Analyzing, Not Psychoanalyzing

In what way is ego analysis not psychoanalysis? To begin an answer, we need to ask what exactly is psychoanalysis, and the answer depends on who you ask and when you ask it. For a recent answer there is Victoria Hamilton’s (1996) report of her interviews with 65 American and British analysts in four major cities between 1988 and 1990.She found that “the area between [their] avowed theoretical orientation” and actual clinical practice “reveals the muddled overlaps and uncomfortable, precarious coexistence of parts of belief systems.” She argues for what she calls “pluralism,” seeing it as a healthy tolerance of differences, as contrasted with the “strife [that] is rife among psychoanalysts of different persuasions, each arguing for the ‘right’ developmental theory and therapeutic action.”

Among these scattered views Hamilton discerned two explanatory models, British Object Relations and the Relational, Self Psychological model. The popularity of both of these models can be seen as a reaction to the previous model, known by many as drive theory or classical psychoanalysis, although it will put this conceptual map in sharper relief to call it the conflict model.

Following Freud’s usage, it has become customary to speak of “the negative therapeutic reaction.” The term was meant to refer to a crisis resulting from a groundswell of negative reactions to psychoanalytic interpretations. Two such crises were fateful in the career of psychoanalysis, one well known, the other not yet labeled as such. The first was reported by Freud in the twenties—for which he coined the term—and the second hit in the fifties through the seventies and is responsible for the decline in popularity of psychoanalysis and intensive therapy.


Originally, Freud thought that repressed instinctual drives (needs, wishes, especially sexual) are transformed into anxiety, literally poisoning the system, and so repressions have to be lifted, expunging the anxiety they were thought to create. The technique of interpretation was to point out hidden wishes in associations, slips, and dreams. These were so-called “naming” interpretations, and this was id analysis, the technique of making the unconscious conscious. The analyst was a “libidinal detective” (Sterba, Reminiscences of a Viennese Psychoanalyst).

For example, in treating Dora, Freud told her, to her embarrassment, that running her finger in and out of her little coin purse was the expression of a masturbatory impulse. He was pointing to a sign of repressed and transformed libido. Nowadays, we would no longer think, “Aha! We caught you, the fact that you are embarrassed means that you, Dora, really do want to masturbate.” We would think that her embarrassment simply reveals that she felt scolded by Freud, chided, belittled, put down. Another example (discussed in Vignettes) is from a case of Alexander’s in which he says that his interpretation “called the patient’s bluff,” whereupon the chastened man left the session with “manifest embarrassment,” which in those days was the proof that an interpretation had struck home. Fromm provided a similar example in declaring that a patient may respond to an interpretation by saying, “Nonsense,” but then, the next day or an hour later this person has a severe depression, because he could not stand this sudden truth. Now, if this whole thing is so repressed, why does he react that way? Something in him has heard it.

The patient’s, anxiety, anger, or guilt was taken as confirmatory. You can picture the analyst at that time thinking, just as did Fromm,”Why else would a person get upset?” This is the idea that if you feel guilty, there must be something to feel guilty about. When a patient “protested too much,” as we say, this proved that he or she had been caught red-handed. The negative therapeutic reaction was a crisis waiting to happen.

Id analysis is the version of analysis that the public still has. The analyst is thought of as a wiseguy who trips you up, who catches you in slips and denials. The id analytic assumption is still alive in the popular mind—that when someone responds by being defensive, anxious, or guilty, this confirms whatever has been said about them.

Psychopaths, so-called, could not be treated, nor could anyone else who, on being shown repressed impulses, would be unmoved—would not consciously repudiate them. Although patients were expected to feel terrible in reaction to having repressed wishes exposed—as unearthed in associations, slips, and dreams—some patients felt so bad that Freud came to think of them as “beyond the pleasure principle,” that is, as masochistic. It was obvious that not everyone could tolerate id analysis, and he had been using trial analyses (1912) for screening purposes. He revised his criteria of analyzability and proposed that some people are intractably masochistic, that is, are too far “beyond the pleasure principle” to be analyzable.

This would have held the line for the original theory, but Freud eventually was discouraged. He came to believe that the problem was not wholly in the patient, and that the theory was seriously flawed—a considerable accomplishment since he had not believed that psychoanalysis was in any way lacking or incomplete. Quite the opposite. As Jones (1957, p. 265) put it, just a few years earlier, that is by 1915, Freud thought he had a “well-rounded account” of psychoanalysis, and even “felt he had completed his life work.”

There was not the slightest reason to expect that in another few years Freud would have produced some revolutionary conceptions which necessarily had the effect of extensively remodeling both the theory and the practice of psychoanalysis.

Freud’s explanation for his revision was that it came out of his experience treating patients who got worse instead of better, who showed the “negative therapeutic reaction.” Everyone has accepted this explanation, but if you think about it, it minimizes Freud’s accomplishment, since in those days patients were expected to have negative reactions: to get defensive, angry, to protest, be shocked, argue. Indeed, feeling better was not the objective, or at least not feeling that much better. Neurotic misery was to be replaced by ordinary human unhappiness, Freud famously declared.

A familiar way to identify the revised theory is by reference to Freud’s reformulation of the relationship between anxiety and repression. He had thought that we come into the world biologically antisocial, and that all of us suffer from the necessity to repress our instinctual make-up. This creates ordinary human unhappiness—his tragic view. What causes neurotic anxiety is excessive or overzealous repression. But he came to realize that anxiety (shame, guilt) is itself the problem, causing wishes and needs to be repressed, just as we now take for granted is the case, but just the reverse of what Freud had at first thought. This was such an extraordinary reversal that it has yet to be fully grasped. Freud himself had trouble fully letting go of his earlier model. But—never one to mince words—he then declared that “repression has the power of instinct.” Which was to say that we can be just as driven by anxiety, shame, and guilt as by what Freud had considered the ultimate motives, sex and self-preservation. Voila! The superego.

Fenichel, the dean and encyclopedist of psychoanalysis in the forties, named the new model “ego analysis,” and the old, classical model, “id analysis.” It was not called “superego analysis” because what now was to be analyzed was the interplay between the ego and the superego. In the earlier model Freud had depended on the patient’s good conscience to consciously repudiate what was freed from repression. This was the conscience of common sense, not to be confused with the superego. The superego is not simply or not only the conscience, but is a relationship we have with ourselves, an internalized version of our significant relationships, what Freud grandly referred to as the precipitate (lasting impressions) of abandoned (now not present) object (people) cathexes (connections). In other words, once he discovered the superego, he really discovered it.

It now should be clear why there was “not the slightest reason to expect” this revision, even though what was left out of this “well-rounded account” was the superego. Having been so thoroughly taken in by the superego, yet ultimately able to see through it, Freud was the first to realize that it is in the nature of the superego to be practically undetectable. This is why, to get ahead of my story, in many ways it is still undetectable.

Astonishingly enough, there still is no consensus (and very limited discussion) among analysts about whether the new theory was a radical revision or whether it was merely an elaboration. There still is no consensus about the implications for treatment, and little discussion of that either. Part of the problem is that these changes came late. Freud was 70 when his most complete revision was published (in 1926 in Inhibitions, Symptoms and Anxiety, earlier translated by Brill as The Nature of Anxiety, a title which may better convey its import). His papers on technique had all been published ten years earlier and all his cases had been published years earlier also. In his later years he devoted himself to broader issues.

Clinically, there was, of course, the general idea that analysts were now to pay more attention to analyzing resistances, defenses, and guilt, but most of them thought they already were doing enough of that. According to Sterba many even resented the new model, having become comfortable with the old. Id analysis is seductive; it hardly taxed the analyst. (Drinking coffee and eating behind the patient had been standard practice for the relatively unoccupied analyst, Stone, in Langs & Stone, l980, p. 107.)

Although Freud’s refinement of the content of interpretations was far-reaching, little has been written about it, and there is no generally agreed upon version, partly because psychoanalysis has functioned as a guild (less so in recent decades), its concrete operations being an arcanum passed on in training analyses and supervised (control) analyses and not dependent on being public. This is especially obvious from the analytic journals, in which it is hard to find much about what analysts actually do.

