The Stirling County Study: a case study of interdisciplinarity and its effects on the history of psychiatric epidemiology

Abstract

Epidemiology of mental disorders emerged in the post-1945 era at the intersections of different areas of knowledge. Given its ambitions, the Stirling County Study provides an instructive case study. It is also a good example of how the epidemiology applied methodological skills from social sciences. This paper aims, first, to reconstruct one of the first episodes in the development of psychiatric epidemiology. Its second purpose is to provide a detailed description of interdisciplinarity at work, and to examine its effects. After explaining some of the major features of the Stirling County Study, I emphasize the links between some of the first results, particularly regarding young people as a population at risk, and the job market after the Great Depression.

Introduction

The Stirling County Study has been a very ambitious psychiatric epidemiology research project, situated in Nova Scotia, Canada. It was initiated at the end of the 1940s and involved the collaboration of more than 100 researchers from a wide variety of academic disciplines. There are therefore many ways to investigate its history, and no single one is sufficient; thus, the historical problems and questions with the study merit enumeration. One perspective I usually favour when presenting the Stirling County Study in lectures is to examine its place in the history of sociological studies in the Chicago tradition, particularly the ecological paradigm of urban studies. The historian Matthew Smith explores the resources offered by this approach in his contribution in this issue (Smith, 2023a), so I will leave my discussion of this perspective to the end of this article. Other possible historical analyses concern, for example, the contributions of the study’s first leader, Alexander H Leighton, to the field of American anthropology in the 1930s; his indebtedness to the Swiss-American psychiatrist Adolf Meyer, whose student he was; ¹ and how his studies in biology may have impacted his conception of the ecology of the communities he studied. From another perspective, his collaboration with the Bureau of Indian Affairs – he worked on Native American reservations as a young anthropologist – established his enthusiasm in applied anthropology (Guerrier, 2007). Other lesser-known facets are no less interesting: for instance, Leighton’s relationship to photography. How common is the knowledge that he counted among his collaborators and students one of the first professors of visual anthropology, John Collier Jr, ² and that he himself was an author of documentary films?

From a methodological point of view, I first read the sources, including the scientific literature published by the actors, the archives in Nova Scotia, and also secondary literature. I also conducted a number of interviews, including one with Jane Murphy, one of the study’s leaders. As this corpus is too vast for a single article, I have decided to separate several topics for analysis in separate articles and in a forthcoming book. Thus, I have chosen to restrict myself to an in-depth account of a single dimension of the Stirling County Study: the role of interdisciplinarity at the beginning of the study, from the point of view of the history of science. Indeed, the disciplines involved determined the framework of analysis. The dominance of certain disciplines over others shaped the specific design of the research, the choice of variables, and the data collected. To put these into context, I will first explain some of the characteristics of the original team that conducted the survey. Then, to interpret the results, I will use Canadian social history. Finally, I will mention other interdisciplinary configurations at work on research teams during the same period.

A rapidly evolving state of the literature

Psychiatric epidemiology is a little-known field that has rarely been investigated by historians of health, as historical accounts of psychiatry usually focus on individuals, whether it is the careers of physicians or the voices of patients. Social groups, populations and communities – not to mention social classes – are the poor relatives of this literature. However, its historiography has been marked recently by two publications initiated by the anthropologist Anne Lovell (Lovell and Oppenheimer, 2022; Lovell and Susser, 2014). I will not summarize them here, but build on the material relevant to my purpose, noting that these analyses not only focused on North America, but also positioned the history of psychiatric epidemiology within colonial history and post-colonial issues. The Canadian context is certainly specific, but it is wise to remember that the leaders of Quebec’s sovereignty movements often raised the banner of decolonization to claim their fight for independence. ³ To avoid confusion, it is also important to point out that the Acadian country where the Stirling County Study ⁴ took place is a bilingual French and English region that is not part of Quebec. This ancient settlement area had a different history after its integration into the English-speaking province of Nova Scotia, where Acadians represent only a small demographic minority, far removed from other francophone communities.

A paradox of the collected publications edited by Lovell and her colleagues is the omnipresence of the Stirling County Study as a historical milestone, yet there is no analysis of this study. Geographical remoteness is one factor, as Nova Scotia is far from the major North American urban centres: travelling there to conduct research is a long journey and expensive for non-local historians.

Lovell and Oppenheimer (2022) note that, in the twentieth century, there was not one unified psychiatric epidemiology, but several different ‘regional formations’ or ‘regional historical formations’ (pp. 6, 47). The latter resulted from the heterogeneity in preliminary questions, tools and types of expertise, but also economic constraints, political objectives of control, reformist ideals, networks, etc., or what Gaston Bachelard thematized in philosophy and the history of science in terms of ‘regional epistemology’ (see Braunstein, 2013; Lovell and Oppenheimer, 2022: 1–52). Not all epidemiologists considered pioneers used the same techniques in mental health or the same frame of reference.

