“TO DEFINE TRUE MADNESS,” said Shakespeare’s Polonius in Hamlet, “What is’t but to be nothing else but mad?” He may well have been right. But that doesn’t seem to be the view of the specialists paid to treat the mentally ill. For more than four decades, leading American psychiatrists have been obsessed with the problem of diagnosis. The ever-fatter volumes of the Diagnostic and Statistical Manual (DSM) of the American Psychiatric Association have produced enormous profits for the organization that publishes them. The third edition brought in $9.3 million. The fourth was still producing $5 million a year more than a decade and a half after its publication in 1994, and its most recent incarnation, the fifth edition of 2013, sold $20 million worth of copies in its first year. Less clear than their profitability is whether any of these documents has advanced our understanding of mental illness or brought us closer to identifying diseases that actually exist in nature.

Allan Horwitz’s new book, DSM: A History of Psychiatry’s Bible, seeks to understand where this obsession with diagnosis came from. It turns out that addressing the problem of classification provides a great deal of insight into the evolution of the profession. Though Horwitz is a sociologist by training, his approach is chronological and firmly historical, retaining throughout a focus on context: the professional, economic, and cultural forces that shaped successive understandings of mental illness, and thus successive iterations of psychiatry’s diagnostic systems. He shows these to be as much the result of social processes as scientific ones. The key takeaway: The growth in the scientific understanding of mental disturbance has played an insubstantial role at best in creating the labels we use to distinguish among various forms of mental illness.

Psychiatry as a specialty first emerged in the early 19th century, and for more than a hundred years was mostly content to speak of mental disorders in terms of broad categories inherited from the ancient Greeks: mania, melancholia, dementia, and idiocy. Some “alienists,” as psychiatrists were called, invented more elaborate lists of mental disorders, but none was broadly adopted, in part because the men running the vast network of asylums that increasingly confined those with serious mental illness saw little use in distinguishing its varieties. This started to change, argues Horwitz, at the fin-de-siècle, perhaps because of growing scientific knowledge in general. As many as a fifth or quarter of male admissions to asylums at the time exhibited not just delusions of grandeur and hallucinations but also ominous neurological complications: they lost the ability to speak or walk, became riddled with bedsores, choked to death on their own vomit. Their decay was usually swift. Called general paralysis of the insane (GPI), this condition was initially thought to be the end state of all psychoses. By the early 20th century, however, it was becoming clear that it was in fact a disease with a quite distinct pathology and that it was caused by a bacterium: the bacterial origins of syphilis were uncovered in 1905 and the distinctive spiral-shaped bacterium that causes the disease was detected in the brains of those with GPI in 1913.

Horwitz is right to highlight the importance of this development, though in my view he does not pay sufficient attention to a development that flowed from these findings. Besides the impetus it gave to classifying mental disorders, the discovery of an organic cause for GPI and the emergence of a treatment (which, incidentally, won its inventor a Nobel Prize for infecting patients with malaria) launched an orgy of experimentation on patients in the 1920s and 1930s: surgical excision of teeth, tonsils, and colons; insulin coma therapy; various shock treatments; and lobotomies, to mention only the most well-known examples.

Some historians have suggested that a second source for American psychiatry’s increased attention to diagnosis originated in Europe. Beginning in the 1890s, the German psychiatrist Emil Kraepelin was inductively developing a distinction between what he held to be two distinctive forms of psychosis, based on their course and outcome: dementia praecox (later re-labeled schizophrenia), and manic-depressive psychosis. These broad divisions would exercise lasting influence over 20th- and 21st-century psychiatry, but as Horwitz rightly notes, their influence on its American practitioners before World War II was marginal at best.

War certainly shaped psychiatry, with the shell shock epidemic that accompanied the horrors of trench warfare during World War I encouraging a broader view of mental illness: mental breakdowns might have their origins in psychological and social trauma. World War II was even more important, transforming the locus of care from traditional mental hospitals to out-patient practices catering to a far broader patient population. Relatedly, psychoanalysis became the currency of the profession’s elite, dominating American culture for several decades.

