Title: Organizational Attractors: A Chaos Theory Explanation of Why Cultural Change Efforts Often Fail
Abstract: Organizational change and development researchers and practitioners are concerned with whether or not the value systems found within complex organizations can be changed.(1) This issue has been highlighted by Moch and Bartunek's (1990) extensive case study of a change effort which was designed to institute a more participative working style in an organization with an autocratic working style. Essentially this was a cultural change effort. This effort failed and its failure was attributed to different cognitive frameworks, defined appropriate relationships, behaviors, and values within the organization. These cognitive frameworks were found to be very resistant to change on a deep level. Surface elements related to them could be changed slightly as long ago the changes did not violate beliefs held at the deep level. Attempts at deep-level change were strongly and successfully resisted by actors within the organization. This article proposes that the failures of such change efforts are better understood in light of recent developments in chaos theory and evolutionary theories of change for biological systems. The application of these developments highlights the nature of complex systems and the difficulty inherent in changing them at a deep level. This article will show how this application helps explain the role of organizational culture in organizations and the way in which organizational cultures change. Their application clarifies the importance of understanding the cultural context in which a change effort is occurring and provides preliminary guidelines for the organizational change practitioner. NEW DEVELOPMENTS IN CHAOS THEORY Massarik (1990) has shown that chaos theory has implications for organizational change and development in a discussion of different levels of chaotic change which may be seen in an organization. Chaos theory also has implications for why complex systems are not easily changed. The essence of the ideas underlying chaos theory is that, for complex systems, the exact prediction of events and causal relationships which is impossible on a micro-level is practical and possible on a macro-level (Gleick, 1987; Briggs and Peat, 1989). Therefore, while prediction of any one event may be difficult, the prediction of the general pattern of events that will occur within a system is predictable. Many complex, seemingly random systems have attractors which define the system on a macro-level (Gleick, 1987). The patterns of these attractors are defined in a n-dimensional Euclidian space. The prediction of the state of the system after Time 1 is possible for points close in time to this initial measurement. Even in the short term, however, the location of the system at any point in this space rapidly becomes random and unpredictable. This is termed the butterfly effect (Gleick, 1987; Briggs and Peat, 1989). Longitudinal study of a system is used to define the underlying pattern of macro-order which is represented by a shape within this space. In the long term, the possible responses of a system are defined by this shape. This is illustrated in Figure 1 (here in 3-dimensional space). Figure 1 shows the development of an attractor shape over elapsed time and repeated measurements. The system's attractor is represented by the final illustration in the sequence. The state of the system may shift from any point on this attractor at Time 1 to any other point at Time 2. Therefore, prediction of the state of the system in the short term remains difficult and appears random. The attractor, however, represents stable boundaries for the system. Long-term predictions may be made that the system will remain within the boundaries set by the attractor. These boundaries represent a feasible set of alternatives for change from Time 1 to Time 2. The system will be highly resistant to change from this attractor shape to another attractor shape. The system, even after drastic perturbations, will return to this attractor with high probability. …
Publication Year: 1993
Publication Date: 1993-10-01
Language: en
Type: article
Access and Citation
Cited By Count: 64
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot