[From Rick Marken (2001.08.09.2150)]
So far there's only been one comment (Bill Curry's very incisive
observation about "boids") re: my "Control Theory Glasses" paper.
I really would like to hear other comments and/or suggestions. Does
it seem like a good way to present the PCT view of behavior?
Best regards
Rick
--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: marken@mindreadings.com
mindreadings.com
[From Kenny Kitzke (2001.08.10)]
<Rick Marken (2001.08.09.2150)>
<So far there's only been one comment (Bill Curry's very incisive
observation about "boids") re: my "Control Theory Glasses" paper.
I really would like to hear other comments and/or suggestions. Does
it seem like a good way to present the PCT view of behavior?>
Of course, I gave you my main comment personally at the conference. But, I
shall share some comments with the net.
I liked the way you explained how, depending upon what psychology theory
glasses you are wearing, you see different things when you observe and
experiment and try to explain what people are doing.
And, the rather non-confrontational approach of not directly maligning the
other glasses may get you published, which may be your goal along the way to
a higher purpose?
I will be eager to see if it is published and how many readers it influences
to visit the web sites, run the demos and test for the controlled variables
essential to seeing behavior through PCT glasses. It just might influence
some minds to try some new glasses!
I applaud your contribution. At the current rate of gaining PCT acceptance
as science, some new attempts and approaches seem in order. And, isn't that
PCT in action?
Kenny
[from Jeff Vancouver (2001.08.14.1200 EST)]
Rick,
You asked for feedback on your glasses paper. Naturally, I have some.
First, I found it potentially very interesting. I am not sure what
perceptions you are trying to move with this product, but my feedback
regards increasing the impact that it might have in a scientific community.
So take it for that value.
The first issue relates to the label for the non-control theory lens. By
referring to it as causal, it implies that control theory is non-causal.
This is not the case. In control theory, causality is circular. It seems
the issue is more about open-loop vs. closed loop or linear vs. circular
causality. This notion that control theory speaks to causes, but in arcs
that connect in a loop, is critical. One of the main points I have said
before, and it is a major issue in this paper, is that I believe many
(particularly more scientifically-oriented) psychologists are aware that
behavior emerges from a circular causality process. Yet, that does not
prevent them from examining specific hypotheses about arcs in that loop.
Hence, you must demonstrate that current theories are either a) missing the
point by looking at what they think is an arc, or b) missing the point
because they are not even thinking in terms of the arc (i.e., there is no
circularity acknowledged in the body of work). By "missing the point," I
mean they are making incorrect inferences and creating theories with poorer
predictive or proscriptive power than a control theory lens would product.
As it stands, you claim that "the dominant prescription in psychology has
been and remains the causal prescription." Yet, a psychologist can easily
dismiss that claim (if it is translated to open loop). For instance, Kurt
Lewin, who is considered the father of social psychology, was into
cybernetics, and Herbert Simon, who is considered an elder if not the father
of cognitive psychology, was as well. They may have gotten some details
wrong, but that requires picking on the details, not the perspective.
Further, you do little to counter this issue. For instance, the Mother
Goose example was great, but Lorenz and Tinbergen made their explanation in
1938. The question is whether anyone still buys that explanation today.
Using Social Science Citations index you should be able to find modern
references to it. If you can find such a reference in a well-respected
psychological journal (like Annual Review) that speaks to the (positive)
legitimacy of the explanation they gave, that would be powerful. I actually
think that you will find modern references of that type. I am not arguing
that all scientifically-oriented psychologists or their theories reflect the
circular causality understanding. But reviewers will dismiss your work as
irrelevant if you do not put it into the current context.
Toward that end, you do cite Bargh (and somebody). As you know, I have
found Bargh's work a good target. It is current, popular, but causally
misguided (probably). Pick on him.
Interestingly, I think the issue of the boids illustrates my point. It was
pointed out to you that the boids now include negative feedback loops. That
was "a seen cockroach" to your argument. That is, it turns out that the
"causal" lens is no longer the perspective used. The potential is that
there are 10 more cockroaches hidden from your view (e.g., Lorenz and
Tinbergen's explanation was replaced by yours or one like it, with data,
years ago, but no one pointed this out to you - it was hidden from your
view).
Again, I know that the argument of the loop still rages in the motor control
literature. Use it (note however, that PCTers are not the only ones arguing
the closed loop side). Loop structure certainly rages in my literature. I
use it (my modus operandi is to find someone prominent saying control theory
cannot or does not explain this and demonstrating that they are wrong).
That is why I think the fact that Bill P.'s model reproduces the squiggle
that so intrigued me at the conference is so cool. I suspect, and talking
with others has confirmed, that we would expect a more complex mechanism at
work. That we are wrong is cool to us. Now we just have to find out if my
friends and I are just dupes, or if this mistake can found in the current
psychological literature.
That I was duped, by the way, is not surprising given your Willie Mays
example. You acknowledge in the end that the catch was miraculous because
there was no input in the end. This is one of the types of problems I am
trying to deal with: how do control systems handle missing inputs. You may
not like my explanations because they are unnecessarily complex. Fine,
demonstrate with simple models.
{I do not like the subheading "getting the point" - it is pejorative.)
Simon's view was that the control models are useful and simple, but only go
so far. He was trying to understand the role of thinking and running
through if-then programs (or productions). Important in that issue was
conscious experience and attention. Hence, the question that your balancing
acrobat raises is: is it true that the wire balancing requires conscious
experience that is not normally required, and if so why?