Moreover, epidemiology is a science whose quest for objectivity is reflected in statistical measurements and parameters. These techniques were used for the census of the population and the administrative control of minorities, foreigners, the colonized, the poor and the marginalized. The emergence of a new discipline requires creativity and innovation, in addition to employing useful practices from established disciplines. For example, one of the steps in constituting a body of knowledge in psychiatric epidemiology consisted of abandoning asylum statistics (e.g. prevalence of a disorder among those treated) in favour of measures of ‘true prevalence’ (Lovell and Oppenheimer, 2022: 32) in the general population. Another distinct step was the ‘replicability of measures’ (p. 15), which was a relatively new consideration in psychiatry generally.

These two dimensions are at the heart of the Stirling County Study, the aim of which was to achieve a milestone in the stability of the techniques used across studies. This study was clearly positioned to be the first to use ‘replicable’ methods, i.e. measurements that could be ‘repeated’ or ‘reiterated’, ⁵ and measurement tools that could be used in other studies, as well. Here, the experimental science model was certainly a source of inspiration for epidemiological methods, even if the ‘laboratory style of reasoning’ ⁶ is, of course, far from the study of mental health in communities.

Additional characteristics were highlighted by Lovell and Oppenheimer (2022), such as the coupling of quantitative statistical data and qualitative ethnographic observations, a genre developed primarily in ‘exotic’ colonial societies. Physicians from metropolitan areas relied on racial anthropology and the methods of the humanities and social sciences to produce knowledge about subjugated peoples, imposing their own cultural and medical categories on social groups dominated by force. Other tools, such as the ‘key informant’ or ‘informant survey’ technique (p. 28), and those from psychometrics, developed by psychologists (e.g. screening questionnaires), played a significant role, as seen in the Stirling County Study and other contemporaneous mental health epidemiological studies.

Another less specialized literature is also important. There were already accounts about the history of psychiatric epidemiology in Canada from the point of view of the epidemiologists themselves. ⁷ The history of mental health in Nova Scotia also includes the Stirling County Study, although most often not from the point of view of the history of science, but rather as an example of local challenge to what was ultimately viewed as a foreign model, i.e. American (Fingard and Rutherford, 2005, 2008). Finally, as the leaders of the Stirling County Study invested much of their research activity replicating their analytical framework and measures in other sociocultural contexts, it is worth noting the ongoing research of historian Matthew Heaton, contemporaneous to mine. Heaton’s research includes the Aro-Cornell Study in Nigeria, also led by Leighton in collaboration with the Nigerian psychiatrist Thomas Adeoye Lambo (Heaton, 2022). Last but not least, Matthew Smith has just published a history of American social psychiatry (Smith, 2023b), which includes a chapter on the Stirling County Study in relation to another study conducted in the USA, the Manhattan Midtown Study. As the present article was written before the publication of the book, I invite readers to refer to it directly.

Interdisciplinarity in the history of psychiatric epidemiology

At a time when an epidemic such as Covid-19 is making news, it is important to place epidemiology within a historical perspective that includes non-virus-related issues. The Stirling County Study was launched 75 years ago (1948) in Nova Scotia by the psychiatrist and anthropologist Alexander H Leighton (1908–2007), with the goal of assessing the ‘effects of sociocultural environment on the prevalence of psychiatric disorders’ (Leighton, 1959a: 9–10). It was contemporaneous with other pioneering epidemiological studies on chronic diseases, such as the Framingham Heart Study, which was also started in 1948 (Giroux, 2013; Oppenheimer, 2010). Both studies are still ongoing. But the history of the Framingham study is well known, while the history of the Stirling County Study has never been described. However, the scientific landscape has changed: psychiatric epidemiology was a new field in the post-war period, and not an international discipline as it is today.

In what way was this study foundational? It was one of the first community-oriented research projects in mental health, as opposed to statistical studies based on psychiatric hospitals. Before Leighton’s project, several community studies were already known in the scientific literature, and he commented on their results: the Wellesley study (Massachusetts) done by Erich Lindemann, Ernest Gruenberg’s investigation with regard to the aged in Syracuse (New York), etc. However, the Stirling County Study had a specific research design. Led by anthropologists with the collaboration of more than 100 researchers from the 1950s on, it is an interesting example of interdisciplinarity in mental health research. Leighton wrote that among his aims were: ‘an epidemiological survey based on psychiatric method, and the carrying out of anthropological and sociological studies’ (Leighton, 1959a: 7), and ‘a blending of medicine, anthropology, sociology, psychology, and statistics’ (Leighton DC et al., 1963: 24).

