How and Why Do We Write the History of the Social Sciences?

Why should we write the history and sociology of the social sciences? Some have suggested that putting science under the sociological microscope is self-indulgent and dangerously relativist. Others murmur that only those who can’t do science study science. Ernst Wilhelm Eschmann, a Nazi sociologist, wrote in 1934 that a science that makes itself into its object of study, that studies “its relations and boundaries with other sciences, its epistemology, methods, and history,” represents “the symptom of a profound sickness of an entire culture,” a “pathology of scientificity.”¹

Few nowadays would be included to agree with a Nazi scientist. Yet these criticisms should not go unanswered, especially in an age when scholars are insistently called upon to demonstrate the usefulness of their work. The historical sociology of social science is useful. It is a necessary part of all social science. Before outlining the usefulness of this apparently useless form of knowledge, and sketching some of the methods currently being used to carry it out, I will briefly sketch its emergence.

The social study of knowledge and science did, as Eschmann’s comments suggest, emerge during a period of profound sociopolitical and cultural crisis in the first decades of the twentieth century. This period saw the stunning collapse of empires after World War I, the German economy’s plunge into hyperinflation, the hydra-like rise of fascist movements in Europe, and the first signs of organized resistance to colonial domination in Africa and Asia. There was also a widespread “devalorization of objective and rational life which … declares science to be bankrupt,” as Gaston Bachelard observed in 1938.² The skepticism about science was an international movement, but it was stronger in Europe than the United States and especially powerful in Germany and the late Austro-Hungarian Empire. The precedents here included Karl Marx’s critique of political economy as an expression of capitalist class interests and his more general argument that social existence determines consciousness. Nietzsche described scientific ideas as instruments of a will to power. Social class, political power, and religion were central explanatory factors in the nascent sociology of knowledge. Discussions among participants in the Budapest “Sunday Circle,” who included György Lukács and Karl Mannheim, circled around the idea of the dependence of knowledge upon social position.³ A number of social scientists and historians in Weimar Germany (Ernst Grünwald, Karl Mannheim, Max Scheler, and Alfred Weber) developed the approach that came to be known as the sociology of knowledge.⁴

A key development in this intellectual movement was the turn to analyzing the physical and natural sciences sociologically. The idea of explaining science sociology had also emerged before World War I. In Elementary Forms of the Religious Life (1912), Émile Durkheim traced the basic epistemological categories of thought, and modern science, including time, space, number, cause, and force, to religious social practices and structures. George Sarton created the journal Isis, dedicated to the history of science, in 1913; this was followed by Osiris (1936) and Journal of the History of Ideas (1940). But the history of science was not “established in academic departments, centers, and programs in Europe and North America” until the 1950s and 1960s. The early historians of science were less oriented toward explaining science sociologically than to celebrating its inexorable march toward perfection and linking it to “something called ‘the modern mentality’,” whose home was in the West.⁵ Some of these founding narratives of the history of science were also ambivalent about the destructive aspects of scientific modernity, as Lorraine Daston points out, but this did not lead their authors to a full-fledged sociological account of scientific change until the 1960s.

This story of the rise of a more theorized history of science sometimes overlooks a more eclectic set of thinkers who began to develop sociological approaches to the history of science between the wars. The journal Science and Society, created in 1936, was “dedicated to the growth of Marxian scholarship” and announced a special interest in work that “illuminates the interdependence of science and society.⁶ One contributor was Dutch mathematician Dirk J. Struick, who analyzed the “sociological stain” said to shape the development of even the purest forms of mathematics.⁷ In a series of articles published in the Austrian socialist journal Der Kampf, philosopher Edgar Zilsel analyzed the social roots of scientific causal thinking and the concepts of scientific progress and physical laws in religion, legal forms, state politics, and class relations.⁸ Philosopher Gaston Bachelard analyzed epistemic obstacles to scientific knowledge rooted in unconscious emotions, some of them linked to scientists’ anxieties about social status.⁹ American sociologist Robert K. Merton redeployed Max Weber’s Protestant Ethic thesis by explaining the rise of science in seventeenth-century England partly in terms of Puritanism.¹⁰

The cultural revolutions of the 1960s gave new impetus to the history and sociology of science. A signal event was Thomas Kuhn’s The Structure of Scientific Revolutions (1962). New journals were created: Archive for History of Exact Sciences (1960), Minerva (1962), British Journal for the History of Science (1962), and Journal of the History of the Behavioral Sciences (1965). Cheiron, the International Society for the History of the Behavioral and Social Sciences, was founded in 1968.

