Mounting evidence attests to a decline in psychoanalysis which we believe cannot be reversed either by public relations or advertising. Empirical psychoanalytic research to test the tenets of psychoanalytic theory and practice offers the best hope of reestablishing respect for the integrity of psychoanalysis and for the effectiveness of psychoanalytic treatment, both in the general public and in the scientific community. Although contemporary psychoanalytic research increasingly generates creative studies, empirical work on a much broader scale is necessary. To create the required cadre of researchers, we must reconceptualize the role of a psychoanalyst to include proficiency in both research and clinical analytic activities. Moreover, the development of such professionals requires that psychoanalytic education utilize joint teaching by researchers as well as by clinicians. Furthermore, since the conflicting goals of a training analysis may compromise both the optimal therapeutic benefit for the candidate as well as the candidate's capacity to develop an inquiring, critical, research attitude toward the training analyst's psychoanalytic values and techniques, we propose that the candidate arrange a personal analysis unconnected with the institute.
Mounting evidence attests to a decline in psychoanalysis in the Western world. For example the APsaA (American Psychoanalytic Association) analytic practice survey of 2006 demonstrates that the average number of analytic patients per member continues steadily to decline. Followup studies of recent analytic graduates at three institutes, Columbia University Center, New York (Cherry et al., 2007), Cincinnati (Kaplan et al., 2008) and Los Angeles (Pretsky et al., 2007) indicate that four out of five analytic graduates devoted almost all their practice to other-than-four-
We would like to thank Judith S. Schachter, M.D. for her many helpful suggestions as well as her careful editing.
Faculty, Columbia University Psychoanalytic Center for Training and Research.
session per week analytic treatment. The number of new candidates continues to decline; several pairs of institutes have merged in the hope of attracting a combined number of new candidates sufficient to start a class.
Serious problems with psychoanalytic education have been identified in Anglo-American and European countries (Wallerstein, 1972, 1991, ; Kernberg, 1986, 1992, 1996, 1997, 2000, 2004a, 2004b, 2006, 2007; Thomä, 1993; Thomä & Kächele 1999; Kirsner, 2000, ; Gabbard, 2003; Glick, 2004; Levy, 2004; Reeder, 2004; ; ; ). Psychoanalysis has scant presence in universities or in departments of psychiatry or of psychology; university libraries are subscribing to fewer psychoanalytic journals. Psychiatric texts pay little attention to psychoanalysis. Fewer publishers are printing psychoanalytic books.
We believe that neither public relations, nor advertising, nor outreach community programs can reverse this decline in psychoanalysis. The only effective response will be analytic research programs which by testing the tenets of analytic theory and practice will reestablish respect for the integrity and effectiveness of psychoanalysis in both the general public and in the scientific community.
It is beyond the scope of this paper to present a comprehensive survey of contemporary psychoanalytic research. A good deal has been accomplished in tackling the issues set out more than fifty years ago in Kubie's (1952) “Problems and techniques of psychoanalytic validation and progress”. The Menninger study (Wallerstein, 1986) set the stage for concerted efforts systematically to study outcome. The report on the Ulm conference on “Psychoanalytic Process Research Strategies” (Dahl et al., 1988) opened the arena for systematic process studies. Westen (1999) concludes that there is overwhelming empirical substantiation of the psychoanalytic hypothesis that much of mental life involves a dynamic unconscious. That confirmation, however, fails to resolve conflicting theories and practices of different “schools” of psychoanalysis.
