Tuesday, Sep. 8, 2009          

  Home Page 

  About the MPC 

  MPC Members 




  Board Minutes 

  Study Groups 

  Helpful Links 

  Discussion Papers 

  Photo Gallery - photos courtesy Suanne Zager 

  Questions? Contact
  the Webmaster.


Discussion Papers

Discussion Paper: Analyzability or the Ability to Analyze
Author: Bertram P. Karon, Ph.D.      Click here to view other discussion papers.


Bertram P. Karon, Ph.D.

Michigan Psychoanalytic Council

Dept. of Psychology, Michigan State University
East Lansing, MI 48824
517-355-2159 (office)
517-353-1651 (secretary)


Analyzability is the wrong question. The right question is what needs to be done to make this person analyzable. Change is produced by insight and by internalization of the analyst. It is essential for an analyst to be confused; it is not “accurate empathy” which cures patients but our attempt at accurate empathy (whether or not we are successful). Analysts value patients who keep growing, who are different from them, and who can do better than they can. It is important for the analyst to be tolerant, kind, and stubborn. Examples are given of patients who would ordinarily be considered unanalyzable, for whom an analytic process was the only means of escape from the hell in which they lived. One was an “incurable” schizophrenic; the other, a chronic patient with multiple, severe diagnoses with homicidal and suicidal crises despite medications. Both responded well to interventions that only a psychoanalyst would make.



It is traditional for psychoanalytic institutes to have a course in analyzability. The notion is that there are certain patients who are analyzable and certain patients who are not analyzable. In essence, the question is whether the patient is good enough to be analyzed; that is, whether the patient has a high level of ego functioning and object relations, which will make a therapeutic alliance easy to achieve, and a full transference neurosis easy to develop, easily tolerated, and easily resolved. A great deal of thought has gone into what aspects of ego functioning are most relevant and how best to measure them. On the other hand, history shows how vain this search is.

It is the wrong question. The right question is what needs to be done to make this person an analyzable patient.

The problem of the analyzable patient is not just a problem for institutes and training. Today we hear analysts decrying the absence of suitable patients. You cannot make a living as a psychoanalyst; there are just too many psychoanalysts and not enough psychoanalytic patients. And it is true, if by a suitable patient you mean someone who does not have serious problems, has no serious ego deficits or characterological problems, makes no phone calls, or emergencies, or unexpected crises, comes in four or five times a week, and pays $140 per session, indeed there are very few such patients.

But if a suitable patient is someone who is hurting badly and is willing to work because they are desperate, and who may have to be seen on a sliding scale, who has crises, and provides challenges, then there are patients and they are interesting patients, and they are treatable patients.

In the 1950's, it was traditional to rule all borderlines as unanalyzable; it was believed you could precipitate a psychotic break by uncovering, and if you precipitated a psychotic break then there was nothing you could do. Yet, today, it is not unusual to hear that the borderline patient represents the domain where psychoanalysis is uniquely effective even from professionals who prefer to treat their patients with medication or make-shift therapies.

Of course, even by the 1950's, Karl Abraham had treated manic-depressives in Berlin; Editha Sterba had successfully treated two autistic children in Vienna; Sullivan and Fromm-Reichmann and their students had analyzed schizophrenics; and the Kleinians, with different theory and technique, had treated both schizophrenics and manic-depressives successfully. Bettelheim, Ekstein, and Anthony had already separately written about pioneering work with psychotic children. Aichhorn had worked with delinquents. The list goes on: Lidz, Boyer, Giovoccini, Benedetti, Rosenfeld, Eissler and many others. There was an exciting growing body of literature dealing with difficult patients that all analysts knew, but of which most analysts seemed afraid.

Amongst those who derided psychoanalysis, the analyzable patient was referred to snidely as the YAVIS-patient (that is, young, attractive, verbal, intelligent, and successful). A myth developed that only such people were treated by psychoanalysts. The equally hostile current phrase is “the worried well.” Of course, this was never the truth. Frequently, only the patient and the analyst knew the hell in which the patient lived before analysis. Nor should anyone else know. Nonetheless, the kernel of truth in these criticisms lay in the search for the “analyzable” patient. In the words of Reuben Fine (1990), “It is clear that among analysts, a considerable caution has arisen concerning which patients are analyzable and which are not. This caution suggests that the image of `analyzability' has become an ideal fiction that few patients can live up to” (p.506).

