There Are Only Three Kinds of Psychotherapy

The mainstream of our profession has ceased paying much formal attention to the help our patients get from the doctor-patient relationship itself. But patients are stubborn. They still expect to have helpful relationships with us! Witness the ill-informed boyfriend of a public-clinic patient who asked me, at the close of a medication visit, "But when is she going to see a psychiatrist? She has things to talk about!" Equally disturbing was the well-meaning enthusiasm of a young psychiatrist interviewing for work at the same facility when I told him that I use my psychotherapy training in many a patient encounter there. "Oh, like if there's time, you might do a little cognitive therapy on 'em?"

We psychotherapists have ourselves to blame for this state of affairs, having husbanded our subspecialty's development in two narrow directions while psychopharmacology and neuroscience have dramatically broadened their own worldviews. Some of us have manualized and proceduralized psychotherapy (or rather, certain psychotherapies) in imitation of procedural, technology-driven medicine, so as to be good citizens in the world of indications, interventions and outcomes. Most new psychiatrists have had some exposure to these techniques, but few seem inspired by them.

Another potentially influential group has circled the wagons for psychoanalysis and psychodynamic therapy. The large body of literature this group continues to generate takes many fascinating forms, from the quasi-literary to the quasi-neurological, but few psychiatrists-in-training read it.

Neither of these two currents in psychotherapy convincingly illuminate, or render more effective, our work with the entire spectrum of patients we see in everyday general psychiatric practice.

What sense can we possibly expect our trainees to make of this Tower of Babel? Unless they are unusually persistent, they will either stick to the DSM and the meds and/or develop their own ways of conceptualizing personality and human nature, often with the help of pop psychology or New Age spirituality. In both cases, this dooms them to repeat sundry mistakes of the past, because an intellectually rigorous psychology of relationship, cognizant of boundaries and power arrangements, is missing. Should those of us who still try to teach psychotherapy in this inhospitable clime simply accept that the next generation will have to reinvent the wheel, or can we do any better for them?

We can, and it isn't all that hard. A global overview of the psychotherapeutic enterprise, based not upon theory but upon what therapists actually do, is easy to provide and reveals that there are really only three kinds of psychotherapy. I will begin my description with the rather typical odyssey of a real patient at a residency training program.

This patient's prolonged panic attack began when her mother arrived from out-of-state at the local bus terminal in a hospital johnny, seeking her care. Unable to cope, the patient was brought to the emergency department by her husband. She was so desperately anxious as to admit to suicidal ideation: the "Open sesame!" of our triage culture. At age 31, this secured her first psychiatric hospitalization and evaluation.

The patient settled down quickly with the help of anxiolytics and an alternative placement for her regressed, mentally ill mother. But during her brief stay, an alert inpatient resident noticed that she engaged in several classic compulsive rituals, including much hand-washing and showering. Referral was made for outpatient psychotherapy with another resident who would obtain supervision in the cognitive-behavioral treatment (CBT) of obsessive-compulsive disorder (OCD) in addition to the usual medication trials.

The outpatient resident got more than he had bargained for. It was difficult for him to concentrate on response-prevention and thought-stopping techniques under the silent, watchful eye of the patient's husband -- and she insisted upon his presence in order to "feel comfortable." The resident kept trying, and one day when the husband had to leave the session early, the patient volunteered that in her own mind, the pervasive sense of dirtiness that kept her always washing had something to do with repeated episodes of sexual abuse by her mother in childhood. There was also fear of retribution by her mother for "not being a good girl," which transferred in the present to her husband. His idea of good behavior was that she never leave their apartment alone.

It became clear to this intelligent resident that, beyond the illness dimension of Axis I OCD, a relational dimension would have to be entered to treat this patient successfully. A newcomer to the incoherent shipwreck of non-psychopharmacological psychiatry, he understandably groped for some of the remaining scraps of a psychological outlook that are still afloat. Childhood trauma...pathological dependence...Axis II borderline personality disorder (BPD)? Well, close enough to try dialectical-behavior therapy (DBT), opined the supervisor.

