I am grateful for this opportunity to state some of my views informally and more personally than is usually appropriate for a journal article. Originally, my friend and colleague Irwin Hoffman was to interview me, but I preferred the more organized form of an essay. For some time Irwin and I have been in continuing dialogue; his hand is here as well as mine. That doesn't mean, though, that he agrees with everything I say.
I was introduced to analysis in college by way of Freud's first set of Introductory Lectures, which fascinated me. I resolved to become a psychoanalyst (to find help for myself, of course, even if I did not know that then). I went to medical school, the obvious route to becoming a psychoanalyst at that time. My analysis as a candidate was at the Menninger Clinic, where I had the great good fortune to become a protegé and later a colleague and friend of David Rapaport. It is noteworthy, and perhaps peculiar, about our field, which emphasizes history, that so great and relatively recently deceased a pioneer in psychoanalytic theory is so little referred to any more.
I have had experience in private practice (although never full time), psychoanalytic hospital practice, and medical school positions. I was given an NIMH lifetime research scientist position after 22 years in the field. Such lifetime positions are no longer available. They were tied to the institution, so I lost mine when I left Downstate Medical Center in Brooklyn after five years. I have now been at the University of Illinois Medical Center in Chicago for almost 20 years and am Emeritus.
My analytic training at Menninger's was classical, but over the years, I have changed my mind about a number of classical tenets. The most important single theoretical change is my abandonment of the classical Freudian metapsychology for a hermeneutic one. It was through Rapaport that I became involved in Freudian metapsychology, a subject that I realize in retrospect I do not find personally congenial. I have likened it in its worst forms to hair-splitting Talmudic disquisitions. I am sometimes asked how I could change my views so radically as now to be in opposition to Rapaport's. My answer is that he systematically pushed metapsychology to the point where its inappropriateness to psychoanalytic data became brutally clear to several of us who had worked with him. His tragically early death at 49 prevented him from moving on to another intellectual stage. I am convinced that he would have done so because, however strong his fervor for what he believed, he had an open mind.
My position is that psychoanalysis is not a natural science in the sense that physics, chemistry, and biology are but, rather, is a science of human meanings. I call it a hermeneutic science and I am told that is a contradiction in terms, that hermeneutics is a flight from the obligation of an empirical science to validate its propositions. I know that hermeneutic psychoanalysis is usually referred to as a discipline rather than a science, but I believe that characterization avoids the issue of its epistemological status. The very fact that in speaking of hermeneutic science I use what is generally considered to be an oxymoron shows my ignorance of philosophical hermeneutics, but I know no other term that refers to meaning and interpretation and that I can oppose to Freudian metapsychology. If the conclusion of philosophical hermeneutics is that validation is impossible or unnecessary, I can only say that that conclusion does not hold for psychoanalytic hermeneutics. A related conclusion is that I do not understand why a reason cannot be a cause, although again that will likely be considered philosophically naive.
I see the natural science position as applied to psychoanalysis as aborting the search for meaningful explanation by resting on pseudoexplanation. With hermeneutics the way must be opened to a reconsideration of the psychoanalytic theory of motivation, which in Freudian metapsychology is frozen in the dual instinct discharge theory.
In addition to the fear that the hermeneutic position makes psychoanalysis an art rather than a science, a main obstacle to the acceptance of the hermeneutic position seems to me to be the fear that such a position ignores the centrality of a human being as a biological organism and with that the abandonment of a unique contribution of psychoanalysis to human knowledge, the discovery of unconscious bodily fantasies. I have already said I do not believe the hermeneutic position is antiscientific, nor do I believe that the hermeneutic position need abandon the body. In brief, in a scientific hermeneutic psychoanalysis, the body retains an important centrality, but it does so in terms of its meanings to the person rather than as material body parts. To say this is to do no more than restate the principle of psychic reality. That conclusion still leaves open the question of the role of bodily fantasies and drives in motivation, since they could still be the basis of motivation in terms of their meaning. The hermeneutic position is therefore neutral with regard to the role of bodily fantasies and drives in motivation. Important though these bodily fantasies and drives are, I believe that their overall role in motivation is still unclear. Their role will differ in different people, in different stages of development, and in different stages of an analysis.
