Psychiatry’s Strange Objects: Modernism and the Language-Games of Diagnosis

The era of modernism was also the moment when psychiatry became modern—at least this is the story that psychiatry has often told about itself. With the development of new nosological taxonomies, severe forms of mental illness were identified, described, and organized into distinct categories such as manic depression and dementia praecox (later schizophrenia) by psychiatrists Emil Kraepelin and Eugen Bleuler. These systems became the foundations for the diagnostic vocabulary of contemporary psychiatry. Psychiatrist Nancy Andreasen has observed, “Most of the discoveries of modern psychiatry represent a series of footnotes to and amplifications of Kraepelin’s textbooks.”[1] However, uncertainties about the status of these diagnostic entities and methods also persist—doubts about how speech acts of diagnosis function, about the status of the entities that such acts summon into the world, and whether modern psychiatry has ever been as modern as it has claimed. These doubts found vivid expression in literary modernism, including works by Virginia Woolf, Anna Kavan, Samuel Beckett, Thomas Mann, and others. In what follows, I consider conceptual sources of diagnostic instability, modernist responses to problems of psychiatric diagnosis, and the challenges and opportunities that these formations present for scholars of modernism today.

My title does not refer to the eccentric behavior of psychiatric patients but to the fact that psychiatry has failed to reach consensus about what its diagnostic categories refer to, how these entities emerge, or how they should be known. As Andrew Lakoff writes, “Psychiatry, whose objects of knowledge emerge in the encounter between patients’ subjective reports and clinicians’ interpretive schemes, has had a difficult time shifting the disorders under its purview into stable things in the world.”[2] Perhaps such problems persist because these objects do not readily submit to a single epistemic or ontological framework but emerge through a plurality of incommensurable language-games.

With the phrase “language-games,” I suggest that Ludwig Wittgenstein may provide a resource for thinking through the confusions that arise from unrecognized differences across linguistic practices of diagnosis. The principal corrective of Wittgenstein’s Philosophical Investigations (1953) is the observation that language is not reducible to a single function but is comprised of multiple language-games that operate according to distinct rules with varied effects. Following this approach, we might sidestep questions concerning the ontological status of historical diagnostic entities in order to consider the ways that illness and diagnosis were often enacted within not one but many language-games.

Diagnosis has often functioned as not simply a neutral representation or explanation of a mental phenomenon but as a performative speech act: it not only describes but configures difference into recognizable forms. Diagnosis inaugurates a process through which a person will be made to participate in certain forms of rhetoric, protocols of treatment, and modes of discipline. Michel Foucault argued that the modern taxonomic impulse rendered a speciation of human types.[3] Given the constant revision and disputation of these taxonomies, it follows that a patient presenting with consistent symptoms could be transformed into many subspecies of the human—or dehumanized in many ways—within a single life span.

Late nineteenth-century clinicians could not agree whether an array of symptoms should indicate a diagnosis of epilepsy or hysteria. Mark Micale argues that with the introduction of new nosological categories such as dementia praecox and schizophrenia, hysteria virtually “vanished into a hundred places in the medical textbooks”—its symptoms were regrouped and redistributed across the newer German and Swiss taxonomies.[4] These taxonomic redistributions have continued over the last hundred years. The Statistical Manual for the Use of Institutions for the Insane (1918) recognized twenty-two disorders; the successors to this manual—the notoriously revised editions of the DSM—would multiply this number to what is currently around four hundred conditions. Ian Hacking writes,

One taxonomy replaces another, to the point that we simply do not know what hysteria was any more. It was [Thomas] Kuhn’s idea that when taxonomies are dumped in the course of revolution, a concept (or linguistic entity, a lexical item) from an old taxonomy could not be translated into a concept in the new taxonomy. Hence old ideas and practices become literally unintelligible to new-wave thinkers. Or, as Kuhn puts it, the old and the new taxonomic structures and languages express incommensurable ideas.[5]

