Abstract
This study proposes an agent-based model of the impact of research success on the structure of scientific communities. In the model, heterogeneous scientists scattered about a ‘social landscape’ influence each other through networking. Peer networks are allowed to change based on the accumulated achievements (or prestige) of researchers. The dynamics of these networks are illustrated. The framework is then adjusted to allow for interdisciplinary practices (modelled as network links to more distant peers on the social landscape). Separate disciplines are shown to collapse into a single, large scientific network. Managing growing research networks, therefore, becomes a concern.