapsychology in analytical practice. Freud was able to formulate a theory of the very early stages of the emotional development of the individual at a time when theory was being applied only in the treatment of the well-chosen neurotic case. (I refer to the period of Freud's work between 1905, the Three Contributions, and 1914, Narcissism.)

For instance, the part of theory that concerns the primary process, primary identification, and primary repression appeared in analytical practice only in the form of a greater respect that analysts had, as compared with others, for the dream and for psychic reality.

As we look back now we may say that cases were well chosen as suitable for analysis if in the very early personal history of the patient there had been good enough infant-care. This good enough adaptation to need at the beginning had enabled the individual's ego to come into being, with the result that the earlier stages of the establishment of the ego could be taken for granted by the analyst. In this way it was possible for analysts to talk and write as if the human infant's first experience was the first feed, and as if the object-relationship between mother and infant that this implied was the first significant relationship. This was satisfactory for the practising analyst, but it could not satisfy the direct observer of infants in the care of their mothers.

At that time theory was groping towards a deeper insight into this matter of the mother with her infant, and indeed the term 'primary identification' implies an environment that is not yet differentiated from that which will be the individual. When we see a motherholding an infant soon after birth, or an infant not yet born, at this same time we know that there is another point of view, that of the infant if the infant were already there; and from this point of view the infant is either not yet differentiated out, or else the process of differentiation has started and there is absolute dependence on the immediate environment and its behaviour. It has now become possible to study and use this vital part of old theory in a new and practical way in analytical work, work either with borderline cases or else with the psychotic phases or moments that occur in the course of the analyses of neurotic patients or normal people. This work widens the concept of transference since at the time of the analysis of these phases the ego of the patient cannot be assumed as an established entity, and there can be no transference neurosis for which, surely, there must be an ego, and indeed an intact ego, an ego that is able to maintain defences against anxiety that arises out of instinct the responsibility for which is accepted.

I have referred to the state of affairs that exists when a move is made in the direction of emergence from primary identification. Here at first is absolute dependence. There are two possible kinds of outcome: by the one environmental adaptation to need is good enough, so that there comes into being an ego which, in time, can experience id-impulses; by the other environmental adaptation is not good enough, and so there is no true ego establishment, but instead there develops a pseudo-self which is a collection of innumerable reactions to a succession of failures of adaptation. I would like here to refer to Anna Freud's paper: 'The Widening Scope of Indications for Psycho-Analysis'.2 The environment, when it successfully adapts at this early stage, is not recognized, or even recorded, so that in the original stage there is no

—————————————

1 Contribution to the Discussion of Problems of Transference. 19th International Psycho-Analytical Congress, Geneva, 24–28 July, 1955.

2 J. Amer. Psychoanal. Assoc., 2, 1954.

- 386 -

feeling of dependence; whenever the environment fails in its task of making active adaptation, however, it automatically becomes recorded as an impingement, something that interrupts the continuity of being, that very thing which, if not broken up, would have formed itself into the ego of the differentiating human being.

There may be extreme cases in which there is no more than this collection of reactions to environmental failures of adaptation at the critical stage of emergence from primary identification. I am sure this condition is compatible with life, and with physical health. In the cases on which my work is based there has been what I call a true self hidden, protected by a false self. This false self is no doubt an aspect of the true self. It hides and protects it, and it reacts to the adaptation failures and develops a pattern corresponding to the pattern of environmental failure. In this way the true self is not involved in the reacting, and so preserves a continuity of being. This hidden true self suffers an impoverishment, however, that results from lack of experience.

The false self may achieve a deceptive false integrity, that is to say a false ego-strength, gathered from an environmental pattern, and from a good and reliable environment; for it by no means follows that early maternal failure must lead to a general failure of child-care. The false self cannot, however, experience life, and feel real.

In the favourable case the false self develops a fixed maternal attitude towards the true self, and is permanently in a state of holding the true self as a mother holds a baby at the very beginning of differentiation and of emergence from primary identification.

In the work that I am reporting the analyst follows the basic principle of psycho-analysis, that the patient's unconscious leads, and is alone to be pursued. In dealing with a regressive tendency the analyst must be prepared to follow the patient's unconscious process if he is not to issue a directive and so step outside the analyst's role. I have found that it is not necessary to step outside the analyst's role and that it is possible to follow the patient's unconscious lead in this type of case as in the analysis of neurosis. There are differences, however, in the two types of work.

Where there is an intact ego and the analyst can take for granted these earliest details of infant-care, then the setting of the analysis is unimportant relative to the interpretative work. (By setting, I mean the summation of all the details of management.) Even so there is a basic ration of management in ordinary analysis which is more or less accepted by all analysts.

In the work I am describing the setting becomes more important than the interpretation. The emphasis is changed from the one to the other.

