Methodological holism and causal explanation: lessons from Durkheim and Weber

In this thesis I argue for methodological holism in sociology. I make this argument in the limited context of the provision of causal explanations of social phenomena as classically understood; these are phenomena like marriage, a rugby match, or an economic depression. I claim that if we are going to causally explain social phenomena, then the causes we adduce as explanations must be holistic social phenomena, i.e., they must be social phenomena understood as wholes, rather than as aggregations of individualist phenomena such as the motivations and beliefs of the individuals involved. My argument for this claim has two parts. Firstly, I argue that if a causal judgment, C caused E, is to be explanatory, such that C explains E, then there must be a certain kind of justification for the judgement that C caused E. This justification must be the kind of justification that supports inferential relations between the parts of the judgement that refer to events as causal relata. Secondly, I argue that in cases in which a social phenomenon is the explanandum, only a causal judgment in which another social phenomenon is the cause can appeal to the right kind of justification and thus provide an appropriate explananda. Consequently, social phenomena can only be causally explained by other social phenomena. In Chapters 1 to 4, I make both parts of this argument in the context of the philosophically rationalist interpretation of causation favoured by several key figures in the history of sociology and philosophy; firstly, as it is formulated in Kant’s reply to Hume, and secondly, as it is adopted by the classic sociologists Durkheim and Weber. In the final chapter I develop some particular applications of my argument using concepts taken from social system theory.