The future of an illusion

This article originally appeared in The Times Literary Supplement (reproduced with permission)

By Andrew Scull

Fights over how to define and diagnose mental illness are scarcely a novel feature of the psychiatric landscape, but their most recent manifestation has some unusual features. For more than a decade now, the American Psychiatric Association has been preparing a new edition of its Diagnostic and Statistical Manual (DSM), the fifth (or by some counts the seventh) edition of that extraordinary tome, each incarnation weightier than the last. Over the past two years, however, major attacks have been launched on the enterprise, replete with allegations that the new edition shows signs of being built on hasty and unscientific foundations; that it pathologises what are everyday features of normal human existence; and that it threatens to create new epidemics of spurious psychiatric diseases. These verbal assaults have come in substantial part from an unexpected quarter, however: not from the ranks of the anti-psychiatric chorus, Szaszian, sociological or otherwise, but, amongst others, from the editors-in chief of DSM III and DSM IV (Frances 2009; Spitzer 2009).

These are psychiatrists whose previous work, to be sure, will be modified and superseded in the new edition, but they are also men whose careers were built upon their unswerving commitment to the underlying logic of creating a nosological Bible. Their critiques have spawned claims from the ruling psychiatric oligarchy that they are motivated by pique at seeing their creations cast aside, or perhaps, as some have suggested, even by the loss of royalties the editor-in chief of DSM IV will suffer when his version of the classificatory system is rendered obsolete (Schatzberg, Scully, Kupfer, and Regier 2009). But their criticisms have already forced a delay in the publication of DSM-5 (the pretentious resort to Roman numerals to designate successive editions of the manual having finally been abandoned).  And they have helped to intensify a renewed crisis of psychiatric legitimacy.

Some historical context is in order here. Complicated nosologies were a feature of nineteenth century psychiatry. They proliferated endlessly, and seemed to be of little clinical use, not least because they were so hard to operationalize. The first generally accepted sub-dividing of the psychoses emerged in Germany in the late nineteenth century, at the hands of Emil Kraepelin, who claimed to have developed inductively a distinction between two basic sub-types of serious mental disorder, dementia praecox (soon relabelled schizophrenia), and manic depressive psychosis – for Kraepelin a sort of residual category for psychotics who didn’t manifest the symptoms or have the hopeless prognosis of dementia praecox, and something generally regarded at the time as a more hopeful diagnosis.  It was testament to the wide and continuing influence of Kraepelin’s endeavours that the revolution in psychiatric nomenclature launched by DSM III in 1980 is commonly referred to as the neo-Kraepelinian revolution in psychiatry. Both enterprises sought to transform a disorderly chaos of symptoms into an orderly list of illnesses.

As its title indicates, DSM III had some predecessors. American psychiatrists had constructed two previous official diagnostic systems of their own, small pamphlets that appeared successively in 1952 and 1968.  Both set up a broad distinction between psychoses and neuroses (roughly speaking, between mental disorders that involved a break with reality, and those that, less seriously, involved a distorted view of reality), and they divided up many of hundred or so varieties of mental illnesses that were recognized in accordance with their alleged psychodynamic etiologies.  In that respect, they reflected the dominance of psychoanalytic perspectives in post World War II American psychiatry. But diagnostic distinctions of the broad, general sort these first two editions set forth were of little significance for most analysts, focused as they were on the individual dynamics of the particular patient they were treating. The first two DSMs were therefore seldom consulted and were seen as little more than paperweights – and rather insubstantial paperweights at that. DSM II was a small, spiral-bound pamphlet running to no more than a hundred and thirty-four pages, and encompassing barely a hundred different diagnoses that were listed alongside the most cursory of descriptions.  It sold for a mere three dollars and fifty cents, which was more than most professional psychiatrists thought it was worth.

It was precisely that lack of concern with diagnostic categories and the sense that questions of nomenclature were supremely unimportant that led psychoanalysts to view the formation of an American Psychiatric Association task force on creating a new edition of the DSM with a complacency that verged on contempt. It would prove a stunning political miscalculation. The task force quickly came to be dominated by its chairman, Robert Spitzer, and by a group of biologically-oriented psychiatrists who liked to refer to themselves as DOPS (data oriented people), which was an interesting conceit, since data and scientific evidence had remarkably little to do with what emerged from the committee’s deliberations. Instead, their work product had much to do with the preferences and prejudices the self-anointed DOPS shared. These were psychiatrists, many of them hand-picked by Spitzer, who preferred pills to talk, and for whom creating a wholly distinctive new approach to the diagnostic process became a decisive weapon in their battle to re-orient the profession. Psychoanalysts had placed but a single member of their fraternity on the major committee, and he was so swiftly marginalized that he ceased attending the sessions at which the proposed changes were discussed and finalized.

