Therapist – client complementarity

By Dr. Matthew Bennett copyright 1995

According to Wolitzky (1995), there are two important curative factors in psychoanalyis: insight and the relationship. This observation suggests that in addition to the therapeutic interpretations offered by the therapist as a way of helping the patient achieve greater awareness, the quality of the client-therapist relationship, or alliance, itself contributes to the positive effects of psychotherapy. Greenberg and Cheselka (1995) defines the therapeutic alliance as “the part of the therapist-patient relationship that enables them to work collaboratively” (p. 72).

Wolitzky (1995) identifies alliance formation as the process whereby a therapist develops a sense of the patient’s threshold for anxiety and narcissistic injury, and thereafter abstains from exceeding those optimal levels. “The considerations,” he states, “come under the heading of clinically informed common sense, clinical wisdom, and tact” (p. 32). Common sense or not, recent research has indicated that the quality of the alliance is the best predictor of successful therapeutic outcome (see Safran, Crocker, McMain, and Murray, 1990). Naturally, this finding has sparked a flurry of research interest in exactly what it is that constitutes successful therapeutic alliances. 

Another recent avenue of research interest has focused on the implications of attachment style theory for dynamic psychotherapy. Attachment theory, developed by Bowlby, indicates that psychological and social development occurs within the crucible of interpersonal relationships from early in life. Although Bowlby’s original theoretical work concerned children and early development, more recently attachment theory has been applied to the study of adults and the enduring attachment styles which survive into adulthood from childhood. These enduring interpersonal styles, known as working models, serve as a kind of cognititive filter which profoundly influence the nature of adult relationships. And what is psychotherapy, note Henry, Schacht, and Strupp (1990) but “simply a structured relationship between two people” (p. 768)? Thus some researchers became interested in the adult attachment working model as a variable mediating the quality of the therapeutic alliance.

Dolan, Arnkoff, and Glass (1993) investigated client-therapist complementarity in terms of attachment patterns and internal working models. In several case studies, these researchers observed that therapists actually alter their therapeutic posture to take into account patient attachment characteristics. They note that the various attachment styles are associated with differing interpersonal approaches such as degree of trust or expressed independence. Thus it would seem that the “clinically informed common sense” described by Wolitzky should include an appreciation of such interpersonal factors as internal working model and attachment style, which may serve as a foundation for an appropriate therapeutic stance on the part of the therapist. Kiesler (1986) and Henry, Schacht, and Strupp (1990) confirm the clinical utility of complementarity to positive therapeutic outcome.

It may seem that the development of such a carefully calculated therapeutic stance renders the technique incompatible with traditional psychoanalytic therapy. Does a calculated therapeutic stance based on the patient’s internal working models collide with traditional wisdom concerning transference and countertransference issues? Greenberg (1995) warns that “if the patient’s transference and the therapist’s countertransference mesh in a way that recreates familiar patterns, the patient will be locked into old modes of experience, and old satisfactions, and growth will not occur” (p. 72). Safran, Crocker, McMain, and Murray (1990) echo this concern, indicating that “disconfirming the client’s pathogenic beliefs is a central mechanism of change in psychotherapy” (p. 155). Traditional psychotherapy has addressed countertransference issues with “technical neutrality and relative anonymity” on the part of the therapist (Wolitzky, 1995). Thus it might seem that the approach recommended by Dolan et al (1993) and others might well contaminate the therapeutic relationship, threatening key psychoanalytic concepts such as unconscious intrapsychic conflict, free association, and interpretation of transference and countertransference material (see Gaston, 1990).

Yet careful consideration of the phenomenon of calculated client-therapist complementarity reveals that is not necessarily incompatible with even classical psychoanalytic theory. Freud himself wrote that “the first aim of treatment is to attach the person of the patient to the person of the therapist,” a process which he termed the “friendly and affectionate aspects of the transference” and which he distinguished from other aspects of the therapeutic relationship which might produce distortions (quoted in Gaston, 1990). Advocates of calculated complementarity based on assessments of the internal working model caution that therapists must avoid interpersonal patterns which have proven historically detrimental to the patient (Dolan et al, 1990). They also suggest that although the therapeutic stance can be consciously shifted according to gained knowledge of interpersonal patterns evidenced by the patient, that these shifts are in a perpetual state of dynamic flux and must reflect an alert sensitivity to changing patient needs (Greenberg and Cheselka, 1995). In this way, even the carefully constructed elements of complementarity proposed by Dolan et al (1993) seem to amount to what Wolitzky would label clinically informed common sense.

Furthermore, therapist-client complementarity, whether based on attachment style or on some other salient variable, must be distinguished from mere uninterpreted countertransference, the bugbear of psychoanalysts everywhere. Strupp and Binder (1984) identifies countertransference in relational psychotherapy as an event in which the therapist is “continually pulled into reactions that conform to the roles designated by the patient’s maladaptive interpersonal scenarios” (p. 142). Wolitzky (1993) offers the more traditional vision of countertransference: “The presumption is that undetected (and therefore unmanaged) countertransference reactions have a detrimental impact on the patient” (p. 37).

In light of knowledge of the various detrimental results which have long been associated with uninterpreted countertransference, and yet mindful of the utmost importance of the quality of the therapeutic alliance, it is important to distinguish between uninterpreted countertransference and the kind of therapeutic complementarity that has been shown to enhance therepeutic outcome. Greenberg and Cheselka (1995) approximate this distinction in stating that there is no once therepaeutic stance that can be effective with all patients: “Some patient, for example, will be comfortable with the benign distance, even aloofness, of the classical psychoanalyst….but the same posture will remind another person of a depressed, unavailable parent” (p. 74). In this sense, complementarity is not really countertransference at all but an almost automatic and yes, common sensical shifting of therapeutic approach that the intuitive clincian can effect deftly and naturally. It is not based on hidden prejudices or past experiences of the therapist, but on empirical understanding of an important kind of typology. Complementarity is a context for the therapeutic process designed to enhance its effectiveness; it is not a content of therapy. It is a way of establishing a safe environment in which the content of therapy can be effectively analyzed. Dolan et al (1993) indicate that complementarity based on the internal working model of the patient results in positive modifications in level of expressed empathy and affect, activity level, pace of work, and emotional depth. Yet perhaps the most important distinction which separates complementarity from uninterpreted countertransference is that it is a conscious process undertaken by the therapist: since it is a conscious (even if relatively automatic) process, the therapist can manage its execution and effects without allowing it to take on the problematic outcomes of uninterpreted countertransference.