Two very simple examples are now discussed. The first is an account of the response to a community level problem, with special emphasis on knowledge-sharing issues. Hopefully, the analogy to ants’ nests and small-worlds phenomena is apparent.

Comfort (1994) argues that the citizens’ response to a oil spill near Pittsburgh in 1988 was a self-organised one, as the situation developed too rapidly and was too complicated for a simple top-down leadership response. She reports that the crisis began with a four million gallon diesel fuel tank collapsing. This resulted in a seventeen-mile-long emulsified oil and water mixture flowing down and over the locks and dams of the Monongahela River, extending bank to bank. The river provided drinking water for the Pittsburgh metropolitan region but the risk of damage to water filtration systems made the water authorities shut down the water intakes, resulting in a lack of water for either drinking or fire suppression. For two weeks alternative arrangements had to be organised, requiring the coordination of 25 different types of organisations – public, private, and community non-profit. The zoo, the fire service, medical services, the coastguard, hazard waste services, car washes, and bottled water companies all had to be coordinated.

One can easily imagine groups of concerned persons establishing informal clusters around their particular concern, or expertise. The bottled water people may be one cluster, the fire services another and so on. Most of their knowledge sharing would be within their cluster, perhaps on a one-to-one basis. Every now and then one of these clusters would need information from another concerned cluster; for example, the bottled water people might need an estimate of how long the crisis would last, or need to know how to get access to extra transport, or bottle manufacturing facilities. They would then use their ‘weak link’ to make contact with another cluster, as they would need to keep an overall appreciation of what was going on. The whole system would only work if there were both the locally knowledgeable clusters and the presence of weak but effective inter-cluster links.

One knowledge-sharing centre handled an estimated 37 000 thousand incoming and outgoing messages during the crisis; averaging 154 per hour, 24 hours a day, seven days a week. This would have been a fraction of the knowledge sharing involved. Comfort (1994) emphasises the need for a dynamic decentralised information system, one that provided up-to-date local and overall information as the situation changed, able to record messages asynchronously and then supply relevant messages to inquirers at a later time. The danger was that critical information would not be stored and located effectively and so not be correctly identified due to the sheer mass of messages generated. Achieving this was not possible through one hierarchical knowledge sharing hub. A self-organisation or small-world knowledge-sharing system was required; one that needed to use human memory and awareness.

Another simple but familiar wicked problem example may help. A university is made up of numerous groups undertaking research in their own discipline area. This typically involves small groups ranging in size from perhaps only two to laboratories containing 10 to 20 or more members. These groups know much the same ‘stuff’, the discipline-specific research methods, the literature and the worldwide experts in their field. View these research groups as small-worlds knowledge clusters. The strategic imperative, or common purpose some of these wicked system clusters may appreciate is the need for multi-discipline research to provide a comprehensive research effort to deal with wicked problems such as poverty, terrorism or natural disaster response. This common purpose may spread through the weak inter-cluster links. The Research Dean may also further encourage this concern by allocating increased resources to multi-discipline research solutions. Each cluster has specialist knowledge and any excessive attempt to insist all its members spend a significant proportion of their time getting to know other cluster’s research in detail may distract them from developing their own knowledge. However, these clusters do need to be ‘weak linked’ both with each other and with clusters knowledgeable about research resources. These weak links will need to be synchronous and asynchronous (stigmergy), using web pages, internal research newspapers, publication listings, signage, question-asking software like ‘askme.com’, web publishing of seminar PowerPoint slides, financial rewards and Listserv public acknowledgement of achievements. All of these are examples of the asynchronous (stigmergy) weak linking. Telephone lists by knowledge area, cross discipline coffee groups and conferences are examples of synchronous weak linking. The role of the Research Dean is merely to provide effective responses to those that do multi-discipline research compared to those that do not, and to encourage weak linking (not strong linking) between groups that would not normally even appreciate each others existence. Given the common purpose and the presence of weak links, the self-organisation perspective anticipates that members of the clusters will knowledge-share and self-organise an appropriate response for the improvement of the knowledge holding of the entire university.