Thinking about Methods: Longitudinal Research and the Reorientation of the Postwar American Mental Sciences
No observation and no longitudinal study has replaced or . . . will replace the value of the psychoanalytic observation proper for the study of child development. [Yet longitudinal research is] the method to show how various phases of the past were interrelated; to see the life history as a whole, as it is organized by the personality and in turn has organized the personality.
—ERNST KRIS, “Notes on the Development and on Some Current Problems of Psychoanalytic Child Psychology” (1950)
If many of the contributions to this collection represent attempts to find meaning about postwar American society in the responses of Yale University Child Study Center participants, this essay seeks to do similarly for intellectual culture by looking at the types of questions investigators asked. Why did researchers, and in the case of what follows, those interested in the study of children, elect to use certain methods? What counted as data, and why were these data arranged and discussed in the ways they were? What was the relationship between researchers’ decisions to use a certain method and the underlying assumptions that organized knowledge across the fields in which they worked? Such questions are broadly interesting, historiographically salient, but above all relevant for those wishing to understand the Yale Longitudinal Study. Like much of the work carried out in the human sciences in these years, investigators involved in the Yale study gave “considerable attention . . . to methodology,” even organizing a section of their archive under that heading.1 Historians such as Jamie Cohen-Cole, Joel Isaac, and Rebecca Lemov have all recently shown that in the years around 1940 this was a fairly common stance among those who worked in the American human sciences, leading to a much greater reflexiveness about and innovation toward methods. As the quotation that introduces this chapter from Ernst Kris shows, such an orientation was equally present within the Yale Longitudinal Study. It proved crucial in shaping the study’s early architecture and eventual trajectory, and, ultimately, even had a slight bearing on how the field of child psychiatry emerged and the field of psychoanalysis was reorganized.
Unlike most histories of child psychiatry, which tend to prioritize the institutional hardware that marked the formation of the field, this chapter follows a different set of coordinates. It focuses on the methods that practitioners applied to children in the years around midcentury in an attempt to consider the relationship that existed between this nascent field’s epistemic ideals and the technical dimensions of its research program. The chapter therefore attempts to locate the Yale Longitudinal Study in the broader context of changes within the mental science disciplines, while illuminating specific aspects of this study, such as its methodological and epistemological attributes. It argues that child psychiatrists embraced longitudinal research because they regarded it as a means of accurately tracking and representing the psychological phenomena in which they had become interested, including ego formation, development of individuality, and mother-child interactions. Such considerations about how accurately a particular method described social phenomena were, as the historian Joel Isaac has shown, symptomatic of the “tool shock” that the American human sciences experienced more broadly in these decades. The proliferation of new tools, like longitudinal research, helped to provoke questions about investigators’ root conceptual commitments, and to stimulate discussion about the relationship between tools and theories. Unlike the disciplines that Isaac and other scholars have examined from this perspective, which took up such questions in the seminar room, graduate training program, and textbook, child psychiatry had little equivalent apparatus, and relied greatly on participation in research projects to stabilize any such theoretical and technical adjustments. The longitudinal study was popular among child psychiatrists because it filled the gap between the prior lack of means of studying children and the later development of subspecialist residency programs.2
Children first became a serious subject of inquiry within the mental sciences only in the early years of the postwar period. Before that point, most psychiatrists associated childhood with mental deficiency, rather than mental illness, posing questions about children inspired by the recapitulation perspectives of an earlier generation of natural scientists.3 The introduction that the pediatrician Edward Parks wrote for the first edition of Leo Kanner’s landmark text Child Psychiatry (1935) was typical of such attitudes, referring to children as “simple,” defined by “brevity of experience, the ignorance of life, the simplicity of thought.”4 Such assumptions pervaded the mental sciences in these years, underpinning practitioners’ tendency to characterize the mentally defective as “childlike,” as well as the American Psychiatric Association’s characterization of “senile psychoses” as “childish emotional reactions.”5 This compact between age, psychological development, and mental disorders lingered through at least the mid-1940s, though such views began to be displaced thereafter, for reasons I will elaborate later. What was overwhelmingly treated with skepticism before the war then—that children had a capacity for intellect, an independence of action and thought—became the constituent parts of childhood at the dawn of the Cold War. References to children’s “personalities” swiftly increased in both popular and specialist journals, and as a report by the Group for the Advancement of Psychiatry put it in 1950, “the child began to be thought of as a child and not as a miniature adult.”6
The reasons for this shift lay in a host of factors but foremost were the rise of ego psychology and neo-Freudian precepts, developments with which Ernst Kris, the study’s intellectual progenitor, was closely associated. But these changes also owed much to the growing place of epidemiological techniques in the health sciences, psychiatry’s changing focus away from pathology to what Kris’s wife, Marianne, called “‘varieties of health’” or normality, as well as a series of more nebulous changes, including the drift of American medicine away from the notion of a universal human subject to “niche” categories such as children and adults.7 All of these factors were at play in the Yale Longitudinal Study.
Although it might have gained a name from a book published in 1935, and had been recognized officially by the International Congress of Mental Health in 1937, it took another decade before the field of child psychiatry fully emerged in the United States.8 Practitioners and gate-keepers writing in the early years of the postwar period regularly noted the “rather recent and very rapid” emergence of child psychiatry, while admitting the slightly longer development of child development studies, which were “concerned with problems of the mentally defective child.”9 Of course child psychiatry owed more than just its name to what the Johns Hopkins émigré psychiatrist Leo Kanner wrote in Child Psychiatry. But from the day of its publication sympathetic and interested commentators adopted that work as their lodestar, reflecting the field’s nascent development in the interwar decades.
Anna Freud’s work The Ego and the Mechanisms of Defense (1937), which was first published in German in 1936 and given to her father on his eightieth birthday, won a similar place among some Continental psychoanalysts with interests in these questions, especially Ernst Kris, the leading force behind the eventual Yale Longitudinal Study. Kris celebrated The Ego and the Mechanisms of Defense for introducing “a silent change . . . in the method of observation,” and praised the work for encouraging practitioners to interpret material “in light of the patient’s whole history.”10 Both of these factors would be constituent parts of the Yale study. They would also find their way more broadly into the American mental sciences through the journal Psychoanalytic Study of the Child, which Kris helped to establish in 1945 with Anna Freud and Heinz Hartmann.
