Methodological Individualism between Explanation and Theory Building

The present work aims at criticising the foundationalist assumptions of Methodological Individualism (MI) in the social sciences by showing that they depend on one, but notnecessarily the only, conception of the role of explanation in the construction of overall sociological theories. Philosophical arguments have been supplemented with a bird’s-eye survey of contemporary social research in order to identify such conceptions in common strands of empirical research. This method tries to comply with the need for a tighter interaction between philosophy of science and working social science [8, 9]. In current social research, a sophisticated restatement of MI, known as “structural indi-vidualism”, has been introduced by so-called Analytical Sociology (AS), which is gaining more and more popularity among social scientists [5]. The main idea that characterisesAS is that explanations of social phenomena ought consist of detecting and detailing the mechanisms that generate them, that is, by referring to individuals, their properties, actions, and relations to one another [4]—the “social cogs and wheels” of the mechanism [2]. As such, AS shares with more traditional versions of MI [12] the requirement thatall acceptable explanations of social phenomena be reduced to individuals and their properties. Sofar,mostdebateonMIhingedonestablishingthemostadequatelevelofexplanation [11], either by deriving MI from social ontology or by explicitly defining causation—and hence explanations—in terms of individual action. In contrast, we argue that MI could be better assessed by shifting the focus from ontology and epistemology to the nature and scope of sociological theories. Reductions to the individual level of the kind advocated by methodological individualists crucially presuppose a particular conception of the role of explanations in the construction of more general theories which is but one of the possible alternatives. One of these is the model of explanation that, we contend, underlies many surveyed empirical studies which draw on different sociological frameworks, such as comparative-historical and comparative-institutional research programmes. Explanations of this kind seek to establish causal relationships between the occurrence of a particular event (the explanandum ) and the earlier occurrence of certain historical or institutionalphenomena (the explanantia ) without necessarily further reducing any of them to their components. This model of explanation, we argue, is functional to building sociological theories that make abstraction from the particular historical or institutional contextsof study in order to identify recurrent patterns of causation between general types of phenomena. Examples from comparative political economy provide a case in point: although case-study based, such research provides social scientific explanations applicableto all instances of the same phenomenon—such as, say, the emergence of a certain variety of educational system in Western countries with coordinated market economies [1]. According to AS, on the contrary, to explain a social phenomenon is to identify a social mechanism. The stylized “social cogs and wheels” that make up the mechanism resultfrom the dissection of the explanandum into its constituent entities and their abstraction fromthoseelementsconsideredtobeoflesserimportance[4]. Therefore, whatmechanism- based explanations actually refer to is a stylized version of the phenomenon; mechanisms can thus be said to define an entire class of phenomena analogous to the explanandum . Indeed, to the extent that mechanisms are made up of stylized entities, they might as well be extended to explain phenomena belonging to different classes, insofar as thesecan be reduced to the same stylized constituents. This model of explanation supportssociological theories whose goal is to build models of causal mechanisms so general asto “cut across” the different types of phenomena and be therefore applied to explain events of very different kinds. Individual-level reductions, in short, ought not be seen as attempts to arrive at more exhaustive explanations of particular phenomena—or classes of phenomena—than non-reduced explanations. Quite on the contrary, reductions of thiskind are best understood as the building blocks for more general theories whose scope liesin their abstractness and hence in their broad applicability. This argument is corroborated by the actual analysis of various empirical studies run by AS scholars in order to explain particular historical events by means of general models of mechanisms. Among the most revealing, “rational imitation” models of mechanisms [4] have been extensively applied to explain phenomena as diverse as the emergence of social movements [3, 6] and the participation into a bank run [10]. In conclusion, reductions of explanations to the individual level—like those advocated byMI—arejustifiedbytheneedtobuildsociologicaltheoriesthatembracehetereogeneousvarieties of phenomena. Yet since this conception of sociological theories is far from beingthe only one, the methodological individualist assumption that all explanations ought be altogether reduced to the individual level amounts to a form of what some authors [7] aptly defined as “microchauvinism”.