This unusual juxtaposition deserves clarification. In a classic text in the history of science, Wolf Lepenies explained that attention to the interdisciplinary scientific environment is needed to understand the characteristics of a preliminary phase of a new field of knowledge. These nuances will be lost if it is regarded as a discipline: ‘If a historical study of sciences wants to reconstruct necessary developmental stages that are now outdated, it cannot very well limit itself to one discipline or merely put together several disciplinary reconstructions in a merely additive manner’ (Lepenies, 1978: 56–7). This reflection makes it possible to examine the interrelationships between the sciences and their effects. The specific organization of the collaborations implemented by Leighton was first inspired by the dynamic psychiatry of Adolf Meyer and the social anthropology of Bronislaw Malinowski, which certainly impacted the design of his survey.

After studying biology at Princeton University in New Jersey and at Cambridge in England, Leighton completed his medical training with an MD (1936) at Johns Hopkins University in Maryland, where Meyer promulgated a model in which the patient was thought to have a dynamic relationship with their environment. To explore this dynamic more thoroughly, Meyer encouraged young psychiatrists to take an interest in the human and social sciences. This led Leighton to become an anthropologist and, as was traditional, he conducted his first field studies on Native American reservations. He worked among the Navajo (Diné) and among the Inuit of Alaska. He then assumed responsibilities as a scientist during the war and was one of the main social science researchers involved in running a Japanese Relocation Center in Arizona, based on the administrative model of John Collier at the Bureau of Indian Affairs – a model which raises many questions today (Guerrier, 2007). Leighton was also a member of the team that went to Hiroshima in 1945 to study the consequences of the atomic bomb. This series of responsibilities explains his rapid ascent in academia. He had already published several books about these studies when he was appointed Professor of Anthropology in 1946, and later, Professor of Social Psychiatry, both at Cornell University. It was here that he launched the Stirling County Study in the late 1940s.

The Stirling County fieldwork was conducted with an applied anthropology methodology inspired by the functionalism of Malinowski, according to which every cultural element fulfils a function in society (Malinowski, 1944). But the design of the study also incorporated the approach of the Chicago School of Sociology, which had already inspired studies on schizophrenia and the organization of social groups in urban areas, especially the study by Robert EL Faris and H Warren Dunham (1939). ⁸ Based on my reading of Leighton’s publications, I propose that he was progressively more ‘in dialogue with’ Hollingshead and Redlich’s (1958) study in New Haven (Connecticut) than influenced by Faris and Dunham’s study, given the greater similarities between the Stirling County Study and the former (Hughes, Tremblay, Rapoport and Leighton, 1960: 398).

Interdisciplinarity at work: 113 main contributors, 18 socio-professional categories

Leighton deemed that functionalism could also take into account the dynamic relationship between mental disorders and the sociocultural environment. The basic hypothesis of the Stirling County Study was therefore that ‘noxious aspects of sociocultural disintegration’ (Leighton DC et al., 1963: 388) increased the risk of psychiatric disorder in communities, especially the most reactive anxio-depressive disorders (not focusing on schizophrenia). The dependent variable of the Stirling County Study was formulated in terms of ‘community integration-disintegration’. Notably, Leighton referred to French sociologist Émile Durkheim’s concept of ‘anomie’, translated by American sociologist Leo Srole as normlessness and as a ‘general sense of the futility of life’ (Leighton DC et al., 1963: 384; Srole, 1956). Finally, the Stirling County Study independent variables reflected those of social epidemiology, namely: (1) sex, (2) age, (3) retirement status of household head, (4) occupational disadvantage, (5) marital status, (6) education, (7) number of moves after age 20, (8) number of moves before age 20, (9) Acadian-English typology, ⁹ and (10) private religious participation. The epidemiological perspective was defined as ‘finding out how psychiatric disorders are distributed in the selected sociocultural categories’ (Leighton, 1959a: 304).

Besides verifying the basic hypothesis, the goal of the study was to assess whether these independent variables were relevant statistical indicators. Nova Scotia county selected by Leighton for the study still remained very rural after the war, with approximately 20,000 inhabitants clustered in small towns and hamlets. Leighton’s choice was based on the county’s relative socio-economic homogeneity, with fishing, woodworking and farming still producing a significant portion of revenue alongside service and supply businesses and tourism. But two different cultural groups coexisted in the county: the Anglophone and Protestant majority, and a Francophone and Catholic minority, the Acadians. Put simply, Leighton took on the challenge of selecting a diverse region in terms of cultural values, which provided the opportunity to establish interesting cross-cultural comparisons. These were not based on a simplistic dichotomy. In the six towns chosen, both among the Anglophones and the Francophones, certain communities were prosperous and others were impoverished and falling apart.