Too late, realization dawned among the psychoanalystic elite that the new nosology would have profound effects on the future of psychiatry, and the very terms in which the broader culture conceptualized and thought about mental illness. The speculations about the psychodynamic etiology of the psychoses and neuroses that had been central to the first two editions were stripped out of the new nosology, along with all traces of obeisance to psychoanalytic doctrines. The distinction between psychosis and neurosis was abandoned. In their place, the task force adopted a seemingly simplistic and radically revamped approach to distinguishing among sub-types of mental illness. A succession of studies during the late 1960s and 1970s had demonstrated the extraordinary unreliability of psychiatric diagnoses. Many of these studies had been conducted by the profession itself (including a landmark study by Cooper et al. (1972) of differential diagnosis in a cross-national context), though the study that drew most public attention (and inflicted most damage on psychiatry’s public image) was an experiment using pseudo-patients conducted by the Stanford social psychologist David Rosenhan, whose results appeared in Science (Rosenhan 1973).  Whatever its methodological flaws (and they were considerable), Rosenhan’s study was widely seen as confirmation of psychiatry’s diagnostic incompetence.

The documented failure of psychiatrists to agree on what was wrong in any given case before them proved a great embarrassment to the profession. Lawyers used the lack of consensus to cast doubt on the profession’s claims to expertise (Ennis and Litwack 1974), and drug companies seeking homogeneous populations on which to conduct clinical trials for new psycho-pharmaceuticals expressed their frustrations at psychiatry’s shortcomings in this regard. As drug development proceeded, the need to standardize the patient population on which new drugs were tested had become more pressing. And as new drugs seemed to have an effect on some, but not all, psychiatric patients, it became commercially attractive to try to distinguish different sub-populations among the mentally ill.

Unable to demonstrate convincing chains of causation for any major form of mental disorder, the Spitzer task force abandoned any pretence at doing so. Instead, they concentrated on maximizing inter-rater reliability to ensure that psychiatrists examining a particular patient would agree on what was wrong.  This entailed developing lists of symptoms that allegedly characterized different forms of mental disturbance, and matching those to a “tick the boxes” approach to diagnosis. Faced with a new patient, psychiatrists would record the presence or absence of a given set of symptoms, and once a threshold number of these had been reached, the person they were examining was given a particular diagnostic label, with “co-morbidity” invoked to explain away situations where more than one “illness” could be diagnosed. Disputes about what belonged in the manual were resolved by committee votes, as was the arbitrary decision about where to situate cut-off points: i.e., how many of the laundry list of symptoms a patient had to exhibit before he or she was declared to be suffering from a particular form of illness. Questions of validity – whether the new classificatory system was really cutting nature at the joints, so that the listed “diseases” corresponded in some sense with distinctions that made etiological sense – were simply set to one side.  If diagnoses could be rendered mechanical and predictable, consistent and replicable, that would suffice.

DSM III’s triumph marked the advent of a classificatory system that increasingly linked diagnostic categories to specific drug treatments, and an embrace on the part of both profession and public of a conceptualization of mental illnesses as specific, identifiably different diseases, each amenable to treatment with different drugs. Most importantly, since the insurance industry began to require a DSM diagnosis before agreeing to pay for a patient’s treatment (and the preferred course and length of treatment came to be linked to individual diagnostic categories), DSM III became a document that it was impossible to ignore, and impossible not to validate. If a mental health professional wanted to be paid (and could not afford to operate outside the realms of insurance reimbursement, as most self-evidently could not), then there was no alternative to adopting the manual. In subsequent years, particularly once antidepressant drugs took off in the 1990s, biological language saturated professional and public discussions of mental illness. Steven Sharfstein, the then president of the American Psychiatric Association, referred to the upshot of this process as the transition from “the biopsychosocial model [of mental illness] to… the bio-bio-bio model” Sharfstein (2005:3).

