The Problem with Prediction: Contingency, Emergence, and The Reification of Projections

Abstract

Here we assess the limitations of “social physics” methods for making predictions of future social conditions in order to contribute to the development of an alternative model of scientific sociology. First, we discuss the complexities inherent in causal analyses of social phenomena, where social organizing principles change historically and causal forces interact to generate outcomes. Second, we analyze the assumptions that underlie prediction in the social sciences by using the population projections of demographers as an illustrative example. Demographers typically make projections assuming the future will be the same as the (recent) past. This approach neglects the possibility of abrupt historical change, contingent events, and emergent social structures. Third, we argue that historical contingency is an essential and necessary consideration in any analysis of the world. A historical materialist approach provides the basis for better formulating nomothetic and idiographic modes of explanation and appreciates the important connections between the two. Sociological inquiry should be an attempt to distinguish between truly nomothetic processes and emergent historical background conditions that only appear nomothetic in a specific historical era. We close with an argument for the importance of developing a sociological science that runs parallel to nonmechanistic approaches in the biological sciences.