CHAPTER 9

Alienists Work to Systematize Haphazard Causal Data, 1854-1907

What asylum statistics asserts about heredity is at bottom relatively worthless. From the considerable differences among institutional tables, we can much more readily draw conclusions about the observers than about the observed.

—Paul Samt (1874)

My examination reveals a hereditary taint on the side of an alcohol-dependent father who ended his life by suicide after a business collapse.

—Thomas Mann, Felix Krull

Andrew Scull remarks that the teeming statistics of the asylum age “tell us more about the confiners than the confined.” Asylum doctors themselves, as we have seen, were among the harshest critics of their own numbers. Still, it was comforting sometimes to experience a number as a simple fact, a moment of release from the burden of interpretation. Journal editors in Germany, who could be brutal in their attacks on naïve number-mongering, nevertheless used figures from asylum reports as filler. A report having come to them from Ohio, Venice, Petersburg, or Colditz might be summarized compactly as the numbers of patients admitted, cured, improved, unimproved, and dead, then reduced still further into a cure rate. Causes also were reduced to nuggets of data. A snappy paragraph in the AZP in 1850 passed on news that mental illness affected 70 of the 8,000 inhabitants of the Faroe Islands, a ratio of 1:110, and provided a short list of causes, with heredity in the lead at 22. Their source judged this to be an undercount.1

In insane asylums, as in so many institutions, official sanction gave numbers a validity that did not depend on independent truth. The pronouncement of a cure meant the patient would be discharged, and if a brother and sister were each diagnosed as mentally ill, this was relevant data for measures of the force of heredity, even if doubt remained. Asylums struggled at every level with data management, which became more complex because it was more consequential when a decision how to classify had immediate consequences for patients and their families. Once these numbers left the sites of decision, to be accumulated in books, files, or cabinets, they could be treated as neutral data. Techniques for merging and processing such data grew increasingly elaborate over time, especially after 1870. In German lands, they provided a basis for the study of heredity in medical research institutions, notably the new system of psychiatric clinics. In Britain and America, where nonmedical statisticians and geneticists became the most prominent researchers on medical and psychological heredity, records of institutions made the work possible. In many cases, doctors had already created elaborate databases unbeknownst to the scientists, who welcomed the numbers as manna from heaven, the answer to their prayers.

The history of data and recordkeeping is a very human history, involving decisions with welcome or distressing consequences for people at moments of particular vulnerability. The data may subsequently be converted to infrastructure, the basis for bureaucratic and scientific work, again with hopeful or terrible implications. Accessibility and manipulability were greatly enhanced in the late nineteenth century by filing technologies and then by electronic ones. The basic structure remained intact throughout the following century.

Knowable Family Histories

Within Europe and North America, including institutions serving Europeans in the colonies, it was possible to impose a grid of patient characteristics, outcomes, and causes. Non-European populations did not fit well, even when the effort was made. The idea that insanity might be altogether different among these peoples, or even a disease of the civilized only, lingered on. The English in India found enough lunacy there to justify a system of asylums, yet their cases diverged from the European norm. The leading cause, far outpacing all others, was ganja (cannabis). Heredity scarcely came into the picture, and the doctors never explained why. We may suppose that local informants did not volunteer such causes or tell stories of manic uncles and dotty grandmothers. British medical officers seem to have had little access to information on families or, indeed, on any etiological factors.2 While an interest in ancestry and lineages is common among the peoples of the Earth, the ability and inclination to share knowledge of family illness is anything but routine. The first state asylums in California, as we have seen, also lacked information on heredity. It is no surprise that the segregated state asylum in Milledgeville, Georgia, in the final decade of the nineteenth century recorded fewer causes and much less hereditary illness for its black patients than for white ones.3

