PREFACE  (The preface from the 1999 addition is in Times Roman font. Material added to the 1999 edition is in this [Arial] font.)

The editor of this re-issue has asked me to write a new preface. As with the 1999 edition itself, I have revised the preface mainly by addition, since the original still stands.

This book concerns the relationship between what is called mental illness in our society and the larger context in which it occurs. If the science of “ mental health and mental illness” is to advance, it may need to consider not only the microworld of biology, but many larger worlds, notably, emotions, relationships and social systems. The individual self is not made up only of physical systems, but also emotional and relational ones, and is deeply and intimately involved in social systems. Perhaps we need to conceptualize the basic problems in this area in a new way that would integrate individual and social systems. Labeling theory, which concerns the social system, can be integrated with a theory of emotional/relational and social dynamics that occur within, between and around the self.

To simplify her task, when God created an onion, she probably started with the inner core working outward. However she could have also started with the outer skin, working inward. Even if you are not God, it doesn’t seem to matter which way you start, as long as you realize an onion is not just the inner core, nor is it only the outer skin. We should describe each of these layers, and at least sketch out theories that relate them one to another.

Biopsychiatry has so far convinced the public that the innermost core of the self, an individual physical system, is all that matters. The labeling theory of mental disorder (Scheff 1966; 1983; 1999) proposed the opposite trope (metaphor), that the societal reaction, one of the outer layers of the onion, was more important. But the truth of the matter is that most of the onion, the great bulk, is neither core nor skin. We need to move beyond both of these tropes, into the layers and layers that make up a complex phenomenon.

The latter chapters of this book apply new concepts to specific cases, indeed to actual dialogue within cases. For example, Chapter 9 provides a specific moment in a psychotherapy session that implies an intermediate layer, the complex family system dynamics between a mother and her two daughters that lead to the labeling of one of the daughters within the family.

Tom Szasz is one of the earliest debunkers of the idea that psychiatry should be based solely in biology. He is in the great tradition of rebels, those who have attempted to overthrow the status quo. In this light, Szasz has taken what seems to be an absolutely necessary first step, challenging the trope of "mental illness." If he hadn’t existed, we would have had to invent someone like him.

But Szasz has not offered any sufficiently detailed alternatives. Art and science, as William Blake urged, needs to deal with "minute particulars." Perhaps the next step is to establish a model alternative to the myth of mental illness, one that encompasses the minute particulars of social, psychological, and biological reality. The later chapters in this book concern one such alternative, a theory of the emotional/relational world as it has been described in Retzinger (1991), Scheff and Retzinger (1991) and Scheff (1994; 1997).


The grip that established metaphors and myths have even on science has been nicely caught by the philosopher Quine (1979):

The neatly worked inner stretches of science are an open space in the tropical jungle, created by clearing tropes [metaphors] away" (1979, p. 160.

That is to say, it often happens that before science can be applied, an obstructive metaphor has to be overthrown. The history of the physical sciences is full of examples. For example, progress in the astronomy of planetary motion was delayed for over a century because of the trope that the earth was the center of the universe. This idea is a correlate of universal ethnocentrism: the conceit that we human beings are the center and purpose of the cosmos.

Astronomers, like everyone else, took for granted that planets circled around the earth. In the l6th century, Brahe had made a very accurate charting of the transit of Venus. But he could not plot the shape of the orbit because he assumed Venus was orbiting around the earth. Kepler, who inherited and added to the data after Brahe’s death, was equally puzzled for many years.

The idea of a logocentric universe was so ingrained that Kepler hit upon the solution only inadvertently. In a largely irrelevant burst of creativity, he devised a geometric model of the planetary orbits based on transparent solids representing cubes, spheres, and polyhedrons. The model was absurd except for one feature; in his playful excitement, Kepler had inadvertently placed the sun, rather than the earth, at the center. But the "error" made him aware of a new way to compute the orbit.

Similarly, Einstein began work on the theory of relativity with a joke concerning persons passing each other on trains, trying to determine their speed relative to each other. He realized intuitively that this situation challenged the ruling trope that time and motion were absolute. Although he had a doctorate in physics, Einstein knew little mathematics. Apparently he had found it boring, and missed or slept through many of his classes. He had to get help to put his anti-trope into mathematical form. Like Kepler, he was a trope-clearer.

