CONTROL MECHANISMS (Problems in)
| Collection | International Encyclopedia of Systems and Cybernetics |
|---|---|
| Year | 2004 |
| Vol. (num.) | 2(1) |
| ID | ◀ 689 ▶ |
| Object type | Human sciences, Methodology or model |
K. BOULDING states (1952) “In organisations of any degree of complexity there is not a single variable to be controlled but a large number of variables, and the effectors which affect one may also affect others. If each variable to be controlled has an isolated effector which is capable of controlling it, the multiplicity of variables presents no problem. We do not in general find this to be the case, however. Action which is taken to influence one variable will almost always have some effect on the others. If the effectors are sufficiently confused, the problem of control will become insoluble: in endeavoring to stabilize one variable, we inevitably unstabilize others; acute cycles may be set up and the organization may even desintegrate” (1952, p.389).
These effects appear clearly in the systems dynamics models created later on by J. FORRESTER. Interactions between effectors, simultaneity of effects in different parts of the complex system and time lags (delays) are very difficult to modelize satisfactorily… and to control.
As a result, in L. DOUGLAS KIEL words: “The traditional ”engineering“ view of public policy interventions intended to produce desired outcomes appear to better represent an abstract Newtonian vision, rather than the uncertainty of actual experience with such efforts. In short, the dynamic and nonlinear relationship between many variables in social systems reveals that policy decisions result in an enormity of unexpected outcomes” (1992, p.38).