The theory of self-assembling games studies how evolutionary games emerge from basic interactions, and explores how once formed, the structure of a game may evolve on continued play. In Self Assembling Games, Jeffrey A. Barrett offers an incisive analysis of hierarchical, dialogue, and network-style games, illustrating how trial-and-error learners coevolve meaningful language and inferential capacities, how social structures that allow for more reliable learning emerge, and how learning itself develops in more sophisticated ways throughout this process. In a philosophical exploration of self-assembling games' impact on animal and human behaviour, Barrett follows a continuous philosophical thread, providing a framework for a thoughtful pragmatic account of empirical inquiry. In demonstrating how simple agents may evolve meaningful language, useful social connections, and better ways of learning as they succeed and fail together in action, Self Assembling Games offers readers new ways of considering game-theoretic ritualization at large.