Fenichel's (1945) discussion of analyzability is as sensible as any. Indeed, he is more sensible than most such discussions because he explicitly stated that these were general guidelines and that there are frequent examples of their being wrong. Let us examine his criteria:

1) Age between 16 and 40, although he notes successes younger and older do occur. Of course, even then child analysis was accepted as real; and one of Freud's most serious mistakes was the notion that patients over 40 were not analyzable. Most of us today have worked with patients in their 50's and 60's, and Riess (1992) reported working psychoanalytically with patients into their 80's and 90's.

2) Feeblemindedness is a contra-indication, although he notes that pseudo-feeblemindedness may disappear in analysis, and that some truly feeble minded patients may nonetheless benefit from analytic therapy. Sarason (1968) has reported analytic therapy being very helpful for people with low IQ's whose IQ does not change but whose life adjustment typically does, although analysts and therapists find them uninteresting, which is not the same as untreatable.

3) Unfavorable life situations. Here, Fenichel is different from Freud, who, as quoted by Rollo May (personal communication), said he would still analyze because "it is better to go down in fair fight with destiny than to be a neurotic." Of course, all of us know that what may seem like unalterable bad circumstances may be changeable when the patient becomes more effective.

4) Triviality. Some symptoms are not worth the trouble.

5) Disturbances of speech (although he mentions cases where the analysis was carried out in writing temporarily).

6) Secondary gains which make the symptoms too valuable to lose. Certainly this is still a valid concern. However, he here includes the fears of artists of losing their creativity. He suggests that more artists have been helped to overcome blocks to productivity than have had their art interfered with. I have worked with three artists and one writer, and the art played no role in the analysis -- it was the one area of their lives which was least problematic. They did not want nor need help with their art. Almost everything else in their lives was a disaster. Luckily for them, this analyst knows so little about art, that it would be not have been possible to intelligently intrude in their work. Nonetheless, their colleagues felt their art improved after analysis.

The one instance where an artist asked about art was a chronic paranoid schizophrenic artist who explicitly asked about the connection between "shitting and painting" after being in a phase where he could only paint in browns and blacks. After our discussion of gifts to his mother, he was able to paint in color and never asked another psychodynamic question about painting.

7) Urgency. It may be felt that relief is urgently required and that analytic technique is too slow. Of course, one deals with an emergency in the way most likely to succeed. Hospitalization is a resource. Medication, if you believe it is the best way to handle a crisis, is a resource. But you have to deal with the fact that medications have been oversold, are not nearly as effective as the manufacturers say, have more side-effects than the manufacturers say, and habituate. Moreover, they frequently represent the analysts' wish for omnipotence, or for relief from having to think about the meaning of difficult patients, or the need to devalue the analyst's abilities (from the analyst's childhood), or the wish to increase one's income (medicating patients is more lucrative than analyzing) or ingratiate oneself with biologically-oriented colleagues. The worst parody of the criteria of analyzability came from the chief psychologist of the most psychoanalytic hospital in Philadelphia in the '50's, a hospital where all the psychiatric residents were psychoanalytic candidates and most of the psychiatric staff were analysts, and almost all the in-patients and 50 outpatients were shocked every Monday, Wednesday, and Friday. Common in Philadelphia at that time were psychiatrists who were analysts for prestige with a shock practice for income. "It depends," I was told, "what you mean by psychoanalysis. We mean the treatment of the classical syndromes four or five times a week on a couch. What do you do with people who have something else wrong with them or who can't afford it. Either you don't treat them or you shock them. We shock them. We think it's kinder."

At a recent meeting, a psychoanalyst discussed obsessively the parameter of being more active and kindly to a patient who was anxious, but mentioned putting the patient on medication as if it were a trivial or automatic decision that did not have serious transference and countertransference implications. Incidentally, it has been reported that now 30% of training cases at one very prestigious medically dominated psychoanalytic institute are on medication. It is interesting that recovery in training cases at that institute has been reported to drop from 69% before 1987 to 39% from 1988-1992. It was a training analyst at that institute who described, in a colloquium on sexual addictions, treating an attractive young woman who was a gym instructor who slept with a different man each day. She would not lie on the couch, and this analyst could not figure out why. One day she came in and said she was ready for the couch. She was wearing a mini-skirt and the top buttons were unbuttoned exposing her breasts. She crossed her legs and she was wearing no underwear.

