This proposal is an exploratory study of the effects of network topology on the dynamics of collective action. The one-year project will involve three core activities: a series of computational experiments using agent-based models, curriculum development for a new graduate seminar, and organization of an international interdisciplinary workshop. The proposal shifts attention to the configuration of social ties through which participation in collective action can propagate as a cascade. The proposal also bridges the disciplinary gap between the study of dynamical systems and game-theoretic models of collective action, by using computational methods developed for modeling cascades on large networks to explore how the structure of social ties can affect the spread of participation in collective action in heterogeneous populations of boundedly rational actors. The proposed workshop will promote the exchange of theoretical ideas, skills, and methodologies between social and natural scientists working on similar problems but with minimal mutual awareness or appreciation of potential synergies and no common language for communicating ideas or results. In the longer term, the graduate curriculum developed in this project will help to train a new generation of social scientists in the use of computational modeling techniques that are just beginning to migrate over from the natural and computational sciences.