A. HISTORICAL BACKGROUND AND FOUNDATION OF THIS RESEARCH
This book is about small group (or face-to-face) communication and the analytical operations on which the views in the following pages are based. The study is not primarily about psychotherapy, even though a sound motion picture of a psychotherapy session provides the illustrative data. Certainly the orientation is never psychological or sociological in the classical sense. It is, rather, a study of behavioral integration — a study of how communicative behaviors are integrated to enact social process.
A basic tenet of the method is that the meaning of an event may be abstracted by seeing it in context. This research and viewpoint have evolved in a context of major change in the sei־ enees of man. Since the contextual changes which give this book its orientation are not universally known, and because these operations and the assumptions behind them cannot be taken for granted as they might be in a more traditional method, I must begin by describing the background to this study.
When this analysis began, a decade ago, only a few researchers were doing work with this general orientation. Since then, a general theory of communication has been taking shape through the shared effort of a number of theorists, and more and more people have come to use context analysis or similar approaches. I have been continuously in touch with many of these researchers during this decade, and this volume is my attempt at making a synthesis. I would like to emphasize, however, that the ideas at issue reflect the joint contributions of a number of linguists, anthropologists, sociologists, social biologists, and psychological scientists.
I have not wanted to write an abstract account of methods and theory. It seems more useful to illustrate principles with some data, even though it is data from a single transaction. After all, a cardinal principle of behavioral systems approaches is that we show behavioral events and not simply present our beliefs about them. Therefore I will describe specific behaviors from the filmed session at the end of each chapter. Since I am not sure that we can adequately describe our own research operations, I relegate my metaconceptions of the method to Appendix B.
Two major changes in orientation and a series of minor ones have occurred during the history of this research. We started with a psychological approach to psychotherapy. In the mid-fifties, a social-science era, we enlarged our field of observation in order to understand processes at the social level of organization. Then during the late 19 50s we shifted step by step to a focus on direct observation, cinematographic recording, and behavioral analysis. Each of these developments was guided and spurred by an increasing reliance on systems concepts.
We did not necessarily know what we were doing in making these changes. At least, we did not always have words for the methods or conscious rationale for making the changes. In the main, we simply were dissatisfied with the research methods we had been using and designed other approaches which might eliminate the difficulties. But our methods merely reflected trends of the social science and behavioral science phases of the systems movements. Soon these approaches appeared also in the research of other behaviorally-oriented workers.
Thus, the account of the research projects reported here represents an evolution of operations and views which have become general during the last generation. In this Introduction I will describe my own research experiences and relate them briefly to broader trends, and thus provide an introduction to the principles of context analysis and the theory of human communication elaborated in this volume.
Classical Research in Psychotherapy
In the early 1950s a group of us began to do research in psychotherapy. We were faced with certain classical problems of psychological research. We relied either on subjects and judges to interpret the events of psychotherapy, or on our own experiences as participant observers. We had reservations about the objectivity of subjectivist reports, a problem which was then being recognized. A series of research strategies were developed to deal with it.
The first method we tried was very popular at the time. We conducted structured interviews with therapists and patients, or else we had them fill out rating scales and questionnaires. Then we tried to determine statistically whether the various raters had achieved consensual validation.
This procedure caused us endless difficulties. We were never certain that various subjects and raters attributed similar meaning to the items. We could obtain a statistically significant consensus only among raters who had been trained in psychoanalysis by the same training analyst. And a broader doubt always hung in the background. Suppose various subjects did agree. Consensus in a belief system may mean only that members of culture share the same myth — a hundred or a hundred million Americans can be as wrong as an/ one person. Our objective was twofold: we were trying to objectify belief systems about psychotherapy and we were trying to find out what happened during psychotherapy.
In 1955 we had an opportunity to take a step forward. Dr. John Rosen invited us to observe at first hand his method of psychotherapy for schizophrenic patients, and we received a large grant from the Rockefeller Brother Fund to carry out research on the process. As a result, a group of experienced psychoanalysts, psychologists, and social scientists watched sessions of Dr. Rosen’s method of direct analysis every day for three years and discussed the observations in great detail.