And so, aided by the husband's periodic departures after check-in, the resident soldiered on with chain analyses of anxious thoughts and other such techniques in a DBT-like individual therapy (the skills classes, of course, were unacceptably anxiety-provoking). Eventually the patient stopped coming, and when the resident finally called, she told him that she was afraid she would always be a disappointment to him and was wasting his valuable time. Could she see the hospital resident who had originally admitted her, "just to talk?"

To the second resident's chagrin, the patient proceeded to settle in comfortably with the original resident, attending regularly although she (the hospital resident) was even more of a therapeutic neophyte and knew no specific techniques. In fact, she said, "I have no idea what I'm doing!" But she is clearly doing something specifically effective for the first goal of any therapy: keeping the patient in the room. The patient is comfortable enough to dispense with the husband's presence. A second "safe place" is emerging that can compete with the safety of the husband's total control, and within which she can begin to take possession of herself.

We can imagine where things might go from here if the utilization reviewers are kept at bay. Perhaps patient and resident (or the resident's successor) will feel ready to retry the response-prevention strategies that failed at first, building on the strength of a working relationship. Or perhaps they will work directly on that relationship itself, when the idealized "safe" therapist inevitably falls from grace and disappoints (or enrages) the patient. Ideally, the patient will eventually work on both behavior and relationship, although not necessarily with the same therapist.

This patient exemplifies the everyday complexity of the real-world population psychiatrists treat. Hers is not a pure form of a DSM-described illness that dutifully improves like the cases in academic studies ("OCD+CBT=OK," "BPD+DBT=OK" and so on, as I like to joke). Instead, she is the usual biopsychosocial layer cake (Table 1) of Axis I syndrome (bio-), relational disturbance (psycho-) and developmental damage (-social). Because of this, benign power -- here, the doctor who tries to do response-prevention -- is not a viable reality for her. Before she can get help, she needs help with getting help. And that is what the resident who "doesn't know what she's doing" is giving the patient.

In the real world of clinical practice, based not on theory but on the unspoken power narrative embedded in the therapeutic relationship, there are only three kinds of therapy (Table 2): supportive, directive and relational. Our patient began her odyssey in the directive column, where most contemporary psychiatric treatment starts, first with the CBT, then DBT. When she came back, it was to the supportive column, albeit to a therapist who had not yet learned its language. If all goes well, she will finish in the relational column, probably availing herself again of directive techniques along the way.

The underlying premises of each kind of therapy establish a power structure within doctor-patient relationships -- a structure that doctors may not have thought through in advance and that they might not endorse if they had. With deliberate exaggeration, therefore, let me explicitly voice the power narrative of each of these three forms of therapy to show the differences between them.

When I speak of power, I refer to the therapist's power to define the nature of reality. Thus, in directive therapies, the unspoken narrative is, "I, the therapist, am powerful and know what is or isn't dangerous [in anxiety disorders], is or isn't negativistic [depression], is or isn't real [psychosis]. I know the diagnosis and causes of your illness, and I direct you to change your thoughts/feelings in the following way," and so on.

Such a narrative reiterates the existing social status quo and asks the patient to adjust to it. This is often a good and necessary thing, as with a potentially violent paranoid man in the emergency department, or a widowed woman whose grief has extended into major depression. But because many damaged patients are unable to let themselves be helped by a benign authority -- unable, in D.W. Winnicott's words, to "use the object" -- directive therapies don't work as often as their academic proponents would like to think.

While the supportive therapies are centered on the person rather than the symptoms, their power narrative has a remarkable underlying similarity to that of their directive cousins. "I, the therapist, am powerful-powerful enough to keep the world at bay. I will create the environment and give you the nurturance that you need. Let yourself unfold under my protection, bask in my empathy," and so on.