This history still does not explain why anyone should be interested in studying the social sciences. There are two sorts of justifications: some intra-scientific, and others of more general interest.

First, the historical sociology of social science helps social scientists understand their own quasi-spontaneous orientation toward specific theories, concepts, and methods. This form of “self-reflexivity” should not be imagined along the lines of a confessional; it is better described as a practice of self-objectification, in Pierre Bourdieu’s words. A historical excavation of meaning allows us to better grasp the buried significations and connotations of our current scientific language, which is often the result of earlier scientific battles.

One example of the hidden sources of present-day conceptual language is the phrase social science. Different actors and institutions have preferred different genus labels. In battles over the scientific direction of UNESCO starting in 1946, Americans argued in favor of the label social science, which at the time suggested applied, mainly quantitative social research, organized around the individual person as the fundamental unit of analysis. French scholars pushed for UNESCO to adopt the label human sciences (sciences humaines), which for them encompassed philosophy and the humanities.¹¹ Back in the U.S. at the same time, however, the phrase behavioral sciences was being used for the Ford Foundation’s program to fund the disciplines that had hitherto been known as social sciences. This was intended to avoid any confusion of social science with socialism and to emphasize the same set of approaches associated with “social science” at UNESCO. In Germany, the central debate pitted the traditional idea of Geisteswissenschaften against social science, which was usually rendered as Sozialwissenschaft. The new universities created after 1918 in Germany were the first to have “social science” divisions. The reassignment of sociology from the Geisteswissenschaften to the social sciences was largely completed between 1933 and 1945 due to the Nazis’ preference support for applied research.¹² The history of social science can make these sorts of histories part of the self-objectifying practice of social scientists.

A second justification for this form of research is that excellent social theories and concepts have been collectively forgotten and can be recovered and reintroduced. Serious research on Marx or using Marx’s ideas was largely excluded from professional social science in the U.S. and Germany before the late 1960s.¹³ Durkheim and Weber were widely dismissed as outmoded compared to American-style social science in France and Germany, respectively, after 1945. Zilsel was largely ignored, like most of the refugee sociologists from Nazi Germany, but has now been recovered.

A third justification for the history of social science is provided by Lothar Peter.¹⁴ Modern phenomena such as individualization, secularization, and capitalism may become more invisible as they become more universal, entering common sense and the taken-for-granted. Earlier observers who had a foot in two different social worlds and worldviews may have been able to perceive these nascent social processes more clearly than we are today.

The fourth justification for the history of social sciences is that it can help us understand the conditions in which social science flourishes or stagnates.¹⁵

Finally, the history of social science contributes to explaining (and thereby, perhaps, solving) various social problems. The social sciences permeate our social existence. They influence the configuration of our economies, social policies, educational systems, and foreign affairs. Here it is important to distinguish between the intentional application of social science to policy and the unintentional impact of social science on social practices.

Areas in which social science has been deliberately used to guide policy include eugenics and social insurance, labor market policies and poverty relief programs, and most recently, schemes intended to “nudge” individuals toward desired behaviors.¹⁶ One exceptionally well-documented instance of social science being directed toward government policy is modernization theory, which emerged in response to Harry Truman’s program for the development of previously colonial areas. Sociological ideas informed guided projects of community development, land reform, dam building, and the agrarian “Green Revolution.”¹⁷ Similar interventions were undertaken with social scientific guidance throughout the French and British African colonies during the postwar period under the guise of colonial development.¹⁸

Counterinsurgency campaigns have also commissioned social scientific research. During the Algerian Civil war, social scientists helped to design and run the resettlement camps created to undercut the uprising. The entire process was overseen by the Specialized Administrative Sections of the French Army, some of whose officers studied “Muslim sociology” and some of whom were charged with studying indigenous villagers before and after resettlement in order to discover how to make the camps more viable.¹⁹ A number of these officials were trained by the former colonial military officer and Durkheimian sociologist Robert Montagne at the Parisian Centre des hautes études d’administration musulmane (Center for Advanced Study of the Administration of Muslims).

