Psychotherapy Integration Papers

Influential Readings on Psychotherapy Integration   

 CyberPsych Site Index




Transference, Schema, and Assimilation:

The Relevance of Piaget to the Psychoanalytic Theory of Transference

Paul L. Wachtel

Reprinted by permission from The Annual of Psychoanalysis, 1981, 8, 59-76. Copyright© 1981 by International Universities Press.
Article copies are available from International Universities Press Document Delivery Service: 1-800-TELEIUP E-mail address:

Editor's Note: The article "Transference, Schema, and Assimilation: The Relevance of Piaget to the Psychoanalytic Theory of Transference", by Paul Wachtel, was published on The Annual of psychoanalysis, 1981, 8, 59-76, and reprinted in "Action and Insight", by Paul Wachtel, 1987, New York: Guilford Press. In the introductory remarks to the latter, the Author states that "Both the reality basis on which transference reactions are based (cf. Gill, 1982; Hoffman, 1983) and the patient's role in bringing about that very behavior on the part of the therapist (and on the part of other significant people in the patient's life) are considerations that have been central to my evolving position over the years and are evident in considerable degree in the chapters that follow." These themes are central in this article, that recasts these questions in terms of the Piagetian notions of assimilation and accommodation. The Author introduces the chapter: "It develops a theoretical strategy I first attempted in Psychoanalysis and Behavior Therapy, but it spells out the clinical and theoretical implications in considerably more detail. The chapter begins by examining how transference has traditionally been conceived by psychoanalysts, highlighting the usual emphasis on transference as inappropriate to the situation and the problems with such formulations. It then goes on to show how the phenomena usually discussed in terms of transference can be understood in terms of Piaget's conception of schemas; transference reactions reflect schemas in which assimilation predominates over accommodation. This view, as the chapter develops, points the theorist or therapist to the necessity of seeing what role accommodation may play in the process as well, since from a Piagetian perspective pure assimilation without accommodation is an impossibility. Once the question is framed this way, considerably greater appreciation of the therapist's contribution to the patient's transference experience is afforded, as well as a more complex picture of just how the predominance of assimilation occurs and of how the therapist can play a more transformative role in the patient's life. This perspective also helps to illuminate ways in which the transference experience can be a disguised and highly symbolic representation of what is actually transpiring between patient and therapist without awareness."

The article is amazingly fresh after twenty years. The Piagetian schema, whose usefulness to shed light on the transference phenomena has been keenly shown by Wachtel, has a growing appeal in the integrative field. The concepts of assimilative and accommodative integration, derived from Piaget, allow for a new understanding of the basic integrative processes.


The Freudian concept of transference originated in observations of disturbed adults, obtained in the context of therapy, and was an attempt to account for certain distortions in their perception of reality. The Piagetian concept of schema derived from observations of healthy children, obtained in the context of research, and was an attempt to account for their increasingly accurate perception of reality. Two more disparate origins would be hard to find. Yet, I will argue, the Piagetian concept can provide a very useful and clarifying perspective on the phenomena to which Freud directed our attention.

Freud's first reports of transference phenomena were published in the Studies on Hysteria (Breuer & Freud, 1895/1955). There he referred to the patient establishing a "false connection" (p. 302) between the doctor and a figure from the past. Transference reactions were "a compulsion and an illusion, which melted away with the conclusion of the analysis" (p. 304; italics added).

In his discussion of the Dora case, Freud (1905/1953) elaborated somewhat on the concept and introduced an interesting complexity, related to the major theme of the present chapter. Though transferences were generally to be regarded as "facsimiles" which "replace some earlier person by the person of the physician," some were found to be "more ingeniously constructed" and "may even become conscious, by cleverly taking advantage of some real peculiarity in the physician's person or circumstances and attaching themselves to that" (p. 116).

In "The Dynamics of Transference" (1912/1958) Freud states that "The peculiarities of the transference to the doctor, thanks to which it exceeds, both in amount and nature, anything that could be justified on sensible or rational grounds [italics added], are made intelligible if we bear in mind that this transference has precisely been set up not only by the conscious anticipatory ideas but also by those that have been held back or are unconscious" (p. 100).

Three years later, in further considering the idea that transference reactions should not be understood as real reactions to what is going on, Freud (1915/1958) discusses as an "argument against the genuineness of this love . . . the fact that it exhibits not a single new feature arising from the present situation, but is entirely composed of repetitions and copies of earlier reactions, including infantile ones. We undertake to prove this by a detailed analysis of the patient's behavior in love" (p. 167).

He then goes on to indicate, however, that in thus proceeding "we have told the patient the truth, but not the whole truth" (p. 168). The argument that transference love is not genuine because it is a repetition is weak, Freud says, because "this is the essential character of every state of being in love. There is no such state which does not reproduce infantile prototypes" (p. 168). Of particular relevance to the arguments to be advanced in the present discussion, Freud notes that the difference between transference love and what we call normal love is one of degree, and further adds, "it displays its dependence on the infantile pattern more clearly and is less adaptable and capable of modification; but that is all and not what is essential" (p. 168; italics added). Thus Freud indicates here that the processes responsible for the emotional and perceptual phenomena we label as transference are essentially the same as those in all relationships between two people, differing only in degree.

