In this chapter I offer a theoretical perspective on the challenges faced by archaeologists in addressing micro-scale social interactions with network concepts. I do this by sketching three ontologies – very loosely defined as Melanesian, Classical and Gothic – which allow us to imagine the possibility that social interactions may be locally conceived in ways that do not lend themselves to being abstracted in network terms. If, for example, people are socially constructed as dividuals, with their identity distributed across relations with other people and with things, then it could be difficult identifying separable entities for the purposes of network analysis. I argue that this kind of challenge needs to be tackled through the development of models that problematize these kinds of micro-scale phenomena.