Psychiatric breakdowns among American troops during World War II, notes Horwitz, vastly exceeded the rate of shell shock during World War I. Relabeled combat exhaustion or combat neurosis, such breakdowns strained military morale and preparedness, and gave new prominence to military psychiatrists, now trained with great rapidity to treat the legions of men undone by trauma. Whereas the sources of mental illness had traditionally been viewed as the product of biological factors, now it seemed they could be situational, a view reinforced by the psychotherapies employed with varying degrees of success to return the psychologically disabled to the front. The chief military psychiatrist, William Menninger, a psychoanalyst of sorts, amplified this psychodynamic turn: these “were ordinary people, their problems stemmed from [an] external precipitant, and their conditions were not chronic.”

The military likes manuals, and Menninger duly produced one, emphasizing the role of life history and the social environment in the genesis of mental disorders, and giving pride of place to psychoneurotic and behavioral disorders, the complaints that would become the bread and butter of out-patient psychiatry in the postwar years. Menninger’s military recruits to psychiatry formed the nucleus of the growing office-based profession, soon joined by others trained with moneys supplied by the Veterans Administration and the newly established National Institute of Mental Health. Psychoanalysts dominated these training programs and they also accounted for the growth of psychiatry in medical schools, which meant, as Horwitz points out, that the old-school psychiatrists were soon marginalized, becoming relics of an earlier biological tradition.

It was the new breed of psychoanalysts who wrote the first two DSMs, in 1952 and 1968. The two volumes shared fundamental characteristics: they were short, spiral-bound volumes, each running to little more than a hundred pages; presented only cursory lists of diagnoses; and, with the exception of a handful of organic psychoses, emphasized the psychological origins of all forms of mental disorder. They shared another notable characteristic: they were treated as irrelevant by most psychiatrists. DSM II cost $3.50, which was more than most practitioners thought it was worth, and if they owned one, they generally treated it as a paperweight.

For psychoanalysts, “the central aspect of psychiatric diagnoses at that time was their very unimportance […] as they saw it, all neurosis was caused by remarkably similar unconscious conflicts.” In the words of William Menninger’s brother Karl, the most popular analyst of the postwar age, “There is only one class of mental illness — namely mental illness.” Nowadays, with the advent of managed care and some degree of insurance coverage for the treatment of mental illness, diagnoses — and labels — do matter enormously, if only as a way for psychiatrists to get paid and patients reimbursed. But those consulting psychiatrists in office-based practice in the 1950s and 1960s didn’t have those incentives: they were paid directly for their services, and patients’ help-seeking was not motivated by the need for a specific diagnosis.

With startling speed, however, the psychoanalysts who embraced Polonius’s perspective on madness got their comeuppance. Within the space of less than a decade, a profession that disdained diagnosis embraced it, making it central to both clinical practice and psychiatric research. Equally remarkable, in a psychiatric world turned upside down, the psychoanalysts who had seemed so securely in charge of the profession, their perspective widely embraced in both academia and popular culture, found themselves defenestrated. Psychodynamic perspectives fell by the wayside. Biochemistry replaced the Oedipus complex. Neuroscientists and geneticists soon ruled the roost. Drugs took the place of talk therapy, which was increasingly the province of the non-medically trained (and more poorly paid) clinical psychologists and social workers. And the public learned, through the propaganda machine that is Big Pharma, that deficiencies of dopamine and serotonin were the source of their mental anguish, not the malevolence of their refrigerator mothers and ineffectual fathers.

Horwitz does a generally excellent job of analyzing how this psychiatric revolution came about. Analysts were oblivious to a set of changes that were covertly undermining their dominance. Where patients themselves had paid for psychotherapeutic interventions in the 1950s and 1960s, from the mid-’60s onward, they increasingly relied on insurance coverage, with such reimbursements increasing from 38 percent to 68 percent over a 15-year period. Skeptical of treatments that not only lasted for years but had little empirical proof of efficacy, insurance companies sought briefer treatments for circumscribed conditions, preferring for instance the time-limited, laboratory-tested cognitive behavioral therapy (CBT) to oft-interminable analyses, particularly since CBT could be administered by clinical psychologists at a greatly reduced cost.