While a growing interest in children certainly existed among mental-health officials before midcentury, before that point most who worked within the mental sciences did not take children seriously as subjects in their own right. There were a number of reasons for this outlook, including a pervasive belief that children lacked sufficient intelligence and independence of thought, and that there was an absence of individuality. The prominent New York psychiatrist Pearce Bailey wrote in 1922, for example, that it was “well known” that “distinct types of psychoses are practically unknown in childhood, the period before the intelligence is fully developed.” This was hardly a revelatory view, and essentially restated what Adolf Meyer and others had contended in earlier years, it being widely believed that “well-established mental disease, aside from idiocy and imbecility, is not frequent among children.” It was not just a lack of intelligence that prevented children from experiencing mental disease, though. In his earlier study on group psychology, The Crowd: A Study of the Popular Mind (1895; U.S. edition 1921), the French sociologist Gustave Le Bon described another strand in such conceptions. Drawing a comparison with crowd behavior, children, Le Bon argued, were also characterized by “impulsiveness, irritability, incapacity to reason, the absence of judgment and of the critical spirit.” Such conceptions meant that before World War II, American mental health officials intervened in the lives of children only to treat what they thought of as mental deficiency or feeblemindedness.11
Only gradually did these precepts loosen their hold on psychiatric knowledge, as the mental sciences were reoriented during the war and early years of the postwar era, not least by the introduction of new modes of explanation and the arrival of hundreds of European émigrés. One index of these shifts is that in the fall of 1948 Leo Kanner finally revised his original 1935 study, rather than just republishing it. The second edition of the work was, in the words of Kanner’s colleague at Johns Hopkins, John Whitehorn, “a wholly revised edition,” in fact a “wholly new work.” Whitehorn welcomed the changes that Kanner had made to the study, celebrating the author’s “more fully matured perspective.” But like most reviewers, he did not comment on the work’s most striking feature—the author’s inclusion of a new chapter on “Personality.” References to such a concept had been sparse in the original edition. In fact the index to the 1935 volume contained just one mention of the term, and there was no substantial discussion of it. Yet in the work’s second edition, Kanner devoted over twenty pages to the discussion of children’s personalities, explaining that in the intervening period “the vista [of psychiatry] has been expanded by interest in the attitudes and resulting behavior of persons as dynamic, clinically observable, and continually active molders of children’s personalities.”12
That the personalities of children began to warrant lengthy psychiatric dissection in the postwar years was indicative of the greater conceptual depth emerging in the study of children. Accepting that children had something as potentially coherent as a personality, and that these personalities could be studied, was a striking change from prior precepts, and one that resonated across a host of fields and disciplines. Such shifts in thinking were equally evident in the pages of Child Development, the journal of the Society for Research in Child Development. The number of articles written on children’s personalities that appeared in that periodical more than trebled in the postwar decades, rising from just fifteen between 1930 and 1945 to over fifty between 1945 and 1960. The few longitudinal studies around children that spanned the inter- and postwar decades also revealed this change. The Oakland Growth Study that began in the early 1930s at the University of California, Berkeley’s Institute of Human Development, performed a follow-up experiment in 1953–54 that focused, for the first time, on participants’ personalities and included “personality inventories, and a psychiatric assessment” as a part of the battery of examinations it used.13
The Yale Longitudinal Study was a further manifestation of these trends. The study’s intellectual bearings, though, were also taken from a series of prior reorientations within psychoanalytic theory and technique, not least the elevated importance of the ego, the wish to understand the effect of social and cultural factors, and the greater attention paid to thinking about analytic inquiry. These were the chief concerns of many émigré analysts and some of their American colleagues, especially those affiliated with academic institutions, like Kris. It was not long before their work in these years became known by the broad title of ego psychology. As other contributors to this volume point out, Kris was one of the leading figures within this movement. Yet unlike a number of his counterparts, especially Erich Fromm and Theodor Reik, Kris’s efforts to tidy up psychoanalytic theory were advanced with exemplary caution. Typically he modified Freudian precepts by first presenting his audience with a relevant passage from Freud’s own work that seemed to authorize the refinements he was proposing, and then framing his thoughts as a gentle and legitimate “enlarge[ment of] the scope of psychoanalytic therapy.” One of Kris’s main contributions to the field was the work he did in expanding analysts’ toolboxes, an accolade he shared with many equivalent figures in sibling disciplines, including the anthropologist Clyde Kluckhorn and the sociologists Robert K. Merton and Talcott Parsons. Like these other scholars, Kris’s interest in introducing new tools, such as “psychoanalytic observation, experimental procedure, and systematic observation,” was coupled to his ideas about the role that theory played in the discipline. Kris believed that theory saturated “the everyday tools of all analysts,” and, citing the work of the British physicist J. J. Thomson, suggested that theory was “not a creed but a tool.”14 It was these tool-like properties of theory that he said allowed investigators to distinguish preconceived notions from hypotheses, to test and validate long-held precepts, and so forth. His conception of theory as a tool clearly influenced discussions among Yale investigators, as well as within the discipline more generally.
The years around midcentury were therefore marked not only by a change in the way the mental sciences conceived of childhood, and the altered values given to interpretive concepts, but also by the introduction of new lines of investigation. The field of child psychiatry was created as epistemic and methodological attitudes within the mental sciences were recalibrated around midcentury—when earlier developments crystallized in centers of psychoanalytic importance, like New York and Chicago, as well as London.
Longitudinal studies were hardly a new development in the postwar period. Physicians’ practice of keeping and updating patients’ files was one prior, rather rudimentary example of such methods, as contemporaries well knew. A bibliography compiled by two medical students in 1959, and dedicated to Kris and to Milton Senn, a pediatrician and the director of Yale’s Child Study Center, argued that the systematic recording and analyzing of data over time began at least in the late 1920s, with the various studies that Arnold Gesell conducted on human growth.15 Given his influence within the early child-study movement, and especially within New Haven, it would be tempting to follow this genealogy and claim a role for Gesell in shaping the later Yale study. But to interpret events in this way is to overlook the local institutional rupture that occurred within New Haven’s psychiatric community in the years immediately preceding the commencement of this study, as well as to mistake the very different purposes to which longitudinal methods were put in the medical sciences in the years following World War II.16
One aspect of postwar research into children that distinguished such programs from their predecessors was their deliberate invocation of “longitudinal.” Before 1945 few psychiatrists’ projects used that term in their title or literature; afterward, it was pervasive. The effect was contagious, with studies begun in earlier decades that had taken time as a variable gradually adopting the rubric of “longitudinal.” The acclaimed Fels Study carried out by researchers at Antioch College in Yellow Springs, Ohio, and begun in the late 1920s, formally located its research within this bracket in the postwar years. Only in 1958 did the Fels Study publish a report with the word in its title. It was a symptom of the moment. Between the early 1950s and mid-1970s, longitudinal studies proliferated across the human and medical sciences. In North America, new longitudinal studies were begun to investigate a variety of conditions, diseases, and statuses, the best known being the Framingham Heart Study (begun in 1947), the Twin Cities Business and Professional Men Study (1947), the Stirling County Study in Nova Scotia, Canada (1948), the Midtown Manhattan Study in New York (1952), the Duke Longitudinal Studies of Normal Aging (1955), the National Institute of Mental Health Human Aging Study (1955), and the Baltimore Longitudinal Study of Aging (1958).17
But we should not be misled by the profligacy with which the term was used in these years. After all, longitudinal studies differed greatly in size, shape, and sort. The common usage of “longitudinal” in the title of such studies tends to conceal the variations that existed between them. One striking point of difference was the scale at which longitudinal studies were organized. Such studies tended either to be very large or quite small. The Framingham Study into cardiovascular disease that started in Massachusetts in 1947 is a good example of one of the period’s largest programs, with 5,200 participants.18 In the mental sciences, Alexander Leighton’s Stirling County Study and Thomas Rennie’s Midtown Manhattan Study, which enrolled approximately 1,000 and 1,700 participants respectively, should be cast in a similar light. The longitudinal child study at Yale should not. As other contributors to this volume note, the Yale study initially enrolled sixteen families, though this figure fell to ten, and then to six when some lost interest and when others were no longer needed. Accordingly it was one of the smaller longitudinal studies undertaken in these years, similar in this respect to a parallel study begun at the Child Health Center of the University of Washington in Seattle that enrolled twenty families, but qualitatively different from Stella Chess’s medium-sized New York study, which began in 1956 and enrolled 110 children from “middle-class urban and suburban” families.19
The small number of participants enrolled in studies like those at Yale and the University of Washington allowed practitioners to obtain their data with, as Senn later wrote, “intensive and personalized concentration on each family.” Where Chess’s staff for the larger New York study examined participants every three months using a predetermined interview, practitioners at Yale saw their subjects twice a week for a number of hours at a time. The Yale study thus showed that longitudinal methods were not just used for “quantitative stud[ies], involving large numbers of people,” though it was such studies that would later define popular understandings of the term.20 (Other points in the taxonomical placement of longitudinal research that would later become crucial—between cohort and panel studies, for example—mattered less when the Yale study began, and therefore to the present discussion.)