Yet only a year later, in his Introductory Lectures, Freud (1916/1963) repeats the kinds of statements which were the basis for the prevalent tendency (discussed below) to treat transference reactions as something quite distinct from "realistic" reactions to others. There, discussing transferences, Freud wrote that "we do not believe that the situation in the treatment could justify the development of such feelings. We suspect, on the contrary, that the whole readiness for these feelings is derived from elsewhere, that they were already prepared in the patient, and, upon the opportunity offered by the analytic treatment, are transferred on to the person of the doctor" (p. 442). He goes on to say, discussing negative transferences, that there can be "no doubt that the hostile feelings towards the doctor deserve to be called a 'transference,' since the situation in the treatment quite clearly offers no adequate grounds for their origin" (p. 443). And in discussing how to deal with transferences in treatment, he says, "We overcome the transference by pointing out to the patient that his feelings do not arise from the present situation and do not apply to the person of the doctor, but that they are repeating something that has happened to him earlier" (pp. 443-444).

Increasingly, this latter emphasis became the standard and predominant psychoanalytic view. Transference reactions were regarded as inappropriate and unrealistic. They were not to be viewed as responses to the current reality of the analyst or the relationship he had established with the patient, but had to do with something in the patient's past which was being erroneously transferred to the present context. Despite developments in ego psychology, which have alerted us to the complex interaction between long-established psychic structures and current environmental input, contemporary formulations and definitions continue to treat transferences strictly as unfounded departures from reality. Greenson (1967), for example, states unequivocally that "transference reactions are always inappropriate" (p. 152). And Langs (1973a) says that "to identify a fantasy about, or reaction to, the therapist as primarily transference. . . we must be able to refute with certainty any appropriate level of truth to the patient's unconscious or conscious claim that she correctly perceives the therapist in the manner spelled out through her associations" (p. 415; italics added).

At the same time, however, it has been increasingly recognized that the actual behaviors and attributes of the analyst do play some rote in evoking transference reactions. Informally, it is frequently pointed out that transference distortions often have a reality "hook" or "peg" on which they are hung; and, more formally, Langs (1973b) has referred to "reality precipitates" of patients' transference fantasies. Macalpine (1950) has pointed to the particular features of the psychoanalytic situation which foster and call forth regressive transference phenomena. Gill has even suggested--correctly, I think--that unless the analyst acknowledges the role of his own behavior in evoking the patient's reaction, the patient's ability to accept and make use of the analyst's interpretation will be severely limited.

But the continuing emphasis on viewing transference as a distorted or inappropriate reaction, a displacement of something from the past, has required the introduction of a number of other concepts to account for the patient's ability to react to the realities of the treatment situation-- to accept, for example, that the analyst's silence is a technical part of the procedure rather than a deprivation aimed specifically at the patient and designed to hurt or punish him, or to recognize after a while that the analyst does value him even if he does not give overt reassurances. Such concepts as the "therapeutic alliance," the "working alliance," and the "real relationship" are designed to address these aspects of what occurs during the course of an analysis (Greenson, 1965, 1971; Greenson & Wexler, 1969; Zetzel, 1956).

Since the same patient who is able to continue to cooperate in the analysis because he recognizes the technical nature of the analyst's silence may also fantasize that the silence is really a sadistic act or a retribution for sins, it is usually suggested that the transference and the working alliance may exist, as it were, side by side, proceeding apace as two different features of the therapeutic process. Such a way of conceptualizing does acknowledge the complexity of the patient's manifold levels of reactivity to what is happening, but it also creates serious dangers of reification, in which the transference, the working alliance, and the real relationship are separate and discrete "things." Schafer (1977), from a slightly different perspective, has also addressed the at once realistic and unrealistic aspects of the patient's reactions to the analyst. He suggests that Freud had not quite reconciled two varying views of transference. "On the one hand, transference love is sheerly repetitive, merely a new edition of the old, artificial and regressive . . . and to be dealt with chiefly by translating it back into its infantile terms. . . . On the other hand, transference is a piece of real life that is adapted to the analytic purpose, a transitional state of a provisional character that is a means to a rational end and as genuine as normal love" (p. 340). He notes that integrating the two perspectives is an important theoretical problem, and suggests that one major obstacle to such integration is the tendency to draw too sharply such distinctions as "past and present, old and new, genuine and artificial, repetition and creation, the subjective world and the objective world" (p. 360), etc. As we shall shortly see, applying the perspective of Piaget's theory to these questions helps to transcend these dichotomies and to foster the integration of the different views of transference.


Piaget's concept of schemas, characterized by the two basic functions of assimilation and accommodation, seems particularly useful for understanding the diverse phenomena of transference and other more or less closely related relationship phenomena. Piaget's work in general highlights the active role of the developing individual in shaping and defining his experiential world. Neither as children nor as adults do we respond directly to stimuli per se. We are always constructing reality every bit as much as we are perceiving it. This emphasis on the importance of evolving structures which mediate the individual's experience and behavior is quite compatible with the psychoanalytic view. Both theories suggest that man is not stimulus-bound, that he does not just reflexly respond to external stimuli, but rather selectively organizes and makes sense of new input in terms of the experiences and structures which define who he is.

The concept of transference was an attempt to come to terms with an extreme version of this tendency to experience events in terms of structures and expectations based on earlier experiences.

The observations which generated the concept seemed to suggest such an unusually strong role for internal mediating structures that the reality of who the analyst was or what he was doing was virtually ignored by the patient. The tendency to perceive the present in terms of the past became, in certain affectively laden areas of experience, so acute that it seemed to override all evidence of the analyst's actual neutral, investigative role. The difficulty with the concept of transference as it is usually formulated is that it is so exclusively focused on distortion, on the lack of perception of the real characteristics of the analyst. It is for this reason that the observation that the patient does also recognize the analyst as a real person in a professional helping relationship to him must be represented by a completely different concept than that of transference (e.g., therapeutic or working alliance, or real relationship). It is difficult, therefore, to know quite where or how to fit the observations that transference reactions do seem to have a "reality peg." The difficulty, discussed above, in integrating the varying perspectives on transference phenomena is a result of this dichotomous theorizing.

From a Piagetian perspective, one can readily see a continuity between those phenomena usually described as "transference" and those designated by terms such as "therapeutic alliance" or "real relationship" (or indeed, more generally, between transference phenomena and the accurate gauging of other people's motives and characteristics which facilitates effective adaptation). Transference reactions, in Piaget's terms, may be seen simply as reflecting schemas which are characterized by a strong predominance of assimilation over accommodation. The experience with the analyst is assimilated to schemas shaped by earlier experiences, and there is very little accommodation to the actualities of the present situation which make it different from the former experience.

In part, of course, such a way of talking about transference phenomena is simply a translation from one language system to another. But it is a translation that has some important implications, both in terms of pointing inquiry in somewhat different directions and of facilitating the integration of varying views of the phenomena of interest. Perhaps most importantly, once one views these phenomena in terms of schemas, one is confronted with the idea that schemas can never be characterized only by assimilation. Assimilation may at times predominate over accommodation, but there can be no such thing as "pure" assimilation--or, for that matter, as "pure" accommodation (Piaget, 1952, 1954).

However necessary it may be to describe assimilation and accommodation separately and sequentially, they should be thought of as simultaneous and indissociable as they operate in living cognition. Adaptation is a unitary event, and assimilation and accommodation are merely abstractions from this unitary reality. As in the case of food ingestion, the cognitive incorporation of reality always implies both an assimilation to structure and an accommodation of structure. To assimilate an event it is necessary at the same time to accommodate to it and vice versa.. . . [T]he balance between the two invariants can and does vary, both from stage to stage and within a given stage. Some cognitive acts show a relative preponderance of the assimilative component; others seem heavily weighted toward accommodation. However,"pure" assimilation and "pure" accommodation nowhere obtain in mental life. (Flavell, 1963, pp. 48-49)

Transference, seen in this tight, can be understood as the result of a state of affairs in which assimilation is strongly predominant, but is nonetheless not inexorable. Some accommodation to the actual details of what is being experienced, and to how they differ from those of previous experiences assimilated to that schema, must also occur. Since assimilation is strongly predominant, it does not take a particularly dose fit to activate the transference schema. So two very different analysts may, in separate analyses with the same patient, be subjectively experienced in very similar fashion by the patient. The schema is easy to activate, and it does not change very readily despite the lack of fit. But since "pure" assimilation cannot occur, it is not completely arbitrary. The range of activating events is wide but nonetheless does have some bounds. The occurrence of transference reactions can seem at times to be almost completely the playing out of an internal dynamic, so striking and deviant can it be from the reality of what is going on between patient and analyst; but it is never completely unrelated to what is transpiring. This is what Gill is calling our attention to in his emphasis on the importance of acknowledging the analyst's role in eliciting such reactions, and this is why a perceptive observer can often find a "reality peg" or "hook" in even the most extreme transference reaction.

This perspective also shows us why Gill's emphasis on the finding of a reality peg in no way undermines the important clinical core of the concept of transference; nor does it ignore Freud's insights about how the continuing effect of the patient's childhood way of experiencing reality is revealed in the transference. If anything, it provides a basis for making the traditional psychoanalytic formulation even more powerful by making it more precise: It points us to ask in all instances of transference precisely what aspect of the analytic situation or of the analyst's behavior or characteristics led to the occurrence of this particular transference reaction at this particular time. Since the predominance of assimilation is emphasized, no loss of the role of intrapsychic factors or the patient's unique individuality is entailed by this particular kind of effort to relate the patient's behavior to events currently going on about him. The schema notion implies responsiveness to environmental cues without positing stimulus-bound, slavish reactivity to environmental events. Thus, one can avoid the pitfall of the false and limiting dichotomy between understanding in terms of intrapsychic factors or "psychical reality" and understanding in terms of the "actual" situation, and appreciation of reality factors can enhance rather than compete with a psychodynamic perspective. An understanding of what particular features of the situation bring forth the transference reaction can in this way be seen as a legitimate part of what is pointed to by psychoanalytic understanding, rather than as the undermining or watering down of that understanding. A broader and firmer base is thus provided for the psychoanalytic view, which also gains increased power and utility.