Simultaneously, drugs were acquiring unprecedented importance. The first anti-psychotic, Thorazine, had been introduced to the marketplace in 1954, and together with a host of copycat medicines, transformed the treatment of schizophrenia. The invention of the minor tranquilizers — Miltown and then Valium and Librium — had similarly revolutionary effects on the therapeutics of less-disabling mental disorders. And the discovery of the first anti-depressants in the late 1950s opened another enormously profitable market. But in the wake of the thalidomide tragedy of the early 1960s — a drug marketed for morning sickness that produced an epidemic of badly deformed babies (largely averted in the United States when the FDA refused to license the drug), federal regulations mandated tighter oversight of new drugs coming to market. Though far from a panacea, as later events would show, FDA regulations now insisted on clinical trial data showing efficacy and safety, and, most significantly for the profession’s evolution, required that new medicines’ efficacy be tied to their demonstrable usefulness in treating defined diseases. This policy meant that drug companies had a crucial interest in how mental illnesses were defined, which did not bode well for psychoanalysts, who almost by definition disdained differential diagnosis.

On another front, a powerful anti-psychiatry movement, some of it emerging from within psychiatry itself, characterized the 1960s. The renegade psychiatrist Thomas Szasz had denounced mental illness as a myth with no biological reality. It was merely, he asserted, a mask for social control carried out by his malign colleagues. Horwitz acknowledges Szasz’s influence, but in one of his few lapses, fails to explore broader antipsychiatry forces. From the opposite pole of the political spectrum to Szasz, for example, the Scottish psychiatrist R. D. Laing proposed that schizophrenia was a form of super-sanity, and he proclaimed that the world was mad, not psychiatric patients. Meanwhile, Horwitz’s sociological brethren insisted that mental illness was about nothing more than labels, and mental pathology the product of the deforming and dehumanizing impact of the mental hospitals in which psychiatrists locked up and mistreated their charges.

However ill-assorted and often incoherent these criticisms, they enjoyed surprisingly broad resonance, creating a negative portrait of a profession whose basic competence was now held up to ridicule. Horwitz would have done well to acknowledge just how powerfully Miloš Forman’s film version of Ken Kesey’s novel One Flew Over the Cuckoo’s Nest consolidated these themes, creating in the public mind an indelible portrait of a bumbling profession. He also ought to have mentioned the well-publicized intervention of one of the leading figures in the new mental health bar, Bruce Ennis, who poured scorn on the claims of psychiatrists to expertise in the courtroom. Drawing upon an array of academic studies, Ennis zealously pointed to their inability to diagnose reliably the patients in front of them, and compared the probative value of their evidence to the results of monkeys flipping coins in the air: heads, you’re mad, tails you’re sane.

Ennis’s devastating analogy had its roots in the academic literature of the time. A handful of psychiatrists had begun to examine the reliability of their colleagues’ diagnoses, with uniformly depressing results. Horwitz does pay some attention to this literature. Potentially most damaging of all was a comparative study, published in 1972, of diagnoses in London and New York by a team headed by the British psychiatrist John Cooper. Schizophrenia, they showed, was diagnosed far more frequently in the United States than in Britain, and manic-depressive illness far more frequently by British psychiatrists. New York psychiatrists diagnosed nearly 62 percent of their patients as schizophrenic, while in London, only 34 percent received this diagnosis. And while less than five percent of the New York patients were diagnosed with depressive psychoses, the corresponding figure in London was 24 percent.

However damning these and other findings were, none of them attracted much attention beyond a handful of fellow-professionals, which may explain why Horwitz underplays their significance for later developments. In fact, however, they prompted some psychiatrists to focus on diagnostic issues. What Horwitz doesn’t minimize is a paper that appeared in the prestigious general science journal Science in January 1973, “On Being Sane in Insane Places,” by the Stanford social psychologist David Rosenhan. It purported to report the results of an experiment designed to test psychiatrists’ ability to distinguish the mentally ill from the sane. Rosenhan claimed to have recruited carefully screened pseudo-patients who were instructed to show up at a variety of mental hospitals claiming to hear voices saying “empty,” “thud,” and the like. They were otherwise to behave perfectly normally. All of them, he reported, were promptly admitted, and all but one were given the devastating diagnosis of schizophrenia. On average, it took nearly three weeks for the subjects to be released (one remained institutionalized and given antipsychotic medication for 56 days), at which point most were given the diagnosis “schizophrenia in remission.”