The incommensurability and unintelligibility of older taxonomic entities (such as neurasthenia, railway spine, fugue, etc.) are challenges not only for historians of literature and culture but also for medical practitioners; taxonomies did not neatly succeed one another but often competed and overlapped. The lack of consistent methods and shifting diagnostic nomenclatures during the era of modernism were hardly resolved by the middle decades of the twentieth century. Historian Anne Harrington observes that as late as the 1960s, “when American psychiatrists were asked to independently diagnose the same patient, they tended to agree on the diagnosis only about 30 percent of the time.”[6]

In the era of modernism, psychiatric diagnosis was deployed in multiple ways, according to a variety of criteria, and with varied effects. There was profound disagreement over what kind of phenomena diagnosis of mental illness referred to (behavioral changes, altered phenomenologies, neuro-anatomical abnormalities), how it should be identified (longitudinal studies, symptom checklists, anatomical markers, family histories), and whether diagnosis entailed a particular form of treatment (pharmaceutical, surgical, electro-convulsive, or talking cures). Should diagnosis be simply descriptive, or should it venture into etiological and causal explanation? If so, should the “cause” take the form of neurological, psychological, genetic, or social explanations? Can words like “diagnosis,” “cause,” and “treatment” operate consistently and sensibly across such a range of frameworks, or do such broad uses of these words result in what Wittgenstein and his followers referred to as “category confusions?” Within the psychiatric “establishment,” itself unstable and contested, the diagnostic terrain was constantly shifting beneath the feet of clinicians and patients, and this pluralization of language-games was and continues to be one source of doubt about the status of psychiatric diagnosis.

Modernist-era diagnosis and treatment of mental illnesses extend beyond the Kraepelinian, Bleulerian, or Freudian methods that largely dominated the nascent field of psychiatry. Micale writes:

[I]t would be a mistake to exclude from historical consideration theories and practices that the medical sciences of our own day judge to be wrong, silly, unscientific. To be specific, hypnosis, somnambulism, psychical research, magnetotherapy, metallotherapy, dream interpretation, mediumistic psychology, automatic writing, faith healing, and spiritualism were all subjects of keen and widespread interest.[7]

A line between strictly “scientific” psychiatry and para-psychological practices of the period cannot easily be drawn. Claims of scientificity made by neurologists and psychiatrists outstripped progress, and speculation was often presented in the trappings of science already achieved. To this end, early psychiatry frequently invoked the language of psychological or neurological “mechanisms.” However, such discourse was often based not on empirical evidence but rather on analogical links were analogies to the natural sciences—precisely the kind of uses across distinct language-games that, Wittgenstein warns, often result in metaphysical confusion.

If the multiplicity of competing language-games of diagnosis was one source of confusion and doubt, another related source of instability emerged through what Hacking has described as “the looping effect of human kinds.”[8] Being ill is not only a biological but also a social process: one learns how to be a patient according to the protocols of one’s clinic. Patients often align themselves with the diagnostic nomenclature, descriptive vocabularies, and explanatory frameworks of their doctors. Therefore, as the mind sciences deployed increasingly mechanistic explanatory language to psychiatric disorders, many patients would regard themselves as dysfunctional, neurological machines. In her memoir of schizophrenia, Operators and Things (1958) Barbara O’Brien writes, “to a psychiatrist the unconscious mind, in insanity, is a machine out of working order, no more, no less.”[9] In the early decades of the twentieth century, the nervous system was often described as a network of electric impulses, and patients often felt that their minds were radio or telephonic devices receiving pernicious thoughts via transmissions at a distance.[10] A generation later, O’Brien felt that she had been reduced to a mainframe computer operated by her doctors. Through looping effects, diagnostic language-games not only classify or explain but give form and content to the altered states that they purport to describe.