The behaviour of the analyst, represented by what I have called the setting, by being good enough in the matter of adaptation to need, is gradually perceived by the patient as something that raises a hope that the true self may at last be able to take the risks involved in starting to experience living.

Eventually the false self hands over to the analyst. This is a time of great dependence, and true risk, and the patient is naturally in a deeply regressed state. (By regression here I mean regression to dependence and to the early developmental processes.) This is also a highly painful state because the patient is aware, as the infant in the original situation is not aware, of the risks entailed. In some cases so much of the personality is involved that the patient must be in care at this stage. The processes are better studied, however, in those cases in which these matters are confined, more or less, to the time of the analytic sessions.

One characteristic of the transference at this stage is the way in which we must allow the patient's past to be the present. This idea is contained in Mme. Sechehaye's book and in her title Symbolic Realization. Whereas in the transference neurosis the past comes into the consulting room, in this work it is more true to say that the present goes back into the past, and is the past. Thus the analyst finds himself confronted with the patient's primary process in the setting in which it had its original validity.

Good enough adaptation by the analyst produces a result which is exactly that which is sought, namely, a shift in the patient of the main site of operation from a false to a true self. There is now for the first time in the patient's life an opportunity for the development of an ego, for its integration from ego nuclei, for its establishment as a body ego, and also for its repudiation of an external environment with the initiation of a relatedness to objects. For the first time the ego can experience id-impulses, and can feel real in so doing, and also in resting from experiencing. And from here there can at last follow an ordinary analysis of the ego's defences against anxiety.

- 387 -

There builds up an ability of the patient to use the analyst's limited successes in adaptation, so that the ego of the patient becomes able to begin to recall the original failures, all of which were recorded, kept ready. These failures had a disruptive effect at the time, and a treatment of the kind I am describing has gone a long way when the patient is able to take an example of original failure and to be angry about it. Only when the patient reaches this point, however, can there be the beginning of reality-testing. It seems that something like primary repression overtakes these recorded traumata once they have been used.

The way that this change from the experience of being disrupted to the experience of anger comes about is a matter that interests me in a special way, as it is at this point in my work that I found myself surprised. The patient makes use of the analyst's failures. Failures there must be, and indeed there is no attempt to give perfect adaptation; I would say that it is less harmful to make mistakes with these patients than with neurotic patients. The analyst may be surprised as I was to find that while a gross mistake may do but little harm, a very small error of judgement may produce a big effect. The clue is that the analyst's failure is being used and must be treated as a past failure, one that the patient can perceive and encompass, and be angry about. The analyst needs to be able to make use of his failures in terms of their meaning for the patient, and he must if possible account for each failure even if this means a study of his unconscious counter-transference.

In these phases of analytic work resistance or that which would be called resistance in work with neurotic patients always indicates that the analyst has made a mistake, or in some detail has behaved badly; in fact, the resistance remains until the analyst has found out the mistake and has tried to account for it, and has used it. If he defends himself just here the patient misses the opportunity for being angry about a past failure just where anger was becoming possible for the first time. Here is a great contrast between this work and the analysis of neurotic patients with intact ego. It is here that we can see the sense in the dictum that every failed analysis is a failure not of the patient but of the analyst.

This work is exacting partly because the analyst has to have a sensitivity to the patient's needs and a wish to provide a setting that caters for these needs. The analyst is not, after all, the patient's natural mother.

It is exacting, also, because of the necessity for the analyst to look for his own mistakes whenever resistances appear. Yet it is only by using his own mistakes that he can do the most important part of the treatment in these phases, the part that enables the patient to become angry for the first time about the details of failure of adaptation that (at the time when they happened) produced disruption. It is this part of the work that frees the patient from dependence on the analyst.

In this way the negative transference of 'neurotic' analysis is replaced by objective anger about the analyst's failures, so here again is an important difference between the transference phenomena in the two types of work.

We must not look for an awareness at a deep level of our adaptation successes, since these are not felt as such. Although we cannot work without the theory that we build up in our discussions, undoubtedly this work finds us out if our understanding of our patient's need is a matter of the mind rather than of the psychesoma.

I have discovered in my clinical work that one kind of analysis does not preclude the other. I find myself slipping over from one to the other and back again, according to the trend of the patient's unconscious process. When work of the special kind I have referred to is completed it leads naturally on to ordinary analytic work, the analysis of the depressive position and of the neurotic defences of a patient with an ego, an intact ego, an ego that is able to experience id-impulses and to take the consequences.

What I have described is only the beginning. For me it is the application of the statements I made in my paper 'Primitive Emotional Development' (1945). What needs to be done now is the study in detail of the criteria by which the analyst may know when to work with the change of emphasis, how to see that a need is arising which is of the kind that I have said must be met (at least in a token way) by active adaptation, the analyst keeping the concept of Primary Identification all the time in mind.

- 388 -