Response to Zepf and Gerlach

I appreciate the interest and dedicated study of Siegfried Zepf and Alf Gerlach in their commentary (JAPA 61/4, pp. 771–786) on my paper “Suicide Prevention for Psychoanalytic Institutes and Societies” (JAPA 60/4, pp. 707–719). Their agreement with major conclusions and recommendations formulated in my paper of course gives me satisfaction. In fact, to my reading, much of what they disagree with is based on a misunderstanding of my position. To begin with, they agree with doing away with the privileged status of the training analysis system; they agree with a careful selection of supervisors in terms of realistic assessment of their interests and capabilities for that function; and, importantly, with the fact that in essence the “training analysis” does not differ in quality from the therapeutic analysis. They also agree that my proposals would counteract the “corrupting narcissistic grandiosity induced by the privileged status currently accorded the ‘training analyst’ as superior psychoanalyst . . .” (Kernberg 2012, p. 714; Zepf and Gerlach 2013, p. 773). They also agree with my proposal that certified psychoanalysts, all of whom are authorized to psychoanalyze candidates, do not need to be members of an institute nor of the society to which the institute belongs. However, I question their expansion of this view to suggest that analysts should not be permitted to give lectures in the teaching of candidates they analyze: in my view, this would make analytic training in smaller communities impossible.

The authors also agree with my critique of the “ecumenical tolerance” of contradictory theoretical and clinical approaches and the need to carry out scientific research to determine the truth value of incompatible or contradictory approaches (Zepf and Gerlach 2013, p. 774). Here, however, the authors make a fundamental error in their reading of my paper. They suggest that my proposal for tolerance of different theories and approaches and an open exploration of them all stands in contradiction to my recommendation that incompatibilities be resolved by scientific research. Their statement that I advise us to “tolerate contradictory theories regardless of their truth” (Zepf and Gelach 2013, p. 779) is unfortunate. It seems to me obvious that I was proposing tolerance and open examination as a first phase to clarifying in depth what the issues and controversies are before submitting them to scientific research regarding their truth value. Open exploration of contradictory theories is an essential precondition for submitting them to scientific research.

Finally, the authors agree with my proposals regarding the selection and rotation of lecturers and the avoidance of ossification of the curriculum; the invitation of scholars from neighboring sciences to teach seminars; giving preference to scientifically interested candidates; and promoting a candidates’ organization at all levels of the institute’s educational commitments. They agree with adopting a research orientation for the institute and installing a department of research.

Their main critique is directed against what they perceive as my proposals in the area of research. I regret having to say that they distort fundamentally my view of the nature of the research and the methodology of research to be stimulated and developed as part of the functions of psychoanalytic institutes and societies. I certainly stressed the development of empirical research and the exploration of research methodology throughout the mental health sciences and, I would add, throughout all the boundary sciences of psychoanalysis, which include both psychological and sociological sciences, as well as neurobiology. However, at no point did I suggest that research should be limited to empirical research and to methodologies “imported” from those scientific fields. Nowhere did I state that empirical research should be the only research acceptable.

Zepf and Gerlach also imply that research methodologies in the boundary fields of psychoanalysis obey natural laws of those sciences that do not apply to those of psychoanalysis, which is obvious. However, the assumption that research methodologies developed in the psychological, sociological, neurobiological, and mental health sciences are necessarily incompatible with a function to explore the psychoanalytic field is clearly unwarranted. Methodologies from one field may be adapted to explore questions in another field, or inspire new ways of exploring the issues under consideration. We already have significant research in certain areas of psychoanalytic theory, on aspects of defensive operations and transference analysis, that illustrates the usefulness of such “imported” methods, creatively modified and applied to new subject matter. For example, I wonder whether the authors, with their background in the Saarland, are familiar with Rainer Krause’s application of research methodology from contemporary studies of affect expression and communication in his elegant demonstration of the mechanism of projective identification in empirical studies of breakdown of the communicative process under particular circumstances (Krause 1997, 1998). Do they know of Per Høglend’s demonstration (Høglend et al. 2008, 2011), with a relatively simple but innovative instrument, that transference exploration as a key technique in psychoanalytic psychotherapies may make no major difference in patients with relatively mild disorders, but is related to significantly better outcomes in the psychoanalytic psychotherapies of patients with more severe illness? Høglend thus contributed significant empirical evidence to the earlier conclusion, drawn in the same direction, by the Menninger Foundation Psychotherapy Research Project, in which we used the “imported” instruments of multi-scalogran analysis and facet design to compare simultaneously a large number of variables within a relatively small group of patients (Kernberg et al. 1972). In the Institute for Personality Disorders at Weill Cornell Medical College, we have been able to compare, in a randomized clinical trial, transference-focused psychotherapy, supportive psychotherapy, and dialectical behavior therapy (Clarkin et al. 2007), demonstrating that only transference-focused psychotherapy improved borderline patients’ Reflective Function (Levy et al. 2006).