The literature of the era took the measure of psychiatry’s claims to the status and authority of science and often staged these looping effects of human kinds. In Mrs. Dalloway, Septimus Warren Smith invokes the injunction that, “one must be scientific, above all scientific”—a trace of his doctors’ rhetoric that becomes lodged within his delusional monologue.[11] Looping effects also include varying degrees of instability as patients may not always understand what their doctors want from them. It is not clear that Septimus understands what it would mean to be “scientific” about one’s madness, but the same was often true of clinicians: Dr. Bradshaw, Septimus’s Harley Street “specialist,” reflects on “this exacting science which has to with what, after all, we know nothing about—the nervous system, the human brain” (Mrs. Dalloway, 96).

The doubts raised by Woolf about psychiatry’s claims to the status of a modern science intensified a generation later in the late-modernist fiction of Anna Kavan—an underappreciated, experimental writer whose short story collections, Asylum Piece (1940) and I Am Lazarus (1945), reflect her experiences as both a patient and nurse in several institutional settings. The latter work begins from the perspective of an English doctor visiting an asylum on the continent: “[H]e distrusted this insulin shock treatment there had been such a fuss about. Why should putting imbeciles into a coma make them sane? It didn’t make any sense. He did not think and he never had thought that there was a cure for an advanced dementia praecox case like you Thomas Bow.”[12] Kavan’s piece expresses skepticism about the categories and brutal psychiatric treatments deployed in the 1930s whose mechanisms of effect were not understood even by practitioners, and this skepticism is recursive: the visiting doctor’s critical assessment is followed by the frank admission that he “was not a very good doctor” (I Am Lazarus, 7). Kavan’s work leaves no solid ground from which claims about mental illness may be made aside from a thick description of patients’ experiences as they are transformed by the language-games and looping effects of the asylum.

The gap between science and scientism is a subject which Wittgenstein took up in several lectures on psychoanalysis: “there is no way of showing that the whole result of analysis may not be “delusion.” It is something which people are inclined to accept and which makes it easier for them to go certain ways: it makes certain ways of behaving and thinking natural for them. They have given up one way of thinking and adopted another.”[13] While Wittgenstein’s immediate target is psychoanalysis, his observation may be extended to other modes of diagnosis and speculative explanation that had little therapeutic effect beyond transforming patients into subjects who adopt particular accounts of themselves and their symptoms.

However, while he compared the “explanations” of psychoanalysis to those of mythology, Wittgenstein was also intrigued by the practical work that such practices might perform for individuals as such a language-game “makes it easier for them to go certain ways.” He suggests that, in such contexts, a “statement is not right or wrong, but may be practical or impractical. Hypotheses such as ‘invisible masses,’ ‘unconscious mental events,’ are norms of expression. . . . We believe we are dealing with a natural law a priori, whereas we are dealing with a norm of expression that we ourselves have fixed.”[14] Some psychiatrists admitted similar doubts about the status of their own diagnostic entities. Of the state of psychiatry in the 1940s, historian Anne Harrington writes, “The whole thing was a giant diagnostic mess, and thoughtful people knew it. In 1946 the endocrinologist R. G. Hoskins—whose career was largely devoted to pursuing the biological basis of schizophrenia—could nevertheless still ask: What is it? ‘Is it an entity, or mayhap, merely a semantic convention?’ Yet the diagnosis continued to be used, because there was nothing better to replace it” (Harrington, The Mind Fixers, 140, emphasis in original).

If diagnosis might be regarded as “not right or wrong, but may be practical or impractical,” its practical effects may not reside only in a recommended course of treatment. Sociologist Nikolas Rose observes:

Each category, and each array of categories, provides a language for speaking of our distress, making it thinkable in a way that also provides some kind of account of its nature, origins and likely implications. . . . No wonder patients and their families sometimes become attached to their diagnostic labels, and protest if for some reason they are eliminated altogether from the official psychiatric lexicon.[15]