The recognition of insanity as a persistent public-health menace did not translate automatically into higher reported rates of hereditary causation. The rare historians who pay any heed to the ubiquitous tables in asylum reports might well anticipate, as I did, a progressive increase in the figures for heredity. If there is any such trend, it is weak and inconsistent, overwhelmed in every institution by the effects of changes of administration. At Hanwell near London, where the male and female divisions had different superintendents, the distributions of causes for men and for women diverged radically.4 A few medical officers did not even include heredity as a cause, and the rest might or might not call specific attention to it. Hereditary influences on insanity had long been familiar to medical and lay reporters alike, to the point that case reports from early in the nineteenth century often mentioned casually when a patient was not hereditary. In 1885, in a fifty-year report from the asylum at Winnenthal in Württemberg, the assistant physician commented explicitly on shifts in recording practices. The effects of hereditary disposition, he remarked, have drawn much attention in recent years, yet these results were unavoidably inexact.5

By 1870, most asylum doctors were required to fill out a form containing specific questions about family background or hereditary causes. The rising medical hereditarianism of the late nineteenth century, reflecting new conceptions of professional responsibility and a growing sense of medical failure, was closely allied with new data practices, including case records as well as numbers. These were developed most impressively in Germany, where data on insane heredity grew to monumental proportions. Asylums and special schools were turned into sites for purposeful collection of hereditary data. Biometry and Mendelism, so often construed as the foundational methods of hereditary science, depended utterly on new techniques of data recording and filing.

English Cases

Assigning causation, if it went beyond merely recording a declaration by the patient or “friend,” required some interpolation or untangling of the evidence. A young woman, Elizabeth T., admitted to Bethlem in 1854 was assigned “anxiety of mind” as cause, with a clear “no” for “relatives similarly affected.” But after a report from a family member cast doubt on this, she admitted that her brother had died insane and that her own terrible anxiety arose from a fear that she would follow him. She was, perhaps, no victim of heredity but of hereditarian fears. Another new patient, Elizabeth Y., mentioned her niece, P, under “relatives similarly afflicted,” and gave, as her own cause, “anxiety about an insane relative.” That relative was P, who had been discharged from Bethlem just a few weeks earlier. Did P succumb to reciprocal anxieties, niece and aunt each dragging the other down by their bootstraps? But the problems of Elizabeth Y. went deeper. According to the medical certificate, she fancied that her food was poisoned, that she was threatened with violence by an unknown person or “stilettoing” by a neighborhood Italian, and that Papists wanted to confine her. She also charged her husband “with acts of adultery with a Servant Girl.” For this charge there was no basis whatsoever, reported Dr. Albert Hall of Hounslow. How did he know? “This information I had from her husband.” The minutes of the Bethlem subcommittee record that on 9 June 1854, this man, Thomas Y.—and not any conspiracy of Papists—had presented a petition for her confinement, supported by his own account of the causes and by two medical certificates. Did anyone imagine that he could be an interested party? A Bethlem physician subsequently interviewed Elizabeth Y. for a case history, but only after the legal process was complete.6 Superintendent Charles Hood always insisted on a spirit of openness in his asylum, not for the sake of the patients but to promote statistical science. “The facts which occur, and the information which may be obtained in this Institution, ought not, I conceive, to be confined within its walls; it is the interest of science, and the benefit of humanity, that they should be made known.” It was safer to publish statistical nuggets than to air family disagreements.7

The English system advanced toward nominal uniformity through parliamentary acts and resolutions of the Medico-Psychological Association, which might in turn be incorporated into rules sent out by Commissioners in Lunacy. Case books and admission registers in England had grown uniform by the 1870s to the point that they could be printed by medical stationers and sold from catalogues. At Hanwell, the case book devoted two folio pages to each patient. The first provided spaces to fill in basic information required at the time of admission and to affix the medical certificate. The bottom of the page had more little spaces for notes of the medical interview with a relative or acquaintance. The second page was set aside for case notes, which came to include more frequent and longer entries, in accordance with advancing legal demands. The doctors now included a case history for each admitted patient, which often, however, was nothing more than the items on the printed form strung together into a paragraph. It was like an Office of Medical-Legal Circumlocution. The law insisted on proof that admitted patients were genuinely ill and that the asylums were providing medical treatment. Medicine complied, legalistically, the letter followed and the spirit forgotten.8