In other ways too, the manner in which practitioners at Yale and elsewhere approached children departed from how an earlier generation of investigators had done so. Prenatal registration was a major point of distinction between the two. A planning memo from April 1951 confirmed that such goals were present in the Yale study from the beginning. “Gathering of data starts at the earliest possible stage, i.e. the prenatal interview,” the memo read. “These data and their connection with the later behavior of the child can be used to increase the predictive powers in future cases,” it continued.21 Such priorities had been absent from earlier investigations. The Oakland Growth Study carried out in Berkeley in the 1930s had drawn most of its samples from the city’s fifthand sixth-grade elementary schools. By contrast, it was de rigueur for postwar researchers in this field to observe their subjects from before birth. This change in practice stemmed largely from what historian Carl Degler has called the “triumph of culture” in the human sciences—a shift that led to the general thought that, as one committee pointed out in 1960, cultural conditioning “commenced in the womb and continues through the intimate relationship of mother and child.” The second edition of Kanner’s Child Psychiatry had conveyed this shift in opinion by lessening the first edition’s heavy reliance on physiological explanations. “The advance of psychiatry has brought an increasing appreciation of what happens to children in the process of development, over and above that what goes on within them instinctively,” Kanner wrote in 1948.22
On occasions in the early 1950s, those affiliated with the Yale study framed their work in similar terms, that is, as an intervention in debates over the roles that nature and nurture played in shaping human development. In late 1952 those who sat on the study’s research team characterized their work as being concerned with “the development of personality as an interaction between the child’s equipment and the effects of the environment.” The study’s psychoanalysts placed these claims into their own register, describing their research as attempting “to distinguish as early as possible between those sectors of the ego which develop autonomously and those sectors of the ego which develop as defenses.”23 When Anna Freud visited New Haven in October 1952, she urged those involved in the Yale study to think of “ego development [as] both inward and outward.”24 This newfound emphasis on sociocultural factors was, as Deborah Weinstein has shown, a common feature of much work in the postwar human sciences, but it was more elevated among American psychoanalysts than their British counterparts.25
The decision to begin the Yale study prenatally was also inspired by researchers’ attempts to supplement psychoanalysis’s reliance on the “reconstructive method” with other types of information, including contemporaneous observation and prediction.26 Such practices were more broadly evident within the mental sciences in these years. A number of Yale psychiatrists began experimenting with electronic recording equipment in the mid-1940s, for example. Their efforts won the endorsement of the analyst Lawrence Kubie at an American Orthopsychiatric Association meeting at which Kris also spoke, in February 1946. It was at that meeting that Kris proposed other ways of expanding analysts’ toolboxes. These included the idea that the psychoanalytic interview be supplemented by and compared to “other types of observation and better checks on the interview procedure.”27 Elizabeth Lunbeck’s recent work on another Austrian analyst, the émigré Heinz Kohut, reminds us that Kris was not alone in attempting to bring observational techniques to psychoanalysis at this time. But the point I wish to emphasize here is the manner in which the Yale study attempted to interweave these various types of data, imagining longitudinal research as a neat way of making “observational and reconstructive data . . . comparable.” As Kris told a meeting of analysts in Detroit in 1950, “Only the systematic longitudinal study of life histories, combined with attempts to predict at each point all that can be predicted about future developments, seem to meet the requirements of the moment.”28
There were other reasons why child psychiatrists in these years started to embrace prenatal registration and contemporaneous observation. For instance, inaccuracies would develop when patients either deliberately or mistakenly remembered past experiences. Writing in 1959, one analyst criticized the practice of taking information from parents on the basis that “mothers’ accounts are neither complete nor entirely reliable.”29 Stella Chess made a similar claim in the American Journal of Psychiatry the following year. “The inaccuracies of recall by patients in psychiatric treatment has been repeatedly observed,” she complained.30 But this shift was also given greater force by a prior conceptual shift. “Retrospective data are insufficiently powerful for the delineation of the processes involved in the development of individuality,” Chess added in her article. By the late 1970s, some within the mental sciences were describing retrospective methods as “much maligned.”31
Longitudinal research was also regarded in these years as providing a framework for coordinating different types of data. In the mid-1960s, the prominent American analyst Roy Schafer would conceptualize longitudinal studies as making it possible for researchers “to compare propositions based on different fields of observation,” while prompting them to clarify how propositions formulated in the course of such research “are related to specific areas of theory formation.”32 Anna Freud had made a broadly related point during her visit to Yale, commenting, “One must get the interaction of a multiplicity of factors to be able to trace the development of a personality.” There was, therefore, a clear symmetry in the value that child psychiatrists attached to composite methods, like longitudinal studies, and the subject of the Yale study, that is the personalities of children, which they also understood to be “interdependent,” interacting entities. Both the ego and the longitudinal study drew information from a number of sources and organized it through what were imagined to be analogous processes of integration. Part of this field’s attraction to such research methods, then, lay in the way these tools were imagined to mirror the social phenomena they were purportedly representing.33
In addition to their use of observation, researchers with the Yale Longitudinal Study also recorded the testimonies of children. This was a remarkable departure from how research on children had previously been conducted. As the historian Alice Boardman Smuts observes, “Before the 1930s reports on patients in children’s hospitals contained comments by parents, other relatives, and social workers, but never by the children themselves.”34 Indeed, where the Yale study’s use of observation distinguished it from previous psychoanalytic practices, an interest in children’s testimonies similarly set it apart within the field of child developmental studies. Children were now speaking subjects within the mental sciences, and the weight of attention of the Yale study in the mid-1950s fell on the discursive interactions between investigators and subjects. The shift was evident in a number of similar studies at this time, notably the University of Washington’s investigation of the “mother-child relationship.” It received equal encouragement at Yale Medical School, as some of the records published in this volume show, with their attempts to capture the children’s voices, and to report conversations between subjects and investigators.35
In addition to the hundreds of observational reports that the study’s social workers filed, the physical examination forms that pediatricians completed, and the psychoanalytic case notes written about the children’s parents, by the mid-1950s Yale researchers were also gathering data from instruments developed in psychology. In fact the proliferation of new tools within the American human sciences in these years was certainly captured in studies like this one, which made extensive use of the Wechsler Adult Intelligence Scale (WAIS), the Thematic Apperception Test (TAT), and the ubiquitous Rorschach Test. That they did so was a further symptom of what Rebecca Lemov has rightly called the human sciences’ “overweening love of data.” Some of the issues raised by this habit—for both investigators and later archivists—are addressed in Chapter 6.36
