The Cumulation Problem in Political Science

Abstract

Abstract There is no purpose in launching new theories if they do not lead, over time, to a cumulation of knowledge on the subject. Brave conjectures must be followed by sober analyses, and these must cumulate over time. This chapter begins with a discussion of what counts as progress in political science, and in the social sciences more generally. Next, it discusses several widely regarded agendas for achieving cumulation—centered, respectively, on laws, mechanisms, theory, and inference. It argues that none of these traditional approaches is likely to go very far on their own. The final section explores another approach, dubbed standardization. Cumulation does not just happen, especially if scholars are working hard to distinguish their own work from everything and everyone that came before. Some degree of standardization is needed. And this, in turn, requires coordination among scholars working on a subject.