It was hard to make these inscriptions consistent without a standard as to what made a case hereditary. In the Hanwell asylum books for 1873–1874, we discover that the friends of Emily G. rejected any role for heredity. When asked about relatives affected they mentioned two brothers and a son. By contrast, the sister of Sarah G. said her case was hereditary because of her mother’s cousin. The daughter of Clara R. denied hereditary influence, despite a sister who was nervous and had severe headaches, while George M. considered the suicide by hanging of a maternal grandfather sufficient to mark Maria M.’s case as hereditary. Stephen R. was pronounced hereditary without any mention of relatives, and William S. because of a father and cousin who died of softening of the brain. Although William E.’s illness was not hereditary, it seemed relevant to the friend that his mother was paralyzed and his brother was “ultra religious, crying & praying a good deal.” The friend of Christopher C. left open the question of cause: “Great aunt had a daughter who died insane & other relations from the stock, but whether the taint was through marriage is uncertain.”9 Two decades later, in 1893, the husband of Mary F. said her disease was not hereditary, despite a sister committed to an asylum, while the melancholic Ada B. required no more than a “light-headed” mother to earn an emphatic “Yes!” to heredity. The medical superintendent of the female department complained in 1876 of all the friends who held back information or who were too dull and stupid to provide “anything approaching a correct account of the antecedents of the patient.” Despite these drawbacks, he claimed to have reliable histories for 25 of the 111 new admissions for the year, “and from these we gather that hereditary taint was the most frequent cause.”10

At other times and places, for example, at Bethlem in 1894, the doctors did not even ask if the case was hereditary but made their own judgment based on what they could ascertain about relatives. The physicians were always alert to families with more than a single member in an institution. The more serious made efforts to correlate cases from multiple institutions. Around 1890, physicians at Hanwell were actively pursuing such data.11 Occasionally, we find that a doctor has corrected the original record after learning of the confinement somewhere of a near relative. Emily S., admitted to St. Luke’s Hospital in January of 1876, was entered as not hereditary, but the doctors reversed this a year later, with the comment: “A brother has lately been admitted into one of the Middlesex County Asylums. Some few months back a younger brother was drowned, this in all probability was accidental.”12

The Futility of Standards

The Committee on Statistical Tables of the Medico-Psychological Association recommended in 1882 the registration of at least one cause in addition to heredity or previous attacks. This change, it stipulated, should not add to the burden of recording. Indeed it would not, if data entry was treated as an empty ritual. Practical administrators were more concerned to learn which among the new patients would be dangerous or suicidal.13

The English form of statistical standardization seemed often to reduce to specifying questions on mandatory forms while providing little or no guidance on the structure of appropriate answers. It reached a pinnacle of pointlessness in the consolidated reports of the Metropolitan Asylums Board from 1888 to 1913. These publications were credited to a statistical committee, which took a professional interest in the numbers and rearranged the forms often enough to obscure the results. Its responsibility extended to institutions for imbeciles, which had no tradition of registering causes, but not to insane asylums, which did. By far the most common cause given was old age, which could be asserted without external evidence. Congenital patients were assigned causes like teething and epi.tify. In 1897, any cause could be registered as either predisposing or exciting. By 1905, there was a third column, “predisposing or exciting,” for cases in which these could not be distinguished. Although causes were to be classed either as moral or physical, one of the asylums, in Darenth, assigned only physical causes. While “Unknown” was the most common entry, sometimes overwhelmingly so, the list of causes kept expanding, requiring ever more folds in the tables provided to sum up years of results for multiple asylums. In the report for 1913, there are fifteen classes of etiological factors with up to ten subcategories. Hence most of the boxes, even on the summary table combining all institutions, were empty. Nonzero entries would spring up in some category for two or three years then melt back into the void, leaving a strong impression that such entries registered the crotchet of some physician and were nothing like neutral data.14

The London tables, already too large to be reproduced usefully in a book like this one, are scarcely in the same league as the monumental volumes of Prussian asylum statistics from the last quarter of the nineteenth century. The Germans made an excellent show of medical and bureaucratic efficiency in the registration of hereditary data. There were more than a million spaces to be filled in, most of which required sorting as well as tallying. In contrast to all the blank spaces in the London tables, virtually every space in the Prussian ones was filled in, year after year. Yet these incomparable tables seem to have held very little medical or scientific interest.