Like Kepler and Einstein, Szasz attacked a ruling metaphor in Western society. His target is the idea of mental illness, and its planetary system of concepts (symptoms, patients, hospitals, drugs as medicine, etc.). He was the first psychiatrist to forcefully and repeatedly claim that these concepts fit reality poorly. Since our society still accepts the myth of mental illness, Szasz’s attack is ongoing.

Accompanying Szasz’s challenge was the work of others in the anti-psychiatry movement: Goffman, Laing, my own labeling studies, and many others. These studies, like Szasz’s, were a step forward. But they were not sufficiently detailed, on the one hand, nor theoretically framed, on the other, to get to the core of the problem. Neither Szasz nor Goffman or Laing sought the equivalent of the mathematical formatting that enabled Kepler and Einstein to replace an obsolete view with one that better fit reality.

Although Goffman’s approach is sociologically sophisticated, it does not contain a theory of mental illness. He defines his terms conceptually at most, without attention to the problem of goodness of fit to actual instances. Laing’s work is psychologically sophisticated, but involved even less conceptual development. Szasz, finally, uses no concepts; his approach is stated entirely in vernacular words. This approach makes it easy for anyone to understand, even laypersons. But it is much too narrow and simplified to use for analyzing and understanding real cases, each of which is apt to be quite complex, like most human conduct.

Szasz’s reliance on vernacular words reduces his theory almost to caricature. For example, the terminology that Szasz suggests as an alternative to "psychiatric symptoms" is "problems in living." If adapted, this usage might help to de-stigmatize the sufferers. But the phrase is much too broad a tool for understanding, since it encompasses an impossibly vast realm of problems. Unrequited love, over-extension of one’s credit, and the physical incapacities of old age are certainly all commonly encountered problems of living, but they are not the particular types of problems that are usually designated as mental illness.

If Szasz had used the terminology "residual problems of living" (problems which don’t have conventional names), he would have come close to my own solution of the problem. In any case, a social theory requires statements of explicit hypotheses, all of which are couched in terms of conceptual and operational definitions. Labeling theory was a step in this direction, but it was too narrow to grasp the intricacies of emotion and relationship that generate functional and dysfunctional behavior.

One obvious limitation of labeling theory was that it dealt only with the societal reaction to residual deviance, but not with the origins of that deviance. It simply bypassed the question of the biological and/or psychological causes of symptoms. In some ways, this orientation was a strength. In some cases of what is labeled as mental illness, there are no symptoms. When I was observing mental hospitals in the 60’s in the US, England, and Italy, many of the patients were unhoused seniors who wintered in mental hospitals. In Stockton State Hospital in 1959, I found that there were many patients with no detectable symptoms who were being used as unpaid laborers. The vast laundry facility was run almost entirely by these patients, most of whom were Chinese immigrants to the US who had never learned English. Perhaps symptoms had led to their original hospitalization. But, in any case, symptoms no longer played a role in their "mental illness."

But the single focus of labeling theory on societal reaction is also a weakness, a sin of omission. If human beings are going to live in peace together, we need to better understand the origins of symptoms such as depression, delusion, compulsion and obsession, for example. These symptoms can be found not only in mental patients, but scattered throughout whole societies, even among ruling elites. They are an important aspect of the human condition, and need to be further investigated. Lemert’s study (1962) of paranoia by persons in high positions in organizations provides a model.

Closer to the dialogue that is to be analyzed in this paper, labeling theory had a further weakness: it focused entirely on the formal, official societal reaction to residual deviance. As the dialogue excerpted from Rhoda’s psychotherapy session suggests, there is also an informal process of labeling that takes place within families, before they has been any suggestion of a formal reaction from officialdom. Informal labeling in the family, it would seem, is the next layer down of the onion, right beneath its outermost layer, the formal societal reaction. In his extraordinary paper on the growth of paranoid symptoms in an individual inside of an organization, Lemert (1962) provided an earlier description of the process of informal labeling and its effects. But labeling is but one of many processes that occur within the emotional/relational world.