The theory of motivation remains a major weak point in psychoanalysis. It ranges from the somatic, as in sexual discharge, to the seemingly psychologically autonomous motive of a striving for self-cohesion. I am impressed by the literature that proposes several sets of motivations in hierarchical relationship, without that hierarchy necessarily being the same for all people or even for the same person in different phases of development (see Lichtenberg, 1989).
The implication for therapy is obvious. The therapist must give up the assumption that a particular set of motives will necessarily always rule.
My present view, then, is that meanings can be literal or metaphorical, that both literal and metaphorical meanings can be either bodily or nonbodily, that any one of these four meanings—literal bodily, metaphorical bodily, literal nonbodily, and metaphorical nonbodily—can serve as a defense against any other and can be affectively superordinate at any particular time.
So how is psychoanalysis different from such a hermeneutic discipline as literary criticism, where it is granted that while many interpretations seem clearly improbable, more than one seems clearly plausible, and none is necessarily the singularly correct one? The difference is fundamental and, to me, obvious. In literary criticism the text does not respond, at least not in the way the patient as “text” does. In analysis the patient does
respond, so there must be a way—although that way is doubtless not the way of positivist science—in which hypotheses can be checked.
The most important single clinical change in my views—arrived at through my several experiences as a patient and as a therapist in psychoanalysis and psychotherapy—is that the role of the analyst as a person and as a participant in the therapeutic interaction cannot be overemphasized. The idea that the analyst participates in a therapeutic alliance, or even shows an occasional countertransference, is but a pallid recognition of this role. In any analytic process worthy of the name there is emotionally significant continuous interaction between analyst and analysand. Those who stress this point often make it by saying that the analytic situation must be understood as a two-person psychology rather than the one-person psychology of classical analysis. The common rejoinder is that classical analysis is incorrectly described as a one-person psychology, indeed that the stress it lays on object relations means it has always been a two-person psychology. It is true that classical clinical theory makes much of object relations—Freud said that psychoanalysis could appropriately be called a social psychology—but this emphasis is lost in viewing the analytic situation. There the ideal is the neutral, “detached,” as one of my teachers put it, analyst. It seems to me that the idea that an analyst can be detached or neutral is absurd in the light of what we have learned about the ubiquity of conflicted unconscious motivation in human beings. To think otherwise is to reject what was surely one of Freud's basic discoveries. To express that discovery in the idiom of one of the great revolutions of our own time: “Man/woman is not master/mistress in his/her own house.” The best we can do is to remain alert to our contribution to the transferences. Since we cannot be conscious of what we are unconscious, evidences of our contribution will be indirect. One of the best sources of clues to that contribution is implicit references to the patient's experience of the relationship in his associations.
An interpretation purports to reveal the patient's psychic reality, not “objective” reality. That is old hat. But an interpretation is also an expression of the analyst's psychic reality. It is remarkable that though analytic theory accepts the concept of psychic reality for the patient, it often does not do so for the analyst—at least for his functioning in the analytic situation—except in the abortive form of the countertransference.
I believe that the human being is a social animal from the beginning and that the classical one-person view of the analytic situation is wrong. Classical analysis sees the analytic situation as a closed system with the participants ideally having relatively little interpersonal influence on each other, although the analyst does maintain a humane interest in his patient. In that view, countertransference is a malfunction. A two-person psychology sees patient and analyst as an open system, in an interaction importantly meaningful to both parties. The paradigm of a two-person psychology, however, will not suffice for psychoanalysis. Psychoanalysis requires both one-person and two-person psychologies (Ghent, 1989). While each person in an interpersonal interaction is continually influencing the other, at the same time each starts with a fixed position. Furthermore, important though object relations are, one is more than one's object relations. Psychoanalysis must reach an integration of the one-person and two-person paradigms. “Paradigm” may be considered too grandiose a term for this change. Whether or not it is depends not on the substantive issues involved but on the influence of traditions, priority claims, and the like.
“My position is that psychoanalysis is not a natural science in the sense that physics, chemistry, and biology are but, rather, is a science of human meanings.”
The classical Freudian position sees two-person issues as subordinate to and arising from a one-person base. The opposite position sees apparent one-person issues arising from and embedded in a two-person psychology. Sullivan (1950) sometimes seemed to take this radical position. Or, as I think the most likely, the two psychologies develop in parallel. The concept that both primary and secondary processes have innate roots may be more than only an analogy here. Whatever their hierarchical and developmental relationship, that the psychologies are intricately intertwined seems obvious.
My emphasis on the centrality of the interaction in the here and now has led some to consider me as solely an interpersonalist and thus to have abandoned the classical Freudian position. I have certainly moved from my education as a classical Freudian to a greater emphasis on the interpersonal point of view. I now think much more in terms of an integration of the two approaches—classical and interpersonal—but I suppose my reputation as an interpersonalist is sealed. One of my purposes in writing these remarks is, of course, to change that perception if possible. It is unfortunate that in some writings, especially self-psychological writings, interpersonal is incorrectly regarded as synonymous with manifest, conscious, and “social psychological.” Interpersonal need not have such meanings. It can mean, and in the writings of many interpersonalists and object relationists does mean, the interpersonal interaction as intrapsychically experienced. The interpersonal-intrapsychic dichotomy is false. Emphasis on the interpersonal usually implies not a rejection of the intrapsychic but, rather, a rejection of the psyche as a closed system. Nor is there any incompatibility between an intrapsychic-interpersonal psychology and an important role for the body in the psyche.
I believe, by the way, that Kohutian self psychology, which sometimes strongly emphasizes that it is an intrapsychic psychology (Goldberg, 1986, p. 388), is also an interpersonal psychology—and even more so in practice than in theory (Bacal and Newman, 1990)—but in a limited sector, that of the selfobject transferences. Self psychology emphasizes empathy because empathy is one of the bridges between the analyst's psychic reality and the patient's psychic reality. I speculate that self psychology's claim to being an intrapsychic psychology is to refute the charge that it is not genuinely a psychoanalysis. My own view is that where self psychology departs from classical analysis is in its relegating the body to a defensive role, insofar as it deals with it at all. Whether self or drive should be the central organizing concept of psychoanalysis is probably the most crucial question facing psychoanalytic theory and practice today. Any comprehensive psychoanalytic theory must deal with the relationship between self and drive. An analogous controversy deals with the relationship between the oedipal and the preoedipal.
I have much to criticize about Sullivan too. I have pointed out (Gill, 1983) that in his stance of participant-observation Sullivan saw himself as much more the observer than the participant. He attributed to himself a degree of objectivity that is inconsistent with the stance of constructivism for the analyst as well as the patient, and, even more, he did not attempt to analyze the interaction implied by participation. Of course, since he regarded his participation as “objective,” he would not be likely to analyze
interaction. What I might call neo-Sullivanians have advanced far beyond his position in this regard as well as in the recognition of the importance of the role of the body.
My early enthusiasm for the centrality of the analysis of the transference in the here and now as part of the two-person paradigm led me to overemphasize it at first, even though from my earliest writing on transference I said that genetic and contemporary extratransference work is important, indeed essential, in the therapeutic process. I still think that some things are more important than others and that interpretation of the transference in the here and now should get priority and primacy.
I do recognize that for the analyst always to interpret transference in the here and now as soon as he believes he has become aware of it is a mistake. Considerations of tact and timing are crucial. The analyst must take into account the patient's capacity to hear and make use of an idea the analyst has about the transference, and the analyst must himself be ready to grapple with the issue. The analytic situation is ineluctably an interactive one. It is a series of exchanges. The analyst is, therefore, always engaged in suggestion. If the central feature of analysis is the analysis of suggestion, as I believe it to be—and that does not mean that I am discounting extratransference and genetic interpretations—the analyst committed to that view may become paralyzed by an effort to avoid suggestion. What is required, therefore, is that he permit himself a degree of spontaneity yet recognize that he will be suggesting he knows not what, and only later and only to a greater or lesser degree be able to stand back and take stock.