Anyone tempted to reduce such behaviors to cultural stereotypes should cast a glance over the registration of causes in New York at about the same time. In the report for 1890, every one of the 721 female patients admitted in the two years beginning October 1888 had an ascertained cause. For the male patients in the same period, 406 out of 732 were not ascertained. For 184 females and not a single male, the assigned cause was hereditary predisposition, while 234 female cases, but not one male, were attributed to moral causes. These differences seem to owe more to recording practices than to gender stereotypes. The leading causes in Kings County (Brooklyn) were alcoholism followed by heredity, while in Monroe County, which includes Rochester, both were zero.15

So many boxes filled, but to what end? In social and administrative statistics, as in science, completeness could be its own reward, reflecting on the moral character of the workers as well as on the efficiency of the organization. Also, statistical unification was an element of political unification. The Prussian tables reflected deliberations of asylum directors from all over Germany and had implications for the numerical coherence of the new German state.16 Beyond these moral, practical, and symbolic reasons, there lingered a hope to understand causes, including the effects of an industrialized, urban civilization on susceptibility to insanity in its various forms. Alienists tried to clarify these questions using large-scale statistics.

The dream of international comparability was never abandoned. In November 1879, Ludwig Wille crossed town from the Basel asylum to a meeting of the Basel Economic-Statistical Society to reiterating his appeal to “alienists of all nations” for uniform asylum statistics.17 Belgium revived the campaign for international statistics at a congress of “phréniatrie et psychopathologie,” held in Antwerp in connection with the 1885 International Exposition. Among the leading themes of the congress were uniform statistics and the relations of insanity to criminality. Quetelet’s nation was still admired for its censuses of insanity.18 Ferdinand Lefebvre told of the successes of the Belgian system, and Hack Tuke of the English one. Albert Guttstadt sang the praises of the German model, and in his written report for the doctors back home, explained that the discussions in Antwerp were organized around his presentation. He reassured them that in view of the immense labor they had invested in the existing German statistical system, he had offered no concessions. Even so, he welcomed, in the name of greater uniformity, the appointment of a commission on which he would serve beside Tuke for England, Wille for Switzerland, and others from France, Austria, Russia, the Scandinavian lands, Holland, the United States, and Argentina. Yet none were to enter the promised land of perfect uniformity.19

Forms of Disease

Statistics largely defined the program of disease classification. Abnormalities readily recognized from their physical symptoms, such as cretinism, epi.tify, and, often, idiocy, were easily counted and inserted into tables. Insanity in its narrower sense, which usually presented no stigmata and rarely appeared before adulthood, was a much thornier problem. The ancient categories of mania, melancholia, and dementia, used by Esquirol, seemed too inflexible to mark off a disease so protean. Every attempt at a statistical classification ran up against this obstacle. Although asylum care was often based on symptoms, it was widely assumed that a proper diagnosis must provide useful guidance for treatment, and that illness form would correlate with cause. The tabular genre showing disease form on one axis and cause on the other goes back to Holst’s census of Norway in 1828.

The Diagnostic and Statistical Manual of our own day, that inventory of treatable mental conditions, preserves in its name the ideal of statistical compilation. In the late nineteenth century, the push for an acceptable classification of mental illness was already driven by statistical ambitions to apply a consistent standard across institutions and in every nation. Inconsistencies in the assignment of categories seemed more harmful even than doubts regarding their validity. A new classification could be made real only by adopting new standards and new tabular forms.20 And these data forms had legs. By 1892, the German ones were being reproduced in Tokyo.21

Since patients in the late nineteenth century were likely to be sorted into an institution for care or for cure based on an assessment of curability, classification was as much or more about prognosis as about treatment. A common nosology of the 1870s and 1880s distinguished between “primary” forms of mental illness, such as mania and melancholia, and an undivided secondary form (secundäre Seelenstörung). A case that had developed this far offered very little hope and might properly be consigned to a custodial institution. Heredity entered mainly into a different kind of prognosis, which took in the capacity of the patients to marry and have children and how many of their children would be mentally ill. Such considerations were often merged into discussions of disease classification and its implications. An essay on insanity in 1882 with some comparative international statistics integrated disease classification into a mélange of prognosis, heredity, and social policy. The categories are unreliable, the author conceded, yet we need them for planning.22

