Feedback

Feedback is most usefully understood in considering the behavior of complex dynamic systems and in problem solving related to such systems. Feedback expresses the concept of circular causation postulated to operate in dynamic systems and is particularly related to one of two fundamental elements in such systems, namely the informational content of such systems, the other essential component in that description being the material states or elements in those systems. The term came into wide application with the development of the discipline of systems dynamics in 1971 by Jay Forrester but has been shown by George Richardson in 1991 to have originated much earlier, as far back as third-century Greece. Systems dynamics itself refers to the syntax and mathematical modeling techniques developed by Forrester and his students to describe and analyze the behavior of any kind of complex system. Complexity is understood both as detail complexity—referring to the sheer number of elements operating in a system, and dynamic complexity—incorporating the notion of reciprocal causality in determining the over-time behavior of the entire system or of a particular subset of variables. Since systems dynamics is content free, it is applicable to any system. This feature allows the methodology to be applied to any kind of over-time behavior of variables of interest that results from the complex interaction of a set of material and informational flows representing a coherent set of dynamically related elements.

A more general reference to systems dynamics (as opposed to system stability) was initiated with the work of Ilya Prigogine in 1980 and 1984, in which it was shown that open systems—those that are in constant and reciprocal relationship with their environment—import energy from their environment and export tendencies toward entropy. Instead of focusing attention solely on the role of (negative) feedback processes to maintain system stability, attention could now be turned to understanding how, for open systems, this relationship between a system and its environment could be used by that system for the purpose of achieving growth and change. It is now well-known that this property of open systems serves as protection against disorganization and eventual decline. As evidenced by the recent emphasis on seeing schools as open systems, a focus on feedback and the role it plays in problem structuring in educational reform and more generally in facilitating organizational learning in schools and school systems becomes important for this volume concerned with educational leadership.

Underlying the systems dynamics approach to analyzing system behavior or problems in the behavior of particular variables in social systems is a distinctly different ontological perspective. Traditional views of explanation still rely on the use of linear models, in which independent variables explain dependent variables by summing the effects of the independent variables. Alternatively, the function connecting the explanans to the explanandum is linear or can be made so by some suitable transformation. Dynamic systems approaches, however, operate from a different philosophical orientation, namely, the view that systems ultimately consist of stocks, otherwise called accumulations, on the one hand, and flows between these stocks, on the other, and that focus must be placed on the change processes underlying systems behavior, not only or even mainly on system structure. It is the full listing of accumulations or stocks that at any moment describes the state of the particular system. The combination of a system state with the information and material flows of the system and related system goals, the latter either referencing their own or exogenous (desired) levels, describes the relevant system, which when given initial values and then set in motion via some appropriate modeling device allows the system to exhibit the specific over-time behavior patterns observed. These behaviors result from the ways in which stocks are related to each other, with a specific polarity-positive or negative in each case, together with information flows and delays in how quickly changes in specific stocks affect other stock changes. Positive feedback behaviors move the system away from equilibrium, while negative ones restore the system to some equilibrium level when the latter is [Page 384]disturbed from an original position. The totality of the feedback behaviors in a system constitutes the feedback structure of that system. Over time, this feedback structure leads the state variables of the system to exhibit time trends that are either of an exploding or damping variety, or to oscillate in cycles of constant or varying amplitude, the latter in the exploding direction or in a damping set of cycles.

...