Behavior Mod./Coercion

[FROM: Dennis Delprato (950829)]

Darlene Todd (950825)

Darlene, welcome to the amazing world of perceptual control theory.
I say amazing because behavior analysts and just about everyone else
(no doubt including yours truly here) have a devil of a time "going
all the way" with how PCT approaches behavior. It seems to me that
of all the well-developed literatures in behavioral science,
behavior analysis may have come the closest to PCT. A useful review
paper would be to examine the many places in which one finds "feedback"
and even "feedback control," "control systems," and "feedback functions"
in the basic and applied literature of behavior analysis. Yet,
behavior analysts have not taken the final step of discarding one-way
(environmental) causality in favor of circular causality. The basic
explanatory device in behavior analysis, functional relations as in
behavior = f(x), always takes x as an antecedent cause. This is so
even with the currently in vogue "selection by consequences"

If your professors are "liberal," they may permit you to study
the work of an enigmatic figure who a few have dragged into
behavior analysis, not that he considered his work to be another
stream of behavior analysis. I am referring to the "interbehavioral"
literature inspired by J. R. Kantor. Kantor's ideas are not well
liked by mainstream behavior analysts. One basic reason inter-
behavioral thinking is anathema to many behavior analysts is
because it requires circular causality such as functional relations
with _interchangeable_ (gasp!) variables. Interbehavioral psychology
and philosophy can be a useful first step into PCT, especially for
one with leanings toward today's behavior analysis. It is certainly
not a necessary step as indicated by the fine work of Bruce Abbott.

Rick Marken (950825)

If they are, why would they believe that there is
such a thing as "coercion"? You can't coerce an entity whose behavior is
controlled by its environment. A behaviorist who uses reinforcement toget a
kid to eat spinach is not coercing the kid (from the behaviorist's
perspective) any more than he would be coercing a broom to sweep by exerting
force on its handle.

But -- what exactly is the essence of "coercion"? The word itself can apply
to the effect that punishment, negative reinforcment, or even positive
reinforcement has on behavior. Perhaps the reason operant and respondent
behaviorist resist using such terminology is that it is a little fuzzy.
Skinner sought to operationally define behavior analysis, and a term such
as "coercion" may be hard to operationally define (i.e., which procedure
are we referring to?).

Behavior analysts certainly have expressed concern over coercion.
Regardless of the details of definition, they are not going to go
around saying that they don't care about coercive practices or that
they endorse such practices. Skinner has written quite a bit that
deals with coercion and how to avoid it at an applied level. I
believe if he would have been asked if his utopia, Walden II,
minimized coerciion he would have said that it was designed to
do so. The problem is that the few behavior analysts who have
concerned themselves with coercion have only used observations
of its fallout to detect the harmful consequences of coercive
practices. Being good observers, they observed what Skinner
astutely called countercontrol. However, not understanding the
behaver as a control system, they had no idea of what is behind
the countercontrolling behavior. Their explanation is that
countercontrolling behavior is usually maintained by negative
reinforcement because coercion is aversive and the counter-
controlling behavior is followed by reductions in level of
aversive stimulation. Even if this is granted, why is coercion
aversive? Here conventional behavior analytic analysis grinds to
a halt.

As Rick Marken implies in the above, to the behavior analyst
there really isn't coercion--all behavior is forced by the
environment or history of reinforcement. Coercion comes down
to "aversive control." The practical thing to do is to avoid
aversive control as much as possible. Until one gets to PCT,
one treats coercion as aversive control, or even more generally
as a moral, social, or political issue. Coercion has no place
in the science of behavior when the science takes one-way
causality (be it environmental, mental, or chains consisting
of both sources--witness currently popular cognitive-behavioral
theories) as a fundamental given.

Dennis Delprato

[From Rick Marken (950829.1150)]

Dennis Delprato (950829) to Darlene Todd (950825) --

Absolutely EXCELLENT post Dennis.

Heck, you didn't leave anything for me to say (which should make you one of
the most popular people on CSGNet;-))

Really, nice work.