To cite just a few other examples of contemporary research, Blatt and colleagues reported a series of studies demonstrating that analytic treatment is more effective with introjective patients and that supportive therapy is more helpful with anaclitic patients (Blatt, 1992; Blatt & Auerbach, 2003; Blatt & Shahar, 2004). In addition, Leichsenring et al. (2005) and Sandell et al. (2000, 2007) have demonstrated the effectiveness of analytic
treatment for at least certain patients, while Ablon & Jones (2005) have empirically delineated analytic process and emphasized the uniqueness of analytic dyads. A member of the Swedish STOPP group, Falkenstrom et al. (2007), reported that patients in analysis continued to improve after termination to a higher degree than patients in long-term psychotherapy. Milrod et al. (2007) demonstrated that psychoanalytic psychotherapy was more effective than cognitive behavioral therapy in the treatment of panic disorder. Bucci & Maskit (2007) moved us to a new generation of psychotherapy process research which found that linguistic measures agreed with clinical ratings based on a psychoanalytic perspective. Kächele, Schachter & Thomä (2009) have reported what are probably the most intensive and comprehensive empirical studies of the analytic treatment of a single patient (for a summary see Kächele et al., 2006).
In addition, International Psychoanalytic Association and APsA are enhancing its research activities. Each provides an intensive research training course, and APsaA, in addition to its Fund for Psychoanalytic Research, has committed $100,000 over five years to a study of the effectiveness of analytic treatment, and arranges poster sessions of research projects at its meetings. APsaA's Board on Professional Standards has had a Committee on Research Education for almost five years, and representatives of this Committee have recommended research courses and research opportunities at four different institutes that they site visited.
It is the thesis of this paper that these research activities and plans, in Wallerstein's (2007) words, “fall[s] far short of what should be possible and what is required if we are to live up to our obligations as a science” (p. 968); Luyten (2007) agrees. Actually, the number of senior scientists engaged in these kinds of formal studies is small. Name a few and you have named them all: Ablon, Bateman, Blatt, Bucci, de Maat, Fonagy, Kächele, Kernberg, Krause, Leuzinger-Bohleber, Luyten, Sandell, Target, Taylor, Waldron and Westen and their collaborators are the key players. Our task includes transmitting and implementing into clinical practice what has been demonstrated by research (Talley et al., 1994).
Before proposing a research-oriented conception of psychoanalysis, we want to acknowledge those philosophers and psychoanalysts who believe that psychoanalysis is not a science, cannot become a science and must be one of the humanities. Wittgenstein and Rorty believe that psychoanalysis should abandon evidential pretensions (Ahumada, 2004, p. 584).
Green (2000) feels that research is likely to be of little value to psychoanalysis. Perron (2006) cites a French study of evidence-based treatment (INSERM, 2004) which he believes does a disservice to psychoanalysis. Cruz Roche argues that the methodology of natural science confers an air of rigor, but actually loses touch with the subject of psychoanalysis. Bose (2003) urges exploration more akin to the methods of the humanistic disciplines. Leupold-Lowenthal (1978) suggests that psychoanalysts feel like outsiders, excluded from the scientific community, and therefore respond with feelings of contempt for those who marginalize them.
To some, a hermeneutic conception is an alternative to a scientific view. Wallerstein (2006) criticizes the hermeneutic position, and supports Grünbaum's (1984) critique of Habermas' (1971) position that only the patient has the required privileged access to the ultimate validation or refutation of psychoanalytic hypotheses (see in the same vein Thomä & Kächele, 1975; Kächele et al. 2009, Chap. 2). Grünbaum asserts that analysts traditionally and acceptably can and do interpret against the patient's judgment and in the face of his or her denials. Eagle (1980), furthermore, argues that hermeneutics fails to provide a way to resolve differences in clinical inferences by different analysts. Wallerstein concludes that since science is defined not by its subject matter but by its method it is in no way enjoined from dealing with unconscious feelings, meanings, qualities or unique individuals.
Clinical, hermeneutic narratives rarely disclose the unexpected, but scientific studies are capable of generating surprising findings. Puschner, Kraft, Kächele & Kordy (2007) studied 144 patients in psychoanalytic therapy and 472 patients in psychodynamic psychotherapy. During the pre-treatment period, a small number of observations indicate that psychological distress (measured by SCL-90 and GSI) declined more quickly from acceptance for treatment to the start of treatment, than during the treatment itself.