The problem of reaching consensus, however, remained with us. Each of the observers noted different facets of the complex processes of psychotherapy and used a different frame of reference to express his observations (Scheflen 1958). Consequently, the various researchers disagreed strongly, and each published his own report (Brody 1959; English, Bacon, Натре, and Settlage, 1961; Scheflen 1961). Not only did we disagree about conclusions, but we disagreed as well about what actually took place.
Again we attempted to supply our observer reporters with objective measures and definitions. We worked out careful operational definitions, frameworks, and terms, and again we applied rating scales and other instruments. We made the chain of subjectivist reporting more direct, but we did not solve the problem of abstraction.
We employed still another strategy. We isolated, counted, measured, and correlated variables of behavior themselves. We studied head nods, foot wiggles, noun-verb ratios, and the like. But the scientific (or scientistic) gain in objectivity afforded by these measurements did not bring us closer to a view of the larger behavioral or communicational processes themselves. We had achieved the dilemma that characterized many sciences of man in the 1950s: we were caught between subjectivism and reductionism. The conceptions of experienced clinicians could capture a view of the whole but these overviews were not replicable or explicable. On the other hand, the behavioral tidbits we measured brought us a degree of objectivity about the tidbits but not a picture of psychotherapeutic processes as a whole.
Finally, we decided that if we took a motion picture of the psychotherapy session, we could insure at least that all of the observers looked at and noted the same behaviors as a basis for inference and abstraction. When in 19 56 we started to filmrecord sessions, we were not the first. Bateson (1954) had filmed interviews of psychiatric patients in their homes and a group in Chicago had filmed psychoanalytic sessions (Carmichael and Hazzard 19 55). Within a year or so, many investigators had filmed psychotherapy.
The films did not immediately help us. The clinicians did not want to look at them, and we still did not know how to study־ the behaviors recorded on film in a systematic and integrated way. The films began to collect in storage vaults as tape recordings had in the early 1950s. We didn’t realize that we had the makings of an operational breakthrough.1 But an era of direct behavioral research had begun.
Focus On The Social Level
Interest in social-level phenomena virtually exploded in the late 1950s. Science, industry, government, and the people of common culture suddenly became interested in togetherness, interaction, communication, and social relations. Words like talk and answer were replaced in everyday middle-class lingo by communicate and feedback. Social psychology, social psychiatry, social medicine, and social biology developed, and the existing social sciences were greatly expanded.
Those of us trained in the organismic sciences — medicine, psychoiogy, psychoanalysis, etc. — slowly came to a critical realization. Through study of only the individual participants in psychotherapy we could not grasp all of the phenomena of the process. We had to look at the relations between patient and therapist, as well as the organismic processes within these participants.
When we first talked about interpersonal relationships in psychotherapy, it seemed some kind of heterodoxy. Social theory appeared to be somehow opposed to psychological theory as an alternative hypothesis. We used to speak, for instance, of individuai versus social behavior. We did not yet have a clear idea about levels of organization, as the genera! systems concepts were still relatively unknown (Bertalanffy 1950, I960; Redfield 1942; Miller 1965).
Consequently, psychological theorists were trying to deal organismically with social-level concepts like communication. As psychoanalysts, for example, we thought of communication as the individua! expression of instinct or motivation or defense. The experimental psychologist tried to squeeze communicational processes into an S - R paradigm (Percy 1961). Steven (19 50), for example, defined communication ‘as the discriminatory response to a stimulus.’ In each of the classical disciplines, the first consequence of the social science era was the reformulation of phenomena like communication in the traditional framework of that discipline. Consequently, as many theories of communication developed as there were classical disciplines in the sciences of man. 2 We did not realize that a new frame of reference was emerging which would permit an extensive synthesis and blur old disciplinary boundaries.
By 1957 a great many social-psychological projects were underway. Communication was being examined by looking ‘between group members’ rather than ‘into’ individuals. But the long involvement with organismic-centered concepts constrained our efforts. We developed models of communication that focused upon interaction between individuals and we concentrated on groups as collections of individuals. Two popular ideas about communication came into prominence that were still limited by reductionistic viewpoints.