Thanks to remarkable detective work by a New York investigative journalist Susannah Cahalan, reported in her book The Great Pretender (2019), we now know that Rosenhan’s work was one of the great social science frauds of the 20th century. But her exposé, as Horwitz notes, came nearly a half-century after the fact. At the time, the paper’s effect was dramatic. The media pounced with glee. Panic set in at the highest ranks of the profession, with an urgent meeting of the executive committee of the American Psychiatric Profession to decide how to respond to a public relations disaster. It appointed a committee to revisit the problem of psychiatric diagnosis, with the explicit aim of rendering it more reliable.

Whom to entrust with this delicate task? Robert Spitzer was appointed for several reasons. A professor of psychiatry at Columbia, he had an abiding interest in what most of the profession still saw as the most boring of subjects, the classification of mental disorders. He himself lobbied for the position, but probably his decisive qualification was his role in defusing another professional embarrassment. Psychoanalysts had long argued that homosexuality was a mental disease, a position they saw as more humane than the punitive approach to sexual difference the legal system had adopted. Gay activists within and outside the profession launched a vigorous attack in the early 1970s, decrying a stance that rendered their sexuality a mental illness. As the uproar grew, Spitzer calmly negotiated a climbdown. The issue was put to a vote of the profession, and a majority voted to discontinue dubbing homosexuality a disease. The idea that scientific questions should be decided by voting was greeted with considerable public ridicule. This said, an issue that threatened to split the profession had been promptly resolved and did not resurface.

As the architect of this solution, Spitzer was now chosen to tackle the more far-reaching problem of creating a new diagnostic manual. As Horwitz notes, he was greatly assisted by the disdain most psychoanalysts felt for the enterprise, which freed him to compose the committee as he liked: he filled it with colleagues who shared his commitment to reasserting psychiatry’s medical identity. Spitzer was a master politician. Determined to ensure that two psychiatrists presented with the same case would reach the same conclusion, he and his colleagues largely ignored questions of validity — in other words, whether the categories they created corresponded to real and distinct disease entities. They also sought to purge from the manual any traces of Freudian ideology or etiological speculations about the origins of mental illness.

The final draft of the manual had to be voted on by the profession, still heavily psychoanalytic, and some analysts belatedly realized the threat Spitzer’s manual posed. Most of Spitzer’s colleagues on the committee wanted to reject their complaints outright. Recognizing that disaster loomed, Spitzer came up with a compromise. The word “neurosis” had been eliminated entirely from the document’s penultimate draft. But neuroses were the psychoanalysts’ bread and butter. The term now reentered the final draft, albeit in parentheses: “phobic disorders (or phobic neuroses”), “obsessive-compulsive disorders (or obsessive-compulsive neuroses)”. The verbal sleight of hand proved sufficient to overcome the analysts’ objections. (Seven years later, when Spitzer presided over the first revision of his new manual, the material in parentheses was quietly dropped, just as he had planned.)

Nearly a decade ago, Hannah Decker provided an exhaustive (and exhausting) blow-by-blow account of the creation of Spitzer’s DSM-III. Horwitz’s account is more succinct and sharper, and his is the version that all but specialists will turn to. He is very good at showing that “clinical judgment” and political horse-trading were the ruling approach, with voting the crucial mechanism for deciding on the wording of the text. Diagnosis was to rest purely on symptoms, with diseases being established through a “tick-the boxes” approach. Exhibit X number of symptoms from a laundry list, and voilà, a diagnosis is yours. The sole exception to an approach that rigidly eschewed reference to causal factors was post-traumatic stress disorder (PTSD) — a diagnosis lobbied for by disgruntled Vietnam veterans and some supportive psychiatrists, and adopted by Spitzer only after he had modified the original proposal for a “post-Vietnam syndrome” to encompass a far broader list of traumas.

Just how messy and incoherent the behind-the-scenes negotiations were is nicely captured in Horwitz’s discussion of the creation of major depressive disorder (MDD), which eliminated earlier distinctions between neurotic and psychotic depression, lumping together a disorder that was widespread but often not severe with a much rarer incapacitating condition:

Someone who had been severely depressed for years, could not leave her bed, and had continuous thoughts of worthlessness had MDD, as had an adolescent who felt depressed and unable to feel pleasure, had trouble sleeping, and lost his appetite and concentration after his girlfriend broke up with him two weeks before.

As a “major” disease, MDD expanded the market for psychiatric services and for the highly profitable drugs prescribed to treat it. “Between 1987 and 1997,” writes Horwitz, “the proportion of the US population receiving outpatient treatment for conditions called ‘depression’ increased by more than 300 percent.”

