»   The Shape of Complexity»   Feedback Loops»   The Fitness Landscape and Co-evolution»   Paradigm Shift

Feedback Loops

Negative Feedback Loops:  Management systems developed for organizations in the  industrial economy focus on processes of standardization, specialization, economies of scale and expert knowledge. Information is gathered, defined, analyzed and utilized by experts in specialized functional areas. It is applied to the necessary and beneficial processes in order to identify the most efficient and productive techniques and to minimize variances from the standards imposed. Feedback loops are created to identify anomalies and provide opportunity to eliminate them. The objective is to achieve and sustain an effective “equilibrium” state. The requisite knowledge and models are assumed to be available through these experts who can access it and provide definitive direction. Once in place, these standards are expected to be conformed to by all members of the organization until improvements initiated by the appropriate experts supersede them.

In this model, the role of leaders and experts is to direct, control, reduce conflict and set direction. Their goals and “vision” are assumed to be sufficient for organizational direction and motivation. Organizational efficiency and effectiveness are a matter of figuring out or seeking out “best practices” and pushing this knowledge down through the organization. This methodology has been very powerful and it has been the engine for success for countless companies over the past several generations. However, this linear model is predicated on an environment of slowly emerging technologies, the predictability of stable markets, and the rationality of a highly informed observer. Many companies today, including most technology firms, exist in a competitive environment where technology advance and social networking drives rapid product obsolescence, competition comes from unexpected sources, and complex non-linear dynamics reduce all models to partial and highly tentative.

Today’s competitive environment is turbulent. Advances in communications technology drive customer expectations and demands, create rapid product obsolescence, and promote competition from unexpected sources. The norm today is dealing with the non-routine situation rather than the predictable and repeatable processes that can be controlled with negative feedback loop corrections. This demands an ability to respond based upon available information and resources rather than a thoroughly designed and tested model.  

In this environment, the ability to create, capture, and share ideas and information, and leverage these into new knowledge and innovation that can be broadly useful to the organization is critical in adapting to changing markets, to changing resources, and to changing sources of competition. To accomplish this, we need a model that captures the essence of nonlinear dynamics.

Positive Feedback Loops:  In a non-linear system, not all processes have controllable variables. Within organizations, changes in behavior create variances, but many of these variances are not subject to the control mechanisms of negative feedback loops. Although there may be control mechanisms that attempt to eliminate unexpected anomalies (such as an individual’s cultural assumptions that are unsupported within the organizational system) within the “legitimate” system, every organization also contains a “shadow system” that is subject to a different set of rules that factor in flows of emotion, friendship, trust, and other qualities. Interactions within the shadow system take more diverse forms than in the legitimate system, and tend to be nonlinear, according to varying degrees of uniformity versus diversity, conformity versus individuality (Stacey 1996).

Those variances which are attractive enough to collect a consistently growing constituency eventually amplify the variance to a degree that penetrates the legitimate system and destroys the status quo (Maruyama 1963). These attractors can exist in any of the embedded levels. Where they arise is unpredictable. And as they induce movement in one level - movement along a fitness landscape (Kauffman, 1995) to increase the likelihood of success or survival - the impact upon that level and on other levels is also unpredictable. This dynamic is a cogent model of Schumpeter’s (1934) concept of “creative destruction.”

Feedback is positive when it enters the discovery-choice-action loop in a way that amplifies and destabilizes it. Because of this, small events sometimes blow up into large consequences.  It is in the interest of the competitive organization (and of competitive individuals!) to encourage and sometimes even accelerate variances that lead to changes that promote achievement of its goals and objectives. We cannot “make” this amplification happen. It is a condition of the tensions present in the environment contained within the system. Likewise, we cannot stop negative or dysfunctional variances from gaining a constituency. We cannot legislate them out of existence or declare them invalid by fiat. We can only attempt to identify the drivers, often intangible, that promote the disruptive constituencies and try to influence their presence within the system.

The important point for our consideration of governance that the principles espoused by any organization, the degree to which the organization adheres to them, the modeling of those principles by key leaders, and the visible system for distributing rewards are all crucial drivers of behavior regarding the "discovery-choice-action" process within the system. Those drivers construct the mental models consisting of the regularities that an agent has perceived in the complex world it has to operate in, and these models form the basis for the behavioral rules for operating in that world. This is akin to what Kuhn (1970) called a paradigm, Huff (Mapping Strategic Thought, Wiley:1990) labeled cognitive maps. and Stacey (1996) refers to as a schema

Individual agents, whether a company or an individual, select what phenomena to consider in their “discovery” process. The meaning that is attached to the information gathered within that consideration is based upon the implications of those models.  In the case of individuals, we can consider these models the cultural assumptions, norms and values based upon all the social and psychological aspects of their previous and current experience. Each adaptive action taken by an agent, determined by such model, has an effect upon other agents in the system, and perhaps on those in linked or embedded systems, leading them to respond and causing effects that feed back again into the first agent’s discovery process.


For references, see The Library

The Governance Dynamics Institute is a tax exempt not-for-profit organization under section 501(c)(3) of the United States Internal Revenue Code.

  2020 Governance Dynamics Institute
Website Design by James Web Design LLC