Further clarity is also provided by this perspective regarding the question, both substantive and definitional, as to whether transference reactions are manifested only in the analytic situation or go on in the patient's daily life as well. From the present vantage point, one can readily see that all perceptions and behaviors are mediated by schemas which are the product of past experiences and which attempt to assimilate new input to them--as well as to accommodate to their novel features. Understanding just which aspects of the analytic situation make assimilation more likely or help to highlight the way in which it occurs in the patient's mental functioning (cf. Gill, 1954; Macalpine, 1950; Stone, 1961) has important clinical utility.

Such understanding can also shed light on the question of how best to generalize from the data of the analytic session and integrate the formulations such data suggest with those deriving from other



Much of the confusion which arises from the traditional way of talking about transference phenomena is a result of the cognitive and perceptual theory which underlay Freud's theorizing. As Schimek (1975) has recently clarified, Freud's view of cognition was at odds with the essential thrust of the rest of his theorizing, which was obviously strongly dynamic, motivational, and developmental in its emphasis. In contrast, his ideas about cognition, Schimek shows, were based on the simple associationism that one finds among many stimulus-response learning theorists who have been particularly opposed to psychoanalysis. This simple associational psychology has been sharply criticized by Piaget (1952), by critics of behavior therapy sympathetic to psychoanalysis (Breger & McGaugh, 19ò5), and recently even by a number of prominent behavior therapists who have seen the necessity of taking into account man's active role in defining what the effective stimulus is and how it will be experienced (Bandura, 1974; Mahoney, 1974), It would be unfortunate if psychoanalysis, to which such a view is really most alien, were to retain it.

Precisely because psychoanalysis is in its other aspects so strongly a dynamic, motivational psychology, this aspect of its conceptual underpinning went unnoticed for a long time. The stagnant, nonpersonalistic conception of perception and cognition was obscured because dynamic and personal factors were so strongly brought into the theory at the point after the percept or cognition was formed. As Schimek points out clearly, Freud assumed a simple, camera-like registration of reality and formation of memory traces which, again, stored "accurate" images of reality that were somewhere retained in their true and original form--but then he concentrated, in the more important and original aspects of his work, on how these images and representations were transformed or distorted under the pressure of drives and defenses. It was there that the dynamic features of the theory were evident. So powerful and original were Freud's ideas in this regard that it was little noticed that the perceptual building blocks for these dynamic processes were conceptualized by him in a far less dynamic fashion than they were by many academic psychologists in the developing area of cognitive psychology.

In conceptualizing transference phenomena, this camera-like view of perception and memory traces led to the formulation that a fully formed, pre-existing set of reactions is plucked from their original context and displaced from an early figure to the analyst. As Greenson (1967) puts it, "Transference is the experience of feelings, drives, attitudes, fantasies and defenses toward a person in the present which do not befit that person but are a repetition of reactions originating in regard to significant persons of early childhood, unconsciously displaced onto figures in the present" (p. 155).

Such a formulation leaves little room for any accommodation to the reality of the analyst and the interaction. Something static and pre-existing is simply "displaced," moved from one object to another. The postulation of a somewhat malleable and responsive structure, built up on the basis of prior experience, but shaped as well by new experiences that do not quite fit it, would permit a reconciliation and synthesis of observations of "distortion" in the transference, and observations of accurate perceptions and of realistic, cooperative engagement in the analytic process. But a "displacement" formulation, which implicitly requires a fully formed representation to be displaced, ends up leading to the proliferation of separate and discrete postulated quasi-entities--the transference, the therapeutic alliance, the real relationship, etc.

Rather than dichotomizing between perceptions that are accurate and those that are "distorted," the schema notion helps us to see that all perception is a selective construction, in some respects a creative act. It is not arbitrary, but it never lacks the personal element. Even the supposedly "objective" observations that underlie scientific theory building are richly suffused with the idiosyncratic and personal, as modern philosophers of science--Polanyi (e.g., 1958, 1966) in particular--have strongly emphasized. In the perception of other persons, and especially in the perception of their intentions and affective states and qualities, the variability from observer to observer is so great that it is extraordinary that a sharp distinction between "accurate" and "distorted" perceptions could have been retained for so long. To be sure, each patient's experience of the analyst is highly individual and shaped by personal needs and fantasies. But consider the enormous variation in perception of the analyst by those other than his patients--the differences in how he is experienced by his spouse, his children, his teachers, his students, his friends, his rivals. Which is the "undistorted" standard from which the transference distortion varies?

Discussing the phenomena traditionally designated as transference in terms of schemas, assimilation, and accommodation does not present us with such conundrums. It avoids the sharp dichotomizing implicit in most discussions of transference, yet retains the clinical core. To recognize a unity in the modes of apprehending reality that encompasses both the transference perceptions of the analysand and the observations of the physicist or chemist is not to ignore the differences between the two, or to blunt the problematic features of the former. Indeed, it enables us to incorporate the role of the analyst's real properties and behavior not as something which somehow limits, reduces, or "excuses" the patient's highly personal interpretation, but as a way of amplifying it and gaining a finer sense of its determinants.


