What is relational explanation?


Peeter Selg, Tallinn University; Joonatan Nõgisto, Tallinn University

Relational sociologists are united in emphasizing the special role that social relations play in illuminating the social world. However, there has been little explicit attention to how a relational perspective fits with and enables specific kinds of social scientific achievement, such as description, interpretation, or explanation. Methodological development within relational sociology requires greater clarity on scientific objectives along with criteria for success and failure. This paper focuses on explanation as a goal within relational sociology. How can relational sociology offer novel and insightful explanatory knowledge on the why of social phenomena? What characterizes the explanatory commitments of relational sociology and distinguishes explanatory success from failure? This paper explicates a concept of relational explanation – a distinctive form of social scientific constitutive explanation that is characteristic of the research programme of relational sociology. Recent developments on metaphysical grounding and constitutive explanation in analytic metaphysics and philosophy of science are used to construct a concept of relational explanation as explaining interdependent social phenomena through abductive inference to their common ground in features of dynamic and unfolding social relations. Focusing on the inherently abductive nature of relational explanation, the specificity of abduction is then clarified in view of the two prevalent logics in scientific thinking – induction and deduction – to demonstrate the essentially processual and relational character of abduction. Abduction is a processual movement from puzzling empirical phenomenon to theoretical propositions and other observations making it intelligible and then back to the phenomenon through which both the identity of the phenomenon as well as the corresponding theoretical premises and statuses of other observations can change, and the process is never completely final, since the “final” result itself is a part of the process. “Constitutive explanation” is distinct from the more usual “causal explanation,” since it presumes that both the explanans and the explanandum “can be considered separately, but not as being separate,” to use Norbert Elias’s phrase. What makes relational explanation especially challenging, however, is that, it must consider both the explanans and the explanandum as unfolding dynamic processes, not as given “explanatory factors” or “independent variables”, or “dependent variables” or “outcomes”. Thus, relational explanation is temporal and dynamic throughout, and that not only in the more usual “diachronic” sense, but also in the “synchronic” sense: it has to take seriously the ontological presumption of relationalism that every “moment” in the existence of social processes is “synchronously” constituted by its past, present, and future. We base our argument on recent analyses of “outcomes” in processual-relational sociology, and introduce relational explanation as “unfolding” (in the transitive (!) sense of the word) of explanandum, of social phenomena into their constitutive processes that can be considered separately, but not as being separate. We contrast our notion of relational explanation from structuralism, individualism and variable-centered explanations in the social sciences. We also consider its working in classical explanations of social phenomena (e. g. Marx’s explanation of commodity fetishism), and contemporary examples (such as Mark Granovetter’s explanation of embeddedness of economic relations in social relations and Abbott’s use of “linked ecologies” in explaining social phenomena.)

This paper will be presented at the following session: