Wednesday, January 31, 2024

Defining disciplinary research in the social sciences

The "historical turn" in the philosophy of science in the 1960s and 1970s gave most of its attention to the development of the physical sciences -- especially physics itself. (See Tom Nickles' essay "Historicist Theories of Scientific Rationality" in the Stanford Encyclopedia of Philosophy for a detailed account of this development in the philosophy of science; link.) Historian-philosophers like Ludwik Fleck, Thomas Kuhn, and Imre Lakatos studied the development of astronomy, physics, and chemistry as research communities involving complex social arrangements -- networks of practitioners, training institutions, laboratories, journals, and universities and research institutes -- and shifting but shared cognitive frameworks. They argued that scientific research and knowledge always proceeds through organized research communities that regulate the pursuit of scientific knowledge. Kuhn emphasized the specificity and contingency of the cognitive frameworks (disciplinary matrix or paradigm) that guided a research community, insights that in part reflected his own reading of Ludwik Fleck's earlier work on the history of biology and medicine and the idea of a "denkkollektiv" (link, link).  Here is Alexander Bird's description of Kuhn's view in his essay on Kuhn in the Stanford Encyclopedia of Philosophy:

He claims that normal science can succeed in making progress only if there is a strong commitment by the relevant scientific community to their shared theoretical beliefs, values, instruments and techniques, and even metaphysics. This constellation of shared commitments Kuhn at one point calls a ‘disciplinary matrix’ (1970a, 182) although elsewhere he often uses the term ‘paradigm’. Because commitment to the disciplinary matrix is a pre-requisite for successful normal science, an inculcation of that commitment is a key element in scientific training and in the formation of the mind-set of a successful scientist. (link)

Well and good for astronomy and fundamental physics. But what about the social sciences? How should we think about the sub-disciplines of fields in the social sciences like sociology? The idea is sometimes expressed, including by Kuhn himself, that the social sciences are not yet "mature sciences" precisely because they lack strong and definitive paradigms. Here is Bird again in the SEP:

The claim that the consensus of a disciplinary matrix is primarily agreement on paradigms-as-exemplars is intended to explain the nature of normal science and the process of crisis, revolution, and renewal of normal science. It also explains the birth of a mature science. Kuhn describes an immature science, in what he sometimes calls its ‘pre-paradigm’ period, as lacking consensus. Competing schools of thought possess differing procedures, theories, even metaphysical presuppositions. Consequently there is little opportunity for collective progress. (link)

This view perhaps once had a seductive appeal, but it is no longer convincing. The social sciences are different from the natural sciences because social phenomena are different from the phenomena of the world of chemistry and physics, and the unity of science was a dogma from logical positivism that we are well rid of (link, link). Here is how I've tried to formulate the differences that exist between the social sciences and the physical sciences:

Rather than unity, we should expect eclectic theories, piecemeal explanations, and a patchwork of inquiries at a range of levels of description. Some explanatory theories will turn out to be more portable than others. But none will be comprehensive, and the social sciences will always remain open-ended and extensible. Instead of theoretical unification we might rather look for a more and more satisfactory coverage, through a range of disciplines and methods, of the aspects of the social world we judge most interesting and important. And these judgments can be trusted to shift over time. And this means that we should be skeptical about the appropriateness of the goal of creating a unified social science. (Understanding Society; link)

This view of the social world suggests that sociological research does not require strict paradigms or dogmatic commitment to a "disciplinary matrix" of theory and methodological commitments in order to make progress. Instead, sociology should embrace a pluralistic range of research approaches and cognitive frameworks that can address different aspects of the social world. We gain insight and understanding of the social world through overlapping and pluralistic methods and theoretical frameworks.

That said, it is plain enough that there are distinct (sometimes overlapping) research families within sociology. Any large sociology department at a research university will have members who identify with different approaches and methodologies. Social movement researchers disagree with large-N quantitative researchers in many important ways, and an ethnomethodologist might find the work of both these sets of colleagues to be somewhat foreign. How should we characterize these differences across extended research groups within sociology? 

Several ideas are available. Weakly, a group of researchers might be said to belong to a research tradition if they share a number of common assumptions about the nature of the social world, the methodology that is most suitable to sociological research, and some core examples of sociological explanation. Most commonly this situation would arise from the fact of a common genealogy for a group of researchers -- a founder (Durkheim or Tarde, let's say), and a few generations of researchers who followed in their footsteps. But notice how weak this account is. It does not refer to shared research institutions, social networks of researchers, or definitive processes of research evaluation.