Many years ago Engel (1968) emphasized the necessity for psychoanalysis to develop a strong research cadre within its ranks and, therefore, that the clinical goal and the research goal should be held in equal honor within our training programs. He noted that we have not developed more than a handful of skilled scientists. Candidates with research interest were considered over-intellectualized or even obsessional; a candidate questioning and asking for evidence was seen as resistant. Candidates
have little exposure to analytic researchers: “Literally, there is no way in which a young analyst can learn by precept and example and by identification with a researcher how to become a psychoanalytic scientist and carry on research” (p. 199). It is distressing to consider how fitting these statements remain for the present situation.
Despite Engel's clear prescriptions, psychoanalytic education has failed to develop a research orientation in candidates and few candidates or graduates become involved in research projects. Few American institutes teach analytic research or equip their candidates to read and understand research reports. The certification procedure, which is prerequisite for training analyst appointment and presumably assesses analytic competence, now is based on the write-ups of two analytic treatments. Not only is research not included in the certification procedure, but there is no query into whether the candidate can read and critique a research paper.
We believe that the training analysis system, by its nature, creates obstacles that interfere with the candidate becoming interested in and involved in analytic research, and therefore, we present a brief discussion of training analysis. Although we are unaware of any empirical studies of the effectiveness of psychoanalytic training analyses, there is a research literature about training therapy in psychoanalytically-oriented training. Reviews of studies of training therapy conclude there is very little empirical evidence that it has any measurable effect on client outcome (Macran & Shapiro, 1998; MacAskill, 1999). However, length of training therapy related negatively to patient therapeutic improvement (Sandell, Carlsson et al., 2002; ). In the latter study, mean total duration of training therapy was 10 years, and 59% of therapists had psychoanalysis as their most extensive therapy. Thirtyseven therapists with training therapy of 13-14 years had patients with the poorest outcome; 38 therapists with 7-8 years of personal therapy had patients with the best outcome. The authors conclude that whereas current results may support doubts about the value of extensive training therapy, they do not show that training therapy is unnecessary or even counterproductive.
The training analysis is widely considered the most significant part of psychoanalytic education. Of necessity, the training analyst, with whom the candidate will, to some degree identify and emulate, is a clinician, and often a significant power in the institute. Immersed in analysis with a clinician-analyst, dealing with clinical matters in supervision and in didactic courses, and without contact with researchers, the candidate will
most likely become identified with clinical values. Because the training analysis, required for graduation, straddles conflicting goals, on one hand, to teach about analysis, and, on the other, to foster personal development (Lewin & Ross, 1960; Greenacre, 1966; ; Kernberg, 1986, 2006a; Garza-Guerro, 2002; Meyer, 2003; Reeder, 2004) it tends to inhibit the candidate's curiosity and criticism of the training analyst's values and techniques while fostering compliance and conformity (Kairys, 1964; Shapiro, 1976; Greben, 1982-1983; Kernberg, 1986; Richards, 1997; Reeder, 2004; ). This interferes with the candidate's developing a critical, research attitude. Thus, the training analyst system is a major impediment in the development of research orientation and interest in candidates. Kernberg (2006c) notes that “In many locales, not only does there not appear to be any interest in fostering the development of psychoanalytic scientists and researchers, but to make matters worse, some institutes adopt a frankly negative attitude toward research in general” (p. 1651).
While recognizing that psychoanalysis has failed to develop a scientific tradition (Michels, 2007), we believe that empirical research is our most effective response to the current decline in psychoanalysis. In order to develop the necessary research cadre we propose that the conception of a psychoanalyst be changed from that of a competent clinician to someone who encompasses both clinical and research expertise, which a few analysts have accomplished. In order to effect such a change, psychoanalytic leadership on every level must be shared equally between clinicians and researchers. This ongoing interaction might benefit each group, and energize a focus on empirically-based examination of psychoanalytic theory and practice. To accommodate this change in the educational and political establishment, psychoanalytic education itself must be appropriately modified.