There is essentially only one source for the technical modifications indicated by the structural model, that is, by the recognition of the superego and by what Gill and I (1989) referred to as the relativity of defense. This is Fenichel’s Principles of Psychoanalytic Technique (1941). Although this classic is required reading in all analytic institutes, it is short (only 122 pages), condensed, and largely abstract. Consequently, these revised recommendations were not at all clear to everyone nor generally agreed upon, partly because of the guild system, partly because of the simplicity and attractiveness of the id analytic model, and partly because the ego analytic model is not only more complex than the earlier model, but is, in some essential ways, counterintuitive. This is why you have not heard much about it.

Since the essential logic and explanatory principles of the id analytic model, also known as the classical model, still have a firm grip on the field, and certainly on the popular understanding of psychoanalysis, it may be clarifying to let that be psychoanalysis as a way to distinguish it from ego analysis. This emphasizes how the two frameworks are mirror images of one another in basic ways, even though many believe that they are compatible and even supplementary. (I have to add a clarification that risks anesthetizing the reader, that the classical model is also known as American Ego Psychology, easily confused with Ego Analysis).


Now, remember I proposed that there was a second crisis, a second “negative therapeutic reaction,” not yet labeled as such, that has been fateful in the career of psychoanalysis. I think of it as an inevitable consequence of the lack of appreciation of Freud’s original solution to the problem, his revised model of the twenties.

Freud’s revision was an absolutely necessary step toward solving the problem, but it was a stunning reversal that was difficult to assimilate. Although Freud actually apologized for his earlier error, he did not help matters by proposing that the two models could exist side by side, as “points of view,” the topographic, with its quasiorganic the Conscious and the Unconscious, and the structural, with its fluid and interpenetrating superego, ego, and id. It was all too easy to see the ego and the id as just the old Conscious and Unconscious. Many theorists even wrote of Freud that he simply “turned his attention” to the superego.

The difficulty assimilating the new model arises in part from the way that it so unexpectedly transforms the old, as well as from the fact that id analysis, as I am at pains to show throughout this website, despite its exotic content and terminology, is undergirded by commonsense assumptions. To put that simply, id analysis depended on the superego. The guilt-prone, conscientious patient was counted on to repudiate unacceptable wishes. Since this was the solution, this blocked recognition of the pathogenic effects of self-hate, and self-hate is what the newly conceptualized superego came to represent. It is almost entirely invisible, as Freud was the first to see. Recognizing that self-hate is the problem is one thing, detecting its effects is another.

Psychoanalysis was heading for another negative therapeutic reaction since the revised model did not take. Analysts were still insensitive to superego effects (I think this is not putting it too strongly), so that interpretations (and the analyst’s stance) still came as a shock to the ego, that is, had an unempathic effect (the word “empathy” was not yet in common use in the field). It still was id analysis, although less obviously so—and still could be surprisingly, disturbingly crude.

Although the revised model did not take, it would be an exaggeration to say that the id analytic model has been preserved intact. It became what I prefer to call the conflict model, although it is officially known as the drive model, following Greenberg and Mitchell. Most analysts conceive of the change wrought by the structural model as a shift from bringing repressed wishes to consciousness, to bringing repressed conflicts to consciousness. Where the early analyst thought that bringing a content to consciousness was itself the way to free bound energies, the more modern analyst thought that bringing a conflict to consciousness then gives the person the opportunity to make choices.

There is no better way to make clear what this meant in actual practice than to turn to Roy Schafer’s “action language.” He proposed that psychoanalysis has a lot in common with existential therapy: both attempt to show people that they are making unconscious choices, on the assumption that when they recognize that they already are choosing unconsciously, they will be able to take responsibility for their lives by making conscious choices. He even says that if there is no action, there is no change. Hence, interpretations should be couched in the language of action.

Schafer recognizes that there are areas of real helplessness, but he declares that these areas are simply not available to interpretation (1976, p. 361):

Not that all of one’s life is thus unconsciously altered, arranged, and disclaimed, but psychoanalytic interpretation deals methodically and specifically with the large part of one’s life that is one’s own action.

He states this proposition flatly, and obviously feels no need to defend it since he is simply making plain what he takes to be the fundamental psychoanalytic assumption. For a strongly felt objection to this stance, Gill had this personal comment:

One of my analysts once said that if my parents’ behavior accounted for 95% of my troubles and it was my experience of their behavior that accounted for the remaining 5%, it was only that 5% in which he was interested. I see his point more clearly now than I did then, but I still think I was right to become enraged.

Let’s pause here for a moment to reflect. In the fifties and sixties there was an active protest movement against psychoanalysis among therapists (especially those known at the time as the “NeoFreudians”) and academics, led by Rollo May, who was coming from Sartre and his “existential psychoanalysis.” Sartre passionately believed that he was opposing Freud’s view that our actions are determined by unconscious motives for which we could disclaim responsibility. He and many others thought that Freud’s message was that we are ruled by the Unconscious, and so cannot be held responsible for our actions. It clearly appeared from Freud’s writings that this was his position. But this was to make the mistake I referred to earlier, to try to divine the psychoanalytic angle on things from the literature, that is, to assume that psychoanalysis is a social science that exists in the literature and is necessarily wholly public. Psychoanalysis might well be called not only “the impossible profession,” but the inscrutable profession as well.

Of course, Freud did say that the ego is like a rider on a horse who guides the horse in whatever direction it wants to go. What he rarely said was that the gravamen of clinical psychoanalysis is that what you need to recognize is that it is your horse; these are your motives. In a short paper dated 1925, titled “Moral Responsibility for Dreams,” Freud came down hard on this point (which I had missed until coming across it in Jay Greenberg’s work):

Obviously one must hold oneself responsible for the evil impulses of one’s dreams. What else is one to do with them? Unless the content of the dream (rightly understood) is inspired by alien spirits, it is a part of my own being. If I seek to classify the impulses that are present in me according to social standards into good and bad, I must assume responsibility for both sorts; and if, in defense, I say that what is unknown, unconscious and repressed in me is not my `ego’ [better translated here as “self,” as Strachey points put in a footnote], then I shall not be basing my position upon psychoanalysis. I shall not have accepted its conclusions.

As is well-known, Freud discouraged patients from reading the literature, the idea being that it would take the fizz out of the sessions. This set the precedent for having patients come to the message on their own, with only hints from the analyst. (Which makes it hard to argue with, especially if you think you came to it on your own.) No one else but Schafer had the nerve or the wit to make explicit the assumption that the interpretations of the time were wholly based on showing patients that they are responsible for their lives, that they chose it, no one else did.

What Schafer did was to make it more obvious that analysts did not have time for what patients were genuinely helpless about, in addition to having little to offer patients who were genuinely helpless, period—who suffered from so-called ego deficits. (That may be the hardest thing to take responsibility for—that is, to take responsibility for not being able to take responsibility.)

It should not be surprising, therefore, that the analysts of the period were faced with the same dilemma Freud faced. Should they reconsider their approach, or should they avoid patients who can’t (or won’t) take it?

Unlike Freud, they chose to screen out patients. This became such a preoccupation that at the time it seemed as if moszt analysts were more concerned with how to not do analysis than with how to do it. This work went in two directions. The first were contributions to refining the criteria of analyzability, with the differential being between weak egos and strong egos. This was ego psychology (see above).

Blanck & Blanck (1974) were good spokespersons for what most analysts thought at the time (p. 123):

The individual with a less-than-intact ego cannot tolerate psychoanalytic technique, especially abstinence and the so-called uncovering techniques. The modified ego is usually unequal to the task of coping with the drives. Often, one of the most pressing problems in the psychotic or borderline psychotic is the failure of repression. The technical problems that come into focus are strengthening the ego so it can contend with the drives, and taming the drives to make them less formidable to the ego.

This analytic model is especially obvious in the Blancks’ recommendation that with a borderline case:

the therapist responds to the request for a comment, contrary to psychoanalytic technique. In psychoanalysis, the intact ego can tolerate unresponsiveness for the sake of elaborating fantasies. In borderline structures, failure to respond at such a point constitutes abandonment analogous to the unavailability of the mother in the separation/individuation phase.