That the specificity of the treatments was largely spurious, and that the various editions of the Diagnostic and Statistical Manual from the third edition onwards emphasized reliability and essentially ignored the more central issue of the validity of psychiatric diagnoses, proved largely irrelevant to their success in reorienting perceptions, lay and professional alike. Linked to expanded insurance coverage for the treatment of mental disorders, and providing a new grounding for psychiatric authority and a less time-consuming and more lucrative foundation for psychiatric practice, psychopharmacology encouraged a distancing of psychiatrists from the provision of psychotherapy. Each successive edition of the manual, the revised third edition (III R) of 1987, the fourth edition (IV of 1994) and its “text revision” (IV TR of 2000) has adhered to the same fundamental approach, though new “illnesses” have been added on each occasion, and the page count has mounted, like the ‘Yellow Pages’ on steroids, from the 104 pages of DSM I to the 992 pages of DSM IV TR.

Thus we return to the current controversy. The classificatory mania embodied in the various editions of DSM, from III onwards, arose from an attempt to lend an aura of “facticity” to psychiatric diagnoses, and to stave off the ridicule that threatened the profession’s legitimacy when its practitioners were shown to be unable to agree about the nature of the illness that confronted them (or even whether the patient was sick at all). But as “illnesses” proliferated in each revision, and the criteria for assigning a particular diagnosis were loosened (as they will be again in DSM-5), the very problem that had led to the invention of the new DSMs recurred, and major new threats to psychiatric legitimacy surfaced.

As diagnostic criteria were loosened, an extraordinary expansion of the numbers of mentally sick individuals ensued. This has been particular evident amongst, but by no means confined to, the ranks of the young.  “Juvenile biopolar disorder,” for example, increased forty-fold in just a decade, between 1994 and 2004. An autism epidemic broke out, as a formerly rare condition, seen in less that one in five hundred children at the outset of the same decade, was found among one in every ninety children only ten years later. The story for hyperactivity, subsequently relabelled ADHD is similar, with ten per cent of male American children now taking pills daily for their “disease.” Among adults, one in every seventy-six Americans qualified for welfare payments based upon mental disability by 2007.

If psychiatrists’ inability to agree among themselves on a diagnosis threatened to make them a laughing-stock in the 1970s, the relabelling of a host of ordinary life events as psychiatric pathology now seems to promise more of the same. Social anxiety disorder, oppositional defiant disorder, school phobia, narcissistic and borderline personality disorders are apparently now to be joined by such things as pathological gambling, binge eating disorder, hypersexuality disorder, temper dysregulation disorder, mixed anxiety depressive disorder, minor neurocognitive disorder, and attenuated psychotic symptoms syndrome. Yet we are almost as far removed as ever from understanding the etiological roots of major psychiatric disorders, let alone these more controversial diagnoses (which many people would argue do not belong in the medical arena in the first place). That these diagnoses provide lucrative new markets for psychopharmacology’s products raises questions in many minds about whether commercial concerns are illegitimately driving the expansion of the psychiatric universe – a concern that is scarcely allayed when one recalls that the great majority of the members of the DSM taskforce are recipients of drug company largesse. That psychoactive drugs are associated with rising concerns about major side effects (ranging from iatrogenic and permanent neurological damage, through increased risks of child and adolescent suicides, massive weight gain, metabolic disorders, diabetes, and premature death) only compounds the problem.

Relying solely on symptoms and behaviour to construct its illnesses, and on organizational fiat to impose its negotiated categories on both the profession and the public, psychiatry is now facing a revolt from within its own ranks.  The Heath Robinson apparatus that is descriptive psychiatry seems to survive only because it lacks a plausible rival.  It is, however, an increasingly tenuous basis on which to rest claims to professional legitimacy.  Having chosen to erect a vast and ramshackle superstructure on such frail foundations, psychiatry must pray it doesn’t collapse in a heap of rubble.

This article originally appeared in The Times Literary Supplement (reproduced with permission)

Andrew Scull has held faculty positions at the University of Pennsylvania, Princeton, and the University of California, where he is Distinguished Professor of Sociology and Science Studies. He is a past president of the Society for the Social History of Medicine, and has held fellowships from the Guggenheim Foundation and the American Council of Learned Societies. He is the author or editor of more than twenty books, many of them on the history of psychiatry in Britain and the United States. He has lectured on five continents, as well as making many media appearances on programmes dealing with mental health issues. His book Madness: A Very Short Introduction published in 2011.

The Very Short Introductions (VSI) series combines a small format with authoritative analysis and big ideas for hundreds of topic areas. Written by our expert authors, these books can change the way you think about the things that interest you and are the perfect introduction to subjects you previously knew nothing about.

Subscribe to the OUPblog via email or RSS.
Subscribe to only VSI articles on the OUPblog via email or RSS.
Subscribe to only psychology articles on the OUPblog via email or RSS.
View more about this book on the