Nevertheless, the therapeutic situation is not an egalitarian one. In the past, in my enthusiasm for emphasis on the analyst's contribution, I have sometimes written as though the analytic situation were indeed an egalitarian one. The analytic situation is asymmetrical, perhaps most obviously in that the analyst usually tells little or nothing of what goes on in his mind while the patient is enjoined to tell all that is going on in his mind. It is the patient who is to be analyzed, not the analyst. Nevertheless, the analyst must remember that his ideas are expressions of his psychic reality, not of a neutral, impersonal, “objective” reality. The analysis of the patient's transference neurosis should be accompanied by a self-analysis of the analyst's continuous “countertransference neurosis” (Racker, 1968).
In the light of the argument of Freud (and many others after him) that the charge that analysis produces its effects by suggestion is refuted by the analysis of the transference, a word about the relationship between suggestion and transference is in order. Transference is in part the response to suggestion. This is true for both senses in which Freud used the term transference: as influenceability (as used in the “unobjectionable positive transference” and the “transference neuroses”) and as distortion.
In redefining transference (and countertransference) as contributed to by both participants, I have objected to referring to transference as a distortion. Only if distortion is meant to connote selectivity from among various possibilities and emphases is the term acceptable, and then only with the insistence that it applies to the therapist as well as to the patient. My formulation that patient and therapist codetermine the transference, if taken to mean that they contribute equally, seems to be at odds with my describing the analytic situation as asymmetrical and appears to justify the complaint that my emphasis on the contribution of the analyst means that I have lost sight of who is being analyzed, or as one critic put it, I am courting a folie a deux. Strong medicine has its dangers. One can only do one's best to guard against them. The analyst must be ever alert to the possibility that in any individual instance he may be “reading” more selectively than the patient is. In any case, it is always a relative matter. There is no nonselective reading.
“I now think much more in terms of an integration of the two approaches—classical and interpersonal—but I suppose my reputation as an interpersonalist is sealed.”
What about those situations in which the patient seems to be egregiously misreading the analyst? Is it not true that the therapist's psychic reality is more “objective” than the patient's? Probably so in the “average expectable” analytic situation, but that is a statistical statement. It is true that one of the ways in which the analyst is the expert in the situation is that he is convinced that everyone, including himself, operates out of psychic reality. Furthermore, he is familiar with patterns of wish and denial and presumably with his own propensities. Again, there is a danger: The analyst may make use of his conviction of selectivity to rationalize an obsessional inability to take a stand.
Many people are disturbed by the fear that a thoroughgoing acceptance of the principle of psychic reality means that any ascribed meaning or “narrative” is as “good” as any other, or, as the expression has it, “anything goes.” It does not. It is true that no single meaning is correct and all others false. The most comprehensive, consistent, coherent meaning available at the time is the “best,” but, even so, there may be others that contain equally important truths too.
I have suggested that the transference be spoken of as “plausible” rather than “distorted,” but the word plausible worries me some because of its possible connotations. If the analyst tells the patient his idea is plausible, the patient may experience this as meaning that the analyst condescendingly believes the patient's idea is understandable but nonetheless false. By plausible I mean that the analyst sincerely believes that the patient has a tenable perspective, even if it differs from his own.
It is remarkably easy to slip up on this issue of objective reality versus psychic reality. I once expressed the opinion that the classical analytic situation infantilizes the patient. Schafer (1985) questioned that view. He considered it an illustration of what he characterized as my simplified egalitarianism, which he cogently criticized. He said that if the patient experiences it that way, that is a matter of the patient's psychic reality. He was partly right. I was making a statistical statement as if it were universally and objectively true. But he was also partly wrong. Although analysis is surely much more often experienced as infantilizing than not, that is not solely the patient's doing. If the patient does experience it that way, the analyst has contributed to that experience by setting up a situation designed to produce regression. In that sense, the analytic situation is indeed more “egalitarian” than is usually recognized.
The issue of regression remains a most troublesome one for me. The classical position is that the standard analytic situation is necessary for regression to the infantile neurosis and that regression to that neurosis and its resolution are necessary for a “structural” change to take place. But there seems a good deal of evidence that with less frequent sessions, the sitting up position, and at least comparable duration, far-reaching, if not equal, changes can be brought about. How much of the regression in a standard analysis is unnecessary, even iatrogenic? Is it possible that the limits of analyzability are narrowed by virtue of the fact that only comparatively healthy people can withstand
the enforced and perhaps unnecessary regression of a classical analysis?