“Strengthening the ego,” and “taming the drives,” refers to what analysts did when
they were not doing analysis. With patients who were not analyzable you did therapy, and there was a lot of concern with figuring out what the difference was between psychoanalysis and psychotherapy, now a largely obsolete issue. With especially “fragile” patients, what you were supposed to do was to strengthen their reality sense, and you did this by talking to them about their daily lives, having them be diaristic. This was thought to be therapeutic in itself, although if the patient had some obviously distorted idea about someone, you gently showed them how they could look at it differently. But you also spent a lot of time just chatting. I remember one very intelligent paranoid man I saw during this period, just after he dropped out of treatment with a well-known analyst. He said he quit because did he did not get why this therapist insisted on making small talk when he had serious problems to deal with. As in any relationship, it was unclear who was indulging whom. If this approach sounds absurd, I should say that it sounded just as absurd at the time.

In a paper immensely popular during this period, Eissler proposed that one could introduce “parameters” in treatment, if they were “resolved.” To translate: you could deviate from standard procedure if you did it in a carefully delimited way. The paper was popular for two reasons. First, it reassured everybody that there indeed was a standard treatment protocol that everyone followed—a now-preposterous notion— and, second, it reassured everybody that they could deviate from it.

Also, the slogan went the rounds that “not everything is conflict,” meaning that sometimes you have ego deficits (rather than inhibitions caused by conflict), the psychoanalytic term for these, as it were, “real” psychic incapacities that Shafer made clear were thought to lie outside the purview of the psychoanalysis of the time (the conflict model) because they are beyond choice. The work of Hartmann and his colleagues in the early fifties, Metapsychology, that is, theoretical Ego Psychology, also known as “the structural model” was a remarkably arid effort to account for nonconflict related psychic structures. It essentially ran the structure concept into the ground and was, in fact, buried along with it (see Bob Holt on “The Death of Metapsychology”).

So much for trying to hold the line. By this time, the negative reaction to psychoanalysis was growing, and a number of brief approaches sprang up in the late sixties and seventies, all attempting to remedy the situation, and all believing that patients needed to be hit harder, sooner. The revelation of the period was that analysts were wasting patients’ time and money, taking forever to get to the point. Fritz Perls wasted no time: “Don’t say can’t. Say won’t.” If this sounds like Shafer’s action language it is because Perls, out of Berlin, was the past president of the South African Psychoanalytic Institute. In retrospect it almost seems as if the forces of id analysis were mobilizing for one all-out last stand—colorfully typified by the unmasking fervor than ran through these rambunctious therapies. There was Perls in a film on Gestalt therapy telling a soft-spoken woman that speaking softly is really a sign of cruelty. In Gestalt Therapy Verbatim he declared that “The good boy is really a spiteful brat.” (Have I said that “really” is the id analytic word?)

My all-time favorite line from the period was in a paper on an encounter group, in which the leader reported an exchange with a member who, just having arrived for the session, remarked on what a nice day it was. The leader wasted no time: “It’s a nice day outside, but you are not outside. You are here, inside.” From est to Synanon it was open season on patients. Although the Gestalt approach still exists in some form, all the rest are gone, their passing unremarked on. Had id analysts been less scrupulous about choosing patients and had they been as direct in laying on their interpretations, psychoanalysis (that is, id analysis) itself might have been unmasked sooner.

The line finally cracked, and it was Kohut who did it (see The Two Analyses of Mr. Z., Plus Two). Not only were there all these patients who could not tolerate analysis, it turned out that there also were a growing number of analysts who could not tolerate it either. To put it another way, analysts were increasingly uncomfortable with the siege mentality of the classical model (see A Key to Where the Bodies are Buried in Psychoanalysis).


It was Kohut’s genius to tinker with the classical model in such a way as to make the analyst more comfortable, less likely to feel tested and to see the therapy as a power struggle—less called upon to be tough. Kohut exposed, as no one else had, the way that the classical model—and in many ways the standard dynamic model most therapists have been taught, that derives from the analytic model—makes the therapist prone to feel manipulated and tricked, facing patients who are out to defeat the therapist and sabotage the therapy.

Although in the ego analytic perspective Kohut’s solution was faulty, his diagnosis of the condition which Freud had, in its earlier appearance, called the negative therapeutic reaction, was dead-on accurate. What had made psychoanalysis difficult to tolerate and caused widespread disillusionment with its therapeutic results was the effect of unempathic interpretations. Kohut found just the right word. It may be no exaggeration to say that therapists were jubilant. Now it was neither weak, nor naïve, nor a sign of boundary problems for a therapist to be empathic. However, many seemed not to know or even to care what Kohut meant, and “empathic” essentially has come to mean that being warm and sympathetic is in itself therapeutic. For some therapists, at least, the assumption was that this meant being less analytic. (The recent opposition to self psychology is a consequence of the way empathy is misunderstood as something you give, rather than as an experience that may or may not be evoked in you. The same misunderstanding plagues couple therapy.)

As can be seen from his seminal case of Miss F., Kohut himself believed that the analyst ultimately has to draw blood. As I discuss in detail in Ego Analysis vs. Self Psychology, he eventually pursued a line of interpretation that caused severe shock to the ego. He describes his “continuing revelation of her [Miss F.’s] persistent infantile grandiosity and narcissism” as “slow, shame provoking, and anxious.” Does this mean that, as Kohut conceived of it, the only way to completely avoid the negative therapeutic reaction is to avoid interpreting altogether?

If the diagnosis is correct, that the problem all along has been unempathic interpretations, then wouldn’t the cure be to arrive at empathic interpretations? But is that possible? Doesn’t the truth hurt? Don’t we all defend ourselves against knowing aspects of ourselves that we feel ashamed of, that we are afraid are abnormal, immoral, or immature? If those are the questions that spring to your mind then, 1) welcome to the club, and 2) you are in the grip of the logic of id analysis. The ego analytic insight is that it is these worries and reproaches about ourselves that we suffer from.

From the perspective of id analysis that sounds like the familiar and naïve notion that we are all basically wonderful, and if we could only give ourselves a break we would recognize that. It can easily seem as if the issue is whether we think that people are basically good or basically bad. As if it all depends on which existential position we take. But remember that negative self-talk is almost entirely silent and unheard—that the superego is largely invisible. If we fully work through that insight, then we realize that that is the problem, that the not knowing is the problem. Then it is no longer crucial whether the experiences (or, if you prefer, wishes) being denied (that is, defended against, and therefore either unexpressed or expressed in defended form) are good or bad. The problem is only that the person is not making the negative judgment consciously and so does not recognize its effects and cannot find relief from them.

Consider James Gilligan’s Violence, in which he, as a prison psychiatrist, shows how violent men are driven by shame, something every family worker knows about battering and homicidal husbands. But these men don’t know it. Reducing their shame would of course change them completely, and this may be what the born again experience does for those few who experience it. But what is easy to miss is that their sense of shame is itself unconscious and this is the surface of the problem. If they knew they felt this profound shame, they would be affected at least as powerfully as they would be if the shame itself was to somehow be relieved.

These men are already are “free” to be violent. Relieving their shame, not really a possible outcome—but speaking hypothetically—would make them less violent. (The layering metaphor of id analysis and of mental topography makes this aspect of shame counterintuitive, but it also is true that conscious shame feels severely inhibiting, a misleading commonsense experience, since shame works its effects almost entirely out of awareness. For another example of the driving power of shame, much of our sexual behavior is propelled by performance anxiety.)

But these last remarks are, right here, just a place holder. I only want to say enough to indicate that it is entirely possible, although not easy, to arrive at interpretations that have an empathic effect. An illustration of how interpretations that are intended to be empathic may actually not have that effect, or at least may fall short, takes me to the present phase of the psychoanalytic journey.

Let’s keep in mind Santayana’s dictum that what we don’t remember we are going to repeat. That is, in fact, what now is happening as id analysis enters its next incarnation. Once self psychology cracked the line that by now hardly exists in memory, object relations approaches moved in along with it—the Klein/Bion movement. The popularity of object relations concepts, especially containment and the holding environment, zoomed, and for the same reason—being empathic was such an engaging prospect. But now the status of interpretation itself was unclear. Where in the seventies, analysts were beginning to suggest that “not everything is conflict,” now almost nothing is. It’s all “deficits”—developmental arrests and traumas. And to some extent the analytic relationship is, as such, considered therapeutic. But there also is no dearth of interpretations and, in the id analytic style, they are of the naming kind, but intended to be comforting rather than unmasking.