Emil Kraepelin’s proposed divide between a manic-depressive form of insanity and dementia praecox was in part a response to debates on standardization, even as it drew from a wide field of social medicine. His data practices, too, were linked to statistics. He entered patient information on Zählkarte, cards similar in many ways to those used for patient registration and for the census. They provided a fixed array of spaces in which to enter designated bits of information regarding the patient and the course of the disease—a compact repository of cases that he could sort and rearrange in search of coherence. These have been called, in English, diagnostic cards, since that is how Kraepelin used them, with enumeration as only a secondary purpose. Such cards had a wide-ranging role in the information economy of the asylum and clinic. As Eric Engstrom points out, Kraepelin had used them ever since he was a medical assistant in Würzburg in the 1870s, and they reflected Bavarian recordkeeping from that era.23

For the sake of his database, Kraepelin favored a rapid turnover of patients. His was another instance of prognosis giving shape to diagnosis. Dementia praecox was, for him, quite literally a premature descent into the mental weakness of old age. It had a highly unfavorable prognosis, and he dispatched these patients as quickly as he could to a custodial hospital.24 By 1900, he was putting more and more emphasis on social policy, including prohibitions on alcohol, the control of syphilis, and restrictions on reproduction, to combat mental illness. “Finally, the fact should not be underestimated that in the care of the insane in institutions we possess the only means by which it is possible to eliminate the most frequent cause of insanity, namely heredity.”25

Records with No Purpose

Statistics of mental illness provoked some sharp controversies. Paul Samt, quoted in the epigraph, complained of the execrable quality of observations whenever they become too numerous. His real objection, however, was the misguided focus on human actions rather than brain tissues and nerve ganglia. “A mental illness would have a natural scientific explanation if its mechanics were decoded in the region of the unconscious, if we knew the laws by which concrete pathological forms develop in brain matter.”26 An 1866 Lancet review of asylum reports from England and Canada referred to statistics as so much “idle labor” on account of faulty categories. A decade later, a Lancet commission to examine English lunatic asylums wrote of “fruitless researches” leading to an accumulation of “false data.” “A bog, over which myriad bewildering lights dance and delude the traveler, would be a faint picture of the perplexing and impracticable province that lies before the explorer who attempts to investigate the subject of lunacy by way of statistics.” The problem here was the dishonest assignment of exciting causes in order to condemn or “to elicit sympathy for the afflicted.”27 In 1895, the Lunacy Commissioners in Scotland dismissed the annual tables as lacking even “the smallest statistical or scientific value.” This concern was the erroneous merging of distinct populations.28

Readmissions and relapses provoked some of the hottest controversies in asylum statistics. The admission book, which at first appears boring and harmless, could be a cauldron of controversy. Every admission implied, eventually, a departure, whether by cure, death, or “elopement” (escape), whether at the prompting of the medical superintendent, the decision of a court, or the irresistible entreaties of family. A discharge closed the account. What, then, was to be done if the patient returned? It was the same patient, as the records showed, but it was a new case with a new numbered entry, portending a new result. The first outcome had been entered already into the official report. Should this now be altered? What if, instead, a patient were admitted to a different institution, perhaps a custodial asylum? This was not uncommon, but it complicated the metaphysics and introduced the practical problem of information sharing. Hagen’s nightmare, a world central commission, might be the only solution.

