NEED FOR A NEW TYPE OF CONCEPT
Current Trends and Needs in Psychotherapy Research on Schizophrenia [1]

E.T. Gendlin

University of Wisconsin

In 1813 a group of Quakers reported a large incidence of recovery for a hospitalized group of patients who were taken out of relatively inhuman asylum conditions and treated with friendly and respectful human attitudes (Bromberg, 1959). That report did not generally establish it as a fact that certain positive attitudes in human relations overcome schizophrenia. The finding was neither accepted generally as a fact, nor were its loopholes and imperfections followed up with more exact studies. And the same can be said for very, very many other studies concerning psychotherapy with schizophrenics. From 1813 to the present, loose report after loose report — start after start — has been let pass without the need being felt compellingly to deal with it further in specifically controlled researches.

Thus, today, we are still investigating this attitude conception of the cure for schizophrenia, and we are still investigating whether psychotherapy works at all with schizophrenics. People still can say both that it does and that it does not work without feeling the need to cite just what it is in this or that of the many studies which they question or wish to control.

The current trend I see in research into psychotherapy with schizophrenics is the beginning of this sort of research — carefully controlled, designed research; the sort in which the defects can be at least clearly defined, the sort in which it is possible to estimate the degree to which findings are really established, the sort of research upon which successive studies can be built.

I think this is the most important trend I can cite.

As second and third trends I think I see some specific lines of successive investigations developing. One of these concerns the question: "Does psychotherapy work with schizophrenics?" The other question being dealt with is: "What are the effective therapeutic factors?"

The Trend Toward Controlled Research

To estimate whether psychotherapy works, one needs to compare experimental psychotherapy subjects with control sub- [Page 38]jects who do not receive psychotherapy. From Perry Point there is now reported (Fairwheather, 1960) a controlled comparison between analytic therapy patients and a control group which received work assignments and milieu improvement.

At Wisconsin the current investigation sponsored by Carl Rogers, [2] which I am directing, also fits into this line of development. Therapy and control patients are being matched one to one, on age, sex, social class, length of hospitalization, and degree of disturbance.

Why has it been so long until controlled research began? From the study in which I am engaged, I can report at least three good reasons. First, controls (both in the sense of matched therapy and "control" subjects and in the sense of careful "controls" for many factors) are inordinately difficult, time-consuming, frustrating, slow, and costly. It may take hours of effort expended over several weeks before one patient spends one hour with a psychometrist. Often, many such attempts must be made before even one instrument is completely administered to a patient. Meanwhile, drugs, transfer, discharge to faraway towns, and a host of other factors can invalidate much hard work. The amount of time and work required often makes a careful research design seem impossible because the pace of progress is hard to distinguish from standstill.

A second reason why controls are so difficult is that there are some factors which must impose limitations, regardless of how willing one is to expend effort. Let me give one example. Some patients refuse testing, tape-recording, or even psychotherapy. These have to be replaced in the research group, yet such replacement introduces a bias. With very much effort we continue to collect all obtainable data from these dropouts, so that we can later attempt to evaluate the bias they cause. Yet, clearly, such an evaluation will be outside the research design and will lend rigor only to a further study, not to this one.

Thirdly, a difficulty of controlled research is that patients for psychotherapy are selected by the research design, rather than by intuitive criteria or motivation for psychotherapy. In our project, although we did lose some, we attempted to continue psychotherapy with many refusers if we could possibly bring meetings about at all, on the ward, in seclusion rooms, despite strong rejections from the patients, anger, hopelessness, and weeks of silence.

Looking back on this now, we would say that this has been a very rewarding therapeutic experience for us. However, it [Page 39] was only the research design that included such so-called "unmotivated" (Gendlin, 1961b) patients, and only the research design led to our continuing with them so doggedly.

These three factors at least: sheer time-consuming effort, limitations on rigor, and unwilling patients, I can report as difficulties in the attempts at rigorous controlled research.

This, then, is one trend: research investigations are comparing psychotherapy subjects and control subjects who do not have psychotherapy, matched on specific variables, to determine how, and if, psychotherapy works with schizophrenics.

The Definition of Essential Factors

The second line of investigations, which I think I see, moves toward definition of the essential factors which constitute effective psychotherapy.

Here we now have a study and its replication by Whitehorn and Betz (Betz, 1956, Whitehorn), showing that the incidence of success with schizophrenics is predictable from therapist performance on the Strong Vocational Interest Test. Whitehorn interprets this to mean that the more successful therapists show attitudes on this test which are more like a lawyer's. They enjoy finding leeway in the social rules for the sake of individual needs. The less successful therapists show attitudes like printers', tending to impose pre-set patterns on passive objects. Whatever the interpretation, we can now build upon a tentatively established proposition that some attitudinal factors in therapists (as measured by the Strong Test) are effective therapeutic variables. In a sense this is the finding of 1813, but now we have a measured variable of therapeutic attitudes replicated in predictions of successful psychotherapy with schizophrenics.

In research with generally neurotic clients, Rogers and associates (1959 a, 1957, 1960) also found successful psychotherapy associated with attitudes of this sort, attitudes of "genuineness", "empathy" and "unconditional regard". These attitudes have been measured as experienced by the therapist, as perceived (Barrett-Lennard 1959, Truax, 1961) by the client and by the raters of tape recordings. In our Wisconsin project, Rogers now hypothesizes that these same attitudinal variables will be associated with successful psychotherapy with schizophrenics. To analyze for the attitudes in complete tape recordings is an important research trend in itself. The line of development that I am drawing leads from the studies of 1813 — in which such attitudes were found to lead to improvement — to the measured and predicted therapist attitudes as manifested in tape recorded interactions during psychotherapy.

The Measurement of Patient Behaviors

A fourth trend I see — still in the future — is the measurement [Page 40] of patient behaviors — and changes in these — during effective psychotherapy. Study of this question with non-schizophrenics has led to such instruments of patient in-therapy behavior as Chapple's (1956), Leary and Gill's (1959), Matarazzo's (1958), the DRQ (Mowrer, 1953), self-reference analyses (Braaten, 1958, and Rogers' Process Scale (Rogers, 1959 b, 1958, 1961, 1960; Walker, 1960). These will be increasingly applied to schizophrenics as well.

In the Wisconsin study we are using Rogers' Process Scale which draws together seven dimensions of in-therapy behavior indicative of therapeutic movement. Rogers' summary of the continuum which the scale measures is as follows:

From feelings which are unrecognized, unowned, unexpressed, the client moves toward a flow in which everchanging feelings are experienced in the moment, knowingly and acceptingly, and may be accurately expressed. The process involves a change in the manner of experiencing. From experiencing which is remote in time from the organic event, which is bound by the structure of experience in the past, the client moves toward a manner of experiencing which is immediate,... from fear of relationships (he moves) to freely living in relationship. (1961)

These are strange and unusual terms for a research instrument. Yet, one major trend and need I see is for more such research terms. In this I am very likely to be biased since some of the terms and viewpoint in this scale come from the theory of experiencing put forward by Zimring and myself (Gendlin, 1961 a, 1955, 1962, 1960). I think, however, that this theory of experiencing concretizes a trend that is discernible in the discussions in various orientations. To the extent that there is this trend in psychotherapy, it will require new kinds of research variables for the research into psychotherapy. Let me therefore discuss this trend — if it is one — so that I can later talk about the kind of research variables which I think it is leading us to.

The trend I mean is one away from primary emphasis in therapy upon cognitive exploration, insight, verbal analyzing, and toward emotional, affective, interpersonal experiencing. Increasingly, we hear it said that it does not help the individual merely to understand the intricacies of why and what is wrong, and that he does not change due to conceptual insights. Rather, the felt, immediate, experienced events during therapeutic interaction, these constitute concrete events within him, and thereby he may change therapeutically. A moment of concretely felt living in interaction contains many, many potential meanings — and resolutions of problems — and not all of these can, or need be, conceptually insightfully symbolized. They are nevertheless lived. We term such meanings "implicit in experiencing", i.e., implicit in the organism's life process — one process — which can be analyzed in three ways: in physiological terms; in self, or feeling terms; and in interpersonal terms.

When psychotherapy was thought of as chiefly cognitive, it seemed that schizophrenics were not amenable to psychotherapy. Insight, exploration, verbal cognitive analysis, are not usually possible for schizophrenics who are often primarily silent, or so fearful that in answer to verbal explorations they withdraw even further. But, the individual who does actively and cognitively explore himself — of him also it can be said that he isn't much changed by responses to just his concepts or verbal message. Most psychotherapists would agree, though they would say it in many different ways, that it is necessary to bring into the interaction the patient's private referent, the unconceptualized, implicit experiencing which he just then confronts within himself, as he speaks. The therapist cannot really verbalize the client's experienced referent, but he can refer to it, he can point his words at it, he can fashion his responses so that it is clear that he senses the specific inward experiencing, not just the spoken or conceived message. Or, if such pointing within the patient threatens the patient too much, the therapist can express his own inward sense of the present moment, as the therapist's own expression. It will sound different, but it will still implicitly be a response to the patient's present process of experiencing, and it will tend to make events occur within the patient's experiencing.