Although in their most basic form longitudinal studies were concerned with the lives of specific people, part of the method’s function was to compare potentially great numbers of individual records, plotting and illustrating broader trends in the process. As one technical guide put it, the method allowed researchers to accumulate reams of data from which general patterns might be discerned, such that “the curves for individuals [could be plotted] on log paper for comparison with a model curve appearing on transparent paper.” Two members of the Yale Longitudinal Study, Katherine Wolf and Jennie Mohr, attempted to do likewise using bar charts, and their efforts prompted Kris to note how such a representation “may teach us something on one child, but better, it can enable us to compare this child with another child.” In this view, longitudinal surveys mirrored American social scientists’ general enthusiasm for what the historian Sarah Igo has termed “aggregating technologies,” those mass surveys and polls that pervaded Western science and society around midcentury.37
Yet while the use of the longitudinal method fitted neatly with these broader trends, we should not miss the fact that it also potentially stood in opposition to psychiatry’s traditional unit of examination, the individual patient. Accordingly critiques accompanied the emergence of all manner of epidemiological methods, and the profession’s old guard grumbled that such surveying techniques “violate[d] the concept of the complete individuality of man.”38 It was, in part, for this reason that well into the late 1960s the principal officers of the Yale study would insist that their research be seen “in terms other than quantity, averages or predictive values.” “Interviews, conversations, and observation techniques over an extended period do reveal more pertinent information than questionnaires,” Senn claimed in the retrospective report he cowrote in 1968 with Claire Hartford, a staff member at the Child Study Center.39 Yet this public wish mistook the surreptitious ways in which quantification entered and structured such arenas. Frequently, qualitative research simply became enfolded in the increasingly powerful language of statistics.40 In the case of Stella Chess’s parallel longitudinal study in New York, for example, qualitative observations were made to fit quantitative methods. Upon receiving a case report in which it was said a child had “‘cried because he was hungry,’” researchers would process such remarks only “with regard to the fact of crying.”41 Adapting subjective remarks about children’s behavior to fit binary categories—in this case, whether the child cried or not—was one of the ways in which the boundaries between previously held discrete types of research dissolved.
Psychoanalysts’ views about disease and development fitted more comfortably with longitudinal modes of research. Writing in 1952, one of the country’s best-known psychoanalysts, Karl Menninger, declared that mental diseases “always have a developmental history.” Indeed neo-Freudians like Kris, as well as psychotherapists more generally, found much that they could appreciate in the longitudinal method’s historicization of individual lives. When Kris wrote in the pages of the Psychoanalytic Study of the Child that longitudinal “studies would supply data to some extent similar to those to which we are used [to] in psycho analysis,” this is what he meant.42 Longitudinal research could therefore be made to align with some existing psychoanalytic techniques, as well as be made to answer a number of the other administrative, epistemological, and experimental ideals that those interested in the study of children aspired to meet in these years; it is to these latter considerations that we shall now turn.
Explanations for why the longitudinal method suited the needs of the fledgling discipline of child psychiatry frequently rested on shifting modes and valuations of scientific practice. Scholars outside the mental sciences shared such opinions. Nathan Goldfarb, a professor of business and marketing at Hofstra University, praised the method because “in terms of statistical reliability, the longitudinal design is preferable to the use of independent samples.” Similarly, researchers in other countries celebrated longitudinal research not just for what such programs could reveal, but for how they might reveal it. A British child psychiatrist writing shortly after World War II was drawn to the longitudinal study’s potential to “produc[e] a more objective attitude in the study of behaviour, particularly in the earliest years of life.” These were all qualities coveted by the postwar human sciences, and their presence in such research programs allowed practitioners to parry charges about the field’s supposed lack of rigor. Child psychiatrists took up longitudinal studies then as much for what they could reveal about children, as for the garb of scientific credibility that they could drape around their field.43
The first thing that emerges from reading the work of those affiliated with the Yale study is the way they thought of themselves as pioneers. Some adopted this term explicitly, excitedly discussing their involvement in a “pioneer study,” and later referring to the research program as a “giant step” for psychoanalysis.44 The Yale study, which was part of the first wave of sustained research within the American mental sciences, did in fact mark a new era in psychoanalytic, indeed psychiatric, science. Kris’s study began at a moment when, as one contemporary report noted, “psychiatry [did] not have a strong research tradition oriented to systematic, empirical investigation of important problems,” though the creation of the National Institute of Mental Health in 1949, and formation of the Journal of Psychiatric Research some years later, gradually adjusted those conditions.45 To encourage the introduction of research within the mental sciences, many professional bodies held conferences and roundtables throughout the 1940s at which research principles were debated. The historian Harry Marks has demonstrated that these trends were broadly evident in American medicine at this time, and has explained how medical experimentation became synonymous with certain scientific ideals, including the randomization of variables, detachment between subject and investigator, and separation of research and therapy.46
While these ideals found their way into the discipline of psychiatry too, some scholars, like Kris, opposed such notions, arguing that they were inappropriate for the mental sciences. He told his colleagues that they should design studies that would allow for “constant interrelation between theory and clinical observation.” The Yale Longitudinal Study was in fact designed to make this point. Described by Kris as an instance of “action-research,” or an investigation “in which the observer takes part,” it was imagined to have almost as many therapeutic functions as experimental ones. Such studies were by no means uncommon in academic psychiatry in the 1950s, or indeed the 1960s, though they diminished thereafter. In contrast to many medical studies, these projects refused to arrange staff and subjects so that they remained distant from one another, and instead emphasized the need for researchers to develop relationships with their subjects. Staff at Yale “learned to investigate as part of the therapeutic procedure,” Kris said, and were “trained to take our own actions into account.”47
During the study’s planning and active phases, these casual interactions between staff and patients were presented as a necessary facet of such investigations. Yet when Senn, the study’s main public chronicler after Kris’s death, described the relationships between researchers and subjects in the late 1960s, he placed the emphasis differently. Although he admitted such close relationships were inescapable (“it could not be avoided”), he nevertheless felt obliged to reassure his colleagues that “their influence was recognized and reckoned with.” This shift in how Kris perceived and Senn presented researcher-subject relationships was not just due to personal differences between the two men. But it did reveal a change in how objectivity was conceived: now it was organized around detachment and the supposed impersonality of methods, rather than involvement and intuition.48
A number of essays in this volume note that those involved in the Yale Child Study were drawn from a variety of disciplinary backgrounds. Their common participation in the study was symptomatic of the period’s general enthusiasm for interdisciplinary and collaborative modes of research, and a swarm of factors, general and specific, helped to create such conditions. It helped that many of those interested in child psychiatry understood the need for their field to remain “plastic and adaptable,” as Karl Bowman, president of the American Psychiatric Association told his colleagues in 1944—“ready to try out new methods that offered some chance of adding to our knowledge.”49 Twenty years later, the analyst Roy Schafer would still be adamant that “the study of certain structural and maturational aspects of the early ego requires assistance from other disciplines and other methods.”50 Such support for interdisciplinarity had long been evident at Yale, and Chapter 3 of this volume elaborates on the issue. A study that involved pediatricians, psychologists, psychoanalysts, social workers, and eventually nursery school teachers therefore had some precedent at Yale.51 As with the Institute of Human Relations, though, the Yale Longitudinal Study was also encouraged along these lines by those bodies that financially supported its work. When the Foundations’ Fund for Research in Psychiatry and the Commonwealth Fund gave almost $90,000 in grants to those working on the study, they singled out such interdisciplinarity as noteworthy.52
