Psychiatric Deviance: Diagnostic Labeling
Part 1. "On Being Sane in Insane Places"
Introduction to Rosenhan's Study of Psychiatric Labeling
In his book Outsiders (1963), Howard Becker presented a hypothetical example of a law-abiding boy who was falsely accused and arrested as a delinquent to illustrate a central conceptual principle of labeling theory: it's not what you do, but how others define you that makes you socially deviant. Becker speculated that if such "bum raps" can occur in the legal system where rules of evidence and due process protect the accused, then cases of false accusation "probably occur much more frequently in nonlegal settings where procedural safeguards are not available" (1963: 20). A fascinating study by David L. Rosenhan assigned for this unit, "On Being Sane in Insane Places" (1973), offers some compelling empirical support for Becker's speculation. Among other things, this observational study demonstrates that false accusations of deviance-or, rather, false diagnoses-are, indeed, a very real possibility in psychiatric settings.
This and other discoveries came about through the novel but effective strategy Rosenhan's participant observers used to gain new insights into psychiatric social control: they put themselves on the receiving end of it. Rosenhan, a psychologist, and eight other normal people (including three psychologists, a pediatrician, a psychiatrist, a painter, a housewife, and a graduate student) individually went to the admissions offices of 12 different psychiatric hospitals and complained of hearing voices that said "empty," "hollow," and "thud." It is important to note that this particular set of "symptoms" was completely contrived and had never been reported in the psychiatric literature. Nonetheless, in every case these "pseudopatients" were falsely diagnosed as having a severe mental disorder and were admitted to the hospitals on a voluntary basis. Once the pseudopatients were taken to a psychiatric ward, they ceased reporting any symptoms and behaved just as they normally would. To be discharged from the hospital, each pseudopatient had to convince the staff that he or she was really sane.
Some of the results of this study might strike you as startling, even nightmarish. Despite their apparently normal behavior following admission as voluntary patients, Rosenhan and his collaborators were held in these institutions for an average of 19 days. In one case, a pseudopatient was hospitalized for nearly two months! Even though they eventually succeeded in gaining discharges, none of the pseudopatients was certified to be truly normal or sane upon release. In all 12 instances, the observers left the institutions with the psychiatric label "schizophrenia in remission" entered into their permanent records.
Rosenhan's dramatic findings on the ease of admission and the difficulty of discharge in psychiatric institutions became the subject of considerable controversy, as we shall see later. However, observational data gathered by the researchers during the course of their hospitalization are of even greater significance for a sociological understanding of psychiatric treatment as a process of social control. The pseudopatients took detailed field notes on routine patterns of activity as well as unusual incidents they observed on the psychiatric wards. Most of Rosenhan's descriptive analysis of the experience of psychiatric hospitalization is based on qualitative data taken from these notes. Throughout his article, Rosenhan uses specific anecdotes from field notes to typify or illustrate general observations and common impressions reported by the pseudopatients. For instance, consider the following "conversation" between a pseudopatient and a psychiatrist:
If, as Rosenhan indicates, this type of encounter were observed frequently on the hospital wards, what would you conclude about the nature or quality of social interaction between patients and staff? Rosenhan's conclusion, which he attempts to document with this and other anecdotal evidence, is that hospital staff members tend to "depersonalize" patients and avoid meaningful social contacts with them. Rosenhan s skillful presentation of these qualitative data not only provides a rich description of the peculiar features of the process of interpersonal avoidance but also conveys to the reader a subjective sense or feel for its dehumanizing impact on patients. By encouraging an understanding of social control from the underdog's point of view, Rosenhan s qualitative analyses serve as an excellent example of the humanistic approach to deviance research advocated by Becker and many other sociologists within the labeling tradition.
Despite the intuitive appeal of qualitative insights from participant observation studies, hard-nosed researchers often object that impressionistic or "soft" data can all too readily be slanted or selected to fit the preconceived biases of the investigator. Specifically, isn't it possible that Rosenhan chose to emphasize a relatively few, shocking incidents of depersonalization and that he could just as easily have cited numerous anecdotes portraying warm and therapeutic relationships between staff and patients? Or, when vague and implicitly statistical terms like frequently or rarely are used to describe certain activities, isn't it reasonable to ask exactly how frequently or how rarely these actions were observed to occur? Apparently anticipating such objections, Rosenhan instructed his observers to gather quantitative data on a variety of routine behaviors and events that occurred during their hospitalization. Thus, Rosenhan was able to buttress many of his generalizations about interpersonal reactions to patients with quantitative evidence. For example, his article includes a table showing data from four hospitals where pseudopatients kept a count of the responses they received after approaching members of the staff with simple requests for information. When you read the article, look at the percentages of encounters in Table 1 in which staff did not even stop to make eye contact with pseudopatients, and judge for yourself whether Rosenhan is justified in characterizing staff-patient relationships as "depersonalized."
At this point, however, it will probably come as no surprise to you that many psychiatrists do not believe that Rosenhan s conclusions are justified by his evidence. His article became the subject of almost unprecedented controversy following its publication in the prestigious journal Science. In no fewer than 15 letters to the editor of Science (Fleischman et al., 1973) and seven critical essays prepared for the Bulletin of the Menninger Clinic (Wiedeman et al., 1973; Hoizman et al., 1973), psychiatrists voiced their particular concern with Rosenhan's conclusion that "we cannot distinguish the sane from the insane in psychiatric hospitals" (1973: 257). Rosenhan's critics attacked his findings on the false diagnoses of 12 pseudopatients and defended the ability of the psychiatric profession to distinguish various forms of psychoses from "nonpsychotic" (normal?) behavior. At least one critic implied that the pseudopatients may not, in fact, have acted nor been as normal as Rosenhan indicated in his article! At the other extreme, several psychiatrists attempted to account for his findings on admissions and discharges by arguing that the pseudopatients had by chance encountered 12 hospitals with unusually incompetent diagnostic procedures.
Whatever the scientific merits of Rosenhan's observational research, the intense critical reaction to the issues he raised about the validity and reliability of psychiatric diagnoses highlights a crucial ideological implication of his work. By questioning the credibility of diagnostic labels, Rosenhan, in effect, challenges the legitimacy of the medical model of deviance upon which psychiatric social control is based. Indeed, the goal of Rosenhan's analysis is not only to describe the uses of social control in psychiatric settings but also to raise the more troublesome issue of whether individuals should be subject to the potential abuses of being labeled as mentally ill. In this respect as well as in many others, Rosenhan's study falls squarely in the humanistic, value-engaged tradition of Becker and other labeling theorists.