1. The action-reaction approach. People in a group sometimes take turns talking and listening. If the observer looks first to one speaker and then to the next one, he will see dialogue as an alternative, sequential activity. Thus, A speaks and В responds, then A speaks again, and so on. Often В reacts to what A has said. A’s action was seen as a stimulus to B, and B’s action was seen as a response to A. Thus S-R paradigms were wired in series to develop a model of human communication. Such a model can be called an action-reaction model.
This model lends itself to the lingo of information theory and cybernetics (Shannon and Weaver 1949; Wiener 1961). Mr. A can be said to encode a message which he transmits to Mr. В who decodes and transmits this message to Mrs. В who decodes this message and encodes a response to A, and so forth. In this approach the participants in communication are likened to telegraphic or other electronic devices which transmit and receive signals (Cherry 1961). In the 1950s certain social-psychological theorists had such machinomorphic concepts of psychotherapy that they spoke of exchanges of signals between therapists and patients (Colby 1960; Menninger 1958).
2. Social network concepts. Similarly, we could observe a group of people and decide which members talked to or contacted each other, or we could examine research questions such as who talked first or whose commands dominated the others. Such approaches were popular in the sociological sciences in the 1950s and can loosely be caHed network concepts (Cherry 1961; Chappel 1949; Haley 1959).
There are serious limitations to these individual-centered frames of reference. First of all, the interactional model at best depicts certain linguistic events in a formal conversation. In general, people do not take turns talking and listening to each other. They do not respond only to what someone has just said. Rather, they act within broader systems of events — to what has been said hours or even months before, to something unsaid, to what might be said, and to matters unrelated to the immediate transaction. And their nonspeech behavior is continuous. All participants hold postures and facial expressions at all times and they move together.
There is yet a broader issue here: each of these models was based on the concept of a group. Now, a group of participants is necessary for any social phenomena and obviously a group is made up of individual human organisms, but we don’t get very far looking at the matter in this way. If you begin by conceiving of a group and then make an analysis, the first step in reduction will take you back to the individual — back to the organismcentered models that have already proved inadequate.
The Development of a Behavioral Systems Approach
In the late 1950s we at last recognized something which took us beyond these group models to the study of behavior and the current behavioral science era. The recognition was this: A!though, physically speaking, a group is made up of human organisms, social organization and the processes of group are made up of behaviors; more accurately, they consist of integrations of behavior.3
We had already learned homologous operational truth about the organism — that the living organism is not merely a pile of organs or even an arrangement of them as it is on the anatomist’s table. It is an organization of processes among organ systems, i. e. , of organ behaviors.
Similarly, communication is not made up of people or even of individual expressions but of patterned relations among the behaviors of multiple people.
If we were to study communication, then, we had to retrace our steps from the high-level inferences of the psychological and social sciences and get back to the study of behavior itself. We had to examine action, describe it, analyze its form, and try to define meaning behaviorally. The psychodynamicist had to delay his inferences about personality and describe the behaviors on which he based these inferences. He had to describe what others in a transaction could see. And the sociologist had to describe the relations of behavior which brought and held people together. He could not merely classify groups and abstract qualities of relationships.
I think this realization was behind the widespread movement toward behavioral science which emerged about I960. This movement was not merely a revival of behaviorism and neobehaviorism. Behavior has come to be observed in its own right; that is, we study its structure and do not merely make inferences about neurophysiological or cognitive processes.
When we came to recognize the significance of behavior in communication, we discovered that others had long been working in the field.
One discipline had been at work on this very problem for three decades. The structural linguists had developed methods for studying the structure of language behavior (Sapir 1921; Bloomfield 1933). In the early 1950s these methods were not widely known in the other sciences of man and had not been applied to units of behavior other than speech or to units larger than the syntactic sentence. The behavioral science era resulted in a broad dissemination of structural linguistic approaches and in their extension to behavior in general. The methodology of the research reported in this book depends upon a structural linguistic orientation, as do other new methods for the study of behavioral integration.
The methods of the structural linguists are suited to the examination of behavioral structure. In the late 19 50s and the 1960s at least some principles of this method were applied to the study of behavioral integrations larger than the sentence (Z. Harris 1952; Scheflen 1966) and they were applied to behaviors other than language (Birdwhistell 1952, 1966; Pike 1954).