I think that, in different orientations in psychotherapy, and in various different terms, there is some discussion of this emphasis. What is being emphasized is the process of concretely felt experiencing in the patient, and the effect upon his experiencing, which is made by certain kinds of interpersonal interaction.

At Wisconsin we are retaining the essentials and the basic attitudes of Client-Centered therapy, but are altering (Gendlin, 1961; Hart, 1961) the Client-Centered mode of working toward a psychotherapy of experiencing, by that I mean a psychotherapy consisting of the immediately ongoing interaction and feeling processes which are referred to in both persons. In this experiencing psychotherapy, many psychological contents are referred to in their "implicit" or felt form. Yet, we find also, that if we are even more careful not to impose ourselves, not to demand agreement or disagreement, not to demand commitment to therapy or commitment to the formation of a relationship — by leaving the client more free than ever —, we can also express much more of our own ongoing experiencing, voice our momentary feelings — making clear that these are our own —, and give the patient an ongoing interaction of a warm, open, eventful sort. Thus, even though he isn't ready or able to make such an interaction, it occurs if the therapist makes it, and expresses himself openly and non-imposingly in it. The patient, too, then finds within himself an ongoing experiencing process that has something of the openness, honesty, human warmth, [Page 42] and eventfulness which the therapist expresses on his side. There is only one interaction process, yet to some extent at least, its events are experienced in the organisms of both persons.

How Research Can Investigate Psychotherapy

Now to the question I have been building up to: If, for some of us, this is the nature of psychotherapy, then how can research investigate and measure it? I think that it is beginning to be investigated in two ways: first, in physiological terms, and secondly, in terms of increasingly refined observations of tape-recorded in-therapy behavior.

I want to say a few words about each of these.

Psychophysiological studies have been reported (Berlin, 1960; Matarazzo, 1958) showing that different interpersonal conditions involve different autonomic correlates. We have currently also made a start (Gendlin, 1961 c) at finding autonomic correlates of different manners of experiencing (as defined in the Process Scale) and of resulting changes.

Another line of researches shows that psychotic contents can be produced in the laboratory by inhibiting the normal interaction of person or body, as happens in dreams, hypnosis, and laboratory stimulus isolation. Psychotic contents seem to appear when stimuli for optimal interaction are reduced. Also, LSD, Carbon Dioxide, and other toxics produce them. I think it is likely that we will find, also, physical modes of restoring, rather than inhibiting, optimal organismic process, and optimal interaction. Currently, it is true, society seems more to want to tranquilize its schizophrenics — and itself — so as to avoid, rather than restore, personal interaction and optimal physical interaction. Yet, I think physical and psychological avenues of research are increasingly defining one optimal process of organismic life, physical, subjective, and interpersonal.

But, if it is the case that attitudes and interactions are part of restored organismic experiencing, then we need new observable terms to define and measure this restoration through interaction. This brings me to the second avenue of measurement of this process. Rogers' Process Scale, which I have already described, makes a beginning at defining a few such observable terms. Of course, at first, such terms are "intuitive" and "vague". But one moves to the stage of statistically reliable rating scales with them. And then one can further define and differentiate the observations on which the ratings are based, as Hart (1960) and others (Gendlin, Hart; Tomlinson, 1959; Tomlinson and Hart) at Wisconsin, and Holloway (1960) and Zimring (1958-1959) at Chicago are attempting to do now. Whatever the therapist can sense and interact with, that must, in the last analysis, turn out to be something observable. Unless there is a purely mysterious intuition, which I doubt, reliable so-called [Page 43] subjective estimates of behavior will, upon analysis, be shown to consist of more specifically definable variables of external observation, variables on which behavioral research can be based.

In the clinical field, carefully controlled experimental research is growing. At the same time some of us are coming to consider the central aspects of psychotherapy as implicit, sensed and felt experiencing, and personal interaction. I do not think that the prospects of science are less if this should indeed be a less superficial way of considering psychotherapy. Rather, we would need — and I like to think we are tending toward — the kind of theory and the kind of research variables which will give us a science of the concrete experiencing of interacting organisms, in terms of differentiated and defined measurements of observable variables.