During the 1950s, the main researchers in this ‘interdisciplinary study’ (Hughes et al., 1960: 9) numbered 113, which Leighton divided into 18 socio-professional categories: administration (11), anthropology (18), child psychology (1), economic history (1), economics (2), internal medicine (1), interviewing (21), human relations (1), mathematics (1), medicine (1), nursing (1), photography (1), psychiatry (12), psychology (10), social psychology (6), social work (7), sociology (15), and statistics (3). Note that anthropologists and sociologists formed the core of the team. If economists, psychologists, the photographer-anthropologist and the historian are added together, they total 54 researchers in the social and behavioural sciences. Physicians comprised only one-eighth of the team.

In this regard, the study is often presented as Leighton’s, as he was both an anthropologist and a physician, and because he bridged the two major groups among the contributors, i.e. social sciences scientists and physicians. Even if the available literature confirms that he initiated it and conceived of the design, he was not the only one responsible. The adventure began with his first wife, Dorothea Leighton (1908–89), herself a physician and anthropologist and the first author of the last of the three books that presented the survey to the public. ¹⁰ Then, Alexander Leighton handed responsibility over to his second wife and long-time collaborator, Jane Murphy (1929–2021), after they moved to Harvard. She directed the survey until the 2000s. Therefore, it would be more accurate to say that the study was led by two women and one man.

Of course, to mobilize such a large-scale team of 113 researchers, significant means were required, which were allocated by US foundations (Rockefeller, Ford, Milbank, etc.) and the provincial authorities. ¹¹ After the Halifax Explosion caused by a warship filled with explosives during the Great War (1917), resulting in the destruction and the reconstruction of a large part of the city and its area, Nova Scotia had already become a testing ground for large philanthropic foundations. The Stirling County Study was therefore part of a series of surveys of recipients. As historian Leslie Baker (2014) has demonstrated, experts from New England played a strong ideological role in promoting public health policies through surveillance of the population, including psychology and mental health.

Which interdisciplinarity? The priority given to the study of Acadian folklore, traditional values and the social order

To delve deeper into the question of interdisciplinarity, we must specify what the design of the epidemiological research borrowed from each discipline. Here are two examples. Psychologists were asked to construct questionnaires, develop psychometric measures ¹² of mental health, and verify the validity and robustness of these statistical tools. This involved a process of transforming raw, discursive qualitative material into quantitative data, in order to assess ‘the probability that an individual is now suffering or has in the past suffered from some form of psychiatric disorder’ (Leighton DC et al., 1963: 242). The resulting questionnaire was the HOS (Health Opinion Survey), a psychological screening test constructed by psychologist Allister M Macmillan for the study. Field research ‘key informants’ investigations were carried out in 1948–9 (with students) and 1950–1 (with members of the team), but the ‘general survey’ was conducted in the summer of 1952 (Hughes et al., 1960: 64).

Simultaneously, the role of the anthropologists was to document the cultural values of the Anglophone and Acadian communities, prioritizing folklore and the social order, overlooking the fact that social order was being challenged in the 1950s – an emphasis that was criticized by Canadian historian Ian McKay ¹³ in his major work about Nova Scotia (McKay, 1994). Sociologists and economists played a complementary role, less focused on traditions and more on the transformation of social organization in the post-war period. The knowledge that these social science researchers co-produced made it possible to create the independent variables. Without these data, it would be impossible to interpret the results of the questionnaires, which reported the socio-economic status of the communities.

One of the first results of the study can be explained by connecting it to interdisciplinarity. The hypothesis that the ‘risk of psychiatric disorder increases with occupational disadvantage’ (Leighton DC et al., 1963: 294) was shown to be supported by the data. However, the team was surprised by this finding and deemed it necessary to review the calculations, as the study indicated a high rate of mental disorders: 18% (clear psychiatric cases) of individuals who responded to the investigation presented an episode of mental disorder in the course of their lives.

To interpret this unexpectedly high result, he concluded that the study measured not only the prevalence of mental disorders isolated from other social phenomena, but also the global effects of accelerated social change in post-war society. It is ‘the extent and rate of acceleration in change of the sociocultural systems that has a differential effect’ (p. 366). However, it should be noted that the critical literature within the field of epidemiology questioned the high rates of mental disorders obtained in community studies, during Leighton’s lifetime and decades later (Link and Dohrenwend, 1980). According to Lovell and Oppenheimer (1922), these high rates were in part an artefact of this research framework of ‘true prevalence’ in small communities: ‘Such measures were later critiqued not only for their non specificity but also for the overly high prevalence rates they tended to produce, although the magnitude was replicated in study after study’ (p. 33).

