Creating Mental Illness
by Allan V. Horwitz
University of Chicago Press, 2002.
Book Review by Lynn E. O'Connor, The Wright Institute, Berkeley, California, USA.

Creating Mental Illness, by sociologist, Allan V. Horwitz, is a fascinating and scholarly critique of our classification of mental disorders. Horwitz begins by stating boldly that many so-called mental disorders according to our current symptom-based system of classification, are not really mental disorders at all, but normal responses to social stress, relationship problems, work or other problems in living, or social deviance that may be in some cases, culturally supported. He carefully defines "mental disorder" using Wakefield's (1992; 1993) definition: "a mental disorder exists when some internal psychological system is unable to function as it is designed to function, and when this dysfunction is defined as inappropriate in a particular social context". Prior to the 1970s, psychiatry was heavily influenced by Freudian theory, and mental disorders were seen as "non specific reflections of unconscious mechanisms, not as discrete symptom-based diseases." Psychopathology was vague and viewed as occurring on a continuum, with the pathological merging into the normal. The early Diagnostic and Statistical Manual of Mental Disorders (pre DSM-III) focused on underlying conflicts and unconscious processes considered at the time to be central to all people's mental problems. This Freudian system allowed psychiatrists to spend less time treating the severe mental disorders seen primarily in mental hospitals, and more time treating the "neurosis" of normal people, which were, according to dynamic psychiatry, not exactly normal. Horwitz observes that it was the Freudians, through the pre DSM-III era, who first pathologized the normal problems of living along with personality and relationship problems in this vague continuous classification, based on underlying unconscious mechanisms and etiology. However, this system of classification, dynamic psychiatry, wasn't amenable to research, it was unreliable -- one couldn't assume that two different psychiatrists would diagnose the same person similarly. Furthermore, in this system, symptoms were considered unimportant in themselves, instead they too represented unconscious processes. So this system of classification was problematic in an increasingly science-based medicine.
Medicine was no longer accepting "case studies" as respectable science, however they represented the only research that could be done given the psychoanalytic framework. The premises of Freudian theory could not meet the basic requirements of science -- they included constructs and assumptions that could neither be falsified, nor tested scientifically. For example, Horwitz notes "how could anyone be shown NOT to have an oedipal complex when protestations that one had no such desires were taken as evidence of resistance to admitting its presence?… Freud, for example, interpreted his patients' refusal to accept his interpretations of their symptoms as confirmations of his theory of repression." Major concepts in dynamic psychiatry, "especially the unconscious, were inherently not subject to measurement and others, such as the ego, id, and superego were too vague to be operationalized." Therefore, the theory couldn't be tested by scientific method, and was increasingly unacceptable to medicine in general. As the result of this pressure on psychiatrists to get more scientific, to keep up with the rest of medicine that was moving towards empirically based and validated treatments, there was a rapid paradigm shift in the field. It initially centered in a group of influential and respected research psychiatrists at Washington University, Saint Louis, and then extended broadly to the whole field as they impacted the writing of the DSM-III.

The Washington group wanted to classify mental disorders as categorical rather than continuous and vague, such that they could be reliably diagnosed and studied empirically, across samples and populations. Returning to a Kraeplin-like, symptom based system of classification, they developed 14 discrete disease entities, each with distinct observable symptoms. Taking off from the work of the Washington group and extending it to include a huge number of mental disorders (well over 200) the DSM-III, appeared in 1980 and presented a whole new system of classification, based on symptoms without speculation about etiology and avoiding any particular theory of psychopathology or psychotherapy. This new view of mental disorders Horwitz called "diagnostic psychiatry." The new classification system differed dramatically from the old. In dynamic psychiatry mental illness was seen to occur on a continuum and were categorized by vague and continuous unconscious mechanisms; in diagnostic psychiatry mental illnesses were considered distinct categorical entities with specific symptoms. From Horwitz's perspective, this says this made sense for the psychotic disorders, for bipolar disorder, for depression with psychotic features, but not for the myriad of other problems for which people came to treatment.

Additionally, in order to gain the support of the practicing clinicians as well as those in research medicine/psychiatry, the authors of the DSM-III went far beyond the 13 limited categories proposed by the Washington group, and included all of the problems that clinicians were treating, and that had initially been pathologized by the Freudians in the earlier DSM. This included problems in living, responses to stresses, problems in relationships, and the more pervasive personality styles or the "personality disorders".

Horwitz critiques both "dynamic psychiatry" (that is the diagnostic schemas of theories of the Freudians) and "diagnostic psychiatry" (that is the current field which supports a biological and symptom based classification of mental illness). He suggests that many problems that bring people to therapy are neither mental disorders nor diseases, and neither system of classification allows for this reality. Furthermore he says that outside of the psychotic disorders, bipolar disorders and depression with psychotic features, even those that are valid mental disorders, are not really discrete categorical entities but are more realistically seen on a continuum. For example he acknowledges that problems such as suicidal depression in the absence of any external reason, crippling obsessions and compulsions or phobias that prevent a person from living a reasonable productive life, etc. are indeed mental disorders but he insists that these are more accurately described as being continuous rather than categorical.

Now as a clinician I think about this distinction quite a bit. I am rarely comfortable with the categorization of many of the personality disorders, and I agree that personality problems are on a continuum. However there are Axis I disorders (clinical syndromes) that I believe fit the definition of mental disorder and are not so clearly a matter of degree. For example depression (without psychotic features), obsessive compulsive disorder, and panic disorder seem quite discrete or categorical when they appear in the treatment setting. However, I agree that personality disorders and the less severe forms of anxiety and depression may be better described as occurring on a continuum. I often say that in my work as a practicing psychologist, initially I treat patients for mental disorders (discrete, categorical clinical syndromes), then they get better and I become an executive coach, or they only partly get better and I try to help them with these less than discrete personality problems.

Many of the problems we treat, Horwitz defines as problems in living, as non-pathological responses to life stresses such as divorce, loss of job, illness in the family, aging parents, children with difficulties in school etc. To classify these kinds of problems as mental disorders is indeed turning the concept into one that is basically a social construction, developed in response to political, social, and economic pressure. Horwitz rather convincingly presents evidence that some of the "new disorders" such as multiple personality are in fact iatrogenic disorders, that is caused by the therapists. He suggests that many of our more common disorders in fashion today have also been caused by the culture of psychotherapy so pervasive in our society. For example, he discusses how "social phobia" has taken a naturally occurring temperament different, namely shyness, and transformed it into a mental disorder that is claimed to be common and pervasive, and treatable with SSRIs, with a huge profit for the drug companies. Horwitz has a good point.

After taking us through the rapid transformation of our diagnostic system of classification, the modern symptom-based DSM, Horwitz goes on to describe the flaws in the epidemiological research by which we have come to believe that disorders such as social phobia, depression, or sexual dysfunction are far more common than in fact they are. As he explains, when a clinician is making a diagnosis, she asks the client a series of questions related to specific symptoms, if they have ever been experienced and if so, how often, for how long, and how severely, and in what context. For example, the clinician may ask "have you ever felt depressed for more than a two week period?", followed by a question about the context, in order to discover if there was a precipitating factor such as a divorce or a loss. I think we can all agree that for a person to have experience a period of depression, more than two weeks long, following a divorce or some other loss, is hardly a mental disorder. So the clinician may make a diagnosis based on symptoms, however context is the mediating factor. However, as Horwitz describes it, in the epidemiological studies, people are asked "have you ever felt depressed for more than a two week period?" without obtaining the mediating situational factors. Consequently, numerous people become a depression statistic, when they were actually responding to a temporary situation such as a recent loss. As a result, the frequencies of mental disorders that these studies discover in the general population are highly inflated.
The book also covers research on psychotherapy and drug treatment, showing their strengths as well as their weaknesses. Horwitz isn't exactly a social constructionist, he's not claiming that there are no such things as mental disorders, or that mental illness is a myth, or that mental illness is all a social construction. But he makes a great case for questioning our diagnostic system of classification and I came away from reading the book feeling slightly humble. I found Horwitz's representation of biology and the genetics of mental disorders to be more reductionistic than is the actual state of biological psychiatry today----he seemed more down on biological explanations than may be warranted and from my perspective he makes too much of a distinction between the biological and the social, as if they aren't entirely interdependent. But Creating Mental Illness is a great read and I found it thought-provoking and overall, I think the author has made an important observation about weaknesses in our system of classification, and we would be wise to reassess some of what we have accepted as gospel in our field.