Instead of actively reinforcing the social status quo, this narrative provides the patient a temporary respite from it. Once again, this is often a good and necessary ("indicated") thing, as in the resident's case presented earlier. I want to stress, though, that it is a technical approach in that it purports to know what is wrong -- environmental stresses, developmental deficits from childhood -- and to be able to ameliorate or correct it through consciously (even if intuitively) modulated enactments on the part of the therapist. In comparison to directive therapy, the harsh diagnostic language is muffled and there is some implicit faith in an intrinsic healing process, which supportive therapy tries, indeed, to support. But the power arrangement in which a knowing therapist administers techniques to a malfunctioning patient remains the same. I maintain that this is even true of Kohutian self-psychology, which has developed a rich and beautiful theoretical language to describe its own version of supportive therapy. Such language does not change the fact that the therapist largely defines the nature of reality -- an arrangement much more obvious and above-board in the less humanistically couched directive therapies. Of course, what the experienced practitioners from any well-established school of psychotherapy actually do is likely to span all three of the types of therapy in this classification scheme at some time or another; it is only either very new, or very partisan, practitioners who refuse to adapt their behavior to the needs of a reasonable range of patients.

Relational therapy remains to be described, and it is as important not to idealize it as it is not to devalue the other two forms. It does not fit all or even most patients' needs, and its irresponsible application can be not only "contraindicated," but cruel. However, it is the only form of therapy whose power narrative allows both doctor and patient to be subjects engaged in an unpredictable process, rather than an Expert and the object of their technique. "I, the therapist, possess an institutionally sanctioned role, as well as education and experience. But my power and knowledge are limited. I offer no more nor less than an honest professional relationship whose primary purpose is to address your problem(s), but through which I fully expect to question myself, learn and grow as well."

The language of Martin Buber's "I-Thou" dialogue, Jurgen Habermas' "communicative action" or Winnicott's ingenuous term "object use" (which is really about subjects), all aptly describe such a relationship. Therapy takes the form of questions and statements about one's experience of self and other, and its power lies chiefly in the immediate interaction regardless of how much importance the therapist attributes to constructs such as transference, the unconscious, developmental issues, the Gestalt cycle of experience and so on. We all have our favorite constructs, our own ways of skinning a cat.

Because honest reciprocal dialogue fosters autonomy, independent thinking and acceptance of one's own emotional reality, it is subversive to the social status quo and its usual power arrangement of Expert and object. While patients retain, even improve upon, their ability to get help (including medical care) through relational therapy, its premise encourages them to become the ultimate Experts and Authorities on their own mental and interpersonal lives.

Before I conclude, two important corollaries to this triune scheme deserve mention. First is the place of the powerful modality of group psychotherapy. The various approaches to it fall into the same categories: supportive (peer and professional support groups), relational (process-oriented groups) and directive (psychoeducational and skills groups). As in individual therapy, actual practice favors hybrids rather than pure forms. Family therapies, with which I am less familiar, probably follow analogous patterns.

The other corollary involves the role of medication. It may be directive (as with antipsychotics in acute delusional mania, which aim to change mental content) or supportive (as with temporizing anxiolytics or antidepressants when used to prolong a patient's survival in an intolerable home situation: here the effort is to dull mental content). But medications can never be relational, much as patients sometimes attempt to have relationships with us through them, symbolically mediated by "side effects," paradoxical responses and refractoriness to treatment. This is one reason why longer visits and individual attention lead to more successful pharmacotherapy.

Giving psychiatric trainees an overview of these three kinds of psychotherapy demystifies the field and helps them to find their place in it. They can then move beyond the procedural "see one, do one, teach one" mentality of medical school. Instead of learning to "do some therapy on 'em" in only standardized ways, trainees will begin to look at their own natural aptitudes, as well as their intrinsic limitations, and how possibly to stretch them. Otherwise we consign them to the learning of an ever-narrowing, mechanistic approach to human problems.