It is interesting that the study yielded unexpected results beyond the social gradient finding, which Leighton and his team did not predict. Specifically, the most important variables were: gender, because women were more at risk, except in traditional Acadian communities, ¹⁴ which suggested that there was a cultural difference; and age linked to economic migration:

Only in the case of number of moves before age 20 do we find a relationship significant at the .05 but not at the .01 level. We are impressed by the fact that the relationship is maintained in five out of our six age-sex groups, and that it is in the direction that would almost certainly be predicted by sociological and psychological theory. (Leighton DC et al., 1963: 487)

Overall, it is important to take into account the rural exodus of young people, urbanization, industrialization, and the acculturation of Acadians forced to seek work in Anglophone urban areas where the factories and the service economy were found. All in all, the interdisciplinary research design, which focused on sociocultural data and the integrative function of traditional values, made it possible to show the inseparability of assessed mental disorders from societal changes. In this way, with the aim of laying the foundation for psychiatric epidemiology, Leighton was part of a well-known evolution in the history of epidemiology, which saw a global change of perspective from infectious diseases to chronic diseases, also referred to as ‘pathologies of progress’ (Rosenberg, 1998); these were characterized by high-risk lifestyles, such as smoking, alcohol consumption, a high-fat diet and stress caused by acculturation and a more urban lifestyle. This has been a well-known narrative since Jean-Jacques Rousseau: bad civilization. For the history of psychiatric epidemiology, it is worth recalling that migration and the rural versus urban dichotomy had already been studied by physicians (Ødegaard, 1932) who pioneered the epidemiology of mental disorders before World War II.

Finally, the Stirling County Study is an interesting case study for the history of health because it demonstrates that the concept of risk is not unbiased in ‘the moral economy of science’ (Daston, 1995). It is connected to our lifestyles and the value judgements that we make of them. As a matter of fact, in the design of the study, Leighton hypothesized that the independent variable ‘private religious participation’ would be a good indicator of the noxious aspects of sociocultural disintegration. His idea was that the decline in traditional religious practices in small communities reflected sociocultural disintegration, which would increase the risk of psychiatric disorder in communities. But later he admitted he was wrong: that variable was not significant while the ‘number of moves before age 20’ was surprisingly indicative (Leighton DC et al., 1963: 483–9) of vulnerability to mental disorders.

This unexpected finding has to do with the specific design of the study: the interdisciplinary team was led by social anthropologists. Sociocultural traditions, distinctions between French-Catholic and Anglo-Protestant cultures, beliefs and folklore were valued more than the dynamics of labour and social change at work in the Maritimes (Atlantic provinces of Canada) after 1945, even if individual mobility was included in the design of the study: ‘Migration has been a feature of Depressed Area life for many years’ (Hughes et al., 1960: 256). This social change was far from the Acadian folklore. But the result was coherent with the frame of reference and the search for indices of environmental factors: ‘In order to preserve clarity as much as possible, they will be called indices of social disintegration when referring to communities, and ‘stress factors’ when referring to individuals’ (Leighton, 1959a: 330).

Contextualizing the Canadian youth and young workers from a social history perspective

The global results have led to so many publications that it would be pointless to review them all here. The study measures have been replicated approximately every 15 years, and its design has been used in surveys in other countries. As it is not possible to cover all these results, the rest of my analysis will focus on the finding that the ‘number of moves before age 20’ was an indicator of mental disorders; this was later omitted in the survey re-administration. To put this result in context, the role of the historian is to rely on Canadian social history, especially regarding the job market of the youth and the Great Depression (1929–39) in the Maritimes. Historian Cynthia Comacchio ¹⁵ wrote in The Dominion of Youth: Adolescence and the Making of Modern Canada, 1920 to 1950:

The medical profession, despite its dominant role in the period’s campaigns to regulate child health and education, and despite its growing interest in adolescence, was decidedly vague about the relationships between health, class, age, and the nature of work. ‘The damage done to the child from excessive work may be difficult to detect until he has been engaged at it for a period of years.’ Consistent with the large ‘expert’ discourse on adolescence, medical concerns that were explicitly tied to the life stage of these young workers referred to emotional or psychological more than to physical problems. Thus doctors worried about the ‘striking phenomenon’ of youth’s ‘rebellion against authority,’ a phenomenon otherwise regarded as ‘normal’ among adolescents. For young workers, some contended, ‘monotony in this period of life will kill the best instincts of the future citizen, and very often leads to Bolshevistic tendencies, or sometimes to the vicious characters so often encountered in city life. (Comacchio, 2008: 135)

The normative comments about the ‘best instincts of the future citizen’, ‘Bolshevistic tendencies,’ and ‘vicious characters’ of young workers transplanted in cities are amusing today. But Comacchio’s analysis of the economic consequences of modern adolescence in the interwar period painted a rather dark picture. The data regarding this era are numerous, especially since historians have been interested in the consequences of the economic depression in North America, in all its sociocultural facets in the 1930s. As the affected generations are the core population of the sample chosen by Leighton, it is worth delving more into Comacchio’s line of reasoning.