Closer to home, the U.S. Defense Department has funded numerous social science projects since World War II. Efforts have been made to guide foreign policy using game theoretic models and computational systems, such as the Crisis Early Warning and Monitoring System.²⁰ Among the best-documented Pentagon-funded counterinsurgency research initiatives are the Troy (1950s), Camelot (1960s), and Human Terrain System (2007–2014) projects and the current Minerva Initiative (2008–present).

Social science also shapes policies and practices in many indirect ways. I tried to uncover such hidden causal chains in The Devil’s Handwriting, which set out to explain why the German colonial government in Southwest Africa (Namibia) carried out genocide against the Herero people in 1904, while German colonizers pursued more paternalistic and peaceful policies in other parts of their overseas empire. That book shows that the ethnographic and civilizational accounts of European travelers, missionaries, and amateur ethnographers set the basic parameters of formal colonial policies, even if this was the conscious intent of those accounts and even if they were generated years, even decades before colonial conquest.

This finding is not a recipe for preventing genocide. The social sciences are not good at prediction, except with regard to more trivial, repeated events--and even then this is only true within specific time-space constraints. Epochal historical events like revolutions, massacres, and the collapse of states and empires are overdetermined within open systems. While such events can be explained, the idea of lawlike historical generalization is chimeric. What a historical study like this can do is point to conditions that were causally efficacious in one instance of genocide, suggesting that they may be a necessary if not a sufficient condition, and thereby cautioning against them in present. In the present example, these causal conditions had to be identified using the methods of the history of social science.

But what are these methods? How exactly should we investigate the social sciences? Let me just mention three important aspects of this methodological discussion: (1) the need to connect texts and contexts, (2) the importance of using private and public documentation, and (3) deployment of the concept of semi-autonomous social fields.

With regard to the first point: it has become clear that researchers in this area have to attend to the textual, visual, and numerical works––using all of the appropriate methods of interpretation and analysis––as well as the contexts of the production of social scientific works. Relevant contexts range from the most proximate (the scientist’s biography or prosopography), to the intermediate (scientific institutions, disciplines, and fields), and to the more remote contexts that were the focus of the earliest contextual research discussed above.

An example of the importance of taking into account the recursive relations between social science and its context is suggested by the work of French psychologists from the Centre d’Etudes et d’Informations des problèmes humains dans les zones arides, who carried out psychological tests on the Mekhadma tribe in colonial Algeria during the 1950s. The photo on the left reveals that the testing was carried out under the aegis of the colonial military and the surveillance of the Caïd, the semi-autonomous headman on horseback. Other colonial-era researchers had more autonomy from colonial officials and indigenous actors.²¹

Second, it is essential to combine published materials with institutional documents and personal papers, including private correspondence. By establishing political and institutional constraints and private communications and comparing them with published work we can shed light on the question of how scholars deal with pressures from states, funders, employers, and colleagues.²²

Finally, research in the historical sociology of science has benefited from Pierre Bourdieu’s field theory, which gives conceptual clarity to a crucial intermediate social context level located between the micro-level of biography and the macro-level of society.²³ Bourdieu’s field model has helped us understand such phenomena as the failure of some émigré social scientists to transfer their scientific capital from their earlier context to a new one, and the converse. It sheds light on the changing meanings of a given intellectual program as it travels internationally from one context to another. It explains other things: the internal conflicts and power structures within disciplines; the gatekeepers, export-import controls, and protectionisms at disciplinary borders; and the stubborn resistance of some disciplines or parts of disciplines to trends in adjacent fields as well as the seemingly spontaneous openness of other actors to inter- or transdisciplinary interactions. The field approach explains the overall obsession with ranking and ordinalization in contemporary education, research, and (social) science. At the cutting edge of current discussions of field theory are efforts to introduce spatial analysis into the study of fields and to combine it with a more nuanced model of the human subject.