If this were my patient, I would have asked if she had seen the movie, "Basic Instinct." If, as was most likely, she said yes, I would have said, "I wonder if you are afraid you might kill me?" Other psychoanalytic colleagues have told me they would have asked her if she were afraid she might be a lesbian, or most tactfully, if there was something that she did not want me to find out about? Any of these questions might have furthered a psychoanalytic exploration. But this so-called psychoanalytic expert did none of these. He explained that he did not have sex with her (as if that were even an issue) and that he put her on Prozac. (The audience, but not the expert, noticed the similarity to her mother putting her on Ritalin to quiet her down.) When a member of the audience asked if he helped her get off Prozac, he said, "I find I can't stop Prozac," and most of the audience uncharitably concluded he was speaking of himself.

However, there are emergency psychotherapeutic interactions for crisis situations. One can argue that they are psychoanalytic therapy not psychoanalysis. The important thing is to do what is necessary to deal with the crisis, whether it is suicide, homicide, or some other serious issue. But political considerations have clouded the issues.

In the early days of psychoanalysis, Freud and the other psychoanalysts did whatever they thought was necessary for their patients' good. When the phrase "psychoanalytic therapy" was used, it referred to psychoanalysis as therapy. But in the 1940's, the work of Alexander and French (1946) and their co-workers in attempting to speed up psychoanalysis to make it more widely available, stirred up a hornet's nest of political strife. Psychoanalytic institutes split over whether these ideas should be considered a part of psychoanalysis, instead of simply asking of each idea whether it was good or bad, and what were the advantages and disadvantages of each technical procedure. It became important politically to prove that one was practicing "psychoanalysis" and not "psychoanalytic therapy." In the words of a psychoanalytic candidate in the 1950's, now a well-known analyst, "In our institute we practice psychoanalysis just exactly the way Freud did. As a matter of fact, I don't know what Freud was doing, Freud wasn't practicing analysis." I knew that I was interested in the work of Freud, but I was not sure I cared about psychoanalysis as it was then taught in the American Psychoanalytic Association.

When I was first supervised by Richard Sterba, I expected this ego-analyst and student of Freud's to object to my active intervention in a case. I had been treating the patient for several months before beginning the supervision. I recounted how in the third week of analysis, the patient casually mentioned for the first time that her husband beat her approximately once a week. He was a very athletic man, who was probably a borderline disorder with poor impulse control, despite a successful professional career. She said that she left him early in their marriage after a beating, but his father had said to her, "You don't know much about men, do you?"


"Well, men are just like that, and you have to accept it."

"Oh," she said, “I didn't know that."

She went back to her husband and stayed with him for the many years of their marriage.

As is usually the case with abused wives, it was the analyst rather than the patient who considered it a serious problem. I asked for more details. It became clear that he hit her under two circumstances. One, he told her he had done something wrong; she agreed with him and told him how wrong he was, and he hit her. Two, there was an issue between them, and she took the women's magazine view that they should talk it out until they agreed. Eventually to stop the conversation he hit her.

I pointed out the two situations and gave her direct advice: “When he says he's done something wrong, he isn't asking for moral instruction. In fact, you're not even giving him useful information, because he has already told you he's done something wrong. Instead of his feeling guilty, you become his punisher. No one likes to be punished and he hits you. When he tells you he's done something wrong, all you have to do is listen. That's all he wants, and that's all you need to do. When you and he disagree, you and your husband do not have to agree about anything. You don't have to tell him you don't agree if you don't think it's safe. If you think it's safe, tell him you disagree, because you will feel better. But when you have told him once, stop. You've given him the useful information. No one ever changes their mind in an argument. If he likes you, he'll take it into account when he has a chance to think about it. But immediately he will only defend his view, and eventually hit you to stop the conversation. So say it once and stop.”