The question of the relationship between psychoanalysis and psychotherapy has been a continuing one for me. In 1954 I wrote a paper espousing a puristic concept of psychoanalysis. I defined it as the development of a regressive transference neurosis with a neutral analyst and the resolution of that neurosis by techniques of interpretation alone. I suggested that there is a continuum of psychoanalytically oriented psychotherapy more and more closely approximating psychoanalysis. The paper was widely quoted and is still often cited.
But in 1984, consistent with my changed views on the nature of the psychoanalytic situation, I wrote a radical revision of my views. Although I am of the opinion that I was separating psychotherapy and psychoanalysis even more sharply than I had before, Wallerstein (1989, p. 582) has, on the contrary, seen me as blurring them where I had formerly separated them. I believe some clarity can be introduced into our disagreement if I distinguish between a “mode” and a “modality” (see Oremland, in press). Interpretation is a mode; psychoanalysis is a modality. In psychoanalysis, modes other than interpretation, like suggestion and abreaction, are used. It is generally agreed that these other modes are inevitable, even in an analysis, but they are regarded as contaminants. I say, on the contrary, that they appear because any psychological therapy includes a reciprocal affective interpersonal interaction. These differing reasons for why modes other than interpretation are ubiquitous lead to different consequences for theory and technique. A major one is the recognition that an interpretation is, at the same time, a suggestion too.
I argue that the definitive distinction between psychoanalysis and psychotherapy is that the theoretically ideal goal of psychoanalysis is ultimately to make suggestion explicit, whereas psychotherapy pursues this goal only partially or not at all. I recognize that even theoretically this is a misleading formulation because it implies that all the suggestion in an analysis could be made explicit. But that cannot be. The interpretation of a suggestion has its own suggestive meaning, and so on into an infinite regress. Even pragmatically, at any particular time and for any particular issue that goal must be postponed or even dropped because the analyst considers it to be inconsistent with the state of the resistance. Nor should it be forgotten that “suggestion” connotes influence that is unrecognized, whether by patient, or analyst, or both. Interpersonal interaction is not synonymous with suggestion.
I now argue that psychoanalytic technique can be employed at a lesser frequency, with the patient sitting up, and for more limited periods of time and still be psychoanalytic. If I am right, the conclusion seems to me important for the practice of psychoanalysis, which nowadays is likely to include a good deal of work at less frequent intervals, the patient sitting up rather than reclining, and shorter treatment duration. I fear that the classical attitude of the analyst, with all its drawbacks, may be carried over to the “psychotherapy,” and that the tendency to leave suggestion unanalyzed in the “psychotherapy” will carry over to the analyst's psychoanalyses. A consistent analytic technique, therefore—by which I principally mean the analysis of transference, irrespective of frequency or position—should improve both the psychotherapy and the psychoanalysis being done. I grant the cogency of the question as to how much the results of such a limited therapy, when successful, are a “transference cure.” But what are the results of such a therapy when carried out over a longer period of time, perhaps even as long as in a usual analysis? The answer lies in research, not pronouncements, but research not only into this question but also into the role of interpersonal influences in even an apparently successful analysis. These and related questions are taken up in a forthcoming monograph by Jerome Oremland (in press), to which I have contributed a critical evaluation.
“Whether self or drive should be the central organizing concept of psychoanalysis is probably the most crucial question facing psychoanalytic theory and practice today.”
Psychoanalysis has yet to achieve an appropriate balance between the innate and the experiential. Freud's overreaction to his discovery that many of the tales of seduction he had been told were false dogs classical psychoanalysis to this day. In classical analysis the innate overshadows the experiential, while in some other schools of analysis the experiential overshadows the innate. Psychic reality is now essentially defined as the innate whereas it should mean a constructivist integration of the given with the experienced. Only when it is recognized that this is true for analyst as well as for patient, that each constructs the other's psychic reality, that the interpersonal influence of each on the other is the experiential component in forming their respective psychic realities in the analytic situation, and that analyst and analysand can only “negotiate” about their respective psychic realities because neither can claim to know the “objective truth,” will psychoanalysis become the hermeneutic-constructivist science I believe it can and should be. This is not to say that the relationship between the innate and the experiential is always the same. Freud gave us the model for this variable relationship in his concept of the complemental series.