Kris was particularly enthusiastic about what he termed “intrascientific communication,” having long called for “an intense study of the child by a team of participant observers of many skills.” His understanding of interdisciplinarity crucially paralleled his more general ideas about how “to integrate observational data into the general flow of psychoanalytic thought.” Kris had long recognized that it was not enough to bring together a diverse group of investigators, and that any such expansion “must necessarily be based on a more systematic foundation.”53 Operationalizing such ideals proved less than straightforward, though. During the study’s first eighteen months, some investigators were left feeling that their work resembled “a number of isolated sub-studies,” and called on Kris and Senn to reorganize the project so they could “work in a united fashion toward a common end.” The doodle that one member of the Yale team drew during this period captures both their understanding of the study’s future focus—that is, the development of subjects’ personalities—as well as his own frustrations with the study’s lack of coordination. The introduction of weekly research meetings, frequent staff consultations in which a “systematic pooling of . . . data” took place, and an “organized method of working together” helped to alleviate these tensions by building interdisciplinarity into the structure of the study.54
Like many in the human sciences in these years, Yale investigators also took an approach to their research that valorized methods and theories over the rote pursuit of empiricism. This was quite different from those earlier programs of research on children, like the Berkeley Growth Study that Nancy Bailey had headed in the late 1920s, where facts were generally preferred to theories.55 “The first goal of the present collaboration should be to work out methods and not to consider getting at any results,” Katherine Wolf wrote in 1951. One virtue of such research, she added, was that “there is no . . . pressure of time [because] the study is one that will run over a long period of time.”56 The opinion revealed a confidence that such programs would not be hurried along by funding concerns: as Rebecca Lemov has argued, “new funding, new prestige, and new research alliances [in the mental sciences had] pushed open a window once closed.”57 Indeed the Yale study would receive generous grants from a number of sources, most notably the Commonwealth Fund and the Foundations’ Fund for Research in Psychiatry.58 Still, Wolf’s comment was mainly epistemological, and foreshadowed Kris’s own view that the study’s chief purpose was to “pose problems and develop hypotheses.”59
Figure 1: Doodle by a member of the investigative team
The comments of Wolf and Kris thus bear some similarity to what Joel Isaac has described as the postwar human sciences’ preoccupation with “epistemic design,” that is, with devising theoretical frameworks and, secondarily, arranging empirical data to fit those schemes.60 Although Kris resisted the view that “psychoanalytic theory represent[ed] a structure of thought superimposed upon ‘the empirical material’ that would have precedence,” he still regarded the study as a means of “organiz[ing] hypotheses according to their actual importance,” and championed longitudinal frameworks because they allowed for the “structural interconnection of hypotheses.”61 To achieve these ends thought was given to the writing, organization, and cataloguing of meeting minutes and other records. At a general meeting in early 1951, staff decided that they should record all future minutes “in a way that Case Discussions can easily be separated from General (Method, techniques, etc.) and [an] outline of categories for indexing.”62 It was one mundane, but nevertheless revealing, way in which information produced by the study was subtly arranged so that different species of data could be easily distinguished.
Those who were asked publicly to comment on the Yale study were similarly drawn to reflect on the way that longitudinal studies served as “a tool for organizing a mass of information which otherwise remains unmanageable.”63 Across the human and mental sciences, longitudinal studies were therefore imagined as one way of achieving this desired coordination of theory and data, permitting some investigators to pose “unanticipated questions,” and others to find “answers . . . quickly in data collected many years before.”64 Lester Sontag, a physician involved in the Fels Study, commented in the late 1950s that “longitudinal research . . . sometimes must begin with broad goals in mind, collecting data which may be analyzed when specific problems are formulated at a later date.”65 For those in the nascent field of child psychiatry, these were valuable considerations, and researchers’ embrace of this view aligned them with the human sciences’ broader proclivity for a heightened self-awareness about the mechanics of knowledge production.
Writing about the projective testing movement, the historian Rebecca Lemov has noticed how researchers in this period possessed a clear “self-consciousness about method,” a stance she found so pronounced “as to be nearly meta-methodological.”66 Staff involved in the Yale Longitudinal Study also revealed such attitudes toward their work. For instance, they took an interest not just in the data they were able to harvest from the study, but also in the opportunity it presented them to experiment with their methods. A psychiatric social worker involved in the early phases of the Yale study said that the “longitudinal study was of considerable value . . . since she gained increasing skill” in interviewing families. At research meetings in the early 1950s, Kris had expressed such views to others working on the project, describing how longitudinal studies allowed investigators “to develop and to improve [their] methods,” and particularly to determine in “the long run . . . which type of observation yields the best material.” Senn similarly called the study a “unique learning experience” for practitioners.67 Evidently, programs of research like that undertaken at Yale encouraged investigators to think about methods in two ways: as a means of considering the effectiveness of certain techniques in tackling certain problems, and as a way of gauging their own skill in using those methods. Such an orientation was a crucial feature of research in the human sciences during these decades.
To locate the origins of child psychiatry in the early years of the postwar period, rather than in any previous era, is not simply about getting one’s dates right. Rather, it is to associate the foundations of the field with a general reorientation in psychiatric concepts and the outlook of mental health professionals. Child psychiatry in the postwar decades was not defined simply by the greater attention practitioners paid to children; it was also characterized by the view that childhood marked a stage in a person’s life with a mental currency of its own. In their determination to uncouple the study of children from the axioms attached to adults, psychiatrists worked hard to fashion a field with distinct intellectual precepts. Accepting that children had the capacities to be individuals—a view often expressed via references to their ability to develop personalities—was a crucial moment, and one that rippled beyond just the psychiatric profession, as the pages of the journal Child Development testify.
Between roughly 1950 and the mid-1970s the method of choice for many child psychiatrists was the longitudinal study, a technique that was by no means exclusive to that particular arena. To some degree, the method’s popularity reflected a research community awash with federal and philanthropic largesse.68 But the longitudinal study also reflected contemporaries’ interest in epidemiological techniques, preventative forms of intervention, and new ideas about disease etiology. Finally, with their statistical models, and mechanisms for tracking temporal change, longitudinal methods squared neatly with psychiatrists’ wish to be taken seriously as scientists of the psyche. As I have tried to emphasize, the longitudinal method became pervasive among those who called themselves child psychiatrists because of the particular priorities of the mental sciences at this moment. The greater reflexiveness that these scholars showed toward their research was, it has been argued, a widespread feature of research in the American human sciences during the postwar decades. Such reflexiveness even extended to the purpose to which some imagined the material they had assembled would be put, with Senn wondering out loud if they might entice “future historians interested in knowing how a few Americans lived through a critical period in their lives.”69 This volume indicates that Senn’s prediction would eventually be proved correct.