After this one session she was never hit again. She said, "He doesn't hit me any more. But I know how to make him, and I feel like doing it."

"That is why you need analysis." (In fact, she had had a mother who only held her when she was hitting her.)

Her husband said, "All the fun has gone out of our marriage." Eventually, they divorced and she went on to a much better second husband.

But I expected this classical Viennese ego-analyst, teaching me ego analysis, to object to my activity. Surprisingly, he smiled approvingly. He well understood and approved of my feeling that first you keep your patient alive. Nor did I then understand that the rigid "blank screen" technique invented in New York City was a parody of the human psychoanalytic technique as it was practiced in Vienna.

Finally, according to Fenichel, the most important issue and the most difficult to assess is the lack of a reasonable and cooperative ego. This is still the most important issue. However, while it does relate to the probable course and difficulties, it should not be the criterion of analyzability.

Obviously, the best criterion is the response of the patient during the early stages of analysis. Freud's dictum that every analysis should start with a trial analysis is still a good one with the understanding that if things do not work out, the analyst will help the patient find someone more helpful. As I explain it, "Even a motivated patient and a competent analyst may not work well with each other. If the treatment isn't helpful, I will help you find someone else who will help. You never owe your therapy to a therapist, you only owe it to yourself." Rarely is that an issue, but it is usually important for the patient not to feel trapped. The only decision I require of a patient after an initial interview is whether to come back one more time, and after a second interview whether to come back a third time, although I usually tell them what I think they need by the second interview. After the third interview they need to make a decision about a trial of therapy. Equally important with not making the patient feel trapped, is not making the patient feel on trial, that you may decide they are not good enough. With that meaning the trial analysis itself becomes a trauma. It is my experience that the best prognosticator is the feeling of the analyst that he or she can help this person, irrespective of all other aspects of the case.

Freud himself said any investigation which took seriously the concepts of the unconscious, repression, resistance, and transference was psychoanalysis even if it differed from him in every other respect (which, of course, it would not if it took these concepts seriously). Some analysts cite Freud as requiring the Oedipus complex to be the central issue for it to be analysis, not noticing that he had written by 1931 that we must abandon the concept of the Oedipus Complex as the nuclear complex of the neuroses or redefine it to include all relations of the child to both parents (Freud, 1931, p. 226).

Luckily, the culture of psychoanalysis has changed. Largely as the result of Kernberg and Kohut in the United States, it is now fashionable to treat borderlines and other character disorders. Fairbairn was publishing his ideas in the 30's, but only in the 80's and 90's have Fairbairn and Guntrip become fashionable in the United States, with their routinely considering the issues involved with difficult patients. However, Charles Brenner (personal communication) has said it most succinctly: "The difference between neurotic, borderline, and psychotic is simply sick, sicker, sickest, but the mechanisms are the same."

Meanwhile, patients with simple, delimited neurotic symptoms have become scarce. As one candidate at a good institute in California said when she finished her training, “If a real neurotic ever comes in, I will treat the patient for nothing, because I would like once to treat a real neurotic.”

The more serious the problems the more likely the patient is to be willing to make the commitment of time, effort, and resources that analysis requires. The usual psychoanalytic patient has had three to five so-called treatments before applying for analysis, some of which are so destructive it is hard to believe anyone would call them treatment.

Change is produced in psychoanalysis and psychoanalytic therapy by insight and by internalization of the therapist. In the early days the emphasis was entirely on insight, "to make the unconscious conscious" and obviously this is still central to our efforts. Therefore it seemed like the only curative activity was interpretation and only interpretive activity was analysis. A kind of political correctness grew up, where analysts only talked about their interpretations, made fine-grained obsessive distinctions as to which comments were truly interpretations as opposed to other categories of speech, decided that only transference interpretations were mutative, were careful to prove they did not do anything other than interpret, became guilty about ordinary human kindness, and frequently lied about what they actually did with their patients in order to maintain the apparent "purity" of their work. At one well-known institute there were a series of candidates who had psychotic breaks probably because their training analysts at the time were so dedicated to the training analyses being thorough and pure that they were traumatic.