An illustrative vignette from Winnicott offers a thumbnail of this subtly confrontational, id analytic approach. When a patient reported never having accepted his father’s death, and also developing a headache, Winnicott interpreted this as the man’s need to have his head held, as would a child in distress. The patient then realized that there was no one to hold his head and to comfort him in his grief. Winnicott proposed that the correct interpretation provides a “holding” of the patient, which allows him to regress and be dependent on the therapist (Collected Papers, 1954, p. 261).

As St. Clair (1986) put it, paraphrasing Winnicott, “A therapist needs to understand the client in a deep way and convey that to the individual through an interpretive remark [that is, an interpretation of the naming kind].” This followed from Klein’s use of “deep” interpretations, which were thought to cut through ego defenses. (I will get back to this one-liner approach below.) Also,

A client needs the therapist to be omnipotent, to know and tell him what he needs and fears. The client often knows these feelings, but the crucial issue is that the therapist needs to know and say them. The client’s False Self and defenses may distract the therapist, but the therapist must be wary of this and see the central issues without being told.

The patient’s not wanting to know about his wish to be comforted is considered to be merely his false self interposing itself (cf. Horney’s neurotic self vs. real self), merely a cover for the wish. (For more about the unintended pejorative implications of the conception of defense as merely a cover, see The Shame-Blame Reflex). This is an attempt to reach out empathically, but it falls short because it misses the ego— even deliberately. (I will show how that is, but first let me wind up this section with a concluding paragraph.)

That is my review of psychoanalysis, from id analysis proper, to conflict analysis, to self psychology and object relations. It has been all about what ego analysis is not, although this review is the ideal way to introduce what ego analysis is, since without psychoanalysis there would have been no ego analysis.


With that summary done, let’s continue to think about the Winnicott vignette. Note that the analyst must avoid being distracted by the patient’s defenses in order to get directly to the patient’s experience. The ego analyst reverses this rule. From the ego analytic point of view, what distracted Winnicott was this man’s wish to be comforted, and what Winnicott was distracted from was the patient’s shame about it (characterizing his defense as “false” is itself a sizeable “distraction”). That, in a nutshell, is the difference between ego analysis and id analysis.

I should make clearer what a naming interpretation refers to here. It means pointing to a content: “You really miss your father and wish he was here to comfort you.” It is intended to cut through defense—the patient’s feelings and attitudes about the content. An ego analytic interpretation points to the patient’s feelings and attitudes about the content, the patient’s relationship to the content: “You feel like it’s immature to still need your father, like I would never have such a need.” It is not simply naming.

But don’t go away. There is more to be explained here. First, Winnicott’s interpretation could well be relieving. This man would feel sympathized with and also relieved of his shame. This is not id analysis in the old unmasking sense. If you check out the Alexander case in Vignettes, it should be clear that that patient would not feel sympathized with. He was being caught dead to rights. And in A Key to Where the Bodies are Buried in Psychoanalysis there is Freud refusing to be sympathetic (manipulated). Winnicott seemed to be genuinely kindly and in no way paranoid about being manipulated. So his was a very benevolent id analysis.

What is nice about the object relations approach for comparative purposes is that it controls for the less crucial flaws of id analysis: the patient is not to be kept in a state of deprivation (see below for more about this), nor is there any sense of ripping aside the mask (in Winnicott’s terms, the false self is simply bypassed as irrelevant). There are two considerations here. One is whether the man’s shame actually was relieved, and the second is how that is not the point anyway.

First, the patient could still think he was being childish but that Winnicott was a kind and tolerant man, which he undoubtedly was. If so, the relief would not be generalizable. Also, the patient might subsequently think he shouldn’t worry about being childish, and so could be ashamed of still being ashamed, perhaps especially when he was with Winnicott.

But what is more crucial for this man is to be able to generalize from this experience to his relationship to himself. He felt ashamed and as a result had a headache instead of a wish. So what is really important is not the shame or the wish, but that both were not in his awareness. What would really produce psychological growth in this man would be to develop the mental set in which he would be fully capable of being aware of feeling ashamed of missing his father’s comforting presence. This level of analysis is so superficial that it is deep. Neither common sense nor psychoanalysis prepares you for it (although Fenichel argued that the analyst should interpret as close to the surface as possible).

For this man to be able to experience both the wish and the shame about it would mean that he was able to have a more intimate relationship with himself. The ego analyst might say, “It looks like you are ashamed of wanting to have your head held, just as, at your age, you feel like you shouldn’t still be missing and needing your father.” This interpretation would be followed up with at a later date by, “You now feel like you shouldn’t feel ashamed, like it’s neurotic, or I wouldn’t approve of it.”

This man would have had the sense all along that he was ashamed of having this wish. But it could be called preconscious, or even unconscious-conscious —thoughts and feelings of which we are aware, but not conscious—to use that cognitive science distinction. Or, to use another cognitive science distinction, thoughts and feelings that are conscious but not in the flashlight beam of attention.

What follows may be overkill, but I am moved to include another vignette here, from Hanna Segal (1967), one of the original Kleinians. It has the virtue of being straight-up Klein, in that Segal attempts a deep transference interpretation in the first hour of a training analysis. This man spoke of his determination to be certified as an analyst as soon as possible. He also talked about his digestive troubles, and at one point mentioned cows. These associations established the theme of the hour for Segal. Her interpretation was: “I am the cow, like the mother who breast-fed you; you feel that you are going to empty me greedily, as fast as possible, of all my analysismilk.”

Let me quickly interpose a citation from Hamilton’s survey, the one I cited at the outset. One informant commented that “Americans…say it is crazy to interpret transference the moment the patient walks into the room,” the Kleinians’ rejoinder being the “long-standing joke…about Americans’ fear of interpreting the transference” (p. 119).

OK, back to the story. Segal added that her interpretation “immediately brought out material about his guilt in relation to exhausting and exploiting his mother.” Kernberg (1979) offered an “ego psychological” critique of this interpretation, in which his only reservation was that it may have been too deep, “neglecting the patient’s character defenses, thus leading to intellectualization.”

What makes this little drama interesting, and perhaps not merely overkill, is what looks like Segal’s trained incapacity (to use a sociological term) to notice that this poor man, who just is dying to be an analyst and to soak up Segal’s wisdom, feels accused of being rather nasty, exploitive and a drag. Even without the pathos of his “confirmatory” association, it seems obvious enough that analysis-milk does not sound like something you should want (especially not greedily) as contrasted with analysis. And his association makes it sound as if she did not say this to him with a smile. He heard her saying, “You should be ashamed of yourself. Can’t you be a little more considerate?” His retort should have been, “Look, if it is going to take that much out of you, I feel less greedy already.” But I should not get too carried away. This man undoubtedly knew that Kleinians say these things to people all day long; you shouldn’t take it too personally.

Here is much the same interpretation from Erich Fromm, at least as reported in seminar notes by Landis (1981): “You feel that you have to treat me gently and nicely because that’s how you’ll get what you need; actually you’d like to devour me, but you’re afraid that if I knew that I’d be angry.” (From another record of his lectures comes this version: “You are really afraid of me because you don’t want me to know that you want to eat me up.”) Again, it doesn’t sound like you should want to devour (or eat up) your therapist. But Fromm probably did not think he was expressing disapproval, and might well have insisted that what he said should not be taken that way.

By the way, it is an interesting sidelight on patients that Fromm’s patient undoubtedly did not say, “No! Do you really think so? That makes me feel terrible!” Or even, “Do you think that’s bad?” And certainly not, “Dr. Fromm, when you say ‘devour,’ what exactly do you mean?” It requires an ego analytic approach to bring out such reactions.