The Emotional/Relational World

This essay is a proposal for creating a dynamic model of the emotional/relational world, a world that is mostly disguised and ignored in Western societies. A recent cartoon in the New Yorker is apropos. A male client, lying on the couch, is saying to the analyst: "Look, call it denial it you like, but I think what goes on in my personal life is none of my own damn business!" Like much of the best humor, this caption contains a core of truth. We are all trained to ignore the emotional/relational world.

My approach requires integration between the social sciences in general, and between sociology and psychology in particular. It points towards a theory that could lead to the empirical study of actual social relationships, the core subject of all of the social sciences and a crucial subject for the mental health professions. With emotions and relationships routinely disguised and ignored in Western societies, social sciences that also ignore it serve a conservative function, helping to preserve the status quo in the emotional/relational world.

Goffman (All of his books, but especially 1959; 1959a; 1963; 1963a; 1967) was the poet and prophet of the emotional/relational world. But his work is only a beginning, since it doesn’t involve an explicit theory, method, or systematic evidence. Another problem concerns his treatment of emotions. Compared to most social science descriptions, his are three dimensional, dealing not only with thought and behavior, but also with feeling (Scheff 2000a).

But with the exception of one essay focused on embarrassment (1967), he portrayed emotions only by implication. There is considerable embarrassment, shame, humiliation and disgust implied in his representations, but it is seldom make explicit. In this respect, his methods of dealing with emotion are similar to the usual treatment of emotion in Western societies, disguising or misnaming.

On the other hand, there are fruitful studies that go much deeper into the minute particulars of human relationships, revealing whole new realms of information about the filigree of human conduct. One such study, by Labov and Fanshel (1977), two linguists, provides the basis for Chapter 9. They conducted an exhaustive microanalysis of the first 15 minutes of a psychotherapy session. They analyzed not only what was said but also how it was said, interpreting both words and manner (the paralanguage). They based their interpretations upon microscopic details of paralanguage, such as pitch and loudness contours. Words and paralanguage are used to infer inner states: intentions, feelings, and meanings.

With such attention to detail, Labov and Fanshel were able to convey unstated implications. Their report is evocative; one forms vivid pictures of patient and therapist and of their relationship. One can also infer aspects of the relationship between Rhoda and her family, since Rhoda reports family dialogues. Labov and Fanshel showed that the dispute style in Rhoda’s family is indirect: conflict is generated by nonverbal means and by implication.

Indirect inferences, from a dialogue that is only reported, are made in order to construct a causal model. Obviously, in future research they will need to be validated by observations of actual family dialogue. It is reassuring, however, to find that many aspects of her own behavior that Rhoda reports as occurring in the dialogues with her family are directly observable in her dialogue with the therapist. For example, the absence of greeting, and Rhoda’s covert aggression in the dialogue she reports with her aunt can be observed directly in the session itself (not included in chapter 9 but discussed in Scheff 1989).

The limitation of this and similar language based studies is the opposite of that of the work of Szasz, Goffman, etc. Where their work was largely theoretical, the linguists, working inductively, had too little. In particular, they had no theory of emotion and relationship dynamics that would help them interpret the family conflict they reported. Here I will show how a theory of these dynamics might be developed from close examination of dialogue.

The first edition of this book (1966) presented a sociological theory of mental disorder. Seeing mental disorder from the point of view of a single discipline, the theory was one-dimensional. The second edition (1984), except for slight changes, continued in this same vein. Since that time there have been substantial advances in the biology, psychology, and even in the sociology of mental disorder. What is now most needed is an interdisciplinary approach, one that would integrate the disparate languages, viewpoints, and findings of the relevant disciplines. Such an integrated approach would be far greater than the sum of its parts, the separate disciplines. In human conduct, particularly, the vital processes seem to occur at interfaces, in the intersections of organic, psychological, and social systems.

To use E. O. Wilson’s term (1998), what we want is “consilience,” the interlocking of frameworks from the relevant disciplines. Although not using that word, I had proposed a similar interlocking for the social sciences (Scheff 1997) and illustrated what it would look like with several of my own studies.  As Wilson indicates, many of the recent triumphs of the physical and life sciences have been based on the integration of the various disciplinary approaches.