¤

The framework that underpinned DSM-III survived intact in the next two revisions of the new guide: DSM-IIIR (Revised) of 1987, and DSM-IV of 1994. Required to receive insurance reimbursement, DSM diagnoses became a touchstone for accessing various benefits, both educational and financial. They underwrote and shaped psychiatric research and drug company activities, and they entered the legal system. They also framed discussions of mental illness in the media. At each revision, new disorders were added, and definitions of disorders tweaked, almost always in the direction of broadening the criteria for a particular diagnosis. Not insignificantly, the advent of the DSM universe contributed to the rapid demise of psychoanalysis, replacing the phenomenology of mental illness with a crudely empirical approach to segmenting and treating mental disorder.

The DSMs also provided the vector through which vast profits were made by Big Pharma. It quickly discovered how valuable DSM diagnoses could be for securing approval for, and marketing, medications. The Food and Drug Administration (FDA) had permitted some direct-to-consumer advertising in the mid-1980s but greatly loosened its regulation of the practice in 1997. Three years later, by 2000, expenditures on marketing and consumption had soared from $595 million to $2.5 billion, with pharmaceutical “remedies” for DSM conditions accounting for a large fraction of this increase. The startling market success of Prozac and other selective serotonin reuptake inhibitors — sales of Prozac rose from $125 million in 1987, when it was introduced, to $2.6 billion in 2000 — occurred in substantial measure because of the industry’s ability to exploit the heterogeneous MDD diagnosis.

As SSRIs were threatened with the loss of patent protection and associated profits, Horwitz reminds us that they were repackaged as treatments for other DSM disorders, the paradigmatic case being the repositioning of Paxil as a treatment for social phobia, later renamed social anxiety disorder. DMS-IV had also launched a new version of bipolar disorder, bipolar II. DSM-III had created a sharp division between bipolar disorder (manic-depressive psychosis) and depression. Bipolar II blurred that boundary once more, and greatly expanded the population who could receive the diagnosis — from 2.6 percent to as many as 10 percent of the American population, according to some estimates. Before long, the pharmaceutical industry, as Horwitz shows, seized the opportunity to reposition antipsychotic drugs for this vast potential market. Almost simultaneously, largely through the entrepreneurial zeal of the Harvard child psychiatrist Joseph Biederman and his team, the previously unknown condition of childhood bipolar disorder surfaced. A condition that began in childhood and required lifelong medication could hardly be more attractive to pharmaceutical companies. Notwithstanding the absence of data on the efficacy and safety of such therapies or general agreement among child psychiatrists about the very existence of the disease, diagnoses of childhood bipolar disorder increased .5 percent in 1994 to 6.67 percent in 2003.

By some measures, Horwitz asserts, no one can doubt the success of the DSM revolution. For those applauding its arrival, it reestablished psychiatry as a biologically based medical specialty dispensing not talk therapy but medications that resembled the treatments mainstream medicine prescribed. Its broadened definitions of mental disorders also produced epidemiological studies that demonstrated frightening levels of mental pathology in the general population. For clinicians, the categories were essential for securing insurance reimbursement for their services. Meanwhile, “patients gained names for their afflictions, compensation for their visits to physicians and other mental health professionals, and drugs to alleviate their distressing conditions.”

But as Horwitz proceeds to show, there was a major fly in the ointment. The expectation when DSM-III was launched was that new research based on its categories would validate and refine the understanding of mental disorders. Unfortunately, that did not happen, and the problems this caused for researchers became steadily more pressing as time passed. George H. W. Bush had proclaimed the 1990s “the decade of the brain.” The National Institute of Mental Health (NIMH) would pour huge resources into research on the genetics and neuroscience of mental illness. Convinced that mental disorders were brain disorders, academic researchers were sure that they were on the brink of great discoveries about the etiology of the more serious forms of mental illness, a confidence that was further fueled by the decoding of the human genome in 2003.

That set the stage for yet another revision of the DSM. This time, however, those charged with constructing it — virtually all drawn from the ranks of academic research — promised radical change. The new manual would no longer rely on symptoms to delineate the different varieties of mental illness. Instead, it would draw on new scientific knowledge about the pathology and etiology of the various forms of mental disorder. Advances in genetics and neuroscience would surely allow for a map of mental illness that corresponded to real disease entities, thereby quite possibly discarding existing categories like schizophrenia and bipolar disorder if the new brain science revealed that these constructs were built on faulty foundations.