For another pointed contrast with the ego analytic approach, let’s go back to my citation from Blanck and Blanck. Remember their discussion of therapist responsiveness, in which they declared that if the borderline patient asks a question, the analyst answers whereas, in contrast, “the intact ego can tolerate unresponsiveness for the sake of elaborating fantasies.” This refers to the rule of frustration, according to which, if needs are not gratified, the patient will regress to more primitive developmental stages, signified by the fantasies that emerge. This wild and romantic notion emerged from Freud’s experience with hypnosis, in which the patient’s own interpretation of what was going on was not part of the action. Freud even moved to the pressure method to treat patients who were not hypnotizable. He thought that pressing on their forehead (they were still in the recumbent position) might be an alternative way of bringing out repressed contents. Since this also was the period in which he reported finding that his patients revealed a surprising number of rape and incest experiences, it has been suggested that these reflected this experience with Freud.

However this may be, the analyst did not consider all the ways the patient might think and feel about this unresponsiveness. The ideal of the time was what Glover called the mutative interpretation, a one-liner (more or less) that exploded on the patient’s mind, much like dramas in which characters are told something about themselves that suddenly and conclusively shatters their illusions. This interpretation was to be a thing of art that incorporated a childhood reconstruction and a transference repetition into a single intervention. The analyst waited for the opportunity to arise, asking no questions, making no other comments (certainly no “I see’s” or “Uh-huhs”). This is at the farthest remove from ego analysis.

The analyst magisterially assumed that what the patient thought about his austere presence was simply noise. The patient silently thought (while free associating), “Jeez, is this what you have to go through to get analyzed?” The patient might also think that the analyst acted reserved as a mark of his professional status, or even because he just was a reserved guy. When the patient found this austerity maddening, he or she would talk to friends about it and be told to hang with it; it’s the drill. Nevertheless, the patient would at times feel hurt and rebuffed, even condescended to, silently (while free associating). So it looked like any relationship: two people, each in their own world. When the analyst finally let loose with his presumably bull’s-eye interpretation, the patient would silently try to figure out what a patient was supposed to do with it.

What made this approach possible was the analyst’s sense of speaking directly to the patient’s unconscious, listening to the patient’s stream of talk for confirmatory reactions that interpretations kicked up. There was no exchange with the patient, who soon learned that interpretations were not to be hashed over. The patient could experience the freedom of not having to genuflect, not having to enact the usual social niceties, but at the same time this mimicked an ordinary relationship, in which people are unaccustomed to thoughtful give and take.

The Blancks did convey the conventional wisdom of the time, that with a borderline patient, the analyst was not austere, since, “In borderline structures, failure to respond…constitutes abandonment analogous to the unavailability of the mother in the separation/individuation phase.” It is central to the conflict model that the patient either has an “intact” ego or a “fragile” one, and this is your guideline for how “deeply” you can interpret. If you interpret too deeply with patients whose ego is fragile, they may “decompensate” (which seemed true, and many analysts told horror stories at the time, but what was true was that the more disturbed the patient, the more sensitive and the more likely to feel attacked, condemned, or frightened by unempathic interpretations—being miners’ canaries, as it were, since more intact patients were more able to take their medicine). In contrast, the whole ego analytic focus is on ego sensitivity, and you interpret even more deeply with so-called fragile patients. But this is getting too abstract.

The crucial ego analytic point is that therapists too quickly assume that there is no way to talk about an experience. Maybe I shouldn’t say “too quickly,” since the established psychoanalytic position limits what can be talked about because there is no adequate conception of the ego.

It is not just that the classical analyst was supposed to be the strong silent type, at least in the American version. In this regard, relational and intersubjective analysts make it a focus of their approach to not be autocratic and not to assume that the analyst is always right. This avoids the posture of infallibility while preserving in its essentials the model that gives rise to it (for a discussion of a case treated from this model, see the third of the four analyses in The Two Analyses of Mr. Z, Plus Two). The paradigm is much too simple—too undifferentiated—to generate a way to talk to a patient about the experience the Blancks speak of: feeling abandoned by the therapist’s unresponsiveness. In this explanatory model, the experience of feeling abandoned is treated as determined by unconscious forces. What the patient is conscious of is just a cover, a defense. Thus, “borderline structures” are seen, as by the Blancks, simply as a form of ego deficit, and hence as not analyzable—only reparable by corrective experience or just something that the therapist has to work around. The trouble with this conception is that if you think it is true, you have no way to find out that it isn’t.

Here is how that same problem—a patient’s seeming inability to cope with analytic reserve—looks in the ego analytic perspective. It is likely that such patients think they are not supposed to feel abandoned. They might well feel ashamed of being so vulnerable. It also is likely that they would expect the analyst to be offended if this feeling is expressed. Further, they probably expect that to express feeling abandoned would be taken by the analyst as a demand, or worse, as a complaint—not an unreasonable expectation. Given this expectation, the patient may already be angry, hence could only express it as an attack or a demand and this would feel too dangerous. Any or all of this could be readily recognized by the patients or brought to their attention, as well as relieving to have it anticipated and understood.

This picture is an example of “analyzing, not psychoanalyzing.” It should make it clear that the classical analyst’s preoccupation with the “drives” was terribly distracting and nonproductive. The neglect of patients’ ways of relating to their own experience is, in retrospect, truly deplorable. What I mean by deplorable is, again, that analyst and patient would be like any ordinary couple, each in their own world— the patient feeling ashamed and resentful about feeling abandoned and the analyst thinking that the patient is stuck in this preverbal reliving of “the unavailability of the mother in the separation/individuation phase” and thus unable to stand the rigors of psychoanalysis.

The conflict model made it difficult for the analyst to be empathic since he or she operated at such a distance from the patient’s unconscious-conscious. Unless your model prepares you to envision the patient’s inner life, his relationship to himself, you cannot fully identify with him, and without that the analyst’s empathy is necessarily limited. Kohut himself provided an illustration of this.

In his early work in the fifties (with Seitz), Kohut published on “the repression barrier,” a concept that typifies id analytic thinking. The idea was, believe it or not, that psychoanalysis could proceed most fully only if the patient had a good strong repression barrier. You may recall that I have already quoted the Blancks to that effect: “Often, one of the most pressing problems in the psychotic or borderline psychotic is the failure of repression.” Despite his insight into the unempathic effect of interpretations proceeding from the classical model, Kohut could easily revert to the most primitive version of this frame. Here is an example of that kind of thinking from his groundbreaking first book, The Analysis of the Self (p. 224). It is a startling example of Kohut’s continuing uncritical acceptance of id analysis, by no means the only such example, but the most extreme and revealing:

There are, of course, moments in the analysis of some narcissistic personalities when a forceful statement will not come amiss as a final move in persuading the patient that the gratifications obtained from the unmodified narcissistic fantasies are spurious. A skillful analyst of an older generation, for example, as asserted by local psychoanalytic lore, would make his point at a strategic juncture by silently handing over a crown and scepter to his unsuspecting analysand instead of confronting him with yet another verbal interpretation.

From the ego analytic perspective, this intervention is inconceivable, since confrontation is no longer considered to be analytic. This intervention is conceivable (even prized as skillful! Get the “of course”) only when lifting repression is all that matters (id analysis). Notice that the analyst employs his practical joke rather than “confronting him with yet another verbal interpretation,” the implication being that the analyst was wearying of the use of “verbal” interpretations—of attempting to hammer his way through the barrier created by the patient’s not being able to stand being called narcissistic. He surprised the “unsuspecting” patient with a move that the patient had no way to defend himself against or to argue with (or even to grasp: after all, what was the interpretation?).

The first thing the ego analyst notices is this patient’s difficulty arguing his case (people get defensive when they are unable to adequately defend themselves) or being able to say how all the verbal interpretations affected him. Apparently the analyst just kept hammering away without finding out what the patient thought about his doing that.

But the key point from the ego analytic point of view is that the patient’s not being able to stand thinking of himself as narcissistic is the problem and necessitates defense. This was an extraordinary reversal in psychoanalytic thinking. In the original view, certain wishes are repressed because they conflict with universal standards. That approach was easily understood. The revised approach is difficult to understand, even especially so because the original approach is so lodged in people’s minds.