The important contributions to be achieved by studying the efficacy of psychoanalysis as a treatment, a major emphasis in my paper, is denied by Zepf and Gerlach by their assertion that psychoanalysis as a treatment cannot be evaluated objectively, because “psychoanalytic process” is an essential variable that would define this treatment, and because it has been amply demonstrated that it is impossible to reach consensus regarding what “psychoanalytic process” is. In fact, careful reading of their commentary might justify the conclusion that the authors are saying that we cannot do research on the efficacy of psychoanalysis as long as it is not clear what psychoanalysis really is, in terms of the lamentable state of the operational definition of the “psychoanalytic process” variable. I regret that the authors should have resorted to such an eminently inappropriate measure to illustrate the impossibility of evaluating psychoanalysis. From the viewpoint of a scientific research approach to the process of psychoanalytic treatment, the overinclusive nature of the concept “psychoanalytic process” is clear: it implicitly includes, first, the application of a clearly defined technique; second, the competency with which this technique is employed by the particular practitioner; third the appropriate indication or contraindication for a particular patient to this treatment; and, fourth, the extent to which treatment contraindications or severe resistances interfered with the development of a psychoanalytic process in terms of an optimal interaction involving a psychoanalyst providing new sources of understanding to a patient who, in turn, by the method of free association, would be able to employ those interventions to acquire further understanding of his or her unconscious processes. In short, “psychoanalytic process” is a highly complex condensation of variables that would need to be studied independently to evaluate their relationship to treatment outcome. Otherwise, we would be tempted to define what psychoanalysis is with cases that already are showing a successful process of carrying out this treatment methodology.

Fortunately, in contradiction to the authors’ pessimistic and problematic approach to this research issue, two major projects to assess the efficacy of psychoanalytic treatment are presently under way at Columbia University and the University of Vienna, in addition to the very important work regarding the outcome of psychoanalytic treatments carried out in Germany under the overall leadership of Leuzinger-Bohleber and her colleaguess. A Psychoanalytic University College engaged in active empirical research is effectively contributing to reinserting psychoanalysis in a university setting in Berlin, and a related college has been functioning as part of the University of Buenos Aires.

As mentioned before, it is a major misunderstanding by Zepf and Gerlach to imply that I would want to restrict research to empirical research, outcome research, and methodologies imported from boundary sciences. I certainly am pleading for an expansion in this particular, totally underdeveloped realm of the research function of psychoanalytic institutes. This neglect has contributed significantly to the isolation of psychoanalytic institutions from university settings, and to the questioning of psychoanalysis as a science in large segments of psychiatry and clinical psychology, and is a significant factor in the crisis that psychoanalytic institutes are experiencing worldwide: a decline of interest among potential candidates and a decline in the number of patients applying for psychoanalytic treatment. An expanded view of the realm of research in psychoanalysis such as I have suggested is well illustrated in Leuzinger-Bohleber’s major research project (Leuzinger-Bohleber et al. 2002).

I strongly believe in a broad approach to psychoanalytic research that includes expansion of our contacts within the humanities, a growing relationship with social psychology, sociology, political theory, and philosophy, and the application of psychoanalytic thinking to ideology and social processes generally. In this I am in full agreement with Zepf and Gerlach, as well as with the need to encourage creative fundamental research on the psychoanalytic process and the unconscious dynamics of the mind, to advance scholarly research on concept development, and to study of the influence of other sciences and research methodologies concerning the functioning of the human mind in both normality and pathology.

The authors of the commentary falsely attribute to me the belief that a consensus must be established before entering into empirical research. However, one route to consensual definition is research into differences. If one theory is proved superior, competitors will wither. If several are equal, there will be research into shared factors that will further the inquiry. (The alternative is people shouting at each other, but not listening.) The authors suggest that the “mass of inconsistent conceptualizations” must be subjected to “critical evaluation”; they imply that “hard conceptual work” rather than “hard empirical work” is required (p. 783). But hard conceptual work may lead to conceptually unresolvable issues that require research, and open-mindedness until we have the data. Otherwise, conflicting theories become a matter of taste, not of truth.

An antianalytic bias predominates in a majority of psychiatric university settings at this time, expressed in clinical classifications based on surface behaviors that deny the unconscious dynamics and structures discovered by psychoanalysis, and on attempts to establish simplistic linkages between genetic disposition, temperamental and affective neurobiological development, and manifest behavioral traits. In contrast, a small but persistent group of psychoanalytic researchers in the field of personality disorders have gradually been able to influence this state of affairs. They have carried out research demonstrating, on the basis of contemporary psychoanalytic object relations theory, that the degree of integration or fragmenting of the self, the extent to which primitive defensive operations are distorting the internalization of significant object relations, is a basic organizer of impulse-defense configurations and character structure. This research has been able to influence the new personality disorders classification system of DSM-5. In the new proposed DSM-5 Classification of Personality Disorders, for the first time, the degree of pathology of the self and of relations with significant others has been established as the basic criterion determining the overall characteristics and severity of personality disorders. Most of the corresponding research, however, while carried out by psychoanalysts, has been conducted outside the setting of academic institutions. This, I believe, needs to change, in order to strengthen our science, the vitality and depth of the study of psychoanalysis in psychoanalytic institutes, and the future impact of psychoanalysis as a profession and as a science in the twenty-first century.

To conclude, I appreciate Zepf and Gerlach’s concerns, the opportunity to clarify the various points they raise, and their concluding acknowledgment that quite possibly my “proposals to reform the organization of psychoanalytic institutes and adopt a research orientation would make psychoanalytic education more interesting and attractive to candidates” (p. 781).