There is not much new about many of these activities, perhaps because weak linking across clusters is naturally efficient, an unappreciated theory in use (Argyris and Schon, 1978). However, this paper has attempted to make this theory explicit, and provide a clearer picture that makes sense of the phenomena of interest. This, it is hoped, will make the governance of strategic knowledge sharing more explicit. Given the complexity of wicked problems and the creativity needed to respond to them, hierarchical control of either participant’s actions or of their knowledge sharing is considered naïve. The exact opposite of a ‘need to know’ knowledge sharing policy is required. Participants are to be encouraged to decide for themselves what they need to know and to be aware where they can get that information easily, as in the oil spill example.

This paper has argued that strategic knowledge sharing can be viewed as a task in designing small-worlds networks so as to allow a self-organised strategic response to wicked problems. Knowledgeable clusters and synchronous or asynchronous weak links can both identify wicked problems and respond strategically. A designer of this sort of network needs to encourage knowledgeable clusters that are only weak linked together, whether synchronously, asynchronously or both. Perhaps, for a commercial organisation, the designer may allocate resources to encourage a particular common purpose. Two examples of wicked problems being handled by wicked systems were outlined. In one, the wicked problem was a rather obvious oil spill threatening many dimensions of a community. In the other, the wicked problem required innovation and creativity. It is argued here that neither should be managed in a hierarchical sense.

Future research may continue the work to make explicit the design of existing social networks in public and research communities, so as to better appreciate if they follow a small-world structure. The limits of synchronous and asynchronous ‘weak links’ may also warrant further investigation, as might the role of information and communication technology from this point of view. Other suggestions may include: how people communicate, and with whom, when involved in a crisis; and what information they need or could not obtain. How the self-organisation process works also needs to be made more explicit.