Ideally, one might expect to see a fairly even balance between assimilation and accommodation, with neither predominating to any great extent. In that case the individual would be able to be responsive to variations in environmental stimulation, while maintaining a certain consistency and managing to make sense out of new events on the basis of previous experience. The phenomena discussed in psychoanalytic writings under the rubric of transference suggest an imbalance in this ideal relationship, an excessive degree of assimilation that impedes efforts to adaptively gauge and deal with the events of the present. In attempting to account for how this imbalance comes about, two main lines of explanation seem to have developed.

The traditional psychoanalytic explanation stresses the role of repression and other defenses in creating a structural differentiation which, in effect, prevents accommodation. Accommodation per se is, of course, not referred to in most psychoanalytic accounts. Rather, what is stressed is that defenses relegate certain contents and processes to the id, preventing them from becoming part of the ego. Since it is the ego which is the part of the personality which is in touch with the perceptual world and which has well-developed properties of organization and coherence (Freud, 1923/1961), the result of repressing something is to prevent it from being modified by new perceptual input--i.e., to prevent accommodation. This is why the contents of the id are described as "timeless" and why, for change to occur, they must be integrated into the ego, where they are brought into contact with perceptual input and with the demands for logic and consistency. Freud's famous phrase "Where id was let ego be" reflects the view that only when id contents are integrated into the ego can they be modified to conform to current reality demands. If one employs (and extends) the conceptual scheme of Piaget in this context, it can be seen that one effect of defensive processes is to interfere with the accommodation of certain schemas to new input. When manifested as transference phenomena, these schemas are revealed in their original structure as they are applied inappropriately to stimulus objects which would be more appropriately assimilated by schemas which have undergone a developmental evolution.

When viewed in the light of the Piagetian notions of schema, assimilation, and accommodation, some questions are raised about this traditional account of how transference reactions persist in unchanging form. Such an account seems to contradict Piaget's view that accommodation and assimilation must both be present. Now, of course, one need not postulate that transference schemas show no accommodation whatever. Even changes in the particular cues which serve to elicit the transference reaction reflect some degree of accommodation; and the postulation of at least a certain degree of evolution and change in transferential schemas (even apart from whatever change can be brought about by analysis) is not really inconsistent with the traditional psychoanalytic view. Moreover, transference schemas are ones in which affective and defensive processes--which Piaget did not address--are centrally implicated. It is certainly possible that in this realm Piaget's observations regarding the dual role of accommodation and assimilation might have to be modified. The question of precisely how defenses can impede accommodation would seem from this perspective a particularly important one.

A different way of accounting for the apparent lack of accommodation in transferential schemas relies less on structural differentiation and a conception of the id as a zone of nonaccommodation. Instead, one might assume that transference schemas, like any others, will show accommodation in response to clear, disconfirming feedback. In that case, a lack of change would imply that the actual feedback is either unclear or not really disconfirming. To understand how this might happen, it is useful to examine some contrasts between our interactions with the physical world and those with other persons.

The schemas which come to represent the physical world to the child, and which form the basis for much of our commerce with the world, do change a great deal as feedback requires accommodation of extant schemas. (At the same time, of course, this input is also assimilated to those evolving schemas.) Whether one is observing an infant learning to grasp an object, a child learning about conservation of various quantities, or an adult learning to drive or ski, one sees a process, varying in speed and efficiency, in which feedback shapes and changes the existing schema. Why then do the schemas associated with transference seem to change so little, despite their apparently poor match to the input with which they are coordinated?

One thing becomes clear if one pursues this line of thought: For the schemas that represent the physical world, disconfirmation is relatively clear and dramatic. The skier or driver who organizes input incorrectly falls or goes off the road; the infant fails to grasp the object he seeks; etc. In the realm of interpersonal and affective events, it is much harder to know one has been in error. Such events are highly ambiguous, and consensus is much harder to obtain. Almost everyone would agree when you have gone off the road.

That is not the case as to whether you have incorrectly construed anger in another (or failed to construe anger). The ambiguity of affectively laden events and the consequent difficulty in determining when feedback requires accommodation make accommodation far less efficient in this realm and the persistence of old schemas in early form more likely.

It must further be noted that the nature of the affective and interpersonal stimuli which we encounter (and which we must assimilate and accommodate to) is substantially a function of our own actions. This is, of course, true to some extent in the physical realm as well. Driving presents us with different stimuli--and a different adaptive task--if we turn the wheel to the left or to the right. But with physical stimuli the process is not nearly as complex, and the potential input is more predictable and varies over a narrower range. Moreover, it is much easier to know when a change in input is due to our own actions and when it is an independent event--the difference, say, between the variation in direction of a hit tennis ball as a function of one's stroke or as a function of a sudden strong gust of wind.