NOTES
1. Audrey T. McCollum, “A Clinical Caseworker in Interdisciplinary Research,” Social Work 1 (January 1956): 88–102, quotation at 89. On the organization of the study’s archive, see folder 163: “Methodology, 1958,” box 16, series 1, Yale Child Study Center, School of Medicine, Yale University, Longitudinal Study, MS 1967, Yale Archives and Manuscripts, Sterling Memorial Library, New Haven, Connecticut (hereafter cited as LngYC/YAMS).
2. Joel Isaac, “Tool Shock: Technique and Epistemology in the Postwar Social Sciences,” in The Unsocial Social Science? Economics and Neighboring Disciplines since 1945, eds. Roger Backhouse and Philippe Fontaine (Durham, NC: Duke University Press, 2010), 133–64.
3. See David Hoogland Noon, “The Evolution of Beasts and Babies: Recapitulation, Instinct, and the Early Discourse on Child Development,” Journal of the History of the Behavioral Sciences 41 (Fall 2005): 367–86.
4. Edward Parks, “Preface,” in Leo Kanner, Child Psychiatry (1935; Springfield, IL: Charles C. Thomas, 1937), viii.
5. Clemens E. Benda, Malcolm J. Farrell, and Catherine Chipman, “The Inadequacy of Present-Day Concepts of Mental Deficiency and Mental Illness in Child Psychiatry,” American Journal of Psychiatry 107 (April 1951): 722. See also “001–79x. Senile Psychoses,” in Committee on Statistics of the American Psychiatric Association, Statistical Manual for the Use of Hospitals for Mental Diseases (Utica, NY: Utica State Hospitals Press, 1934), 32.
6. Committee on Child Psychiatry of the Group for the Advancement of Psychiatry, Basic Concepts in Child Psychiatry, report 12 (Topeka, KS: Group for the Advancement of Psychiatry, 1950).
7. On the unduly neglected place of longitudinal research in these years, see Tiago Moreira and Paolo Palladino, “‘Population Laboratories’ or ‘Laboratory Populations’? Making Sense of the Baltimore Longitudinal Study of Aging, 1965–1987,” Studies in History and Philosophy of Biological and Biomedical Sciences 42 (2011): 317–27; Marianne Kris, “The Use of Prediction in a Longitudinal Study,” Psychoanalytic Study of the Child 12 (1957): 175–89, quotation at 176. Psychiatry’s growing interest in the normal is charted in Elizabeth Lunbeck, The Psychiatric Persuasion: Knowledge, Gender and Power (Princeton, NJ: Princeton University Press, 1994); the shift from a universal human subject to “niche” groups is explained in Steven Epstein, Inclusion: the Politics of Difference in Medical Research (Chicago: University of Chicago Press, 2007).
8. On the existence of prior groups of mental health professionals interested in children, see Alice Boardman Smuts, with the assistance of Robert Smuts, R. Malcolm Smuts, Barbara Smuts, and P. Lindsay Chase-Lansdale, Science in the Service of Children, 1893–1935 (New Haven: Yale University Press, 2006).
9. Beatrix Hamburg, “Current Literature on Child Psychiatry,” Psychiatry 16 (November 1953): 404–13, quotation at 404.
10. Ernst Kris, “Review of The Ego and the Mechanisms of Defense, by Anna Freud” (1938), in Lottie Newman, ed., Selected Papers of Ernst Kris (New Haven, CT: Yale University Press, 1975), 343–56, quotation at 351. Useful accounts of Kris’s life and influence can be found in Samuel Ritvo and Lucille B. Ritvo, “Ernst Kris, 1900–1957: Twentieth-Century Uomo Universale,” in Franz Alexander, Samuel Eisenstein, and Martin Grotjahn, eds., Psychoanalytic Pioneers (New Brunswick, NJ: Transaction Publishers, 1966), 484–500; Bernard Fine, “Ernst Kris, Teacher—‘How Much There Is to Learn,’” in Edward Joseph, ed., Beating Fantasies: Regressive Ego Phenomena in Psychoanalysis (New York: International Universities Press, 1965), 3–10.
11. Pearce Bailey, “A Contribution to the Mental Pathology of Races in the United States,” Mental Hygiene 6 (April 1922): 379–80; Adolf Meyer, “Schedule for the Study of Mental Abnormalities in Children” (1895), reprinted in Alfred Lief, ed., The Commonsense Psychiatry of Dr. Adolf Meyer: Fifty-Two Selected Papers, edited with Biographical Narrative (New York: McGraw-Hill, 1948), 71; Gustave Le Bon, The Crowd: A Study of the Popular Mind (1895; New York: Penguin Books, 1981), 35–36.
12. John Whitehorn, “Preface,” in Leo Kanner, Child Psychiatry, 2d ed. (1935; Springfield, IL: Charles C. Thomas, 1948), vii; Kanner, Child Psychiatry, 110.
13. Glen H. Elder, Jr., Children of the Great Depression: Social Change in Life Experience (New York: Westview Press, 1999), 17.
14. On ego psychology in general, see Nathan G. Hale Jr., The Rise and Crisis of Psychoanalysis in the United States: Freud and the Americans, 1917–1985 (New York: Oxford University Press, 1995), 232–37; Edward Gitre, “Importing Freud: First-Wave Psychoanalysis, Interwar Social Sciences, and the Interdisciplinary Foundations of an American Social Theory,” Journal of the History of the Behavioral Sciences 46 (Summer 2010): 239–62; Ernst Kris, “Notes on the Development and on Some Current Problems of Psychoanalytic Child Psychology” (1950) in Newman, Selected Papers of Ernst Kris, 54–79, quotation at 55. An example of Hartmann doing similarly can be found in “Comments on the Psychoanalytic Theory of the Ego,” in Heinz Hartmann, ed., Essays on Ego Psychology: Selected Problems in Psychoanalytic Theory (London: Hogarth Press and the Institute of Psycho-analysis, 1964), 113–41. On the role these other figures played in introducing new “tools” to their respective disciplines, see Willow Roberts Powers, “The Harvard Study of Values: Mirror for Postwar Anthropology,” Journal of the History of the Behavioral Sciences 36 (Winter 2000): 15–29; Rebecca Lemov, “X-Rays of Inner Worlds: The Mid-Twentieth-Century American Projective Test Movement,” Journal of the History of the Behavioral Sciences 47 (Summer 2011): 251–78; and Isaac, “Tool Shock.” See also Ernst Kris, “The Nature of Psychoanalytic Propositions and Their Validation” (1947) in Newman, Selected Papers of Ernst Kris, 3–23, quotation at 22. The J. J. Thomson quotation is from Heinz Hartmann, Ernst Kris, and Rudolph Loewenstein, “The Function of Theory in Psychoanalysis,” in Rudolph Loewenstein, ed., Drives, Affects, Behavior (New York: International Universities Press, 1953), 13–37, quotation at 14.