Early nineteenth-century medical explanations offered a convenient evasion of this dilemma. Insanity involved a predisposing cause and a triggering factor. Even when asylum care restored health, the susceptibility remained. Little surprise if the discharged patient, returned to former circumstances, should again fall ill. On that hypothesis, a readmitted patient really was properly a new case. Alienists sometimes defended this view of things, though with hints of a bad conscience. At Siegburg, Maximilan Jacobi was challenged by a commission of investigation, appointed in July 1841 by the estates (Landtag) of the Prussian Rhine Province. They were by no means hostile to Herr Higher-Medical Privy Counselor Jacobi, but found him to be unreasonably insistent on using arbitrary criteria of curability to classify the insane as “decisively unsuited” or “only in the very slightest degree suited” to the institution. They also reminded him that most cures come from nature. Siegburg, they insisted, was not only a medical institution, but an arm of state policy, charged to care for its inmates and not only to deliver medical treatment. His claim of so many dubious cures was complicating state policy, and the commission could not ignore the 86 repeat patients on the books, who, in total, had been admitted 204 times. Proper accounts would reduce the 469 recoveries he claimed to 373. A lot of money could be saved by recognizing Siegburg as a charitable institution, rather than a medical one, and replacing the paid personnel with Sisters of Charity, as in Maréville, France. Jacobi, of course, defended his institution. The Sisters of Charity turned Maréville into a scene of horrors, he said. And insanity is like intermittent fever, striking people repeatedly when conditions are similar.29

Readmission as a statistical problem, though never forgotten, acquired new prominence when Pliny Earle raised it in his annual report for the Massachusetts asylum at Northampton in 1876. The AJI promptly reprinted the report, which was widely noticed, and later Earle published the original essay with seven supplements as a book, The Curability of Insanity. Asylum cures had been greatly exaggerated by the early directors of American and English asylums, Earle argued. He retraced some early claims of superlative results, including George Man Burrows with his 91% cures for new patients in London, the Hartford Retreat at 91.3%, Samuel Woodward’s brief moment above 90% in Massachusetts, and finally John Galt’s achievement in 1842 of a perfect 100% at the Eastern Asylum in Williamsburg, Virginia. Galt must have been a first-rate enterpriser. The key was to choose an advantageous formula. Cures should be divided by discharges, not by new admissions or patient population. It helped to exclude deaths, as Galt did. Perfection now followed as a mathematical necessity, provided the asylum was never compelled to release a patient while still insane. In fact, asylum directors in this age of wondrous numbers were not simply unscrupulous but had faith in their own magic. They received in addition a big assist from multiple cures of the same patient. The most extreme example that Earle could find involved a woman at the Bloomingdale Asylum in New York City, who had been admitted 59 times in 29 years, and discharged as recovered 46 times.30

The critique struck close to home. For 5 of those 29 years, Earle was the Bloomingdale director. When he arrived in 1844, he introduced a medical register with about 30 columns in which to enter information and rows wide enough for this enormous volume to take the place of a case book. One column, headed “No. of admission,” was there specifically to keep track of readmissions. He subsequently tracked down the skeletons in these dusty closets in pursuit of a solution to a distressing problem. Why, decades later, were asylums unable to come close to matching the medical successes of their earliest predecessors? In 1844, he had written that for cases without constitutional eccentricity or weakness of intellect, a director should be able to cure 80%. He claimed to cure patients whose “particular propensities and hallucination” appeared hopeless to the nonalienist eye, such as this woman who arrived in 1846: “God forsaken her, torments of hell upon her! Would not eat because offend God. Smoked one pipe to please this Lord and one to please the Devil. Attempted suicide by drowning, hanging & cut throat.”31 It was an age of giants, but not on account of a few heroic cures. The percentages were the problem. By the 1850s, an 80% cure rate was an impossible dream. There was some consolation in discovering the tendentious accounting through which these men had achieved this irreproducible success. Thurnam, Earle now remarked, had had it right when he declared that for every ten patients, five will recover and five will die, while three of the five recovered will suffer additional attacks. Earle’s explanation of the radical drop in cures was welcome news in London, where Hack Tuke of the founding family of the York Retreat invoked it in his 1881 presidential address to the Medico-Psychological Association.32

There were other repercussions. The superintendent at Worcester promptly recomputed its tables and found all its figures for recoveries to be in error. “It is a sad and almost cruel blow to the worth of the earlier tables of this Hospital.”33 A private institution in Vermont, subsisting mainly on state contracts, had to fend off critics who, ignoring a shift in the basis of calculation, claimed that the cure rate had fallen abruptly from 32% to 8%. Earle commented: “It is said ‘figures do not lie,’ but if manipulated they certainly can be made to misrepresent the truth.” The AJI, in an unsigned review by its editor John Gray, gave credit for these pretended insights to an unnamed local party who had anticipated Earle by a dozen years. There in Utica, it barely mattered, since the rate of readmission there was very low.34

