The Neglected Phenomenon of Negative Feedback Control

In a commentary in Nature, Rodney Brooks1 proposed that something is missing from our models of living and behaving systems. I would like to suggest that it is not something undiscovered that is missing, but something old that has been passed over without sufficient examination. What is missing from most “modern” conceptions of behavioral mechanisms may be a sufficient understanding of a remarkable phenomenon called negative feedback control, reduced to a formal theory over half a century ago. Systems organized to carry out negative feedback control behave in a way that a great many scientists do not believe is possible. Given a specification for some state of affairs, they can continue to produce or reproduce the specified outcome even though the actions needed to do so vary from one moment to another. The actions of such systems are of the type that has been termed purposive, in that they appear designed to achieve some specific predetermined end. They are also of the type that has been termed adaptive, for such systems are able (within limits) to vary their actions in just the way needed to continue to produce a particular outcome despite changes in circumstances. In contrast, what most life scientists seem to believe in can be termed a causal system. A causal system mediates, stands between, causes and effects. The effects created by a causal system are those dictated by its physical structure and external forces or other influences acting on that physical structure. If circumstances change, the effects necessarily change, either because the behavior-causing external forces and influences change or because the structure of the system is changed by other forces and influences. What we see a causal system doing corresponds to what is being done to it; its “actions” are more properly called “responses,” for no action of a causal system takes place without an adequate prior external cause or stimulus. The Neglected Phenomenon of Negative Feedback Control