Chris:
As you can see below, there is still some discussion on CSGnet about the
differences between "conventional" human factors approach to behavior and
Powers's Perceptual Control Theory (PCT) approach.
I was hoping I might be able to get you to clarify your position on this.
The PCT approach specifies that we achieve repeatable outcomes requiring
different motor outputs by sending reference signals to lower-level control
systems. These reference signals are not motor programs but rather
specifications for PERCEPTION. The result is that PERCEPTION is controlled
by varying outputs as needed in the face of disturbances. Output varies as
it must to compensate for unpredictable disturbances and therefore cannot
be programmed or planned. It is the input (perception) which remains
controlled. Is your "motor schema" what PCT sees as a reference signal?
You can see how Powers models this in Demo2, "Closing the loop" where you
can vary the reference signal and see the lower system respond so that its
perceptual signal matches the reference signal in spite of continuous
disturbances.
Personally, I am more interested in more "cognitive" aspects of psychology,
such as language acquisition and use and find that Powers's PCT model has
much potential in these areas although some pretty extensive stretches of
the imagination are presently required to apply it. I am however, quite
bewildered about why his theory is not widely known or used in human
factors research where much less stretching would be involved. If the
experts in HF research discount the PCT model for good reasons, then I
should perhaps better look elsewhere for a psychological model of language
and cognition.
Marken maintains that the problem is that PCT shows that what we see as
behavior is really an organism controlling its perception and that this is
a major paradigm shift from the normal behavioral-science view which sees
behavior as being caused by (or even guided by) perception. If the PCT
view is correct, then (according to Marken) much of psychological research
and theorizing which considers stimuli of some sort as independent
variables and behavior as dependent variables is misleading or at best
useless. It may LOOK like behavior is caused by stimuli, but the PCT
perspective reveals this as an illusion, in the same way that the
heliocentric perspective of the solar system reveals the "rising" and
"setting" sun as an illusion.
Williams maintains that you and some other "conventional" HF researchers
really see things in much the same way as PCT, but have different meanings
for words such as "control" and "behavior."
Would you be able to help us out here?--Gary
···
=========================================================
[From Rick Marken (931028.0900)]
Greg Williams (931028) --
Quoted from Christopher D. Wickens, ENGINEERING PSYCHOLOGY AND HUMAN
PERFORMANCE, Second Edition, HarperCollins Publishers, New York, 1992,
page 451:
"...Thus whatever is
learned and stored in long-term memory cannot be a specific set of
muscle commands but must represent a more generic or general set of
specifications of how to reach the desired goal. These specifications
were labeled by Bartlett (1932) and Schmidt (1975) as a MOTOR SCHEMA."
So a "motor schema" is a "specification of how to reach the desired
goal". Does this sound close to PCT to you? Even if we are kind and
assume that a "motor schema" is a reference signal (the signal that
DOES specify the OUTCOME in PCT) then why call it a specification of
how to reach the desired goal. A reference signal is not a specifcation
of how to reach the desired goal -- it is a specification OF the
desired goal (a particular state of a perception, another little detail
that Wickens forgot to mention). It is both a tenet of PCT and a FACT
OF LIFE that it is impossible to specify how to reach the desired goal;
so the statement not only seems inconsistent with PCT, it is inconsistent
with fact. Was this just a slip of the tongue? Is there a description
of how the "motor schema" model works -- even in the "trivially simple"
case of a compensatory tracking task?
Or am I supposed to be impressed with the fact that " A number of
investigators have pointed out that what is consistent is not the
PROCESS of muscular innervation (and therefore the specific pattern of
neural commands) but the PRODUCT of the response". It's nice that
they noticed this phenomenon. Now what? Looks like the "motor
schema" theory (really, description) of this process is nothing more
than a hand waving attempt to preserve the notion that behavior
(products now, rather than processes) is the RESULT of a LINEAL
CAUSAL PROCESS that begins in the brain (as a "motor schema").
I'm sure you consider this unfair criticism. Can you explain
why what you quoted should be taken as anything other than a flat
out failure to understand the nature of control and PCT?
Perhaps we could just focus on the similarities (bridge building).
Good for Wickens et al for noticing that variable means are used
to produce consistent outcomes. Now what? Should we mention that
we have a way to explain this phenomenon? Or would that be impolite.
Best
Rick