The best way to put the revised approach is to say that repression itself is the problem. Here is the idea: the narcissistic person—that is, the one we call narcissistic —is so condemnatory about being narcissistic (this comes out most clearly in his or her reaction to other narcissistic people) that it is rigidly blocked from conscious awareness. If the person was more able to tolerate this motive, perhaps even to enjoy it, then it could be consciously felt and consequently more flexibly and adaptively acted on. In other words, repression removes the wish or motive from conscious control, and it then is expressed without awareness, meaning compulsively, and defensively (cf. the narcissistic sensitivity to slights). This is what it should mean to say, as above, that “bringing a conflict to consciousness then gives the person the opportunity to make choices” (although the word “choice” is unfortunate since it carries the self-responsibility ethic and actually undoes this insight, then meaning that the narcissistic person would, in effect, sober up).

When the standards are taken for granted, the superego is invisible. We all know that people defend against feelings or wishes because of superego strictures. What is not so obvious is why this was not obvious all along, and in fact was an extraordinary discovery. The reason this point is so important is that even though we know that people defend against feelings or wishes because of superego strictures, we are often working from the old model without realizing it. Another way to say it is that we have not yet worked through this insight.

We still tend to be in league with the superego, thinking that some contents are bad, childish, pathological, and that’s why we have to confront patients with them. Going back to narcissistic patients; they alienate us, and we just think they are too narcissistic. So we silently condemn them and try to nudge them into seeing the error of their ways. This only reinforces their self condemnations, which causes them to more forcibly repress this tendency or triggers narcissistic rage (or both). You sometimes hear it said that therapists cannot help communicating their values to patients. This understates the case; in an unknown number of instances it is the whole of the therapy. (Working our way out of moralizing is the next task of the psychological revolution.)

So we tell the patient “You want to control the hours,” or “You are afraid of intimacy,” and we expect the patient to defend against these tendencies because, as everyone knows, they are bad. And we feel if we can just prove to patients that they are like this, they will correct it, will work on it.

But what we have actually done, from the ego analytic point of view, is to make the patient feel bad about these tendencies, that is to increase ego intolerance. Which brings us to the question: what does it mean to weaken or strengthen a defense?


Which brings us to the prior question: what is a defense? We think of a defense as any way of warding off awareness. Rationalization, denial, isolation, and undoing are prominent defenses. The idea is that they are strong to the extent that they block awareness, weak to the extent that they permit awareness. But if a defense is weak and permits awareness, what does this mean?

Let’s remember that the relationship between awareness and defense became less clear cut with the structural revision. Fenichel talked about the degree of ego intolerance. What we are really talking about is the meaning a content has for the person. The more it means something good or OK the less defended-against it is.

If a content means something bad or not OK, then a defense is a way of trying to make it seem good or OK. Or, it can be a way of not caring if it is good or OK. Or, it can be a way of making the content not seem to exist or be true (denial).

There are at least three relevant variables here. One is how susceptible the person is to feeling bad in general and the second is how good the person is at tolerating feeling bad. The third, and most important, is that a person always has some way of thinking about a content. People always need to know what a content “means,” and also whether it is good or bad. Not enough is made of the inevitability and ubiquity of the need to know whether a content is good or bad.

In a way, strong defenses build you up for a big letdown. Weak defenses prepare you for shocks to the ego, and you don’t need strong defenses to the extent that you can tolerate superego effects. Here is what I mean.

The early analyst was accustomed to contradicting the patient’s conscious experience. What the patient thought could almost be discounted. Kohut’s “skillful analyst of an older generation” silently handed the patient a crown and scepter to “persuade” him that “the gratifications obtained from the unmodified narcissistic fantasies are spurious.” Here we have this patient who stubbornly refuses to accede to interpretations about his narcissism, although apparently unable to be explicit about his reaction to them. He is finally presented with this mocking caricature of himself. How could he not take it as a reproach? How could the analyst not expect that? And where is Kohut coming from? The answer to the last two questions is that given the model of the mind that both the older analyst and Kohut were working from, there simply was a limit to how empathic they could be.

They thought that the analyst had broken through a defense, but from the ego analytic point of view, he had strengthened a defense: the patient would now be even more ashamed of being in any way narcissistic.

The id analytic goal was to make the unconscious conscious, on the assumption that a content that is conscious is not repressed. In the ego analytic perspective, what is repressed often is conscious, as well as expressed. Now, to say that what is repressed is often conscious and expressed is a clumsy and hard-to-digest formula, since the term “repression” like “resistance” and “defense” is a holdover from the layering concept and is hard to adapt to the structural view.

In looking for a good way to say what repression or defense actually means in ego analytic terms, I decided that the best way to say it is that if a wish is defended against, what this actually means is that the person can’t enjoy it. So that the older analyst’s patient couldn’t enjoy being narcissistic.

Similarly, the therapist who says “You want to control the hours,” should be saying something to the effect that the patient is unable to enjoy controlling the hours. Rather than “You are afraid of commitment,” the therapist should be saying that “You are unable to enjoy avoiding commitment.” (What would enjoying avoiding commitment look like? It would look like Kramer telling Seinfeld, “Jerry, marriage is a prison, a man made prison!”)

The ego analytic approach takes these issues up from the ego side rather than from the superego side, the objective being to help people to be better able to avoid commitment, if that is their issue, or to be more effectively resistant to authority, if that is their issue, or to be more effectively dependent, if they have a separation problem. That is, to enjoy these wishes.

In other words, we defend against defenses (“fear of intimacy”) just as we defend against wishes. “You are afraid of intimacy” is no different than any other naming interpretation, any other id analytic confrontation. This holds true no matter how tactfully the confrontation is put (“It seems to me that you have some fear of intimacy. Does this seem possible to you?”). (When Freud said that technique is all a matter of tact, he was talking about id analysis and the potential for narcissistic injury created by id interpretations.)


You are not used to hearing interpretations like “You are unable to enjoy avoiding intimacy (or controlling the hours).” So you can see right away that the ego analytic formula has not taken hold. Therapists are still saying “You want to control the hours” or “You are afraid of intimacy” as if you shouldn’t want to control the hours or avoid intimacy (and as if there was no good reason to)—even though, and here is the crucial point, that just like the old id analyst, they think they are simply pointing out something that is objectively true of the patient. If the patient says “Are you saying it’s bad, and I shouldn’t do it?” they say “No, I’m just saying it’s something you should be aware of.” Like the old id analyst, they would not think they were being superego figures.

In fact, no one can ever feel as objective as the therapist who is unaware of superego effects.

An illustration of this is the following vignette, offered in a workshop on the Masterson approach.

The patient arrived at the office with a folding chair and set it up at a far corner. The therapist responded with this interpretation: “You need to distance yourself to avoid feeling vulnerable.” The patient got “furious.”

An interpretation like this passes without being seen by the therapist or apparently, by the workshop participants, as an admonition (of the you-should-be-ashamed-of-yourself kind). Before discussing this intervention, I’ll review how this vignette looks to the ego analyst. Since the wish to put distance between herself and the therapist was acted out—rather than otherwise brought up—we assume that the patient had to overcome some uneasiness, that she was, at best, apprehensive about needing to put some distance between them. Ironically, it would appear that she felt too vulnerable to even mention that she felt unsafe getting close to the therapist.

We can also assume that the patient could hardly formulate this fact to herself. She must have just felt a generalized apprehension, as if it even was dangerous to be physically close to the therapist—that she would somehow be safer if she was to put some literal distance between them. This implies that she had no way to make sense of her fear. She may well have shared the therapist’s values and felt that she should not need to distance herself, that it was a shameful weakness.

The ego analyst would probably say, “It looks like you’re having trouble being able to keep your distance from me.” But before that he would wait to see what the patient said about it, as would most therapists, and if she did not mention it might make the observation that she had not said anything about it. Otherwise even this one-liner would sound too much like an admonition.

Another way the ego analyst would approach her would be to say that she was not able to have this problem of how to cope with the therapist, that she just felt she had to solve it. What being able to have the problem, that is, to tolerate having it, would look like would be her being able to talk about it. She would not have had to act out the problem if she had been able to talk about it, to say things like, “I don’t feel safe here, but I think I shouldn’t be feeling that way, like that means there is something wrong with me, and also that it would offend you, like I’m not appreciating you.” The more disturbed a person is the more rigid they are, with fewer options, easily feeling trapped, confused, immobilized, or apathetic in relationships.