As Wilson also indicates, there has been very little consilience among the behavioral and social sciences. Each of these disciplines goes its own way, ignoring the adjacent disciplines. Each emphasizes its own virtues, largely ignoring its weaknesses, as in the old song: “You got to accentuate the positive, eliminate the negative, tune in to the affirmative, don’t mess with Mr. In-between.”  Contrary to the song, we must begin to mess with Mr. In-between.

Given the need for consilience, is there any point in resurrecting labeling theory, yet another one-dimensional approach to the complex problem of mental disorder? Before preparing this edition, I have considerable thought to this issue. My decision that the theory still had value was based on the following ideas. First, while waiting for consilient approaches to be developed, headway can still be made with one- or two-dimensional approaches. As will be proposed in Chapter 1, biopsychiatry, an integration of biology and psychiatry, seems to have made many worthwhile advances in the understanding and treatment mental disorder. In the last twenty years, even one-dimensional studies of labeling of mental disorder have made contributions to our understanding, as in the work of Bruce Link and his colleagues. When consilient theories are developed, there will still be a need for approaches that are only one- or two-dimensional.

A second idea may be just as important, that of the devil’s advocate. Biopsychiatry, the dominant force in the field, like all disciplines, accentuates the positive. Labeling theory can be considered to be a counter-theory, critical of the weakest points in the dominant theory, and focusing on issues that it neglects. The two approaches can complement and correct each other, while we are awaiting Mr. In-between.

The original theory of mental illness presented in this book had its high water mark in the 1970’s, if perhaps only as a counter-theory. During that decade labeling was taken seriously in sociology and, to a lesser extent in anthropology, criminology, psychology, psychiatry, and social work. Its status began to wane in the next decade, and by the beginning of the 1990’s it had been all but dismissed by the mainstream disciplines. As we shall see in Chapter 1, there are still proponents of the theory. But the majority of scholars and practitioners have moved on to other interests.

There are two main reasons for the loss of interest. The most important is what is called popularly “the tranquilizer revolution,” and the accompanying rise of biological psychiatry. Beginning in the 1980’s and reaching its peak in the mid-90’s, most social scientists and practitioners formed the impression that the problem of mental illness had been solved, at least in principle, by biological psychiatry.  The public was persuaded by claims that the causes and treatment of mental illness had been shown to be biological. It was thought, and still is by many, that genetic causes of mental illness had been, or would shortly would be, found, and that psychoactive drugs could cure, or at least safely control, the symptoms of metal illness.

The first part of Chapter 1 will be devoted to exploring these claims. It seems now that although biological psychiatry has made advances, in the main its claims have still not been substantiated. These matters are too complex to deal with briefly, so will be raised in the first chapter.

A second reason for the declining interest in the theory were various critiques proposing that since the theory was not much substantiated by empirical studies, it should be abandoned. The most important of these critiques was that of Gove (1980; 1982) in two volumes on labeling theory. As with biological psychiatry, it now appears that the critiques of the labeling theory of mental illness were overstated. In Chapter 1, I will respond to Gove’s critique.

The research on which the earlier editions of this book were based on studies conducted during the period 1960‑1982. Since that time, there have been many extraordinary changes in the field of mental illness: the introduction of psychoactive drugs on a massive scale; the discovery of the neurotransmitters; the hope to find genetic causes of mental illness; the proliferation and development of psychological therapies; changes in the mental health laws governing commitment and treatment; and finally, a sizeable increase in the number and scope of social scientific studies of mental illness. This edition  updates the earlier ones, bringing these changes and their aftereffects into its purview.

In addition to these changes in the field since 1984, there have also been changes in my own point of view since the time of the first edition. First, the changes related to my work on catharsis of emotions, as reflected in the book on this topic (1979). Secondly, my studies of the emotions of pride and shame (Scheff 1990; 1994; 1997; Scheff and Retzinger 1991), and the link between these emotions and the state of the social bond. Third, my interest in connecting the world of every day life to the larger institutions in a society has directed my attention to dialogue as data (Scheff 1990; 1997). Finally, mostly as a result of my dialogue studies, I now think, like Wilson (1998) that it is imperative to integrate the separate disciplines that deal with human behavior.