In your description of TCV, I think you overstate how to interpret the
results. You do not want to say that a failed test means the control theory
lens is wrong, even if you correct what you mean the next sentence (i.e.,
just the hypothesized variable). Also, for the reasons Runkel (1990)
mentioned (e.g., the reference level is changing, reorganization is taking
place, gains are reduced to sequence perceptions), you convey much too much
confidence in negative results of the test.
Let me use your TCV in the real world as an example. Suppose that you got
close to another such that you actually were in their personal space.
However, conflicting control perceptions like not wanting to offend overrode
the desire to reestablish the space perception. The presence of confounds
is a constant concern. {you also used "himself" in that paragraph - APA will
not let you do that.)
Finally, I think your discussion of how S-R models are really negative
feedback loops is telling. I think one response to that by presumed,
non-control theorists is "of course, we know that, what is your point?"
The point, which you obfuscate in the discussion of legitimacy of the
glasses, is to make better predictions or proscriptions. I know you bristle
at this description of science, so let me say is another way. The point of
developing a better understanding of human behavior is so that we can better
get along with each other. That means we will control perceptions of
environmental variables that are less likely to disturb higher-level and
intrinsic variables that others are controlling. That is, we are trying to
reduce the conflict among the control systems within and between
individuals. [As an aside, you played the Roman game of first citizen with
the legitimacy issue. That is, in one sentence you say the views are
equally legitimate and in the next you talk about figuring out which is more
legitimate.] Given the practical or applied end of science, the question
is whether a control theory lens will lead to more accurate specific
predictions or more effective specific proscriptions (and that these are
replicable). That is, the question is does the control theory lens lead to
more useful understandings.
This question has too parts which I think you have not fully realized. The
first is that the word "more" implies a comparison understanding. You claim
that the "causal" lens is a different understanding, but you do not back
this up with _current_ documentation, and you do not demonstrate better
prediction or more effective intervention. What I get a lot is "Oh, I think
you are correct, but how can we use this to help the human race." This gets
us to the applied thread, which I have not yet looked at.
If your paper is meant for the masses, ignore everything I said. I am sure
that I will find it useful for my undergraduates and even early graduate
students. The simple, linear thinking is rampant in these circles. But I
have found, through the pain of well-articulated reviews of my work, that
the assumption that serious psychologists are not aware of the legitimacy of
circular causality is wrong.
Later,
Jeff
[from Norman Hovda (2001.08.14.1610 MST)]
[from Jeff Vancouver (2001.08.14.1200 EST)]
[As an aside, you played the Roman game of first
citizen with the legitimacy issue. That is, in one sentence you say the
views are equally legitimate and in the next you talk about figuring out
which is more legitimate.]
This troubled me a bit during my first read as well.
On my second read I now am disturbed by the "miracle catch" claim.
Does not such a reference to the miraculous do a disservice to your
effort to offer more legitimate explanations? <g> Would it not be more in
keeping with your intentions to offer another possible PCT example?
Could it be that Willie's control system brain may have *estimated* the
trajectory of the ball based on all of the previously observed perceptual
data inputs and placed his mit output at a projected *reference* position
in an attempt to intercept the ball... just maybe... and it happened that
the mit's location was correct (on this propitious occasion) so that the
ball was actually caught?
Best,
nth
[From Rick Marken (2002.09.06.1030)]
In case anyone is still interested in PCT my paper on PCT vs conventional
psychology, called "Looking at Behavior Through Control Theory Glasses" is
now available in the "Review of General Psychology". You can get to the
abstract from the table of contents at:
http://www.apa.org/journals/gpr/currentTOC.html
Best regards
Rick
Cool!
-----Original Message-----
From: Control Systems Group Network (CSGnet)
[mailto:CSGNET@listserv.uiuc.edu]On Behalf Of Richard Marken
MindReadings.com
Sent: Friday, September 06, 2002 1:26 PM
To: CSGNET@listserv.uiuc.edu
Subject: Control Theory Glasses[From Rick Marken (2002.09.06.1030)]
In case anyone is still interested in PCT my paper on PCT vs
conventional
psychology, called "Looking at Behavior Through Control
Theory Glasses" is
now available in the "Review of General Psychology". You can
get to the
abstract from the table of contents at:http://www.apa.org/journals/gpr/currentTOC.html
Best regards
Rick
Rick,
Way to go!
And Happy New Year!
David Goldstein
----- Original Message -----
From: "Richard Marken MindReadings.com" <marken@MINDREADINGS.COM>
To: <CSGNET@listserv.uiuc.edu>
Sent: Friday, September 06, 2002 1:26 PM
Subject: Control Theory Glasses
[From Rick Marken (2002.09.06.1030)]
In case anyone is still interested in PCT my paper on PCT vs conventional
psychology, called "Looking at Behavior Through Control Theory Glasses" is
now available in the "Review of General Psychology". You can get to the
abstract from the table of contents at:http://www.apa.org/journals/gpr/currentTOC.html
Best regards
Rick
from Phil Runkel
to Rick Marken:
WONDERFUL that you got your article into a journal of the staid Amer
Psychol Assn. And good that it connects with demonstrations. I surely
hope someone gets connected. ��Phil R.