Let me present two important ideas which are probably new to you, at least in your reading, but will probably resonate with your experience as an analyst. They are particularly helpful to the analyst in working with borderline and psychotic patients, but they are also helpful in working with neurotics. First, it is absolutely essential for an analyst to be confused; and second, it is not "accurate empathy” which cures patients but it is our attempt at accurate empathy (whether or not we are successful) which cures patients.

It is almost universally accepted that empathy characterizes good analysts. Among psychoanalysts, Kohut (1977) most emphasized empathy as a centrally curative part of psychoanalysis; he also said failures of empathy in a basically empathic relationship were necessary for the process of cure. These failures helped the patient realize that analysis was not an adequate substitute for life.

And yet he was partially wrong. As part of my graduate class in the Psychotherapy of Psychosis twice a week I would treat a patient whom I did not know. Most graduate students in Clinical Psychology have been taught that severely disturbed people do not get better with psychoanalytic psychotherapy; inevitably, however, the students would observe that despite my mistakes the patients talked about the strange things described in my and other psychoanalytic papers, and the patients would greatly improve in front of them.

The late Norman Kagan, as part of his research, videotaped an initial segment of one such psychotherapy session with a 17 year old ambulatory schizophrenic, who was also alcoholic. Kagan immediately replayed the videotape with the patient, stopping it frequently, and asking the patient what was going on at that moment.

The thought processes of the patient were extremely complex and the patient described them in detail. Then Kagan asked: “Do you think Karon understands this?”

“Of course he does,” said the patient.

But the truth was that I did not understand. I understood as best I could the fragments that the patient had told me, but he had only revealed fragments of his conscious thought processes and I had no idea of the complexity of his thoughts and of most of what was going on.

What the patient was responding to as my understanding him was not my correct understanding of him, but my attempt to correctly understand him. I have since realized this is very general. We may or may not be capable of correctly understanding our patients, but we are capable of doing our damnedest to try to understand them. That is usually what they perceive and are responding to when they say and feel we understand them.

When in life do we have a bright concerned person really trying hard to understand us, no matter how confusing, terrifying, or obscure our life might be, or how much our defenses slow the process? Never--except in good analysis.

Of course, sometimes we do succeed at understanding. As the late Richard Sterba used to say to patients: “All I have to offer you is understanding, but that is really a great deal.”

Nonetheless, there is no way for an analyst to escape being confused. Early in my career, I was often intimidated by analysts and other professionals who seemed always to understand exactly what was going on with their patients. But their patients did not seem to benefit very much, and sometimes did very badly with such certainty.

The confused analyst is not only able to learn by not excluding possibilities, but provides a model for the patient; that being confused is tolerable. As I sometimes say to patients who complain that analysis is confusing: "Good. You are not sick because you are confused. You are sick because you are certain of things which are not true."

Valuing confusion does not mean valuing arbitrariness, or the narrative point of view, which, in its most extreme form, holds that any consistent narrative which the patient and the analyst accept is equally good, that objective truth is irrelevant, and unobtainable anyway. I strongly disagree. The patient has lived a life and wants to know what it was. It is frequently possible to reconstruct it. When I have reconstructed it wrongly, even when it was accepted by the patient, it was never helpful. But when all the material fell into place, and the patient improved, the reconstruction turned out to be true. In many cases, it has been possible to validate from external sources, such as family, the truth of such helpful reconstructions of the repressed past.

A supposedly incurable schizophrenic repeatedly hallucinated burning in Hell. When asked about the scar on his hand, he said,

“There's a story my mother tells. When I was 5 years old, we were in a store and she asked me where I had gotten the toy in my hand. I said the lady gave it to me. She asked the lady, who hadn't given it to me. She made me put it back, and she made me apologize to the lady. She then took me by the hand and we walked home.”

“How far was that?”

“Five blocks. And then we went upstairs to the third floor. And she turned on the gas burner, and held my hand in it to teach me not to steal. It left a scar. But it had no effect on me, because I don't remember it.”

“I have a different view. What you can't remember has the most effect on you. Most of us can only imagine burning in Hell, but you have actually been there.”

Insights sometimes, but not always, lead to immediate dramatic change: this one did. His hallucinations of burning in hell ceased permanently.

It may be, of course, that you cannot reconstruct the past. But even an uncertain past is bearable if you can share that uncertainty with an acceptant and tolerant other, and continue to think about it.