With affective and interpersonal events, however, the sorting-out process is particularly difficult. It is very easy to be convinced one has experienced what someone "is like" without realizing how much the experienced property (even if accurately gauged in this or other particulars) is a function of one's own actions when with him. Each of us tends to consistently elicit particular aspects of others' personalities, and must of necessity experience the sum of these elicitations as "the way people are." For relatively healthy personalities, the range of elicitations is fairly wide, and variable enough to be roughly representative and in agreement with the experience of others. But it is important to recognize that none of us really lives in an "average expectable environment." We all experience some particular idiosyncratic skewing of the possible kinds of encounters with others. And this skewing is not just accidental, but is a function of who we are. One of the ways in which consistency in personality is maintained is by the selective choice of situations and interactants and the elicitation of a particular side of those we do interact with. Given who we are, we select and create a particular kind of interpersonal world; and given that world, we experience the need to go on as we have--and thus elicit that same kind of personal world again.

The persistence of transferential schemas, then, with little change over the years despite what one might expect to be considerable pressure for accommodation, can be seen as due both to the ambiguity of interpersonal-affective feedback (making it easy not to notice that disconfirmation or lack of fit has occurred) and to the tendency for events to in fact confirm the seemingly inaccurate perception. If the world were, in effect, to "hold still" for the developing child rather than to change with his conceptions of it, he would change to accommodate to it. In learning about the physical world, this is in fact what happens, and it happens enough in the interpersonal world for most of us not to be grossly out of touch. But to a substantial degree, the world of affective and interpersonal events does not hold still. It accommodates to our initial conceptions and expectations (as they are translated into actions toward others) and short-circuits our accommodative activities in this realm. Our suspicions, and the actions they motivate, lead others to in fact be hostile; our expectations of seductive behavior lead to eroticized interactions with others; our submissive behavior, based on past experiences as well as defensive needs, induces others to expect more compliance from us than they do from others.

By the time the patient comes to see an analyst, he has probably had hundreds of such quasi-confirmatory experiences. I call them quasi-confirmatory because the patient's perceptions are in one sense anachronistic, even if they may turn out to be confirmed. For what happens is that the person encounters another who is initially quite ready to relate to him differently than the patient expects, but who over time responds to the patient's pattern of interaction with an all-too-familiar complementary pattern. What to the patient feels like an accurate perception may be inaccurate as that but fairly reliable as an implicit prediction: This is how the other will act toward him after some time in his interactive field (cf. Wachtel, 1977).


The experience with the analyst is, one hopes, a major and dramatic disconfirmation that can permit accommodation to occur. The analyst facilitates accommodation in at least two ways. First, by interpreting unconscious fantasies (and by establishing the analytic situation, in which such fantasies are likely to become more intense and vivid), he helps the patient to be more aware of both the schemas that guide his transactions with others and the kinds of events that constitute confirmation or disconfirmation of his expectations. Thus he helps reduce the ambiguity which makes for easy assimilation and impedes accommodation. Second, he avoids falling into the complementary behavior pattern which the patient's style of relating has so frequently brought out in others. As I have described in more detail elsewhere (Wachtel, 1977b), every neurosis requires "accomplices" to maintain itself, and a good deal of the analyst's effectiveness may be seen as residing in his ability not to become one more accomplice. Both his neutral, analyzing stance and his skill in spotting and interpreting the patient's subtle and unconscious maneuvers enable him to accomplish this task.

It is not necessary, however--nor do I think it is possible--for the analyst to completely avoid falling into complementary behavior patterns. Wolf (1966) has described particularly well how such unwitting participation in the neurotic pattern can occur. For therapeutic purposes, it is sufficient that (1) the analyst for the most part avoid becoming an accomplice to the neurotic process (in other words, that he do a better job at this than most of the people the patient encounters, even if he is not perfect); and (2) he be able to acknowledge when and how he has acted in a way consistent with the patient's transference expectations and to help the patient understand how such patterns come to be repetitive features of his life. Thus, I would agree with Langs (1973a) that when the therapist's behavior "has been correctly and unconsciously perceived by the patient, his interventions will begin, as a rule, with an acknowledgment of the veracity of the perception and refer to the way it served as a stimulus for the patient's responsive fantasies and conflicts." I would further agree that "once the therapist has acknowledged his contribution to the situation . . . the patient's responsibility for his reactions must be recognized and subsequently analyzed" (p. 430). As a result of the considerations put forth in this paper, however, I would strongly disagree with Langs's contention that this is appropriate only when the therapist's or analyst's behavior has been "erroneous" or that all such occurrences are in fact errors in any useful sense of that term.


Transferences can at times seem quite fantastic. All analysts have seen patients express feelings and ideas about them that seem grossly off the mark and appear to have much more to do with their experiences and fantasies with regard to other--usually earlier figures. In order to understand this common sort of observation from the present point of view, several points must be considered. To begin with, one can recognize that transference reactions are indeed very often grossly inappropriate without drawing a theoretically problematic dichotomy between transferences and realistic perceptions. If one starts from the assumption that all perceptions and actions are mediated by schemas characterized by both assimilation and accommodation, then it would appear that we label as transference that portion of the continuum in which assimilation is predominant. Even in this range, however, assimilation is not inexorable, and a particular schema will be called into play only if there is something in the analytic situation that bears some resemblance to the stimuli which have nourished the schema in the past. Since, however, the dimension of similarity can be a highly personal one, there need not be much of an "objective" similarity. Hence the transference reaction may seem completely arbitrary and brought about by "internal" factors. Examination of what in the analytic interaction elicited it at this point, however, is likely to be richly rewarded, for it affords an understanding not only of the kinds of fantasies the patient is capable of, but also of the conditions for their arousal and the particular difficulties to which they may be related.

