The Social Innovation Network (SInet) was established for cross-disciplinary research on social innovation to 'create better futures for people'. SInet is itself socially innovative since a network is a relatively unfamiliar configuration for a university-wide research unit. A network provides an identity to a research collective that is real, having status and support,but which is fundamentally different to an institute. In a network, connections and flows of knowledge tend to be horizontal not vertical. A network is flexible, reconfigurable, responsive to change and less formal, and has the potential for lower administrative overheads. As knowledge workers, university researchers perform best in an organisation that supports an open culture where knowledge workers are left alone to work, with sufficient support and resources. Their performance is maximised by capitalising on their strengths and knowledge rather than trying to force them into moulds.This paper compares the attributes of a research network to a more traditional hierarchical institute. It asks and answers the questions: what is a research network; why have one for intra-institutional research; and how can it be created, sustained and its value determined? Three theories will be used to (a) provide the reasons and justification for network-centric configurations, (b) make sense of the network-centric paradigm and its characteristics, and (c) understand how to act in a network-centric workplace arrangement. Not everyone is comfortable working in a self-directed network-centric configuration, so will SInet work, and if so, how?