ABSTRACT
All human systems are characterized by purposeful activity (Ackoff and Emery 1972), and it is clear that the phenomena examined here fulfi l this basic requirement of systems theory. In classic systems theory, however, the overall purpose of the system is assumed to be shared by the participants within the system and their interactions are aligned with this shared purpose. What we can observe in the case of the projects studied, however, is that there are many different purposes being pursued by the human participants engaging in the ‘interlocking behaviours’ that characterize human organizing (Weick 1969). This heterogeneity of purpose is diffi cult to incorporate into a classic systems theory approach, but CAS theory can accommodate this, in that agents with different schema are a fundamental aspect of CAS (Anderson 1999). Nevertheless, in order for a self-organizing system to emerge from the existence of heterogeneous agents, there needs to be something that causes the agents to interact. In the 12 cases studied, this impetus for interaction came in the form of policy objectives-whether
legislative or managerial-and the initial stages of systems formation were largely focused on clarifying those objectives that may be jointly pursued, along with establishing the ‘rules of the game’ (Koppenjan and Klijn 2004). Over time, the interaction of agents was also observed to result in self-organization, adaptation, path-dependency, bifurcation and emergence-key dynamics of CAS-each of which is present to varying degrees in the projects studied. Before moving into the CAS-specifi c dynamics, however, it is necessary to review what has been observed so far with respect to the core elements of our basic systems framework: systems; environmental factors; environmental rules; agents; processes; outcomes.