In considering just how unrealistic transferences really are, it is important to recognize that the transference reactions of most interest and concern to the analyst are those involving substantial anxiety and conflict. In such circumstances the patient is highly motivated not to see clearly what he is experiencing. Rather than communicating directly what his experience is with the analyst, he is likely to express it indirectly and symbolically. For defensive reasons, his statement about some aspect of his experience of the analyst may be so oblique it is unrecognizable. It simply sounds like an outlandish and incorrect perception that must really be about someone else. If the analyst is not prepared to translate the symbolism not only into childhood references but also into references to what is currently transpiring, he can easily be persuaded that the patient's reaction is simply a "displacement" from somewhere else and has little or nothing to do with actual occurrences in the analysis.

Thus, if the patient has the fantasy that the male analyst is a woman in disguise or has no penis, or that he is much older than he really is, or is a notorious and immoral seducer, the analyst, feeling secure that the fantasy as stated is not true, may not recognize how it symbolically reflects the patient's reaction to some particular action or pattern of actions by the analyst. Depending on the specific meaning of "woman" to the patient in that context, for example, his fantasy that the analyst is a woman might mean he viewed something about the analyst as weak, or soft, or emotional, or nurturing, or smart or whatever.

It is, of course, important for the analyst to determine the meaning of "woman" to the patient in order to understand fully the transference reaction. But, having done so, it is also important to know just what he did that seemed weak, nurturing, or whatever to the patient; and this not primarily for the purpose of discovering his "error" and attempting to weed it out in the future by more self analysis (though either of these aims is certainly at times appropriate), but rather for the purpose of understanding just what kind of input the patient's schemas assimilate in just what way (for not just any behavior on the analyst's part would get registered as "weak" or as "woman"). Such understanding enables the analyst to apprehend much more precisely how and when the patient's psychic processes create problems for him, and importantly, the range of situations in which problems and misperceptions are not likely to occur. All too often, lack of specificity and failure to understand intrapsychic organizing processes in their situational context interfere with an appreciation of the patient's strengths, of where and how intact functioning is manifested (see Wachtel, 1973b, 1980).

In addition, understanding what behavior of the analyst elicited the patient's transference reaction can enable analyst and patient to explore whether other people in the patient's life have tended to behave as the analyst did, and what meaning the patient has given to their behavior. The analytic work can then examine both the kinds of behavior the patient elicits from others and the impact of such behavior as filtered by the patient's complexly motivated perceptual processes, as well as the way in which this in turn leads to behavior on the patient's part which is likely to again elicit similar behavior from others--thus starting the cycle all over again. One then gets a picture of transference reactions as not just the residue of some early experience which is being displaced or replayed, but as part of a continuous process that has characterized the patient's life for years yet has only become fully explicated in the experience with the analyst. Such a perspective, I would contend, provides both a more complete understanding of transference reactions and improved possibilities for facilitating therapeutic change (see Wachtel, 1977b).


The considerations presented here do not pose a challenge to the basic observations of psychoanalysis regarding transference phenomena. I regard as soundly based on clinical observation such central psychoanalytic tenets as that patients regularly show rather substantial distortions in their perceptions of the analyst; that such distortions are personally meaningful and related to the person's history; and that they are in important ways the product of unconscious conflicts and fantasies.

The present perspective does suggest, however, certain modifications in how we think about our observations, and points toward the inclusion of a wider range of observations than has been typical in psychoanalytic practice. It also suggests that the path between early experiences and later transference reactions may be more continuous than has been typically portrayed; that interactions with many figures throughout the person's life tend to occur in such a way as to confirm and perpetuate the modes of perception and reaction that eventually appear as transferences in the patient's analysis; that transference reactions, even when seemingly unrelated to the reality of the analyst or the analysis, are often symbolic expressions of conflicted perceptions of what has actually transpired, or at least of the personal meanings which actual events and characteristics have had for the patient; that accommodation occurs to such a slight degree in some interpersonal and affective schemas both because of the ambiguity in this realm, which makes it harder to know when disconfirmations have occurred, and because of the reactivity of events in this realm to our own actions: What we expect to occur is likely to happen even if it would not have been likely to occur if it were non expected (and if we did not act accordingly).

The potential value of conceptualizing transference phenomena as reflecting schemas in which assimilation predominates over accommodation has not been exhausted by the considerations put forth here. It is to be hoped that future efforts will carry this work forward.