An important, growing body of literature, to which Hoffman (1987, 1991) is a major contributor, proposes that the change in paradigm can be conceptualized as from a positivistic to a constructionist one, in keeping with major movements in other branches of science as well as the philosophy of science.
There are unclear conceptual issues in the knot of such ideas as a two-person psychology, hermeneutics, psychic reality, unverbalized experience, and the ambiguity of language. For example, does a two-person psychology of the analytic situation, that is that analyst and analysand continuously and reciprocally influence each other, necessarily entail a constructionist position? Does hermeneutics do so? Is constructionism the master concept superordinate to the others?
Clarification of issues such as these continues to be a crucial aspect of psychoanalytic work, just as conceptual clarification has always been. Such clarification has been, and continues to be, done by thoughtful, open-minded clinicians, fed by their clinical experience. Whether such work should be called research and how it relates to the more usual concept of research are not easy questions. It has been said that clinical research can be done only by competent analysts. But with the differences of opinion among us, who is to say who is competent? In any case, it seems clear that conceptual clarity is a necessary basis for research in the sense of the systematic testing of hypotheses.
We need research that is not dependent for its evaluation on the reputation or charisma of the researcher. I have long argued that a major reason for the decline of the prestige of psychoanalysis is its failure to engage in the systematic research on its propositions which any body of knowledge worthy of the name science should do. I do not have the temperament or the skill to stick with systematic research. What I consider the most important contribution I have nevertheless made in such research is the development, together with Irwin Hoffman (Gill and Hoffman, 1982),
of a scheme for coding transference. Its reliability and validity remain to be studied. Under the pressure of third-party payers, momentum has grown to study the efficacy of analysis systematically. To study efficacy without studying process makes no sense. Methods for both are primitive. The development of such methods should be the first order of business. It will be years before any finding in efficacy or process will reach the point where its intrinsic validity will perforce influence practice (or third-party payers!). And, even then, whatever general rules are discovered, they will be applicable only in the context of the particular idiosyncratic analyst-analysand couple. I also believe that the research methods required for our hermeneutic science will be quite different from those used successfully in the natural sciences.
What will happen to analysis in the meantime is unpredictable. But it will never die. It contains too much truth for that to happen. At least it will not die as long as there is democracy anywhere on earth. Totalitarian systems cannot abide analysis because its ultimate goals are to enhance individual human freedom and reciprocally equal human relationships.
George Klein (1976, p. 137) suggested that analysis may continue to be practiced in its classical form only for training and research. Rapaport had the fantasy that there would be a kind of dark ages for analysis and then a renaissance when it would be rediscovered. We may be approaching such a dark age now. Powerful social forces are aligned against our non-“cost effective” psychoanalysis. Biological psychiatry, valuable though it is, and placebo superficial psychotherapies, helpful though they may be, threaten to swamp psychoanalysis. Analysis does not have the protection against these forces it would have if it had a home base as a university basic science with the attendant research structure. I have no illusions about the complexities and difficulties involved in establishing such an academic home base.
I am also convinced that the crucial area where the research must take place is in the analytic situation itself and by analytically sophisticated researchers. And the analysis will have to be recorded, on audiotape or even on videotape. To believe that any other form of reporting who said what is reliable is once again to accept the fiction of an objective analyst. I have no doubt that a recorded analysis can be a bona fide analysis, even though for particular patients or particular analysts the recording may be an insuperable obstacle. Not everything yields to analysis.
I refer to Rapaport (1960) yet again: “Salvation will not come from afar.” It will not come from academic psychology, biology, neurophysiology, or whatever. George Klein (1976) said so too. What we desperately need are career tracks for systematic researchers in the psychoanalytic situation. Perhaps if such opportunities became available, rewards other than those of practice will become more competitive.