As quickly became apparent, such vaulting ambition was an exercise in hubris. The etiology of virtually all forms of mental illness remained as elusive as ever. Genetics, rather than providing increased clarity about different sub-types of mental illness, systematically muddied the waters. Candidate genes that had been expected to correspond to a heightened susceptibility to schizophrenia, bipolar disorder, or depression did not in fact do so. Rather than genes having a specific relationship to separate psychiatric disorders, they were virtually all non-specific, and statistically exhibited a weak connection to mental illness. Nor did neuroscience provide much in the way of useful knowledge. “In contrast to the specific, discrete, and seemingly homogeneous entities found in the DSM, research was uncovering that mental disorders were broad, overlapping, and heterogeneous,” writes Horwitz. As he shows, the attempted paradigmatic revolution simply collapsed, and the upshot was a “new” edition that was little more than a cosmetic reworking of its predecessors.

Actually, the outcome was even worse. The production of previous editions — DSM-III, DSM-IIIR, and DSM-IV — had been tightly controlled by those overseeing the process, Robert Spitzer and Allen Frances. But the two psychiatrists heading the latest review process, David Kupfer and Darrel Regier, provided little guidance or oversight. Horwitz analyzes the ensuing frustrations and chaos, with the American Psychiatric Association ultimately forced to step in to review and, if necessary, veto changes proposed by the task force.

If the psychiatrists reconfiguring DSM-5 were forced to abandon their larger ambitions, they at least hoped to move toward a dimensional rather than categorical view of mental illness — that is, to suggest that mental illnesses existed on a spectrum from mild to serious, rather than being a series of discrete conditions with sharp boundaries. Here, too, they were thwarted. Horwitz plausibly suggests that clinicians, excluded from the ranks of those preparing the new edition, were convinced that insurance companies would decline to pay for treatment of those patients at the milder end of the spectrum. At the 2012 meeting of the American Psychiatric Association, clinicians forced the issue and removed references to dimensional approaches from the main pages to an appendix.

There was one exception to the rejection of a dimensional view, and Horwitz cites this as an example of how deeply embedded the earlier DSM categories have become in psychiatric practice. In the years after the publication of DSM-IV, autism and Asperger’s syndrome were still another group of psychiatric disorders whose incidence had rapidly multiplied. The editor of DSM-IV, Allen Frances, blamed the increase on the broadening of the diagnostic criteria introduced in 1994, a position to which patients’ families took fierce exception. The diagnoses brought their children access to special education, mental health treatment, and disability payments. They quickly formed a highly organized and effective lobbying group against the DSM-5 group’s proposal to collapse these two separate diagnoses (along with such conditions as Rett syndrome and childhood disintegrative disorders) into a single entity, autism spectrum disorder, with individual patients diagnosed as having mild to severe versions. Patients with Asperger’s syndrome objected to being given the more severe diagnosis of autism and parents of autistic children were convinced that those on the lower end of the spectrum would find themselves deprived of the ancillary services they depended upon. Preliminary studies of the likely impact of the new approach showed that they were right: only 60 percent of those with a DSM-IV autism diagnosis would qualify for benefits under the proposed revised system. A public outcry ensued. The upshot: An announcement that all those who already had these diagnoses under the old system would be grandfathered in. No one would lose their earlier diagnosis. A system that was purportedly evidence-based had been shown to be modifiable by outside pressure. As Horwitz dryly comments, “Many diagnoses have become so socially embedded that it is virtually impossible to eliminate them from the DSM.”

That was far from the last of the DSM-5 leadership’s problems. A far more damaging critique emerged from an unlikely quarter, the editors of DSM-III, IIIR, and IV. Robert Spitzer and Allen Frances were intimately familiar with the DSM system, having originally constructed it. Spitzer was angered by what he claimed was the secrecy with which the new edition was being formulated. Task force members had all been compelled to sign confidentiality agreements promising not to discuss internal discussions or how decisions had been reached. It was, Spitzer loudly insisted, a violation of the open discussion that ought to be foundational in a scientific enterprise. From a public relations point of view, Spitzer was extraordinarily shrewd. But, in reality, his point was something of a red herring, since Spitzer’s two task forces had concealed their own horse-trading and politicking, whereas working drafts of DSM-5 proposals had been widely circulated for comment.

