[from Jeff Vancouver (2001.08.14.0930 EST)]
[From Bill Powers (2001.08.10.1937 MDT)]
Thanks for the detailed description. I may see if I can convince one of my
students to take this up as a masters thesis. Besides the issue of getting
a better match to humans, the question that would make the project
interesting is whether the mistake I made regarding what must be going on is
currently made in the psychological literature. Providing an alternative
(and more parsimonious) explanation for a phenomenon that was thought to be
understood in a more complex way is what excites reviewers and editors. The
first step is to document that "a phenomenon was thought to be understood in
a more complex way." That requires looking at the current literature. If
anyone knows of such documentation, I would appreciate the reference (it
will save me or the graduate student lots of time).
Later,
Jeff
[From Rick Marken (2001.08.14.0830)]
Jeff Vancouver (2001.08.14.0930 EST)--
Besides the issue of getting a better match to humans, the question
that would make the project interesting is whether the mistake I
made regarding what must be going on is currently made in the
psychological literature.
The entire PCT research literature (such as it is) is a detailed answer
to that interesting question. The answer is "yes"! Psychologists have
made that mistake and many others that are even more fundamental. These
mistakes are not the result of stupidity. As in the squared circle demo,
behavior often _seems_ to involve complex "learning", "adaptation",
"anticipation", "planning", etc. As I show in the "Blind men..." paper
(http://home.earthlink.net/~rmarken/blind.html) control can _look like_
all kinds of things that it's _not_.
What is disappointing is, now that PCT has shown psychologists that the
"elephant" of control is not a "snake", "rope" or "tree trunk",
psychologists continue to study only "snakes", "ropes" or "tree trunks".
It was great to meet you at the meeting. Thanks for coming.
Best regards
Rick
···
--
Richard S. Marken, Ph.D.
MindReadings.com
10459 Holman Ave
Los Angeles, CA 90024
Tel: 310-474-0313
E-mail: marken@mindreadings.com
[from Jeff Vancouver (2001.08.14.1310)]
[From Rick Marken (2001.08.14.0830)]
What is disappointing is, now that PCT has shown
psychologists that the
"elephant" of control is not a "snake", "rope" or "tree trunk",
psychologists continue to study only "snakes", "ropes" or
"tree trunks".
It was great to meet you at the meeting. Thanks for coming.
It was great meeting you as well. Now back to it:
I am current writing comments on your glasses paper, but your above comment
hits the major issue. You say that psychologists continue to study only
snakes, ropes or tree trunks, but you do not document that. To document it,
I am not saying that you find examples of specific studies (for it is easy
to say "well that was an idiot who did that study [we say it all the time to
ourselves]). I am saying you find examples of well-received programs of
study (multiple researchers building on each others work) that do it and the
PCT explanation that explains it better.
Later,
Jeff
i.kurtzer (2001.08.14.1530)
[From Rick Marken (2001.08.14.1200)]
Jeff Vancouver (2001.08.14.1310)--
> I am current writing comments on your glasses paper
Great. I look forward to seeing them.
> but your above comment hits the major issue. You say that
> psychologists continue to study only snakes, ropes or tree trunks,
> but you do not document that. To document it, I am not saying that
> you find examples of specific studies...I am saying you find examples
> of well-received programs of study (multiple researchers building on
> each others work) that do it and the PCT explanation that explains
> it better.
But we have documented it, and in precisely the way you suggest. We
(particularly Bill Powers) has shown that one _very_ well received
program of study -- the one devoted to the study of the "snake" of
operant behavior -- is misguided because it can't see the "elephant" of
control.
There has not been a series of papers to a concentrated audience about a
particular or number of problems that they have considered worthwhile.
By anyone.
The hit and run, in principle you are wrong approach, has been a pretty
dismal failure.
Not matter how much that should work if...
So has the umpteenth summery/intro approach.
I would suggest that establishing yourself as the foremost expert (such as
baseball throwing) would be the best idea.
Become the expert in something and show how much better your data and
understanding is.
I am not sure that that works but it hasn't been throroughly tried and know
the others don't.
i.
[From Rick Marken (2001.08.14.1200)]
Jeff Vancouver (2001.08.14.1310)--
I am current writing comments on your glasses paper
Great. I look forward to seeing them.
but your above comment hits the major issue. You say that
psychologists continue to study only snakes, ropes or tree trunks,
but you do not document that. To document it, I am not saying that
you find examples of specific studies...I am saying you find examples
of well-received programs of study (multiple researchers building on
each others work) that do it and the PCT explanation that explains
it better.
But we have documented it, and in precisely the way you suggest. We
(particularly Bill Powers) has shown that one _very_ well received
program of study -- the one devoted to the study of the "snake" of
operant behavior -- is misguided because it can't see the "elephant" of
control. It has been hard to show that the PCT explanation is better
than other explanations of operant behavior because operant research is
not done in a PCT friendly way. Operant researchers don't tell us enough
about the experimental situation to make PCT modeling possible or they
use experimental procedures that hide any controlling that might be
going on.
You can't just point to an existing research program and say "See there,
you're studying a misinterpretation. You're looking at control from a
limited perspective and you don't know it". But they are looking at what
looks like a "snake" to them. Telling them that it is _really_ just the
trunk of an "elephant" is not going to be terribly convincing. It looks
like a snake to them.
I think you have to _show_ these researchers that what seems so
obviously like a snake is really no snake at all. This is what many of
our demos (like the Squared Circle) are about. We're showing nice, clear
examples of cases where control looks like (but demonstrably can't be)
the kind of behavior the researcher is studying: snake, rope or tree
trunk. They are demonstrations of principle. A principled researcher who
comes to understand these demos _should_ be able to say what I did (and
what Bill Powers said in the Foreword to _Mind Readings_): "If the
phenomenon you see here really works as this model shows it to work,
then a whole segment of the scientific literature needs to be deposited
in the wastebasket".
But, of course, you can't _make_ people throw away what they think is
real great work and start in a new direction. That's why there are so
few PCT scientists around.
Best regards
Rick
···
--
Richard S. Marken, Ph.D.
MindReadings.com
10459 Holman Ave
Los Angeles, CA 90024
Tel: 310-474-0313
E-mail: marken@mindreadings.com
Tom,
The two attachments are exe and source code files for somewhat improved
version of a program I showed at the recent CSG meeting. The program "Conflict"
illustrates interactions between variables that are in, naturally, conflict.
When distrubances are added to the situation what look like very complex
dynamics are genearted. However, when a trace function is turned on, it is
readily apparent that the dynamics are quite simple.
The program uses ordinary control loops for the outer ( yellow lines ) and
integral control loops for the inner ( red lines ) systems. Adjusting the
slowing, gain, the leakage on the integrator, and fraction of the error that is
integrated ( ACC ) generates some complex interactions.
Diane Gossen thought she could make use of the demo to illustrate how a
distrubance of one of a person's control variables has an effect upon the rest
of the person. I wouldn't want to show the source code for the version I showed
at the meeting, but the latest version is very much cleaned up.
At the meeting, Bill Powers seemed, at least for the time being, to be backing
away from Rick. Same is true of recent stuff on the net. I have hopes that
saner elements may be able to gain control of the situation.
I haven't seen your recent demo's, just heard about them from Greg. I'm working
at putting a course together on control theory. But, putting most of my time on
a book.
I hope you'll think about the meeting next year, it seems likely to be held
again in St. Louis.
best
Bill Williams
ac.pas (50 Bytes)
c.exe (49 Bytes)
···
______________________________________________________________________
Do you want a free e-mail for life ? Get it at http://www.email.ro/