The Flexibility of Files

The British plan for rationalized recordkeeping at the end of the nineteenth century was to have everything in its place in a giant record book. German case records, by contrast, had long been kept in files. If the space set aside for a patient in a case book filled up, there was nothing to do but move to another blank page. The case file could simply grow fatter, almost without limit. Documents could also be separated and recombined. The shift to census cards or counting cards (Zählkarte) made for still greater flexibility. It was not yet the accountant’s paradise, for the cards and files, unless copied, would have to reside in a single location. For the sake of research, Kraepelin fought to keep patient records at his Heidelberg clinic rather than give them up when the unfortunates who did not improve rapidly were moved to custodial institutions.35 The ideal of a Centralstelle, or data center, privileged large-scale investigation and centralized management over local medical use. Despite his inexhaustible appetite for data, Kraepelin complained bitterly of all the meaningless numbers that the administration demanded from him. Yet the abundance of documents in patient files was too much to handle for a census bureau, which liked its information thin.36

A patient brought in to a Bavarian asylum in 1875, as elsewhere, would be promptly examined by a doctor. The history, supplied by an accompanying family member or health official, focused on the origins and progress of the illness and its causes, beginning with a family history. This case history was used to fill out the census card for admission, whose entries must, to a large degree, have structured the interview. The information was then transferred to a bound arrivals book.37 The Bavarians kept detailed accounts of heredity. A patient admitted in January 1877 at the Munich asylum, but sent immediately to the nearby institution of Eglfing, was entered in the arrivals book for women as “direct” in the heredity column. The case history begins:

Frl. Henriette W from Frankfurt am Main. Israelite confession, hereditary burden. The father suffered from periodic epileptic attacks and was in the asylum at Eichberg and Klingenmünster in 1861 and 1862, specifically in a condition of melancholia with suicide attempts. The mother was, according to the illness history a “nervous” lady.

Henriette W.’s mental illness began in puberty when she refused to eat, and so on. Each admission document assigned her a different number, and cross-referencing allowed movement back and forth among them.38

These careful registration procedures enhanced the ability of doctors and statisticians to keep track of readmissions and hence to make patients rather than cases the focus of attention. The files (Akten) provided a ready guide to the entries in admission books. Eglfing File 2413 is the paper incarnation of Ludwig S., who arrived on 1 May 1881 and was readmitted in August 1881, 1884, 1886, 1888, 1890, 1894, 1897, and 1912. Each admission corresponds to an entry number: 3522, 4477, and so on. The father, still living in 1881, had been mentally ill once, while the mother had died in 1865. The file indicates six siblings, two brothers and four sisters, of whom one sister had been mentally ill for ten years, and suggests that the insanity developed from insomnia. The first record in the entry books reduced the family relationship to “direct from Father and collateral a sister.” The second entry mentions that the relatives were melancholy, while the third is a bare statistical entry: “direct and collateral.”

The Aufnahmszählkarte, or statistical admission card, devoted a generous portion of its very limited space to information on heredity. Although some cards were filled out with handwritten particulars, most items could be completed simply by underlining or connecting the appropriate items with a curved line. The information required included the identification by relationship of abnormal or mentally ill family members, including, by 1891, children of the patient. The cards also specified the medical defects or conditions of specific interest, including blood relationship of parents. The two items following concerned criminality and disease form. The bits of added detail must have complicated this statistical registration. Conversely, the thin economy of statistics did not encourage rich detail, though it definitely enhanced data quality. Ludwig’s page in the admission book recorded his transfer in 1904 to another Munich asylum, Gabersee, as well as his subsequent return to Eglfing. Until recently, it had required heroic efforts to keep track of patient transfers. Social and medical investigations of the insanity problem depended on this capacity to track patients and their relatives.39

Catching Up with the Germans

Evolving ideals of recordkeeping in Germany, somewhat paradoxically, were tending to put more responsibility in the hands of doctors in local institutions, for the sake of a form of research that had moved away from the annual report. Large-scale printed statistics, too, gradually lost their appeal. Heroic compilations like the Prussian folio volumes on causes of insanity moldered in libraries. Giant tables, assembled bureaucratically, served budgetary and demographic ends. Research on scientific questions seemed to require more supple tools of information: technologies of sorting and filing. By no means did this imply the death of standardization. It required instead a reshaping of standards into uniform templates for personalized information.