First, social historians recalled that ideologues of the Great War years stressed the need for a certain type of labourer to ensure productivity in Canada and for careful training from early childhood. The public discussion was not only about job-related skills ‘but just as much in terms of character formation, self management, and citizenship’ (Comacchio, 2008: 129). A significant proportion of young Canadians between the ages of 13 and 19 left school for work. The adolescents stayed semi-dependent on their parents, in a state of transition. Concerns about the health, welfare and morality of the youth were related to the worry that they could leave school too early to fill better-paying but temporary jobs. Even if the late-nineteenth-century laws proscribed child labour and established mandatory schooling, work, and not high school, was the main activity of older adolescents in the 1920s and 1930s. While agriculture remained the chief employer of youth, the manufacturing sector ranked second and was a particularly important source of employment for those over 15.

But, secondly, Comacchio pointed out the expansion of amusements and services introducing new jobs and leisure activities:

In 1925, boys as young as eleven were found working until midnight and later in bowling alleys, or as ‘runners’ for local cinemas, delivering films to the railway station after the evening shows. The relative affluence of those who kept their jobs during the Great Depression allowed for growth in such middle-class and largely male leisure activities as bowling and golf, which also led to employment opportunities for youth, many of whom had acquired the ‘family breadwinner’ title by default. (p. 132)

Concerns about juvenile delinquency arose from the moral consequences of youth work and the subsequent susceptibility to diseases, accidents and other moral hazards. Comacchio quoted the surveys about youth labour in the 1930s, done by Leonard Marsh at the McGill School of Social Research; he noticed the impermanence of the jobs dominated by those under 20. The service sector was transitional for young people. In the end, farming remained the main employer of youth. But another important fact is that much of youth labour was unpaid work for the family, while older siblings received wages (Comacchio, 2008: 139), probably causing many tensions, familial schisms, and economic migration.

Lack of training was taken into consideration in the Depression-fighting measures (Dominion-Provincial Youth Training Program) of the Mackenzie King liberal government in Ottawa. But by the end of the 1920s, the official age of completing school (16 in most Canadian provinces) marked the obligatory entry into the adult world of work. Thus, many young workers moved out of unpaid and familial farm-work because it was a transitory feature of the youth labour market. Here we see that Leighton’s variable on the number of moves before age 20 was already part of the interwar period discourses: social and moral concerns about youth became constant after the Great War and the Great Depression.

Integration, acculturation, culture and subculture of the youth

The interpretation of the Stirling County Study results therefore takes into account factors related to age, occupation and gender:

When attention is turned to the association of demographic and environmental factors with psychiatric disorder, a number of features seem clear. Sex, age, and occupational position are all characteristics associated with considerable difference in prevalence. Each of the three has a relationship to prevalence which is independent of the others. Also sex and age are found to have a joint or interactive relationship to prevalence. Education furthermore is a factor associated with degree of psychiatric involvement, but it is probably not independent of occupational position. Migration is important, but only if it occurs before the age of 20. (Leighton DC et al., 1963: 323)

Nevertheless, the indicator of the number of moves before age 20 did not take into account the new teenage culture that asserted itself in the 1950s. Certainly this ‘subculture’ probably supported young people’s integration into modern urban life (Monod, 1968/2006). Taking into account the rapid evolution of lifestyles, there is no reason to consider that their breaks with tradition became an obvious factor in social disintegration. The sociocultural dynamics of the 1950s–60s were far more complex than Leighton could have anticipated in his 1940s frame of reference.

While the design of the survey was conceived in the 1940s, with the idea of ‘two different cultures’ cohabiting in Nova Scotia, the notion of acculturation was already being debated and was controversial in American anthropology. The definition given by Redfield, Linton and Herskovits (1936) was being challenged because it was inadequate for describing the complex process of transition. Moreover, terms relating to cultural contacts, adaptation and migrations remained strongly linked to colonial heritage, as a one-way process (integration) between a dominant and a dominated society. Critics have pointed out that cultural exchanges are much more complex and creative processes that give rise to new and novel cultural forms, with nuances and variations. The anthropological model favoured by Leighton, that of Malinowski, reinforced the normative ¹⁶ framework of this acculturation explanation (Leighton, 1959b). Although Malinowski preferred to speak of ‘culture contact’, his students contributed abundantly to the systematic development of acculturation studies of different traditional societies. This perspective sometimes provoked very negative reactions from other anthropologists, such as Claude Lévi-Strauss (2019).