Now that you have the ego analytic slant, let’s go back to the actual therapist. The first thing she said in presenting this vignette was that she went to the waiting room and was greeted by the patient carrying this folding chair. She told the audience that she immediately felt, “Uh-oh, what do I do now?” (as if any of us would have felt the same way). She felt threatened. That is one of the hallmarks of the id analytic approach. The patient is reacted to as a wily adversary, always liable to get the drop on you. Not an enviable spot to be in. Now you have two people who feel threatened. (Which, I hasten to add, would have not have been a problem if the therapist had been able to think of her reaction as countertransference, that is as a clue to the patient’s struggle. But whether or not this option is available is largely a function of the therapist’s explanatory framework.)

The therapist felt threatened because the patient’s behavior seemed to her unreasonable, like an irrational fear. The Masterson approach is aimed at especially disturbed patients, so she would have expected crazy behavior. She reacted as if she felt the therapy had spun out of control, as if something had to be done right away. Most therapists would not have tried to regain control with an interpretation, but would have just asked about the chair or said something like “I guess you’re feeling anxious.” What makes this therapist’s way of responding of interest here is that it captures the id analytic mindset.

This therapist felt no need to discover why the patient reacted “furiously” to the interpretation. She assumed it was self-evident, taking the patient’s anger as confirmatory. This reading proceeds from the id analytic assumption that the truth hurts, and that given time, it might be best to break the news to the patient gradually and tactfully, but in this emergency it was necessary to shoot from the hip.

The therapist’s automatic assumption was that the patient’s wish to distance or defend herself was just a cover—just fear. It is not seen as genuine. Which means, in some unquestioned sense, that it is in the way. This makes it inevitable that the therapist would not approach this patient’s gesture with what Kohut called “the analyst’s respectful seriousness.”

Let’s suppose that this patient responded to the therapist’s question with a question, asking, “Do you mean that I shouldn’t want to distance myself?” The therapist is very likely to have answered, “No, that’s up to you.”

But the patient hears the meaning correctly. The patient knows that the therapist is pointing out inferior behavior, and since this is exactly what the patient expects to hear, the patient doesn’t question it. She would just resolve to try to not be distant (which usually means to the therapist that the interpretation was effective: the patient now has a choice—in the Schafer sense).

That’s just the problem. If the therapist is not aware of being a superego figure, of making normative statements, then there is no way to pick up on the patient’s hearing interpretations that way. Then the therapist has no way to empathize with the way the patient can feel one-down, although it should be obvious. When one person can say to the other “You need to distance yourself,” it is pretty clear who has the higher status. The patient is in no position to equalize the status relation, as by snapping back, “What makes you think you know so much?”

Not appreciating the patient’s one-down position makes it impossible to see the patient as an informant (see Informants), that is, as striking a blow for patient liberation, if you will allow a little hyperbole here. The therapist was in no position to see the patient’s rather nervy act as a creative solution to a difficult problem.

Of course, the therapist simply took the acting out literally, rather than as a symptom. This means that the interpretation had the impact of a confrontation. Telling the patient “You need to distance yourself to avoid feeling vulnerable” bypasses the patient’s inability to consciously entertain this need, much less to talk about it. To make this a little clearer: the therapist thought that if this patient did not have to avoid feeling vulnerable, she would not need to distance herself by sitting farther from the therapist. The ego analytic version is that the patient would not need to distance herself by sitting farther from the therapist if she was more able to avoid feeling vulnerable, meaning that she was able to talk about the problem (to have the problem).

What can show the therapist that typical interpretations are normative is to ask if the therapist ever would think of helping the patient to be better able to distance him or herself, or to control the hours, or to avoid feelings. Why wouldn’t therapists encourage the patient to be better able to distance herself or to control the relationship?

If a therapist ever does attempt to help the patient in these ways, what comes out is the patient’s norms, and they are the same ones. It will turn out that the patient feels he or she shouldn’t want to control the hours, or to be more detached, to be less assertive, to avoid intimacy, commitment, or relationships—certainly should not be afraid of the therapist or see her as dangerous.

If you asked the patient who brought the folding chair, “What made you angry about my interpretation?”, she probably would say she didn’t know, or if she felt bold, maybe something like, “I don’t see why you have to make such a big issue out of every little thing.” She would not say,

“I feel like I’m not supposed to want to avoid feeling vulnerable. That’s why I need to get as far away from you as possible. Furthermore, the reason I brought my own chair rather than ask you to move the regular one is that you probably would say no, or at least stall me long enough (by asking me why I wanted to) for me to lose my nerve.”

If she could think and say this, she would be enjoying her reaction, in the sense I meant in the previous section. It would be lots of fun—but, of course, the patient had no way of knowing (did not feel entitled to know) that the reason she was angry was that she felt accused (by the therapist’s confrontation) of being weak, neurotic, or childish for trying to distance herself. My assumption is that the patient was furious because she felt scolded or nagged. This would require the ego analytic therapist to then correct the interpretation to “You don’t feel it is OK to distance yourself,” or, “You don’t feel it is OK with me to distance yourself.” (Which can be experienced as another scolding, this time for seeing the therapist as not all-accepting, which, would then, in turn, have to be interpreted.)

This is the ego analytic view of an angry response to an interpretation, that it makes the patient feel accused of not meeting expectations, being strange, unlikeable, depreciated, irritating, immature—something highly unfavorable. The id analytic assumption is that the truth hurts, and so you need to develop rapport and a positive transference before you can make a so-called depth interpretation. The Masterson therapist took that for granted. The patient’s anger was seen as confirmatory, just as the early analyst would have seen it, although with the added idea that this was the reaction of a vulnerable patient.

The idea that some patients are too vulnerable to tolerate interpretation is extremely well-established. All by itself, it means that the shift from id analysis to ego analysis has been partial, at best. It means that interpretations are still confrontations, and that it is not generally recognized that what makes them hard to take is that they feel like criticisms; in other words, they are unempathic. And, I hasten to add, they are especially hard to take because patients do not feel entitled to see them that way. When a patient “decompensates” in reaction to an interpretation the assumption is that the patient cannot tolerate “interpretation,” that it is the content that is hard to tolerate. Therapist can tell patients things like “You really want to devour me,” or suck out all my analyst-milk, and if the patient gets upset it means they can’t tolerate “deep” interpretations. The problem really is that the patient can’t acknowledge feeling hurt and depreciated. Being able to tolerate such interpretations is just as much of a problem, it being an uncritical compliance.

Let me say something here that I have wanted to get said someplace in this discussion. It is an oddity of ego analysis that, although ego analytic interpretations seem simple once you arrive at them and lay them out, this is deceptive. If I were you, I would not expect to be easily able to find the ego analytic interpretation in any given instance. It may, when formulated, seem easy to have come to, but it always is a struggle to reach, and one typically only knows in rough outline what it will be. I don’t know why this is, although I may have been explaining exactly why this is throughout this essay.


The id analyst believes that it is possible to force people to face the truth about themselves. From the ego analytic point of view, what actually is communicated to the patient when you try to do that, is that he or she is immature, weak, crazy, or bad for the reasons given in the interpretation.

“The education of the ego to ever greater tolerance,” as Fenichel put the goal of ego analytic interpretations, means analyzing the superego; the superego in modern terms means internal object relations, internal injunctions, selfobjects, your inner voices, what makes you feel immature, crazy, weak, or bad. Self talk. Nowadays most therapists think in these terms, but this does not necessarily mean that they analyze in these terms.

Now, if you think about what I have been laying out, it should be clear that the notion of being too vulnerable to tolerate interpretation is id analytic. Because if the goal of interpretation is to make the ego more tolerant, then it makes no sense to say that some patient’s egos are too vulnerable to tolerate being helped to be more tolerant. That is the best way to say how far the wheel has to be turned to arrive at ego analysis.

Whereas the id analyst assumes that if an interpretation upsets a patient it may well be confirmatory, the ego analyst assumes that an accurate interpretation should be relieving. Even further, whether or not it is relieving is the criterion of the accuracy of an interpretation.


