Understanding creativity begins with investigating its appropriate unit of analysis: what elements constitute and can sufficiently explain creative cognition, without leaving out any essential aspect? Is creativity sufficiently explained by the study of the human brain alone? Or does it go beyond the boundaries of the skull, such that the investigation of the external representations—movement, body, environment, etc.—can also provide essential insights into the nature of creative cognition?This dissertation investigates the appropriate units of analysis of expert-level mathematical creativity as a canonical example of highly abstract creative cognition. To do so, it draws on mathematicians’ self-report accounts, empirical studies, and formal modeling. It provides both causal and correlational evidence of multiple mechanisms through which external representations contribute to creative cognition. It thus argues that even highly abstract creative breakthroughs benefit—and in some cases arise—from interactions across distributed components including the brain, the body, the environment, and their interactions. Thus, a comprehensive account of mathematical creativity, and creative cognition in general, must go beyond the boundaries of the human skull and embrace external representations as well.