Historian Ian McKay was another critic. He pointed out that the conception of folklore by Redfield (1941, 1947, 1956) was marked by idealization, and that the application of Chicago urban sociology to small communities (Redfield, 1955) would lead to a combination of romanticism and functionalism (McKay, 1994: 89, 104, 323). The use in Nova Scotia of a definition of folklore similar to Redfield’s obscured the industrial past and the existence of a working class, particularly those – including youth – who worked in the mines that shaped the territory and economy of Nova Scotia. Although there were no mines in the Acadian area, it is not possible to cut the area off from the rest of the province, nor to reduce the economy to traditional woodworking and fishing, when researchers observed that young workers were mobile.

Acculturation-related stress among young Acadians in anglophone, industrialized areas with more urban lifestyles, was among the foci that structured the study and led to the observations of traditional Acadian life. Several of Leighton’s PhD students, such as Marc-Abélard Tremblay (1954) and Norman Chance (1957), did PhDs in anthropology with Leighton in the form of monographs about Acadian society. Families and parishes were their focus (Collier, 1957; Hughes et al., 1960), rather than the sociology and the economic issues of youth, which reflected more dynamic notions of culture change. Ultimately, the alarmist rhetoric about youth health and the unfamiliarity with new and flourishing youth culture ¹⁷ in the 1950s had implications for the design and the results of the Stirling County Study, expressed as unexpected results.

For instance, the notions of disintegration or disorganization came from the Chicago School of sociology. An early emblematic study was The Polish Peasant in Europe and America (Thomas and Znaniecki, 1918–20/1996) in the context of extensive immigration from southern and eastern Europe. Thomas had contact with Meyer, Leighton’s psychiatry professor, who used the term disorganization (Chapoulie, 2018: 64, 93), and was interested in social- and psycho-dynamics. The question of adaptation and maladjustment in the USA was approached through the study of the disorganization of some groups, such as the Poles, who experienced difficulties in adapting to new forms of social organization. The notion of group social disorganization was defined as the decline of the influence of the rules of behaviour on the members of the group (p. 89). Norms, disorganization, individualism, demoralization and delinquency were linked and quickly became a focus of research by Chicago sociologists. However, historians of sociology consider this viewpoint to have been widely criticized already in the 1940s and to no longer be specific to Chicago sociology (p. 285). The pinnacle of this type of research was Park and Burgess’s book The City (1925), which focused on social (not individual) control and disorganization, rural–urban transformations, demoralization, crime, etc. It was this perspective that Leighton borrowed: in his frame of reference, statistics recorded social change and the disintegration of certain populations. But notions of assimilation and integration were already being abandoned by a new generation of sociologists in the 1940s (Chapoulie, 2018: 357) because of the value judgements they implied.

For instance, Everett C Hughes ¹⁸ and Howard S Becker developed new sociological research in Chicago to study interaction processes. They integrated the actors’ point of view as an alternative model. Their notions of deviance and cultural dissent moved away from a monolithic explanatory model of social disorganization. But this consideration of the cultural dissent of certain social groups was missing from Leighton’s design. The study’s interdisciplinarity supported a more anthropological approach to Acadian folklore than a sociological analysis of the job market and youth subculture. ¹⁹ In contrast to the dichotomized Anglophone versus Francophone cultures in the core of the survey’s design – trapped in the narrow notions of acculturation and integration that Leighton inherited from his anthropological training in the 1930s – young people leaving rural Nova Scotia were not only placed in the alternative of minority or dominant cultures. Youths also left their community because of the blossoming teenage subculture, which was a transnational movement. It expressed itself noisily, for example, in the form of the fashion for ‘teddy boys’, rock and roll, leisure culture, pop culture, dancing, cars, movies, pursuit of pleasure, etc., which characterized the 1950s and 1960s, when the Stirling County Study started. These trends were familiar to some members of Leighton’s team based on references to the scientific literature (Cohen, 1955). This subculture became a specific culture, that of the baby boom. The new and major cultural phenomenon of youth could not have been anticipated by Leighton’s team, but it could be understood as an integrating factor for the youth, while ‘private religious participation’ was less relevant in the secularizing society in the 1950s and later. The new modes of consumption among young people became mass culture: in the 1960s, counter-culture became culture.

Other kinds of interdisciplinarity in psychiatric epidemiology

Before concluding, let us recall that there have been other ‘regional formations’ in the history of psychiatric epidemiology. Contemporary research teams other than Leighton’s were constituted by interdisciplinary constellations.