Transference phenomena have traditionally been viewed as reactions which are inappropriate and based on the distorting effect of the patient's past. At the same time, they convey an important reality about the patient's life (or at least his subjective life) and--it has been increasingly recognized--an important reality about the therapeutic interaction as well. Integrating these varying perspectives on transference has created some (not always clearly understood) theoretical difficulties. The present discussion has suggested that Piaget's notion of schema, with its stress on the simultaneous processes of assimilation and accommodation, can help to clarify these theoretical issues. By regarding transferences as schemas in which assimilation predominates over accommodation to an inordinate degree, one can incorporate both the traditional clinical knowledge about the distorting effects of transference and an emerging recognition of the importance of what actually transpires between patient and analyst. Such a way of looking at transference both points to and is aided by an understanding of the differences between the ways in which we learn about the physical world and the world of people and emotions. It also leads to a number of other important new questions for psychoanalytic inquiry and new perspectives on psychoanalytic practice.


Bandura, A. (1974), Behavior theory and the models of man. Amer. Psychol., 29:859-869.

Breger, L. & McGaugh, L. (1965), A critique and reformulation of "learning theory" approaches to psychotherapy and neurosis. Psychol. Bull., 63:338-358.

Breuer, J. & Freud, S. (1893-1895), Studies on hysteria. Standard Edition, 2. London: Hogarth Press, 1955.

Endler, N. & Magnusson, D. (1976), Interactional Psychology and Personality. New York: Halsted.

Flavell, J. (1963), The Developmental Psychology of Jean Piaget. Princeton, N.J: Van Nostrand.

Freud, S. (1905), Fragments of an analysis of a case of hysteria. Standard Edition, 7:3-122. London:

Hogarth Press, 1953.

-- (1912), The dynamics of transference. Standard Edition, 12:97-108. London: Hogarth Press, 1958.

-- (1915), Observations on transference-love. Further recommendations on the technique of psycho-analysis. Standard Edition, 12:157-171. London: Hogarth Press, 1958.

-- (1916), Introductory lectures on psycho-analysis. Standard Edition, 16. London: Hogarth Press, 1963.

--(1923), The ego and the id. Standard Edition, 19:3-66. London: Hogarth Press, 1961.

Gill, M. (1954), Psychoanalysis and exploratory psychotherapy. J. Amer. Psychoanal. Assn., 2:771-797.

Greenson, R. (1965), The working alliance and the transference neurosis. Psychoanal. Quart., 34:


-- (1967), The Technique and Practice of psychoanalysis. New York: International Universities Press.

-- (1971), The "rea]" relationship between the patient and the psychoanalyst. In: The Unconscious Today, ed. M. Kanzer. New York: International Universities Press, pp. 213-232.

-- & Wexler, M. (1969), The non-transference relationship in the psychoanalytic situation.

Internat. J. Psychoanal., 50:27-40.

Langs, R. (1973a), The patient's view of the therapist: Reality or fantasy? Internat. J. Psychoanal.Psychoter., 2:411 -431.

-- 1973b), The Technique of Psychoanalytic Psychoterapy. New York: Aronson.

Macalpine, I. (1950), The development of the transference. Psychoanal. Quart., 19:501-539.

Magnusson, D. & Endler, N. (1977), Personality at the Crossroads: Current Issues in Interactional Psychology. Hillsdale, N.J.: Erlbaum.

Mahoney, M. (1974), Cognition and Behavior Modification. Cambridge, Mass.: Ballinger.

Mischel, W. (1968), Personality and Assessment. New York: Wiley.

Muslin, H. & Gill, M. (1978), Transference in the Dora case J. Amer. Psychoanal. Assn., 26:311-328.

Piaget, J. (1952), The Origins of Intelligence in Children. New York: International Universities Press.

-- (1954), The Construction of Reality in the Child. New York: Basic Books.

Polanyi, M. (1958), Personal Knowledge: Towards a Post-Critical Philosophy. Chicago: University of Chicago Press.

-- (1966), The Tacit Dimension. Garden City, N.Y.: Doubleday.

Schafer, R. (1977), The interpretation of transference and the conditions for loving. J. Amer. Psychoanal. Assn.., 25:335-362.

Schimek, J. (1975), A critical re-examination of Freud's concept of unconscious mental representation. Internat. Rev. Psychoanal., 2:1 71 -187.

Stone, L. (1961), The Psychoanalytic Situation. New York: International Universities Press.

Wachtel, P. (1973a), Psychodynamics, behavior therapy, and the implacable experimenter: An inquiry into the consistency of personality. J. Abnormal Psychol., 82:324-334.

-- (1973b), On fact, hunch, and stereotype: A reply to Mischel. J. Abnormal Psychol., 82:537-540.

-- (1977), Psychoanalysis and Behavior Therapy: Toward an Integration. New York: Basic Books.

-- (1980), What should we say to our patients? On the wording of therapists' comments to patients. Psychotherapy: Theory, Research and Practice, 17:183-188.

Wolf, E. (1966), Learning theory and psychoanalysis. Brit. J. Med. Psychol., 39:1-10.

Zetzel, E. (1956), Current concepts of transference. Internat. J. Psychoanal., 37:369-376.

November, 1979


Design and Hosting by CyberPsych