Central to change is the internalization of the therapist. The therapist is internalized into the super-ego so that the patient treats him or herself the way the therapist would, as opposed to the way the parents did. (Of course we recognize that one of the problems with the classical structural theory is that patients frequently have multiple superegos based on internalizing different people.) The therapist is internalized into the ego as a model for the self. The therapy relationship is internalized as a model for what a human relationship might be like. Sometimes analysts worry that the patient will internalize their defects. But, like an adolescent, the patient will eventually discard what is not useful. Further, analysts value patients who keep growing, who are different from them, and who can do better than they can; unlike the parents of most patients--who are often more like the father of Oedipus or the mother of Snow White.

Once internalization is taken seriously as a therapeutic process, it becomes obvious why it is important that the analyst be tolerant of the human condition. It is especially important for the analyst to be kind, so that the patient dares to be kind to him or herself, and even to others. The Stockholm Outcome of Psychotherapy and Psychoanalysis Project (Sandell, Blomberg, Lazar, Carlsson, Schubert, and Broberg, 1997) recently found that psychoanalysis and psychoanalytic therapy were very helpful, were more helpful than alternative treatments, and that psychoanalysis was more helpful than psychoanalytic psychotherapy; if you spend more time, you get more done. They found that psychoanalysts tended to have the same general values about psychotherapy which were different from non-psychoanalytic therapists. But on one dimension, analysts differed from each other: the importance of being neutral versus the importance of being kind. When doing psychoanalysis, analysts with both views were equally helpful. But when doing psychoanalytic therapy, those analysts who stressed neutrality were far less helpful. When seeing patients intensively, the patients usually come to accurately know even “neutral” analysts. But when doing psychotherapy, a neutral analyst can succeed in being “neutral” which may be perceived by the patient as rejection. The more regressed, ego-fragmented, or reality distorting the patient is, the more it is important for the reality of the analyst's kindness to be obvious.

The good analyst is stubborn. There is a part of you that just does not want to give up, no matter what it looks like. Rudolf Ekstein once described the ideal therapist for a psychotic child as "someone who knows as much as possible scientifically combined with an absolutely irrational belief that, no matter what, this child is going to make it." It's not a bad prescription for any analyst.

Indeed, the analyst must not only retain hope, but create hope in the patient. The patient has no reason to be hopeful, and depends on the analyst to create and maintain hope, while often consciously trying to prove how hopeless is the world and worthless is the analysis.

We help patients think about their lives and their contradictory feelings. We allow them to discover their own complexity, and that their feelings are a necessary part of rational thinking. What they cannot remember or think about is important; what is unconscious does not change, but has the most control. What they can think about, they can control. With our help, they reconstruct their lives and their traumas including their fantasies, and they learn the defenses they use. They learn about transference and its ubiquity; and they learn to use their transferences as sources of information. Indeed, they learn to use all of their problems as sources of information that will make their lives better.

In dealing with the more difficult patient, it is important to do what is necessary to cope with crises. You must present as much structure as the patient's defenses demand. You must keep the patient's terror and acting out within tolerable bounds. Specific technical suggestions can be found in The Psychotherapy of Schizophrenia (Karon & VandenBos, 1981) which are useful not only with schizophrenics, but also with borderlines and other difficult patients.

It was Hedda Bolgar who first suggested to me that the Odyssey was a good metaphor for analysis. It is a scary journey of self-discovery which is bearable because the analyst is there.

Psychotic patients are often the most dramatic examples. A patient, termed “incurably schizophrenic” was brought to my office (over the staff's objections) after his wife, on my advice, refused to permit electro-convulsive therapy and had withdrawn him from the hospital. The psychiatric staff told her she was killing him. He was not eating, he was not sleeping, and he was continuously hallucinating.

He was from a middle-class family and had considered himself lucky to have had such good parents, particularly such a good mother. However, even before his psychotic break, he could not remember his childhood before the second year of high school; he did not think this was abnormal.

I immediately stopped all medications and started real treatment -- seven days the first week, six the second, and so on, until a regular three day a week schedule. His wife and friends of the family took turns babysitting with him for the first month or two.