The study of behavioral relations requires that we emphasize operations for examining synthesis. The classical approach of analysis by reduction thus becomes one tactic in an ultimate strategy for reconstructing a view of the larger picture.
Methodologies for such synthesis which developed in the 1950s and 1960s have been described by several authors (McQuown et al. 1969; Pike 1954; Schneiria 1951; Rock 1962; Barker 1963; and Scheflen 1966). The variation of these approaches used in this research has been called variously ‘linguistic-kinesic analysis՝ (Brosin 1967), ‘communication analуsis’ (Scheflen 1906), and ‘context analysis.’4
The operations for synthesis require that we not concentrate on the relations of a behavior to some conceptual system about behavior, but rather that we examine the reiations of one behavior to another and these to a third until we have identified all of the behavioral elements that constitute a single defined subsystem of behavior or change.
Here is an analogy: The classical biologist would observe an animal and assign it to a class of animals in a taxonomy. The social biologist began to observe the relationships of the animal to other animals with whom it lived or related, and he visualized flocks, herds, or prides of animals. Then the behaviorally-oriented social biologist began to observe the calls, or the displays, or the aggressive rushes of animals — the behaviors rather than the animals. By observing these behaviors in relation to other behaviors, he visualized call systems, displays in a courtship dance, and mutual rushes in defense of a territory.
Since the processes of behavioral integration follow the principles of any other system, I will call this general orientation a behavioral systems view.
To study such complexity, we needed a technology. We cannot examine multiple modalities of behavior in detail at a single observation. Armed with eyes, ears, and a notebook, the observer has his hands full merely to hear the speech and to note the gross actions of one participant. The sound motion picture and, more recently, the video tape provided the needed technological means for thorough observations. Given a film record we can go over again and again the events of a transac״ tion, systematically observing one, then another, behavioral modality and testing their various relations until we have described the synthesis of elements in the over-all picture.
The shift to a behavioral focus, the ability to study complex behavioral integrations, and the development of the new recording technologies, all occurred interdependently. None of the aspects of a behavioral systems approach would have been possible without the others, just as the computer and cybernetics could hardly be imagined as separate developments.
All of these developments are offshoots of a systems orientation shift in the subject matter of study which occurred with the coming of the behavioral science era. Whereas researchers were excluded from many confidential activities of society, we now seemed to be welcome and could bring cameras. In the past we had studied the people of esoteric cultures, institutionalized deviants, and subprimates, but in the 1960s we turned to the study of the people next door, to our own urban cultures and middle-class institutions, and to the naturalistic study of our nearest other-animal relatives, the primates.
The behavioral systems researchers of the early 1960s deeply distrusted subjectivist accounts. We had never known how to handle a subject’s account systematically, so we ignored it. We looked at subjects but did not interview them. We did not care what they had to say and did not want to hear it. But a way of dealing with both visible behavior and subjectivist conceptions had to evolve.
M. Harris (1964) advocated that subjects be both observed and interviewed. Now in many projects, video recordings or movies of behavior are made in naturalistic situations and afterward are shown to the subjects, who are then interviewed in depth about their behavior as it appears on screen. An untraditional idea has gradually emerged from the work of Bateson (1955), Pribram (1964), and others, about the relations of cognitive and motor behavior. These matters are discussed in detail in Section В of this volume.
The Research Which Led to This Volume
In 1959 Dr. Birdwhistell and I established a project at Eastern Pennsylvania Psychiatric Institute to study communication in psychotherapy and other transactions through the context analysis of motion picture films. A clinical project at Temple University also was reorganized to study the methods of therapists. We decided that the two projects should corroborate on the analysis of a method. The Temple team, consisting mainly of clinicians, was to make a clinical appraisal, and the Eastern Psychiatric Institute team of Birdwhistell and Scheflen was to make a context analysis.
Experimental sessions, in which two psychotherapists, a schizophrenic patient, and (on two occasions) the patient’s mother participated, were observed and analyzed.
