Structuration theory (ST) and network analysis are promising approaches for studying the emergence of communication networks. We offer a model that integrates the conceptual richness of structuration with the precision of relevant concepts and mechanisms offered from communication network research. We leverage methodological advancements (i.e., stochastic actor-oriented models) to test hypotheses deduced from ST using longitudinal communication network data collected over a two-year period. Results indicate that while structural rules external to a social network play a significant role, internal structural rules that emerge from the aggregate of individual actions during previous time periods also predict current structures, and that the reification influence of the latter is greater than that of the external factors.