A stronger conception of a research community may begin with the weak conception, but then provide a specification of how research proceeds in this field or that. This stronger analysis aims to describe the institutional framework within which "X sociology" is carried out, evaluated, disseminated, and incorporated into the training of the next generation of sociologists in the X approach. What is added here is an account of the institutional infrastructure of the sub-discipline, the institutions and actor networks through which sociological research is carried out and shared. This more elaborate description is one of the guiding assumptions of the field of the sociology of scientific knowledge (SSK) (link).

But this specification remains incomplete in a crucial way: it says little about the intellectual content of the target sociological sub-field, the cognitive frameworks and theoretical ideas to which practitioners within the field adhere and that guide their research. This is the aspect that is the most novel and interesting feature of Kuhn's theory of the paradigm and Fleck's conception of the thought-collective. Kuhn, Fleck, and others in this line of thought emphasize that scientific research is not "blue-sky" theory, but rather proceeds on the basis of a contestable set of beliefs about the nature of the phenomena under investigation. According to this line of thought, the scientist's imagination is framed and directed by a set of assumptions about the world from which the discipline itself discourages deviation. The sub-discipline is indeed "disciplined" to conform to the existing paradigm. 

In the case of the social sciences, the role of "paradigm beliefs" is complicated and ambiguous. On the one hand, various areas of the social sciences are notorious for their rigid adherence to certain methodological principles -- the primacy of large-N quantitative studies, the value of formal models, lack of respect for case studies and comparative studies. (This was the point of the Perestroika debate in political science some years ago, and subsequent methodological debates in sociology more recently; link, link.) On the other hand, there is a very wide range of different assumptions (ontological beliefs) about social entities and processes in the social world across the social sciences and within each discipline. And it is rare to find a high level of consensus about these issues within a social science discipline. (Economics departments in research universities in the United States are an exception, but an unrewarding one: a high level of commitment to a specific set of methods, but a low level of empirical or policy success.) It seems, then, that the full-blooded idea of a "paradigm" as a unified cognitive framework spanning a broad network of researchers is not to be found in the social sciences. So perhaps our assessment of this topic in the social sciences might be: yes to research communities and networks, no to paradigm-driven research communities.

My interest in this question is fairly specific. I've been thinking about the emerging sub-discipline of analytical sociology over the past twenty years, and it would be useful to provide an analysis of the field in terms of the ideas sketched above. Do the manifestos of important advocates like Hedström and Demeulenaere constitute a "disciplinary matrix" in something like the sense that Kuhn had in mind? Do new areas of computational sociology (agent-based modeling, for example) represent something analogous to the body of laboratory techniques and procedures that Kuhn or Hanson included in the disciplinary matrix of physics? Do the strict assumptions of structural individualism advocated by the primary voices constitute a component of a "paradigm" for sociological research that helps to guide productive investigation and theory formation? In short, is analytical sociology a reasonably well-defined research community unified by a specific disciplinary matrix that is enforced by the regulative institutions of allied journals and institutes?

It is useful to compare the field of analytical sociology with other families of research in sociology today. For example, we might consider 

  • "processual sociology" described by Andrew Abbott (link
  • "field sociology" inspired by Pierre Bourdieu (link)
  • "ethnomethodology and microsociology" inspired by Goffman (link) and Garfinkel (link)

Abbott offers a vision of the nature of the social world -- an alternative social ontology -- and his writings provide quite a bit of methodological advice for sociologists interested in exploring these kinds of social phenomena. But it is hard to find the "processual sociology network" of researchers that is carrying out detailed empirical research under this banner. So it is hard to say that processual sociology constitutes a research community at present. The Bourdieu case is different. Many young sociologists make use of Bourdieu's ideas and theories about cultural fields. Bourdieu has a great deal of influence. But it is not obvious that there is a cumulative body of work that "Bourdieu-theory" can claim credit for as a research framework, and it is difficult to articulate the premises of a disciplinary matrix for conducting sociological research for this approach. So here too, perhaps we are forced to conclude that Bourdieu-theory too is less than a research community, in spite of its influence and the frequency of citations of Bourdieu in sociology journals. (According to a table titled "Most referenced authors in 42 sociology journals" reproduced by Gerardo Munck on X (link), Bourdieu is #1 with 9853 citations, and Weber is #2 with 6135 citations!) And though Goffman and Garfinkel are still part of the corpus of "theory" in sociology, it is hard to find instances of research networks proceeding along the lines described by ethnomethodologists fifty years ago. So analytical sociology seems to be a live candidate for an example of a research community organized by an active research matrix, whereas the other examples do not.

(This topic is an appropriate subject for study within the "new sociology of knowledge"; link.)