15. See Alan Stone and Gloria Cochrane, Longitudinal Studies of Child Psychiatry: Abstracts with Index (Cambridge, MA: Commonwealth Fund, 1959). The volume’s foreword was written by Milton J. E. Senn. Other contemporary reviews of longitudinal methods within child studies can be found in Jerome Kagan, “American Longitudinal Research on Psychological Development,” Child Development 35 (March 1964): 1–32. Studies that presented themselves as “longitudinal” in the interwar decades tended to concentrate on children’s physical growth; see William W. Greulich et al., A Handbook of Methods for the Study of Adolescent Children, serial no. 15 (Washington, DC: Society for Research in Child Development, 1938). John Wolfe’s chapter in that collection, entitled “Some Psychological Aspects [of Growth],” made no mention of personality.
16. Kris distinguished between Gesell’s focus on human “maturation” and his own interest in “personality development”; see Kris, “On Psychoanalysis and Education” (1948) in Newman, Selected Papers of Ernst Kris, 36–53, quotation at 51.
17. Lester Sontag, Charles Baker, and Virginia Nelson, Mental Growth and Personality Development: A Longitudinal Study (Washington, DC: Society for Research in Child Development, 1958). In general see Moreira and Palladino, “‘Population Laboratories’ or ‘Laboratory Populations.’”
18. The Framingham Study is described in Thomas Dawber, William B. Kannell, and Lorna P. Lyell, “An Approach to Longitudinal Studies in a Community: The Framingham Study,” Annals of the New York Academy of the Sciences 107 (1963): 540–41. For a general assessment of the emergence of epidemiological methods see Harry Marks, The Progress of Experiment: Science and Therapeutic Reform in the United States, 1900–1990 (Cambridge, Eng.: Cambridge University Press, 1997).
19. The social profiles of the families enrolled in the Yale study are described in Milton J. E. Senn and Claire Hartford, The Firstborn: Experiences of Eight American Families (Cambridge, MA: Harvard University Press, 1968). Stella Chess et al., “Implications of a Longitudinal Study of Child Development for Child Psychiatry, American Journal of Psychiatry 117 (November 1960): 434–41, quotation at 436. See Nikolas Rose, Governing the Soul: The Shaping of the Private Self, 2d ed. (1989; London: Free Association Books, 1999), 186–90.
20. Senn and Hartford, Firstborn, both quotations at 16.
21. Research Meeting: “The Aim of the Study and Its Methods,” April 10, 1951, folder 2, box 23, series 2, LngYC/YAMS.
22. See Elder, Children of the Great Depression, 5; Smuts, Science in the Service of Children, 323. The Fels Study showed an interest in registering subjects prenatally, but it seems that in practice researchers did not end up registering anyone before the age of six. See Sontag, Baker, and Nelson, Mental Growth and Personality Development, compare 13 and 91; Carl N. Degler, In Search of Human Nature: The Decline and Revival of Darwinism in American Social Thought (New York: Oxford University Press, 1991); Richard J. Plunkett and John E. Gordon, Epidemiology and Mental Illness: A Report to the Staff Director, Jack R. Ewalt (New York: Basic Books, 1960), 30–31; and Kanner, Child Psychiatry, 109 (emphasis in original).
23. Collaboration Meeting, November 29, 1952; Memorandum, May 29, 1953, both in folder 9, box 23, series 2, LngYC/YAMS.
24. Anna Freud in “General Research Meeting,” October 13, 1952, folder 4, box 23, series 2, LngYC/YAMS.
25. Deborah Weinstein, “Culture at Work: Family Therapy and the Culture Concept in Post–World War II America,” Journal of the History of the Behavioral Sciences 40 (Winter 2004): 23–46. See especially Heinz Hartmann, Ernst Kris, and Rudolph Loewenstein, “Some Psychoanalytic Comments on ‘Culture and Personality,’” in George Wilbur and Warner Muensterberger, eds., Psychoanalysis and Culture: Essays in Honor of Géza Róheim (New York: International Universities Press, 1951), 3–31.
26. Ernst Kris, “Opening Remarks on Psychoanalytic Child Psychology,” Psychoanalytic Study of the Child 6 (1951): 9–17, quotation at 17. On the formalization of observation in American psychoanalysis see Elizabeth Lunbeck, “Empathy as a Psychoanalytic Mode of Observation: Between Sentiment and Science,” in Lorraine Daston and Elizabeth Lunbeck, eds., Histories of Scientific Observation (Chicago: University of Chicago Press, 2011), 255–75, esp. 266–70.
27. Ernst Kris, “Problems in Clinical Research” (1947) in Selected Papers of Ernst Kris, 24–30, quotation at 30.
28. Kris, “Notes on the Development and on Some Current Problems,” 55.
29. Richard Meili, “A Longitudinal Study of Personality Development,” in Lucie Jessner and Eleanor Pavenstedt, eds., Dynamic Psychopathology in Childhood (New York: Grune & Stratton, 1959), 109.
30. Stella Chess et al., “Implications of a Longitudinal Study of Child Development for Child Psychiatry,” American Journal of Psychiatry 117 (November 1960): 434–41, quotation at 434.
31. Ibid., 434; John S. Strauss and Haroutun M. Babigian, The Origins and Course of Psychopathology: Methods of Longitudinal Research (New York: Plenum Press, 1977), 216.
32. Roy Schafer, “Contributions of Longitudinal Studies to Psychoanalytic Theory,” Journal of American Psychoanalytic Association 13 (1965): 605–18, quotation at 605.
33. Anna Freud in “General Research Meeting,” October 13, 1952, folder 4, box 23, series 2, LngYC/YAMS.
34. Smuts, Science in the Service of Children, 225.
35. See Ann Stewart, “Excessive Crying in Infants—A Family Disease,” in Milton J. E. Senn, Problems of Infancy and Childhood: Transactions of the Sixth Conference, March 17 and 18, 1952, New York (New York: Josiah Macy Foundation, 1953), 138–59. Historian Elizabeth Lunbeck and therapist Bennett Simon provide a sterling example of how such case material might be read as historical records; see Lunbeck and Simon, Family Romance, Family Secrets: Case Notes from an American Psychoanalysis, 1912 (New Haven, CT: Yale University Press, 2003).
36. On the use of these other instruments see Ethelyn Klatskin, “Psychological Examination,” February 25, 1959, volume 1, box 8, series 1, LngYC/YAMS; Lemov, “X-Rays of Inner Worlds,” 276. The pervasiveness of the Rorschach test is discussed in Peter Galison, “Image of Self,” in Lorraine Daston, ed., Things That Talk: Object Lessons from Art and Science (New York: Zone Books, 2004), 257–96.
37. Nathan Goldfarb, An Introduction to Longitudinal Statistical Analysis: The Method of Repeated Observations from a Fixed Sample (Glencoe, IL: Free Press of Glencoe, 1960), 171. Wolf and Mohr were the leading proponents of organizing the study’s data using “quantitative summaries”; see Research Meeting, December 4, 1951, folder 2, box 23, series 2, LngYC/YAMS; Sarah E. Igo, The Averaged American: Surveys, Citizens, and the Making of a Mass Public (Cambridge, MA: Harvard University Press, 2007).
38. A survey of this opposition is found in Paul V. Lemkau, “The Epidemiological Study of Mental Illnesses and Mental Health,” American Journal of Psychiatry 111 (May 1955): 801–809, quotation at 808.
39. Senn and Hartford, The Firstborn, quotations at 16 and 523.
40. On the growing power of statistics in this period see Theodore M. Porter, “Statistics and Statistical Methods,” in The Cambridge History of Science, vol. 7: The Modern Social Sciences, eds., Theodore M. Porter and Dorothy Ross (Cambridge, Eng.: Cambridge University Press, 2003), 238–50.
