Chaos theory, a philosophical viewpoint from physics, is being adopted by some public organization theorists as a new standpoint for understanding and managing organizations besieged by instability and turbulence. A basic tenet of chaos theory, known as mutual causality, denies that an entity behaves or evolves in isolation. Rather, acting, behaving and evolving can only occur through mutual interaction with other entities (Briggs and Peat 1989). Concomitantly, any managerial theory of action consistent with chaos theory must be grounded in the concept of mutual, not linear, causality.

I contend that in the nascent transdisciplinary journey by which chaos theory is moving into public administration as a metaphorical explanation of public organizations, the theory is in danger of being corrupted and bastardized. This is suggested by the fact that a number of managerial applications being inspired by chaos theory promote conventional, linear approaches to action aimed at command and control. Others present only an organizational-level application--rather than the holistic approach that chaos thinking entails. Such linear, bounded applications are inconsistent with the principle of mutual causality. I argue that as a corrective to this emerging trend, public organization theorists need to adopt a more complementary frame to chaos theory, viz., family systems theory, in order that the co-determined nature of human dynamics can be adequately brought to the level of practice.

I present two case studies. The cases illustrate the difference between managing with a theory of action grounded in the principle of mutual causality and managing based on a theory of action undergirded by linear thinking.

The complementarity of chaos theory and family systems theory offers managers of public organizations a more holistic understanding of processes in their organizations. Hence, public administrators who adopt this new, conjoint theoretical viewpoint can develop a different way of behaving in and effectively managing their organizations.