In Creating Mental Illness, medical sociologist Allan Horwitz proposes a definition of mental illness, which he argues is close to the official definition used by the American Psychiatric Association in recent editions of the Diagnostic and Statistical Manual of Mental Disorders (DSM). The definition of mental illness that Horwitz defends depends essentially on the concept of a harmful internal dysfunction, which he explicitly takes from Wakefield's well-known approach (1992). Horwitz argues that the adequacy of such a disputed definition has to be judged by its usefulness and that a central task of this concept is to distinguish between mental disorders and normal reactions to social stressors. However, he claims that modern psychiatry has lost sight of this distinction and has come to include many normal reactions to difficult circumstances among the conditions it classifies and treats as mental disorders. He lays the blame for this trend at the feet of psychoanalysis and symptom-based approaches to classification of mental disorders, both of which tend to blur the distinction between internal dysfunction and normal reaction. Horwitz argues that while psychiatry has moved toward an explicitly neurophysiological understanding of mental illness, it has retained the overly broad definition of mental disorder. He suggests that this serves the interests both of the profession and the interests of the powerful pharmaceutical industry. His conclusion is that psychiatry should narrow its classification of mental disorders to those conditions that are clearly harmful internal dysfunctions, and that as a society we should pay more attention to the social conditions that cause people distress.
Horwitz provides a wealth of detail to make his case. He surveys the history of psychiatry in the twentieth century, [End Page 70] paying particular attention to the influences leading to the publication of DSM-III in 1980. His account tends toward the skeptical, prioritizing sociological and economic explanation over the logic of scientific discovery, and is heavily influenced by Kirk and Kutchins (1992). He documents many cases of surveys that report a large proportion of the population having mental disorders and traces these apparent discoveries to overly-inclusive criteria that depend solely on lists of symptoms. He argues that surveys fail to ascertain whether the symptoms indicate disorders. For example, he suggests that wrestlers who choose to lose weight quickly to qualify for particular weight classes and who then eat large amounts of food afterward would satisfy criteria for bulimia nervosa, when in fact they do not have a mental illness. Horwitz argues that surveys claiming, for instance, that 23 million people suffer from generalized anxiety disorder or that 14% of Americans have an alcohol disorder, tend to massively overestimate the number of people with mental illness. He insists that for a condition to count as a valid mental illness, it should arise "in the absence of any cause that would expectably give rise to them, be of severity and/or duration disproportionate to their precipitating cause, or persist after the causes that gave rise to them disappeared" (98). He argues at length that most of the conditions currently classified as mental disorders do not meet any of these criteria. In his view the clearest examples of valid mental disorders are schizophrenia and manic depression, and he is suspicious of the many other conditions that are often classified as mental illness.
To consolidate his argument, Horwitz addresses the purported biological understanding of mental illness and points out methodological problems in many studies that apparently prove the genetic basis of many mental disorders. He claims that the rates of most mental disorders apart from schizophrenia vary widely across different cultures, and he uses this to underline his skepticism about the validity of the classification of these less serious conditions as mental illness. Turning to talk therapy and medication, he presses the point that therapeutic effectiveness is no proof that the original distress was really a medical condition rather than a symptom of a social problem.
Many of Horwitz's criticisms of psychiatry are thoughtful, well-argued, and reasonable. He poses many important and pressing questions. However, his starting definition of mental disorder is highly problematic, and this flaw undermines the central argument of the book. He shows little awareness...