Frances’s objections were legion, and though Spitzer was eventually sidelined by his advancing Parkinson’s, Frances kept up a barrage of criticism. For the most part, his arguments rested on claims that the DSM-5 group was on the brink of pathologizing ordinary human behavior by inordinately widening the scope of psychiatry, and thus subjecting millions of people to the hazards and side effects of psychotropic drugs. Convinced that he himself had been at fault for the expanded reach of the autism diagnosis, he was determined to stop another such expansion. For example, the task force dealing with depression was proposing to eliminate what was known as the grief exception, the idea that it was natural for those who had lost a loved one to experience an extended period of sadness and related problems. Ordinary and expected sadness, he argued in a series of high-profile op-eds and media appearances, was in danger of being turned into a mental illness. Similarly, one of the proposals considered by the DSM task force was the creation of a new diagnosis, psychosis risk syndrome. Most of the people given this label, Frances contended, would never go on to develop serious mental illness. Yet they were likely to find themselves stigmatized and discriminated against, and they would become an obvious target for the pharmaceutical industry, given the money that could be made by putting them on psychotropic drugs.

These critiques from such apparently well-placed observers received enormous public attention. The response of both the DSM-5’s leaders and the American Psychiatric Association’s president, Alan Schatzberg, only made matters worse. They launched an ad hominem attack on Spitzer and Frances, accusing them of acting out of jealousy at being sidelined, and of being motivated by the residual royalties (about $10,000 a year) they received from the earlier editions. This line of criticism backfired. It turned out that Schatzberg was under attack by a committee of the United States Senate for allegedly concealing his ownership of stock worth perhaps $6 million in a company whose product he was “independently” researching. One outside observer, Horwitz mordantly records, compared the whole unsavory spectacle to the feuds of the Hatfields and McCoys. The back-and-forth delayed the appearance of the new manual by over a year.

Worse, however, was still to come. DSM-5 was finally slated to be unveiled in May 2013. Weeks before its publication, however, the director of the National Institute of Mental Health, Thomas Insel, denounced the whole operation as a catastrophic failure. Far from being the revolutionary document its proponents had promised, it was, he noted, a thinly disguised recapitulation of the basic approach adopted by its predecessors. And that approach, he contended, had proved a disaster. Its continued reliance on symptoms to reach its diagnoses reminded him of pre-scientific medicine. “In the rest of medicine,” he wrote sarcastically, “this would be equivalent to creating diagnostic symptoms based on the nature of chest pain or the quality of fever.” “Patients,” he insisted, “deserve better.” His immediate predecessor at NIMH, Steven Hyman, now running an institute at Harvard, was even harsher: the new manual, he wrote dismissively, “was an absolute scientific nightmare.”

Horwitz points out that it was difficult for the heads of the DSM-5 project to respond to this line of criticism, in part because it echoed comments they themselves had made at the outset of the whole process. They had expected to produce an entirely new diagnostic system reflecting the new science of the brain, with categories rooted in the concrete biology that produced different forms of mental illness. It turned out, of course, that neither they nor anyone else possessed that kind of knowledge. Hence the fudge. Still, something had to be done to mitigate the crisis of legitimacy that now confronted psychiatry. A hastily arranged press conference featuring Thomas Insel and Jeffrey Lieberman of Columbia University, the then-president of the American Psychiatric Association, produced an anodyne bit of public relations-speak: DSM, they solemnly asserted, remained the best and most up-to-date guide for clinicians. Meanwhile, academic researchers would continue to try to track down the biological bases of the various forms of mental illness, and in doing so would not be using DSM categories.

Horowitz tells this sorry tale with skill and panache. While meticulously laying out the evidence for his conclusions, based on wide-ranging archival research and close familiarity with the relevant psychiatric literature, he also makes sure to acknowledge the difficulty and importance of the task confronting a profession grappling with the miseries and mysteries of mental illness. This is not an anti-psychiatric polemic, but rather a sober and careful analysis of the shifting role of diagnostic systems in American psychiatry. It is the best synthetic account of this territory anyone has produced to date.

¤

Andrew Scull’s new book, Desperate Remedies: Psychiatry’s Turbulent Quest to Cure Mental Illness, will be published next spring by Harvard University Press.