The First Publications
The sessions were examined clinically by Doctors O. Spurgeon English, Catharine Bacon, and Warren W. Натре. Dr. Arthur H. Auerbach made a content analysis. In the course of the study, the method of context analysis became clearer and less unwieldy. When the results of all the studies were brought together, it was apparent that the data was of interest to two different groups; psychotherapists, who might be chiefly concerned with the clinical data and conclusions; and behavioral scientists who were interested primarily in research method. We therefore decided to publish two monographs about the study: one describing the Whitaker-Malone method by clinical constructs and content analysis, and the other illustrating the context analysis of the first session. The clinical book, edited by Dr. O. Spurgeon English, was entitled Strategy and Structure in Psychotherapy (1966). The title of the method book was Stream and Structure of Communicational Behavior (Scheflen 1965) — the book upon which this revised edition is based.
B. THE EXPERIMENTAL SESSIONS
We knew that the psychiatrists, Doctors Carl Whitaker and Thomas Malone, had a unique approach to the psychotherapy of schizophrenia. They agreed to become subject-therapists and accordingly spent the first two weeks of October 1959 in Philadelphia being observed and studied. In order to approximate their usual living situation, they brought their wives and younger children along. Both families lived in apartment suites in a center-city hotel. To reduce extraneous pressures, they had few social or professional engagements: each of them addressed open psychiatric meetings only once. Observation of and discussion with them was limited to the research team.
A schizophrenic patient, whom we will call Marge, was selected by the research team from a neighboring public hospital. She and her mother were told that the procedure would be experimental. The patient lived in a residential unit of the Institute for Direct Analysis during the two weeks of the observational sessions. This unit is a three-story brick row house adjacent to Temple University Hospital in an upper-class urban neighborhood. The house had been remodeled previously, with every effort made to keep the atmosphere that of a private home. (Scheflen 1961). The unit was staffed by three young people who had been trained by Dr. Rosen’s direct analytic staff.
As was their usual practice, Whitaker and Malone asked that the patient’s mother be present at the first two sessions. Nine thirty-five minute sessions were arranged on consecutive weekday mornings: two with the mother, seven with Marge alone. After the ninth session, Malone elected to see the patient once for a tenth session. All sessions were tape recorded and observed by the researchers. The first, third, and ninth sessions were recorded on motion picture film. In order to keep some of the variables constant, the filming was done in the same living room of the residential unit in which the films of direct analysis had been taken. After each session the observers met with Whitaker and Malone for two hours to discuss the events of the session and their theoretical premises.
Present at each session were Whitaker, Malone, the patient, (and in the first two sessions) the mother, three or four observing psychoanalysts, and the project administrator.6
The Participants in Session I
When strangers come together for a transaction, information about them is offered to the others. They are often introduced or they introduce themselves. And whenever we give an account of a transaction we begin with an introductory statement about the participants. For instance, we begin a stage play or athletic context or an account of a party by reviewing the dramatis personae. In these introductions something is conveyed about each one’s background, his social position, and maybe about his personality.
Before Session I the researchers talked with Whitaker and Malone and asked them at length about their careers, methods, and plans for the experimental sessions. And several of us interviewed Marge and Mrs. V, obtaining their history and observing their interview behavior. In the preliminary interviews we told each of our subjects something of what we had learned about the others — approximately the information outlined in the description below.
Thepatient, Marge. A 17-year-old girl of Italian extraetion, Marge spoke in a soft, often childish, and barely audible voice. Her south Philadelphia idiom was immediately apparent, and her speech was colored throughout with Italian-American expressions typical of her background. In later sessions a drawl like Whitaker’s was clearly detectable. Physically, Marge gave the overall impression of well-developed femininity. She was rather tall and dark complected, not pretty, though she might be considered ‘kind of cute.’ Her dress and grooming seemed appropriate.
