PCT: What's important

[From Bill Powers (921204.0830)]

The most important concept of PCT is not control of perception or error,
but purposive behavior.

One of the main things that PCT does for us is to get intention and
purpose back into theories of life in a simple form. People like
Brentano and Searle and many others have tried to deal with
intentionality under the assumption that the folk meaning has been
disposed of. So they have come up with vague ideas like "aboutness" and
"directedness" to explain this sense of doing something on purpose,
intentionally. They have purpose (an active process) mixed up with
attention (a form of selective but passive observation). I've read a bit
of the philosophy of intentionality -- it keeps cropping up -- and my
impression is that these people are groping in the dark. Another case of
tossing aside the nuggets on the surface and looking at the ground where
they were for the obscure answer.

The original meaning of purposive or intentional behavior, the one that
conventional science eventually threw out, was that organisms can select
a future state of the world and act to bring it into being. We drive
cars to work in order to arrive at a parking place and get to work on
time. We mix ingredients together with the intention of baking a cake.
We write papers for the purpose of publishing them.

All examples of purposive behavior like these have been argued away by
scientists who didn't know how to explain them. Part of their difficulty
in accepting phenomena of purpose came from flaws in the folk concepts
of purpose and intention. People have the idea that if you have a firm
and forceful intention, it will be more likely to succeed than if you
hold the intention "weakly." I had a dog who had the same idea. When she
wanted to go outside, she would stare fiercely at the back door, and
later at the doorknob, obviously willing the door to be open as hard as
she could. And of course it always worked, eventually, because the
humans in the house liked the effort of getting up and opening the door
better than the alternative. But the dog evidently never connected what
the humans did with what the door finally did. Other dogs would bark or
run to find a human being. This one had a different theory of purpose.

From similar sorts of experience, people have got the idea that purposes

and intentions have some sort of supernatural effect on the world, even
on events that they can't affect by their actions. So this leads to
belief in magic of various kinds -- no wonder science was put off by the
commonsense concept of purpose.

One of the arguments often brought up against the idea of doing things
intentionally (when the debate was still going on) was that intentions
don't always work out. Between the "intentional action" and the outcome
it was supposed to produce, all sorts of accidents can occur to prevent
that outcome. The position against which this argument was presented was
assumed to be that purposes and intentions, by their nature, had some
magical assurance of succeeding, as per the folk and canine

Another argument was that the future effects of purposive behavior would
have to act backward through time to cause the behaviors that led tothem. Behind
this argument was the assumption that all behaviors are
caused by external effects on the organism. Purposes were thought of as
future events stimulating the organism in the present in a particular
way, so as to create the future event. This, too, was assumed to be the
folk concept of purpose, and perhaps it was. People are not generally
very clear about what they mean by "the future."

The idea of intentional actions carried the implication that there was
something different between an intentional action and some other kind of
"ordinary" action. In either case, both science and common sense
considered behavior in terms of action-events that had future
consequences. Do something now; later on something else happens as a
result. The implicit concept was that once you're performed a causal
act, natural processes are set in motion that can no longer be altered.
This concept was behind many objections to purpose, especially
objections that raised the issue of unsuccessful purposive acts.

Control theory would suggest that there is no peculiar quality that
makes one act purposive and another not purpose. It is not the action,
but perception of the outcome that is purposive. Control theory also
shows that if some outcome is to be produced on purpose, it is necessary
to form an inner perception of that outcome, a reference signal, and act
continuously until the perceived outcome matches the reference signal.

Purposes or intentions, as explained by PCT, have no effect on the
external world. Only actions affect the external world. A purpose, even
though it feels directed outward, is really a specification for input.
Only action can alter the perceived world to bring it nearer to the
specification (if th4e world doesn't spontaneously cooperate, by luck).

There is no effect of the future on the present; the reference signal is
a present-time phenomenon that is carried through time in the actor.
Behavior is not a series of events, but a continuous process that is
updated as often as necessary to assure progress toward the intended
state. The actions performed on the way to achieving a purpose will
change with every disturbance that tends to interfere with progress in
the right direction. The only way for an intention to fail to be
achieved would be for external events too powerful to resist to overcome
the organism's maximum opposing effort.

So PCT offers a clear explanation of purpose and intention, one to which
the old arguments simply don't apply. Yet PCT also explains how the folk
concepts of purpose might have appeared to be true under pardonable
misapprehensions about the way the world worked. The mistake made by
science was to argue against bad theories of intention and then assume
that these arguments got rid of the phenomenon of intention, too. The
tacit assumption, also, was that if science in the early part of the
20th Century couldn't explain the phenomena of purpose, they must not
exist. The hubris of our forebears was limitless.

I suggest, therefore, that arguing about whether error signals or
perceptual signals or reference signals are the essence of behavioral
organization is a side-issue of little importance in comparison with
what is really important about PCT. In effect, PCT goes back about 80
years and settles an argument that everyone involved in it at the timesettled
incorrectly. All the philosophical thinking about purpose that
has taken place since then, if it did not include the phenomenon of
control, has been vacuous and amounts to nothing. Vast volumes of words
on this subject can now be consigned to the wastebasket or the delete
key. If we study such words at all in the future, it will only be to
learn how people of good intent can use words to fool themselves into
thinking they understand something.



Bill P.