In America, improved recordkeeping emerged in the 1890s as a key element in a new vision of the medically effective asylum. The annual report of the Taunton State Hospital for 1893 boasted of records made useful at last by prompt recording at the bedside. “A medical record, to be of any value, must be an accurate statement of the condition of the patient, and hence should be written by the physician himself who makes the examination, or at his dictation if written by another.” The doctors were now required to examine each new patient within twenty-four hours of arrival and enter the results in the case book, then to repeat the process every week for the first three months. If he was very busy, he should put aside less important duties for the sake of complete records.40 The inspiration came partly from Adolf Meyer, who arrived in America with a brilliant European pedigree that included education at the University of Zurich, postgraduate work in Paris, Edinburgh, and London, and additional work at Burghölzli in Zurich. He began in 1892 as pathologist at the new mental hospital of Kankakee, Illinois. In his first report to the governor, he explained that diseases run their course under one system in about the same way as another. “The main point is the conscience of all parties is soothed by the idea of superiority of therapeutic skill.” Any method is good that does no harm. He worked there to institute a system of close patient observation and anthropometric measurement, supplemented by examination of relatives to detect signs of hereditary degeneration.41

In recruiting Meyer to the Worcester asylum in 1895, the directors authorized a research tour that began with several institutions in Italy, followed by six weeks with Kraepelin in Heidelberg, brief stays in other German institutions and in Zurich, and a visit to F. W. Mott’s new laboratory in London. Meyer returned with a vision of teams of assistants trained in methods of case taking, including “concise and accurate recording at the bedside of the results of their observations.” The old books, based on recollections of an assistant and recorded weeks after the occurrences noted, would now give way to a corps of interns accompanying the assistants on their rounds and setting down the results of batteries of tests that Meyer planned to introduce.42 In 1901, he was hired as pathologist for the asylum system of the state of New York to set up a training school and work his magic of renewal, substituting for his predecessor’s program of laboratory testing a new focus on clinical observation and information.43

In London, too, at the dawn of a new century, the winds of renewal were blowing. Keeping case books merely to satisfy a lunacy commission or defend against charges of “malpraxis” was one thing; “doing it from a scientific point of view for the advancement of the study of insanity is quite another thing.” These heterogeneous case books contained much of potential clinical value, which might be extracted at the cost of enormous labor. “If, however, it were possible to frame some simple, uniform method of case-taking, and persuade the superintendents of all the asylums to use it, which possibly might be the most difficult of all tasks, then some definite and valuable statistical facts might be obtained.” The author of this vision, Dr. A. H. Newth, envisioned “a simple method of case-taking,” allowing the state of each patient at the time of admission, the progress of the case, the treatment adopted, and the result, to be taken in at a glance. For this, a uniform system would be required.44

Newth condemned the annual reports as costly and time-consuming, useful for nothing except “to light the fire.” His ideal was an efficient system of cases. He especially admired the “check cards” that Dr. A. R. Urquhart culled for private use from the case books at James Murray’s Royal Asylum in Perth, Scotland. While the Perth case books were “models of completeness,” the cards provided a “bird’s eye view” and allowed the rearrangement of patients for new inquiries. The bulky case books brought definite advantages. Urquhart compiled from them family trees for 331 of his patients. Newth, however, was coming to think that “the very completeness of detail in case-taking is a bar to the comparative study of cases.” Keeping these books had also given him a case of “scrivener’s palsy.” He determined to emphasize the minimization of wasted effort, leaving it to Urquhart to balance economy with detail. So much accessible data was grist for the mill of biometricians David Heron and Karl Pearson, at University College, London, who analyzed his data and calculated hereditary correlations. Pedigrees like Urquhart’s were no less suited to hereditary statistics in its Mendelian form.45