[From Rick Marken (930423.0900)]
Ken Hacker [930422]--
It is tempting for those who work on new theories to see those
theories as more explanatory than they really are.
This is certainly a possibility. That is why Tom Bourbon and I
are anxiously waiting for you to submit what I presume will be
some research on an interpersonal communication phenomenon that
cannot be handled by PCT. When we discuss PCT on the net, we
feel free to speculate (fairly wildly) about its possible
applications. But I think you will find (especially if you read
Tom's research articles and my "Mind Readings" collection) that
our "off-net" claims about what PCT can explain are restricted to
the results of the research we describe; and we demonstrate the
explanatory power of PCT by showing that the model behaves just
like the human subjects. If there are behavioral phenomena that
PCT cannot explain in this precise way -- which other models
can -- we REALLY want to know about them. So far, we PCT modellers
have not been able to find ANY such phenomena -- but maybe we are
looking in the wrong place. So you could make a significant
contribution to limiting the grandiose claims of PCT modellers by
presenting just one phenomenon that can be precisely explained by
a non-PCT model and not by a PCT model.
Of course, the more careful Perceptual Control Theorists note that
PCT does not actually undo much of what social and behavioral
scientists have done as much as it forces those scientists to
rethink what they have discovered.
What if the result of this "rethinking" is the discovery that PCT
invalidates most of what social and behavioral scientists have done?
Should these "careful PCTers" keep rethinking things until they come
to the right conclusion (which, I presume, is that the discoveries
of behavioral science need rethinking)? Looks like a potential
infinite loop to me.
Of course,
if A, then B reasoning does not, as he suggests, necessarily imply
causality. There are differences between necessary and sufficient
construct relationships. If A, then always B, despite conditions is
causal and deterministic. If A, then B, can be reversible and stochastic.
I think you'll have to walk me through this again, Ken. Bill made the
same point I've been trying to make for years: the IV-DV approach
to research assumes that DV = f(IV) where f() is a charateristic of
-1
the organism. PCT shows that, for a control system, DV = g (IV) where
g() is a characteristic of the environmental connection from organism
to controlled input variable. If PCT is right about this, the basis
of all behavioral science crumbles. So your paragraph above is pretty
important because it seems to be arguing that IV-DV research is an
appropriate approach to understanding behavior under certain conditions.
I think the only condition that could make IV-DV research appropriate
is that organisms be "open loop" systems. Is this what you are saying
above?
In our discipline, many if not most researchers reject simply causal
explanations of any human behavior. For example, the ones who study the
connections of cognition and communication assume that social behaviors
are indicants of cognitive activities.
This is still a causal model; internal stimuli (cognitive activities)
lead to outputs (social behaviors). Control looks this way to the
third blind man in my "Blind men and the elephant" story. Social
behaviors are indicants of cognitive activities (reference settings)
but PCT shows that you can't tell what constitutes an intended
social behavior just by looking -- because it is perceptions that
are the result of cognitive activities, not outputs. So you have
to do a version of The Test in order to know whether your perception
of a social behavior matches the actor's perception thereof.
They are not studying that stimuli from social behaviors produce
cognitions.
That sounds great; but I would really like to hear a fairly detailed
description of how they study cognition-produced social behavior. I
have a suspicion that they are doing it in the framework of IV-DV
methodology. This was Bill's point; sure, there are many people
(especially the cognitivists but also the interactionists and
whatever) who say "we're against S-R". But the proof is in the
research; it's SR all the way. I can think of exceptions (Piaget,
some nice studies of problem solving based on model testing); but,
by and large, what you find in the journals I look at (which
purport to be very cognitively oriented) is IV-DV research with
high falutin' "non SR" verbalisms attached.
The point here, is that Powers makes a powerful (sorry...) argument
about MANY behavioral scientists still working in an S->R paradigm,
sometimes unknowingly. However, the generalization about all of
them doing this just does not match what people are doing and writing.
I would really like to hear about some research in your field that is
being done in a non IV-DV framework. Just show us.
Best
Rick