In this workshop, we will begin working toward an integrated approach to the study of warfare under decentralized or only loosely controlled conditions, combining empirical data, evolutionary theory, and mathematical models. We envision developing working hypotheses to answer several key questions about between-group conflict in general, and the nature of ‘decentralized warfare’ in particular. These hypotheses will be informed by evolutionary theory using data from diverse conflict settings. Our hypotheses will be articulated in a framework amenable to formal modeling that will point the way toward a multi-level predictive understanding of warfare.