These changes in my point view have had three main effects on this edition. First, they have led me to more strongly emphasize that the original labeling theory of mental illness, as presented in Chapters 3-5 below, is only one of many partial points of view. Each of these points of view is useful, but in the long run, it will be necessary to integrate the differing standpoints, especially the psychological, sociological, and biological approaches.

The second change involves increased emphasis on emotions and social bonds. The original theory was predominately cognitive and behavioral. In this edition, emotions and relationships are introduced, with a special emphasis on the emotion of shame as a key component in stigma and in the generation of the societal reaction to deviance. I now emphasize the role of pride/shame as Durkheim’s “social emotion,” and the interplay of these emotions with social bonds. Since emotions and bonds are biological, psychological, and social, increasing emphasis on the emotional/relational world, largely invisible in Western civilization, may offer a bridge between the disciplines. The original labeling theory was blind to the emotional/relational world; it dealt only with extremes of societal labeling and denial. In this edition, I extend the theory to include more subtle forms of interaction.

Two of the new chapters (8 and 9) illustrate the emotional/relational world by applying labeling theory to the social interaction between therapist and patient. Chapter 8 involves a psychotherapy session between an anorexic woman, “Rhoda,” and her therapist. The patient reports discourse in her family, especially dialogues between herself and her mother. This dialogue suggests that labeling of the patient occurred first in the family, before any formal labeling took place. This chapter points towards a modification and extension of the original theory.

Chapter 9 concerns the first meeting between an outpatient, “Martha,” and a psychiatrist. It turns into a sparring match between the patient, who want to convey her emotional/relational world, and the psychiatrist, who wants to ascertain the facts. This interview exactly reverses the situation between therapist and client from that of the session in Chapter 8. In the latter session, it is the therapist who seeks to interest the client in her emotional/relational world. In the session in Chapter 9, it is the patient who tries to interest the psychiatrist in her (the patient’s) emotional world. Because of her skill and patience, “Rhoda’s” therapist is successful; she introduces her patient to the world of emotions. Martha’s therapist, however, remains oblivious.

With respect to the original theory of labeling, after due consideration in 1999, I revised mainly by addition rather than by making large changes in the original text (Chapters 3-5). A new Chapter 1 took up the issues raised above about the perspective of biological psychiatry, on the one hand, and critiques of labeling theory, on the other.  Because I was unable to find a very concise statement of the theory of social control, I wrote a new chapter for the second edition (Chapter 2), stating the main elements of social control and relating them to deviance and to mental illness.

In the 1999 edition I resisted the temptation to make large changes in the text outlining the theory that was published in 1966 because it is still useful in its original form. Since the discovery of the role of the neurotransmitters, and the impetus to genetic research provided by DNA, researchers who investigate schizophrenia and the other major mental illnesses believe that they  are now asking the right questions, and that knowledge of the causes and cures of the major mental illnesses will be uncovered within their own lifetimes. This research, which grew out of the use of psychoactive drugs, has also convinced many psychiatrists that these drugs not only are important in the treatment of mental illness but also hold the key to the understanding and conquest of these problems. These are heady times for biological theories of mental disorder.

Although their hypotheses are plausible, they are still, at this writing, unproven. To date,  no clearly demonstrable linkage between neurotransmission or genetics has been found for any major mental illness. The idea that the mentally ill suffer from deficient neurotransmission or genes is only a theory. Furthermore, even if the connection were made, most of the basic issues involving the social control of mental illness would remain. Since the connection is still hypothetical, it is premature to discard the labeling theory of mental illness.

The same reasoning applies to what has been popularly called the "tranquilizer revolution."As will be discussed in Chapter 1, even the most useful of the psychoactive drugs do not cure mental illness‑ they alleviate the symptoms in what turns out to be a small proportion of the cases. And again, even if a drug treatment were found that could cure mental illness, the fundamental issues of social control would remain. When the painkilling properties of morphine were discovered, physicians called it "God's Own Medicine" because they thought it was a cure. It took many years to realize that it was only a painkiller. There may be a parallel to be drawn between the discovery of morphine and that of  psychoactive drugs. It has been less than fifty years since the large‑scale use of tranquilizers began. It may still be too early to evaluate their overall effects.

Since the 1999 edition was published, there has been a strong trend in objective studies of the effects of psychoactive drugs. First let me clarify what is meant by “objective” studies. By now (2005) it has become quite clear that most of the clinical trials had been financed by drug companies and that this circumstance compromised the findings. These studies departed from objectivity in many ways. For brevity, only two will be mentioned here. First, it is now obvious that in virtually all of the studies, in order to get funding, the researchers had given control over publication to the funding companies. What happened was that they simply did not allow negative or insignificant results to be published.

A second serious bias was testing only short-term effects, typically one month. Since drug companies did not favor testing effects over longer terms, few were carried out. But all the findings of these few show the same trend, a rapid decrease in effectiveness over the longer term. By one year virtually all drug effectiveness has disappeared. These findings strongly suggest that most drug response is almost entirely placebo effect.

Over the years, there have been many objective (not funded and controlled by drug companies) studies suggesting that the effect of psychiatric drugs is largely placebo. For example, Kirsch, et al, (2002) analyzed the FDA database of 47 placebo-controlled short-term clinical trials involving the six most widely prescribed antidepressants approved between 1987 and 1999. These included "file drawer" studies, i.e., trials that failed but were never published.

What Kirsch and his colleagues found was that 80 percent of the medication response in the combined drug groups was duplicated in the placebo groups, and that the mean difference between the drug and placebo effects was  "clinically insignificant" Other objective (non-drug funded) studies have been published showing similar results for most of the psychoactive drug types. (There are also 9 commentaries accompanying the Kirsch et al article, mostly favorable.) This study would seem to supply a much more accurate picture of drug effectiveness because it includes the “file drawer” results that are rarely tapped in published results. Similarly, the studies of long-term effects (one year or more) also need to be included, since the picture they provide is quite different from studies of short-term (typically one month) effects.

I am not arguing that the neurotransmitter hypothesis is incorrect, or that drugs are worthless; I am only suggesting that it is much too early to discard labeling theory, despite the significant gains that have been made. Some balance is required in evaluating the competing claims of both the somatic and the social theorists. In its heyday, there was a tendency in sociology to overstate the claims of labeling theory. To avoid overstatement, in the 1984 edition I made two changes in the original text. First, I relinquished the "single most important" phrase in Proposition 9, stating instead that labeling is among the most important causes. The issue of the order of importance of the various causes is empirical anyway and should not have been reduced to a theoretical claim.

The second change involves qualifying the contrast between the two poles of the societal reaction. Originally, I called the reaction to deviance that was opposite to labeling "denial"; in this edition I have changed it to "normalization." In fact, denial is only one of many differing ways of reacting to deviance, such as rationalization, ignoring, and temporizing.

In the context of mental disorder it is important to note that treatment is not necessarily a labeling reaction. Labeling, in the sense I use it, always involves stigmatization; there is an emotional response as well as special label. Any form of response which does not stigmatize, such as skillful and humane psychotherapy and hospitalization, may also be a form of normalization. In some ways, the term labeling itself is perhaps unfortunate, since it has become fashionable to apply it to mere classification. What is needed is a more forceful term, one that would connote both labeling and stigmatization, so that a distinction could be made between reintegrative and rejecting classification, as in Braithwaite’s  (1989) approach to crime control.

It may help give perspective if I locate the labeling theory outlined in this book with respect to other “anti-psychiatry” approaches, as they have been called. Like the viewpoints of Goffman (1961), Laing (1967), and Szasz (1961), the theory in this book offers an alternative to the conventional psychiatric perspective. The basic difference from the other anti-psychiatry approaches is that I offer an actual theory of mental illness. That is, I propose a possible social scientific solution to the problem of defining and treating mental illness. The theory is made up of concepts that are at least partially defined, explicit causal hypotheses, and applications to real events. This theory is therefore testable, as Gove and others were able to show in the early critiques of the theory.

Although Goffman’s approach is sociologically sophisticated, it does not contain a theory of mental illness. He defines his terms only conceptually, with little attention to the problem of goodness of fit to instances. Laing’s approach is psychologically sophisticated, but involved even less conceptual development. Szasz, finally, uses no concepts; his approach is stated entirely in vernacular words. This approach makes it easy for anyone to understand, even laypersons. But it is much too narrow and simplified to use for analyzing and understanding actual cases, each of which is apt to be quite complex, like most human conduct.

Szasz makes the case, very well, that the medical model is not appropriate for most cases of what is designated to be mental illness and therefore, that the term mental illness itself is inappropriate. I agree. But in order to make my argument understandable, I have resorted to that inappropriate terminology, only because it is coin of the realm. In this book, it should be understood that every time I use the term mental illness, it should be seen as encased in quotation marks. My own terminology involves a sociological concept, as explained in Chapter 3, “residual deviance.”

Szasz’s reliance on vernacular words reduces his theory almost to caricature. For example, the terminology that Szasz suggests as an alternative to “psychiatric symptoms” is “problems in living.” If adapted, this usage might help to destigmatize the sufferers. But the phrase is much too broad, since it encompasses a vast realm of problems. Unrequited love, over-extension of one’s credit, and the incapacities of old age are certainly all commonly encountered problems of living, but they are not the particular types of problems that are designated as mental illness. If Szasz had used the terminology “residual problems of living” (problems which don’t have conventional names), he would have come close to my solution of the problem. In any case, a social theory requires statements of explicit hypotheses, all of which are couched in terms of conceptual and operational definitions. The labeling theory provides these, the other anti-psychiatric formulations do not.

To illustrate one of the new directions advocated here, I will give one last example of concept development. The field of social work has been showing an interest in extending psychiatric diagnoses to include social dimensions, in addition to individual ones. Karls and his associates (1994) have been attempting to modify the Diagnostic and Statistical Manual (DSM), the bible of practicing psychiatrists, by adding social dimensions that have never been included in the DSM.

The PIE system proposed by Karls et al implies two social systems in the DSM: the subject’s  system of social roles (family, work, and other interpersonal roles.) and any of six larger social systems (economic, educational, etc) in which he or she is involved. This initiative seems to be a step in the right direction, toward overcoming the individualistic bias in the DSM. However, the Karls proposal still seems to focus on individuals, rather than considering the new dimensions to be independent aspects of the situation in which diagnosis occurs.

A more radical break with the labeling aspects of the DSM would be to establish the role and social system dimensions as independent axes, so that what is being classified is not only the individual, but the context. The new DSM might then have only two major axes, one individual, one social and situational:

I Individual: Physical and psychological dimensions

II Social: family, neighborhood, community and societal dimensions.

One outcome of such a system would be a move toward classifying functional and dysfunctional families, neighborhoods, communities, and nations.  For example, a school environment in which the strong get away with bullying the weak could be rated as dysfunctional independently of the individual being diagnosed.  Similarly, any or all of the social systems that foster racism, sexism, homophobia, blind nationalism, etc. could also be rated as dysfunctional.

Such a scheme would raise conceptual and research issues not usually addressed. For example, the whole problem of how individuals can be functional even though involved in dysfunctional interpersonal and/or social systems is only hinted at in current discussions of “resilience.” The classification of solely individual dysfunction leads to the neglect to this kind of question. A whole new system of diagnosis is needed that integrates the DSM with interpersonal and social system diagnosis.

It is my hope that this edition will provide a clear statement of a sociological approach to mental disorder, and at least some small steps toward integrating it with other approaches to the understanding and treatment of mental disorder.


Karls, James, and Karin Wandrei (Editors). 1994. Person-in Environment System: The PIE Classification System for Social Functioning Problems. Washington, DC: National Association of Social Workers Press.

Kirsch, Irving, Thomas Moore, Alan Scoboria, and Sarah Nicholls. 2002. The Emperor's New Drugs: An Analysis of Antidepressant Medication Data Submitted to the U.S. Food and Drug Administration. Prevention and Treatment. 5 (July), 1-11.

Bmi3preface    F4   5980