Since not eating will kill you in 30 days, I did what is recommended in Karon & VandenBos (1981). The second session was at 7:00 a.m at an all-night restaurant. He said, "I can't go in there. They'll think I'm crazy"

"No. They'll think you're drunk."

"I'll vomit"

"You think you're the first drunk who threw up here tonight?"

I ate and we talked about eating. Four days later he was eating and we moved back to my office.

Six months later he was working at an intellectually demanding job. A year later I could say to him:

"Anyone can go crazy under enough stress, but under the normal stresses of everyday life, you will never be psychotic again."

He said, "This is better than I've ever been, better than what I used to call normality, but if you think living like this is good enough for me, you're crazy."

The process eventually became more classical-looking. We moved to a couch and my sitting behind him when it seemed more useful. Patients usually get angry at that change. As another patient said, "You used to tell me what to do. Why won't you now?"

"Your judgement used to be terrible. You needed someone to tell you what to do. Now your judgement is as good as mine, and sometimes better. It would be silly for me to tell you what to do." Patients typically get angry but keep growing.

He saw me for 14 years. He kept raising the stakes. In his third year, he startled me by saying, "I have a book to write and I can't. Is that something you can help me with?"

Somewhat dubiously, I said, "People do go into analysis for writer's block." We spent most of a year on it, and he wrote that book. He has written several since. Others in his field have told me that his professional reputation is based on that first book, now considered a classic in his field. Obviously it was well worth a year's analytic work. He knew that, even if I did not.

In his pre-psychotic period, he had never eaten a meal without nausea. During treatment after a trip to France, he recounted with tears in his eyes, "I can't tell you what French cooking is like. There's nothing like it in the United States."

We went through psychosomatic problems, and then marital problems:

"I could leave her. It would be easy. There are bright, attractive interesting women out there, but it would devastate her. When I needed her, she saved my life. The doctors said shock him, my family said shock him, people in your department said shock him, but she had the courage to defy them and see that I got real treatment. And I just can't do that to her."

He described his need for the last two years of treatment: "I have a teen-age son. When he was a kid, he had a psychotic father, and that was a hell of a thing to do to a kid. And I need help in undoing the harm I did him."

He is now internationally renowned in his field. He recently sent me a magazine article about a prestigious award he received for his scholarship. His therapy did not make him a bright man nor a kind man, but it did keep his brightness and kindness from being destroyed. It did allow him to feel safe, perceive and think realistically and creatively, and use his intelligence and kindness to make his own and other people lives more interesting. That is psychoanalysis.

But difficult cases are not just suitable for experienced analysts. One of our new candidates, who had wavered as to whether to become a candidate at our institute, the Michigan Psychoanalytic Council, or a more traditional institute affiliated with the American Psychoanalytic, was startled by the difficulty of the cases presented at a case conference by our candidates.

"Doesn't anybody in this group treat easy cases?"

"Probably not," I replied.

One of the advanced candidates, Pat Marciniak (1997), had described a patient who had been on five or more simultaneous medications for eight years after a suicide attempt. He had been diagnosed as manic-depressive, borderline schizophrenic, and paranoid, by the medicating psychiatrist. At no time had psychotherapy been suggested. He was unable to work but was informed by his psychiatrist that she did not want to continue to treat him. However, when she discovered he had been referred for, and was receiving psychoanalytic treatment from a social worker, she called him to tell him it was dangerous, to stop and to double his medication dosages. Luckily, Marciniak had already arranged for him to see a rational medicating psychiatrist, who not only approved of psychoanalysis but was himself a candidate at another institute. This psychiatrist helped him by prescribing more rationally, tapering and changing the medications appropriately, and, when the patient requested it, helping the patient stop the medication entirely.

While there were serious physical withdrawal effects, the patient's ability to think clearly greatly improved when he stopped the medication while continuing his analysis. The anti-therapeutic intervention of his previous psychiatrist precipitated a crisis, but it was resolved when it was related to his guilt about no longer conforming to his mother's values.