The commonsense way that we try to change one another is to simply point things out, like you’re inconsiderate, or you’re too self-centered, or too dependent. Or you should have more self-respect. The explicit or implicit assumption is that people will change if they realize that they should. It’s not a closely examined assumption, since most people, if they thought about what they were saying, would know that, for example, you can’t get more self-respect by being told that it is bad not to have it. You get less self-respect.

The explicit or implicit assumption is that what makes people change is to realize that they are being bad. In other words, we assume that the way to change people is to be disapproving. Why do we assume that? Because that’s how our parents treated us and how every else has always treated us.

Why has everyone treated us that way? Because it feels as if people are responsible for the way they are. Which in turn comes from our experience as children of being responsible for everything that happens to us. As children we instinctively feel responsible for ourselves and our fate, and of course, we also are treated as responsible by parents who themselves are struggling with feeling responsible for the way we are.

The big problem is disentangling moralizing from analyzing. The job has certainly not been completed, if it has even begun. Psychoanalysis does not do enough to confront analysts’ commonsense reflex-assumptions.

Now, what is wrong with disapproval? After all, it can change people. What is wrong with it is that it institutes and reinforces feeling bad as the governing relationship with oneself. Why is that bad? Because it is change at the cost of internal stress—the threat of failure and of being unloved or ostracized.

Take a benign example, the CBT device of training a patient to snap a rubber band against his wrist as a way to stamp out negative self-talk. Beating yourself with a rubber band, slapping yourself on the wrist, is a stressful model for growth (spare the rubber band and spoil the person), and for how to relate to yourself.

What becomes apparent when you no longer expect interpretations to be provocative, and when relief from superego tension is your criterion of interpretative success, is that the patient typically reacts as if you are still being an id analyst. It turns out that the patient expects you to be an id analyst, and that this expectation is always present, to a greater or lesser degree, despite your best efforts. It is only when you try to not be an id analyst that you discover how difficult it is not to be. This is, of course, a superego effect and a transference experience, but it also is the effect of constant reinforcement by commonsense moralizing.

What the ego analyst is trying to do is to take people beyond the present level of consciousness that the psychological revolution has reached. Everyone still is afraid that feeling good will lead to stasis, not change. As Kohut says, in effect, by telling his crown-and-scepter story, is that sometimes the only thing that will work is a good swift kick in the pants or, more precisely in relation to his tale, a good swift blow to the ego. And he is talking about patients whose narcissism is severely injured already! Even Kohut thought that sometimes patients need a dose of something stonger than empathy.


Getting back to how therapists often reinforce self-blame: I said that like the id analyst, they fall into becoming superego figures. Another way to say this is that they act-out the countertransference.

Here’s what I mean. An ego-analytic insight about the countertransference is that it reflects the patient’s superego. The patient who is self-rejecting about feeling sad, acts depressed and makes you feel impatient. Why can’t he or she do something to snap out of it? The patient who is self-accepting about feeling sad is someone you just feel sorry for. It’s contagious.

The patient we call narcissistic is unable to enjoy being narcissistic, even though it may look like he or she enjoys it. But, no matter how it looks, if it makes you feel bored, impatient, or left out, it means he or she does not accept it. Otherwise, you would just enjoy it. It would just be contagious.

This insight leads to a different approach to interpretation. For example, take the patient who reacts to whatever interpretation you make by saying, “Yes, but what can I do about it?” The therapist feels impatient and frustrated. This patient is “yes, but-ing.” This generates a negative countertransference. The therapist feels nagged (“You’re not helping;” “You’re not helping!”) and feels guilty about feeling nagged, about not being neutral.

Guilt about the countertransference makes it hard to use. But let’s say that the therapist does not feel guilty about feeling impatient, frustrated, and bored. The egoanalytic therapist knows that neutrality is not an attitude. It’s not possible to hold a neutral attitude, no matter what you say about signal affects. The ego-analytic therapist knows that to be neutral requires an insight.

In the case of the patient accused of yes but-ing, the insight could be that this patient does a terrible job of acting hopeless. In fact, he is acting out his hopelessness, which, to the ego analyst means it’s repressed, defended against, warded off. The patient is not saying “I can’t stand insight.” He’s not saying “I get depressed whenever you point something out because it’s just seems like one more mountain to climb. It just feels like I’ll never be able to overcome it, and that you expect me to since anyone else would be able to.”

Let’s say that eventually the patient is able to say “I don’t want to think about myself. It’s too depressing.” The therapist is happy, even though the therapist originally thought that the reason he couldn’t stand the patient was because the patient refused to think about himself.

This is the ego-analytic insight: that this patient needed to be helped to feel the hopelessness, to feel entitled to feel that he or she hated to think about him or herself. It is extremely easy for the therapist to miss this level of the problem, especially because the therapist reacts against it, feels rejected by it. It just seems like the patient already feels hopeless (the patient thinks that too). That’s how he is acting, but he is acting it out.

So it is easy to miss the self-condemnation. If you do miss it, you are likely to reinforce it. You try to get the patient to talk about himself and than he just feels all the more that he should be able to, and all the more self-condemnatory for not being able to. (I have avoided discussing the recent trend toward disclosing countertransference reactions since that is a project in itself, but I think of it as obviated by the ego analytic approach to interpretation. Similarly, the one-person versus twoperson conceptions look like yet another effort to avoid being unempathic. In the present theoretical ferment even empathy is suspect, being thought of as a reaction you try to have, as I mentioned above, rather than one that either is evoked in you or is not, depending on your explanatory model and the impact of the patient.)

Before the therapist had hit on the neutral interpretation, he was acting out the patient’s self-rejection along with the patient. If the therapist does not hit on the neutral, that is, ego analytic interpretation, he is inevitably stuck in being a superego figure, in working from the side of (that is, being allied with) the superego, whether he wants to or not.

Now, it sounds so far like the ego analyst is always helping people to accept experiences that are blocked by guilt and self-blame. But here is where the relativity of defense comes in (Apfelbaum & Gill, 1989). It is true that guilt and self-blame are often the motives for defense, but it also is true that guilt and self-blame are among the most warded-off and difficult to tolerate of experiences, hence most defended against. So you can be helping people to tolerate their guilt only to find that you are reinforcing their guilt about feeling guilty.

There is an analogy here to panic disorder in which you can easily intensify the panic by reinforcing the panic about the panic. That’s what creates symptom autonomy. When symptoms are thought of as autonomous, it is because they appear to persist even after the apparent cause has been removed. But they are not autonomous; it is just that a causal loop has been created: anxiety about anxiety; guilt about guilt.

To go back to Kohut for a moment. Regarding the countertransference, the ego analyst recognizes that feeling ignored and left out (and hence feeling challenged) by the narcissistic patient is a countertransference reaction. This means that the patient is not comfortable with being narcissistic—is not enjoying it—and so is needing the neutral interpretation that Kohut provided to therapists to relieve countertransference tension, but not to the patient. Kohut told therapists that the patient can’t help it, that is, has no choice, to make them more tolerant of patients, but did not give the patient this interpretation. Narcissistic patients need that interpretation in order to be more tolerant of themselves. They need to be told, for example, that they have no good way to regulate give and take, or that they feel unable to deal with the analyst’s narcissism, and so have to totally give up the floor if they allow the analyst to talk— that it doesn’t feel to them like it’s possible to have two egos in the same room.

Ego analytic interpretations expose the ways we are made helpless by anxiety and guilt, to relieve you of the child’s way of feeling responsible for everything that happens, and to be able to tolerate the reality of powerlessness. Ego analytic interpretations focus on the anxiety and guilt about the helplessness itself that makes people feel like they should be able to take responsibility for themselves.

Another way to make this point is to say that what makes it appear that interpretation must be about choices is the focus on action, as Schafer has brought out, thereby doing us all a service. So when the analyst sees guilt about an action, the idea is that it is the action that is important; the guilt just makes it hidden from consciousness. This is id-analytic logic: defense is just a clue to hidden impulses. From the ego analytic perspective, it is the guilt that is important, not the action. The action just provides a trigger for the guilt. To make this point clear, it helps to think of guilt as a prejudice against oneself, and what the ego analyst is trying to do is to identify the pattern of this prejudice.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s