Marge was the only surviving child of Italian-born parents. A brother had died in infancy. Although the mother had had hál־ lucinations and other psychotic manifestations, the family apparently had been cohesive. The father, it was implied, was jealous and overly possessive. He died when Marge was sixteen. Following his death, Marge became sexually promiscuous, ultimately became pregnant and had an abortion. Marge had been hospitalized at another institution six months before Whitaker and Malone saw her. She had been seen frequently during this hospitalization by a psychoanalytically oriented resident physician and she seemed to be well-acquainted with the current concepts of the etiology of schizophrenia. This was evident in her many statements implying maternal deprivation, which seemed calculated to direct the psychiatrists’ attention to her mother’s shortcomings.
The mother, Mrs. V. At 48 years of age, heavy, stolid, and impassive, Mrs. V sat hunched forward with her knees about six inches apart and her dress pulled well down. She wore a polka dot dress, low-heeled shoes, and heavy-guage stockings. Her hair was worn straight, and she did not appear to be wearing make-up. She presented the picture of a good-natured immigrant peasant.
Mrs. V did not seem aggressive, hostile, or cold, as schizophrenogenic mothers are often considered to be. She was plain and self-contained, yet had a faint pleasant smile and occasionally displayed a hint of friendliness or coquettishness. She did not show the marked shifts in total posture that Marge did. In fact, she moved very little. Her history indicated that she had psychotic symptoms, such as an inability to eat for months, and ideas of hallucinations about the flight from her home. Her affect during the sessions was slightly flattened. At this writing Mrs. V is a patient in a state mental hospital.
She spoke in a typical nonstandard Italian-American dialect. The history, as she presented it, seemed calculated to turn aside blame, cover the family secrets, and make psychosis seem impersonal and unrelated to their emotional lives.
Psychotherapist, Whitaker. Carl Whitaker was born in Raymondville, New York, in 1912. He trained in medicine at Syracuse University and in psychiatry at the University of Louisville in Kentucky. He was formerly Chairman of the Department of Psychiatry at Emory University, then leader of a private group practice in Atlanta, Georgia. He is currently Professor of Psychiatry at the University of Wisconsin.
Large, youthful looking, and bespectacled, Whitaker is generally serious and thoughtful. He has a droll sense of humor, a quiet amiability, and an occasional flair for the provocative and unconventional.
He spoke sixty times during the session, considerably more than Malone, and the content of his statements fell into five clearly discernible types of comments (see Section A). Whitaker spoke only to the women, and at no time did he and Malone speak to each other.
Psychotherapist, Malone. Thomas Malone was born in Mahanoy City, Pennsylvania, in 1919. In 1948 he began to assist Whitaker and subsequently decided to study medicine in order to become a psychiatrist. He graduated from the Emory University School of Medicine in 1953. Malone is currently in practice at the Atlanta Psychiatric Clinic.
Youthful, sturdily built, and good-looking, Malone was more active in general, more verbal in discussion, and more aggressive in manner than was Whitaker. Like Whitaker, he tended to be serious, introspective, and candid about himself and his work. He gave the appearance of a scholarly intellectual and was more inclined than Whitaker toward conceptualization and research interest. In Session I he was serious, untalkative, somewhat stiff, and immobile kinesically. His speech was formal when he was talking to Marge and less so when talking with Mrs. V. In later sessions he was active and informal.
The Basis for Our Study: Communicative Behavior
In communicational theory we say that the behavior of each participant is communicative; that is, it is patterned and structured according to some tradition. If so, it is potentially recognizable and meaningful. Therefore, behavior is the basis for the social processes of communication.
I do not want to suggest an old dichotomy by such a statement. I am not implying that some behavior is communicative and some is not. I do not believe it probable that anyone will behave in a nonstructured and nontraditional way. We learn to behave systematically in becoming socialized and enculturated, and it is very hard to behave in any other way. So it is redundant to speak of communicative behavior, when all behavior is apparently patterned and capable of being recognized and comprehended, i. e. , communicative.
But another type of distinction is critical in communicational theory. While each participant follows some customary blueprint and agenda in guiding his contribution, a considerable range of variation is allowable. Thus a participant can modify, shape, and adapt his behavior to a given situation. Certain unit forms of behavior which are traditional on the social level were employed in Session I (discussed in Part I of this book). There were, of course, individual styles and particular management and adaptions which the participants made in accordance with their own plans, sometimes directly influenced by the special situation of Session I (Part II).