When he started treatment three times per week, he said that he feared a race riot; he had acquired an assault rifle and four thousand rounds of ammunition. He hated almost all of the men he worked with and had been fantasizing killing fourteen of them and then himself. He felt they hated him. He experienced wide mood swings, sometimes was extremely physically aggressive, talked cruelly and grandiosely, and sometimes secluded himself in a basement bedroom refusing to talk to anyone. He sometimes slept with a loaded handgun on his chest with rifles nearby. He also reported washing his hands 20 or more times a day.

After eight months of treatment the patient decided he was ready to go back to work. The first day on the job the other men teased him cruelly about being sick and he described being enraged and humiliated. He called Marciniak that evening, swearing he would go to work in the morning armed, shoot as many as possible and then shoot himself. He was in excruciating emotional pain, panicked and enraged. Marciniak dealt with the emergency by talking to him for two hours that evening. He was exhausted, but refused to change his plans for 8:00 a.m. the next morning.

Marciniak tried to talk to her analytic supervisor, who was out of town, but did talk to another supervising analyst in our group, Ellen VandeMark, who was helpful, but basically allowed the trainee to use her excellent clinical judgement. Marciniak met with the patient early in the morning for a session before work. By the time the hour was over, he agreed these men were not worth dying for, and that as long as he was alive there was the possibility things would get better. He agreed to go home and not go to work. Marciniak described him as being "exhausted from rage and lack of sleep," and herself as "exhausted from sheer fear."

It was at this point the patient decided to stop medication and both homicidal and suicidal crises have ceased since then. He has also described a dramatic clearing of his ability to think. There were severe physical symptoms connected with the discontinuation of Xanax, but they were handled medically. However, the patient understandably resented the emergency room physician's offer to prescribe more Xanax "to make things easier."

With understandable anxiety, the candidate had weathered the crisis and got the patient not to go back to work until he could do so without murdering anyone. The analysis eventually enabled him to go back to work, and then move to a better job (with great difficulty and guilt). Needless to say, his relations with his wife and children also improved enormously. The changes were slow, the treatment was intense and difficult, and the improvement real. The transference bases of his terror, shame, and hatred were examined.

The treatment continues with difficulty, but now his problem is not murder or suicide, but vocational success or failure, marital success or failure, and the ability to come to therapy when things are going badly. He is still an extremely difficult “unanalyzable” patient by many standards, but the work that has already been done is magnificent and it will only take years of hard work for him to go the rest of the way. This is psychoanalysis at its best.


Alexander, F., & French, T. M. (1946). Psychoanalytic Therapy. New York: Ronald Press.

Fenichel, O. (1945). The Psychoanalytic Theory of Neurosis. New York: Norton.

Fine, R. (1990). The History of Psychoanalysis. Northvale, NJ: Jason Aronson.

Freud, S. (1931). Female sexuality. S. E., 21, 225-243. London: Hogarth.

Glick, R., Eagle, P., Luber, B, & Roose, S. (1996). The fate of training cases.
International Journal of Psycho-Analysis, 77, 803-812.

Karon, B. P., & VandenBos, G. R. (1981). Psychotherapy of Schizophrenia:
The Treatment of Choice. New York: Aronson.

Kohut, H. (1977). The Restoration of the Self. New York: International Universities Press.

Marciniak, P. M. (1997). Facets of kindness: Grappling with self-deception in the analytic hour. Unpublished manuscript.

Riess, B. F. (1992). Some thoughts and material on age-related psychoanalysis of the aged. Special Issue: Psychoanalysis of the mid -life and older patient. Psychoanalysis & Psychotherapy, 10(1), 17-32.

Sarason, S. (1968). Psychological Problems in Mental Deficiency. New York: Harper & Row.

Sandell, R., Blomberg, J., Lazar, A., Carlsson, J., Schubert, J., & Broberg, J. (1997, June) Findings of the Stockholm Outcome of Psychotherapy and Psychoanalysis study. Panel at the annual meeting of the Society for Psychotherapy Research, Geilo, Norway.

Versions of this paper were presented as the 1998 George Klein Memorial Lecture, Psychologists Interested in the Study of Psychoanalysis, San Francisco, CA, August 14, 1998, and as an invited lecture at the 1997 International Federation for Psychoanalytic Training, Ann Arbor, MI, October 5, 1997, and at the Triennial Retreat of the Michigan Psychoanalytic Council, Traverse City, MI, June 6, 1998.