41. Alexander Thomas et al., “A Longitudinal Study of Primary Reaction Patterns in Children,” Comprehensive Psychiatry 1 (April 1960): 103–12, quotation at 107.
42. Karl Menninger, A Manual for Psychiatric Case Study (New York: Grune & Stratton, 1952), 18; Kris, “Opening Remarks,” 17. See also Albert Solnit, “Introduction,” Journal of Clinical Psychoanalysis 5 (1996): 102–6.
43. Goldfarb, Introduction to Longitudinal Statistical Analysis, 124; Emanuel Miller, “Child Psychiatry,” British Medical Bulletin 6, nos. 1–2 (1949): 49.
44. McCollum, “A Clinical Caseworker in Interdisciplinary Research,” 90; Ritvo and Ritvo, “Ernst Kris, 1900–1957,” 492.
45. Group for the Advancement of Psychiatry, Some Observations on Controls in Psychiatric Research (New York: Group for the Advancement of Psychiatry Publications Office, May 1959), 538. In general see Bert E. Boothe, Anne H. Rosenfeld, and Edward L. Walker, eds., Toward a Science of Psychiatry: Impact of the Research Development Program of the National Institute of Mental Health (Monterey, CA: Cole Publishing, 1974).
46. Marks, Progress of Experiment.
47. Ernst Kris, “Problems in Clinical Research” (1947) in Selected Papers of Ernst Kris, 24–30, quotation at 24; Kris, “Opening Remarks,” 11–12. For further evidence of how these questions were posed within the study, see “Research Meeting,” March 11, 1952, folder 2, box 23, series 2, LngYC/YAMS.
48. Senn and Hartford, The Firstborn, quotations at 25 and 28. By way of introduction to this subject in the sciences more generally, see Lorraine Daston and Peter Galison, Objectivity (New York: Zone Books, 2007).
49. Karl Bowman, “The Psychiatrist Looks at the Child Psychiatrist,” American Journal of Psychiatry 101 (July 1944): 23–9, quotation at 28.
50. Roy Schafer, “Contributions of Longitudinal Studies to Psychoanalytic Theory,” Journal of the American Psychoanalytic Association 13 (1965): 605–18, quotation at 605.
51. See also Jill Morawski, “Organizing Knowledge and Behavior at Yale’s Institute of Human Relations,” Isis 77 (1986): 219–42. On the importance of interdisciplinarity in America’s postwar human sciences, see Joel Isaac, Working Knowledge: Making the Human Sciences from Parsons to Kuhn (Cambridge, MA: Harvard University Press, 2012).
52. Report of Grant Activities for the Year Ending June 30, 1957 (New Haven, CT: Foundations’ Fund, 1957), copy in folder 4, box 8, series 2, Foundations’ Fund for Research in Psychiatry, Yale record group 37, record unit 484, Archives and Manuscripts, Yale University, New Haven, Connecticut (hereafter FFRP/YAMS). On the role that funding bodies played in encouraging patrons’ interests in interdisciplinarity see Jamie Cohen-Cole, “Instituting the Science of Mind: Intellectual Economies and Disciplinary Exchange at Harvard’s Center for Cognitive Studies,” British Journal of the History of Science 40 (December 2007): 567–97, esp. 573–76.
53. Kris, “Opening Remarks,” 11; Kris, “The Nature of Psychoanalytic Propositions,” 21; Kris, “Notes on the Development and on Some Current Problems,” 72; and Kris, “On Psychoanalysis and Education,” 50.
54. All quotations are from Memorandum to Milton Senn [Copy to Ernst Kris], December 19, 1952, folder 9, box 23, series 2, LngYC/YAMS; McCollum, “A Clinical Caseworker in Interdisciplinary Research,” 91. On the operational difficulties of interdisciplinary research see Cohen-Cole, “Instituting the Science of Mind.”
55. See Elisabeth Lomax, “The Laura Spelman Rockefeller Memorial: Some of Its Contributions to Early Research in Child Development,” Journal of the History of the Behavioral Sciences 13 (July 1977): 283–93, quotation at 287.
56. “Aim of the Study and Its Methods,” April 10, 1951 (emphasis in original). On Wolf’s career and general contribution, see “Katherine M. Wolf: In Memoriam,” Child Study 35 (Winter 1957–58).
57. Rebecca Lemov, “‘Hypothetical Machines’: The Science Fiction Dreams of Cold War Social Science,” Isis 101 (2010): 401–11, quotation at 411.
58. The Commonwealth Fund was the Yale study’s original source of funding. Later the study received a three-year grant of almost $65,000 from the Foundations’ Fund for Research in Psychiatry. This was greater per annum funding than the foundation typically awarded, and rare in that it related to child psychiatry. See Foundations’ Fund for Research in Psychiatry to Katherine M. Wolf, February 12, 1957, folder FFRP, box 39, series IV, LngYC/YAMS. The funding profile of the fund is described in Foundations’ Fund for Research in Psychiatry: Tenth Annual Report for the Years 1953–1963 (New Haven, CT: Foundations’ Fund for Research in Psychiatry, 1963), folder 10, box 9, series 2, FFRP/YAMS.
59. Ernst Kris and Sally Provence, “The Study of Variations of Early Parental Attitudes—A Preliminary Report,” Psychoanalytic Study of the Child 8 (1953): 20–47, quotation at 28.
60. Joel Isaac, “Epistemic Design: Theory and Data in Harvard’s Department of Social Relations,” in Mark Solovey and Hamilton Cravens, eds., Cold War Social Science: Knowledge Production, Liberal Democracy, and Human Nature (New York: Palgrave-Macmillan, 2012), 79–95, esp. 80–81.
61. Kris, “Problems in Clinical Research,” 26.
62. “Instruction for Taking Meeting Notes,” n.d. (ca. January 1951), folder 1, box 23, series 2, LngYC/YAMS. Further examples of this concern for the organization of information can be found in Sally Provenance to all Secretaries, n.d. (ca. July 1957); and “Research Meeting,” December 4, 1951, both in folder 23, box 23, series 1, LngYC/YAMS.
63. David French quoted in McCollum, “Clinical Caseworker in Interdisciplinary Research,” 102.
64. Strauss and Babigian, Origins and Course of Psychopathology, 209–10.
65. Sontag, Baker, and Nelson, Mental Growth and Personality Development, quotation at 13.
66. Lemov, “Hypothetical Machines,” 408.
67. McCollum, “A Clinical Caseworker in Interdisciplinary Research,” 101; “Research Meeting,” February 19, 1952, folder 2, box 23, series 2, LngYC/YAMS (emphasis in original); Memorandum from Milton Senn to “Staff of Longitudinal Research Study,” n.d. (ca. May 1957), folder 23, box 23, series 2, LngYC/YAMS.
68. The role of funding bodies in encouraging longitudinal research is found in Jerome Frank, “Postwar Psychiatry: Personal Observations,” in Roy Menninger and John Nemiah, eds., American Psychiatry after World War II, 1944–1994 (Washington, DC: American Psychiatric Press, 2000), 197.
69. Senn and Hartford, The Firstborn, 525.