In order for psychoanalytic education to enable candidates to develop both clinical and research expertise, we propose that all didactic teaching as well as supervision be conducted jointly. Due to the limited number of analytic researchers, researchers from allied professions might be called upon, to everyone's advantage. (Supervisors could be asked to contribute their fees to provide funds to pay researcher-teachers).
Candidates should be taught both research techniques and clinical techniques, and engage in both research experience and clinical experience (Gerber, 2001). Importantly, this would provide roughly equal opportunity for candidates to identify with a research model as well as with a clinical model in proportions individually determined. (The Columbia Psychoanalytic Center has created a privately funded psychoanalytic research methodology training program with full clinical psychoanalytic training).
A shift to a model in which the candidate engages in a personal analysis de-linked from professional advancement has been recommended for many years. We believe that the conflicting goals of the training analysis may interfere with personal therapeutic goals (many analysts seek additional treatment after termination of a training analysis) and since the candidate's analysis is widely considered the most important part of psychoanalytic education, it is critical that the candidate experience an optimal personal analysis. The candidate's selection of an analyst should be entirely by personal choice, neither limited to designated training analysts nor to analysts associated with the candidate's institute (Kairys, 1964; Thomä, 1993; Thomä & Kächele, 1999) as practiced in the French and Uruguayan models of the IPA.
Kernberg (2006c) prefers maintaining specific quality controls for those authorized to analyze candidates, such as specific years of experience and the assurance of immersion in psychoanalytic work. This is a reasonable argument, but our studies of successful analyses (Schachter, 2005b) lead us to propose that the analyst's personality and potential emotional difficulties with a particular patient trump experience and immersion. We propose that the candidate's choice of a personal analyst be completely unfettered (Kächele & Thomä, 2000). Analysts known to be successful in helping candidates deal with their personal problems could then be chosen in the free market place (Rosenfeld, 2005) rather than because of their status or qualifications. The personal analysis, then, would focus on the candidate's emotional problems, while the institute's program would be free to educate him/her. The personal analysis system, like all systems, has its problems, but the advantages of non-conflicting goals and of minimal theoretical indoctrination outweigh any other shortcomings.
We are under no illusion that our proposed response to dramatically reconceive and reorganize psychoanalysis and education has much likelihood of being implemented, despite the current intense, ongoing controversy about the training analyst system. We propose it because
of our conviction that our existing professional system is unable to cope with the exigencies confronting American psychoanalysis and unable to reverse its diminishing status and prestige. A rejection of this proposal should entail the obligation to present an alternative proposal that is considered equally promising and more feasible.
We are impressed by, and embrace, Popper's (1962) assertion that knowledge grows by a process of falsification; our most useful beliefs are not those that are most strongly verified, but those that have best survived criticism and refutation. We disagree, however, with his conclusion that psychoanalysis cannot be tested empirically and agree with Grünbaum's (1984) argument that, in principle, elements of psychoanalysis are falsifiable. We believe there have been few serious attempts to refute the fundamental tenets of psychoanalysis and that the most compelling evidence of that failure is the fragmentation evident in the persistent, diverse, conflicting views of the basic principles of psychoanalysis. There is no consensual agreement about the definition of psychoanalytic process, about theories of therapeutic action, about the goals of psychoanalytic treatment or about how to assess whether the goals have been reached (Smith, 2007). The failure of dialogue to resolve these disagreements indicates that none have been falsified or rejected. We assert that only empirical research can provide an opportunity to falsify or refute different axioms of psychoanalysis. Such refutation may provide optimal conditions for the growth and development of psychoanalytic knowledge.
In conclusion, we propose that modifying the conception of a psychoanalyst to include research competence as well as clinical competence increases the possibility of developing a cadre of researchers to provide empirical studies to test the tenets of psychoanalysis (Teller & Dahl, 1993).