One Canadian example was epidemiologist HBM (Brian) Murphy (1915–87) in Montreal. He was trained in psychiatry, statistics and sociology, but not in anthropology; his integration into social and transcultural psychiatry team at McGill led to numerous international collaborations in different areas and on all continents. In the USA, some professors of social psychiatry, interested in epidemiology, were classically inspired by the hygiene and social medicine movements. Some American epidemiologists active in the field of mental health were sociologists by training, not physicians. In Europe, the epidemiological studies supported by the Max-Planck-Gesellschaft (Max-Planck-Institute of Psychiatry, Munich) were carried out by West German researchers who went to train in psychometrics in the UK and the USA. For example, Johannes C Brengelmann (1920–99) trained in questionnaire techniques with the psychologist Hans-Jürgen Eysenck (1916–97) in London. The interdisciplinary bases of this work were psychiatry and the psychology of individual differences, without the contributions of anthropology or sociology. Numerous interdisciplinary configurations have existed and have made possible various forms of collaborative research. Their variety makes the historical analysis of the field interesting and more complex, because researchers have not always pursued the same scientific objectives.

Another totally different history of psychiatric epidemiology can be written by looking at the researchers who were inspired by the screening scales and methods developed by Eysenck. In this case, the interdisciplinarity between physicians, psychologists and other professions (e.g. biostatisticians) offered a different case study, in the sense that the interdisciplinarity surrounding Eysenck’s students and collaborators did not include the same knowledge in the human sciences, which distanced psychiatric epidemiology from social anthropology.

Anthropology is itself an ambiguous label. In Leighton’s team, his closest collaborators were his two wives, Dorothea Leighton and then Jane Murphy. However, when Alexander and Dorothea separated in the 1960s, she continued to work and established herself as a specialist in ‘medical anthropology’. Alexander Leighton, on the other hand, never sought to establish himself under this label, but under that of ‘psychiatric epidemiology’. This logic of differentiating professional fields in the process of specialization is not neutral from the point of view of sociology and the history of professions (e.g. as a career strategy), especially since Dorothea Leighton was one of the founders of medical anthropology in the USA. She was the first President of the Society for Medical Anthropology in 1972, a successful section of the American Anthropological Association.

Finally, Chance and Tremblay became professors of anthropology. The latter became a specialist of anthropology and sociology and of health in Quebec. Collier Jr., as mentioned above, positioned himself as one of the founders of visual anthropology (Collier, 1957, 1986) in the academic world, after completing his PhD under Alexander Leighton’s direction. His photographs of Acadian country taken in the 1950s have been exhibited ²⁰ in Nova Scotia. No art historian or historian of anthropology has made the connection between his work – well known in the history of photography – and the history of epidemiology.

Conclusion

The design of the Stirling County Study was dominated by a sort of anthropology more attentive to Acadian folklore and traditions than to social dynamics and the sociology of youth. The study did not set out to emphasize social, economic and health variables in youth. Moreover, Malinowski’s functionalism theory was Leighton’s reference model in anthropology, which gave rise to many so-called ‘acculturation’ studies, thus reinforcing an old model of ‘social disorganization’ in sociology at the expense of more complex and modern models of sociocultural dynamics. Of course, Leighton was not alone in his functionalism, a long-lasting model in social sciences. However, one of the results of the Stirling County Study in the 1950s was to highlight the role of a variable that had not been anticipated in the study design: youth migration. This result is consistent with the social history of youth in Canada during the Great Depression and with the concerns of the public authorities about the morality of youth as they emancipated themselves from traditional ways of life in the 1950s.

Matthew Heaton’s analysis of the Cornell-Aro Study in Nigeria shows that the reification of Leighton’s frame of reference between Canada and Nigeria involved the abandonment of cultural issues (Heaton, 2022: 208, 212) to assert the universalism of psychiatric epidemiology (and of the human mind). In fact, ‘key cultural values’ (Leighton, Clausen and Wilson, 1957: 21) were central to the design of the Stirling County Study at the outset (‘we expect to find differences in the prevalence of psychiatric disorders in the French culture as compared to the English’; Leighton, 1959a: 286). But cultural specificities were progressively omitted in favour of more specialized epidemiological research in the later versions of the survey.

In questioning the interdisciplinarity at work in the ambitious research team assembled by Leighton, it should also be emphasized that interdisciplinarity options and choices would have consequences in the 1960s, 1970s and 1980s and later, when the local population and physicians would hinder or challenge Leighton’s projects in Nova Scotia (Fingard and Rutherford, 2005, 2008; Leighton, 1982). New generations did not easily accept the survey and the creation or use of local clinics (from 1951) for this purpose, rejecting a paternalistic model from abroad, because of a desire for change and autonomy from American scholars.

However, it would be an unhistorical attitude to believe that Leighton did not refine his design for the replication of the study in other decades. Jane Murphy, Leighton’s second wife, took over the direction of the study in 1975, when the couple had already left Cornell for Harvard. The population of Nova Scotia had changed. New generations had populated the Maritimes, including migrants. The scientific team had also changed, shrinking from the 1960s onward, with fewer social scientists and probably more scholars trained in epidemiology.