Given the multiplicity of diagnostic language-games and the looping effect of human kinds, we cannot say in advance how diagnosis functions in every instance. Wittgenstein urges his students to avoid broad theorizing and instead to examine particular cases. For some, such as Kavan’s characters, diagnosis is a process through which difference is defined, reconditioned, and controlled, often in the service of the population rather than the individual. Others actively searched for the diagnosis that would make their experiences recognizable to others and to themselves. Amelia Jones has persuasively argued that we might understand Elsa von Freytag-Loringhoven’s form of Dada through the obsolete diagnosis of “neurasthenia” not in order to retroactively diagnose the Baroness but to see how the category became a resource for the production of her art and identity.[16] Radclyffe Hall’s The Well of Loneliness adapts the sexological category of the “invert,” which long carried the stigma of psychiatric pathology, as a concept which enabled acts of identification and recognition that had not been possible.

Psychiatry has often functioned as a regime for managing perceived social problems by warehousing, sterilizing, and experimenting on bodies regarded as abnormal. Racism, sexism, homophobia, and ableism shaped popular and clinical definitions of abnormality, and many influential psychiatrists (Kraepelin, Bleuler, Meyer) advocated for population controls and eugenics. At the same time, diagnostic and therapeutic techniques of psychiatry were also adapted by psychiatrists, community leaders, and public intellectuals as preventative and reparative resources for those who had been stigmatized and pathologized. Richard Wright helped to establish the LaFargue Mental Hygiene Clinic in Harlem and told a reporter, “the most consistent therapeutic aid that it can render Harlem’s mentally ill is to instill in them what Wertham calls ‘the will to live in a hostile world.’”[17]

Modernist texts often challenge the valences of normality and abnormality or health and illness on which psychiatric taxonomies were founded. While acknowledging that the “law is on the side of the normal,” Woolf expresses fascination with experiences of illness and “the undiscovered countries that are then disclosed.”[18] In The Magic Mountain, Hans Castorp is gradually transformed from a sanatorium visitor to the “horizontal” life of a patient, and he struggles to decide if an expert-defined notion of “health” should be his ideal or whether it would simply be a concession to a “bourgeois affirmation of life” that would suppress the unique insights that might only emerge during illness.[19]

Other modernist writers actively studied and replicated the language of the mentally ill in order to transform the social and aesthetic status of such diagnoses. André Breton trained as a doctor, worked in a neurological ward during the Great War, and later produced with Paul Éluard works such as “Simulation of Mental Debility Essayed,” “Simulation of General Paralysis Essayed,” and “Simulation of the Delirium of Interpretation Essayed.” Samuel Beckett translated these pieces, and his early novel Murphy (1938) invokes the psychiatric language-games and institutions that he studied in the 1930s.[20] Reactionary cultural commentators, such as Max Nordau, anxiously anticipated this convergence of culture and psychopathology and regarded experimental art and literature as potential vectors of contagion through which mental illness could be transmitted from degenerate artist to audience. Nordau expanded the language-game of diagnosis to generations, possible futures, and styles of writing such that modern art and literature could be diagnosed in advance as “degenerate” and “pathological.”

Given the complexities and uncertainties that have plagued psychiatric diagnosis since the era of modernism, what lines of research might such instability suggest to scholars of literary and cultural history? With regard to the mental illnesses of the past, we might bracket questions about the ontology of “hysteria” or “schizophrenia” to reframe such diagnoses as speech acts whose rules and effects were contested by many actors, including doctors, patients, and writers. Rather than broadly generalizing, we might reconstruct what Hacking calls the “ecological niches” that made it possible for certain expressions of illness to emerge in particular moments and places (Mad Travelers, 86). Such niches might include factors such as diagnostic vocabularies, institutional and professional rivalries, discursive techniques, and even epidemiological events. Psychiatric diagnoses may be regarded as hybrid objects—singular points of biosocial entanglement between doctors and patients, language-games and bodies, histories and biologies. Wittgenstein suggests that “to imagine a language means to imagine a form of life.”[21] We might therefore set ourselves the task of reconsidering the looping logic of diagnostic language-games of diagnosis in which not only doctors and patients but also writers and readers participated. To do so would be to recover the forms of life that diagnosis spoke into being.