[From Rick Marken (921213.1000)]
Bill Powers (921212.2000) --
It seems to me that we already have the forum where PCT is
understood and can be developed. Let's forget about the rest of
the scientific world and just do our modeling and talk about the
results here.
I think that psychology is old and tired, and ready to be
bypassed.
Tom Bourbon (921212 22:55) --
Too bad your fate was in the hands of people with scientific bents --
they don't have much use for old fashioned fuzzy heads like PCT modelers.
Maybe we should just accept the fact that they already know everything we
try to say -- and then some. They just don't talk about it -- ever --
until they review a PCT manuscript, that is.
Thank you, my friends.
Your counsul is, as always, right on target. I thought that it would
at least be possible to start an electronic dialog about the issues in
"Blind men" (which you guys delt with experimentally in "Models...");
I didn't think they would be guarding the gates so to this relatively
informal medium so ardently; boy, was I wrong.
So I'll just get back to the balls and the planes (as Bill put it so
well in his wonderful post this morning to John Gabriel et al) and
forget about talking to the "real" scientists (of course, Galileo was
also a fairly abrasive SOB -- and he got noticed [though not always
in ways he would like] -- Gregor Mendel was more polite and look what
happened to him).
Nevertheless, in the spirit of Galileo, I will post a testy reply to
the reviews here. For those who have read these reviews and who
also understand PCT, publishing these replies is a bit like explaining
a joke. For those who don't understand PCT and want to, perhaps these
replies can help you understand some of the fundemental differences
between PCT and other approaches to understanding living systems.
(For those of you who don't understand PCT and don't want to, just
read the reviews and not my replies -- the latter will just be irritating.
···
-------
Here goes:
REPORT #1: Tom Zentall (ZENTALL@UKCC.uky.edu)
This is the only reviewer who recommended publication and for all the
right reasons:
This is the kind of paper that asks to be published in an open
commentary format. It is provocative, relatively simple, and tries to
be inclusive. Even if others find it flawed in some way that I cannot
see, I think it will generate interest in the reader. For this reason,
I would recommend its acceptance.
Hooray for Zentall!!
REPORT #2: ANONYMOUS
the author NEEDS TO OFFER A MUCH MORE GENERAL AND
INTELLIGENT TREATMENT OF THEM. I regret that I cannot be more
positive.
And I regret that I cannot be more intelligent.
The central thesis of this manuscript appears to be that an animal's
behaviour is generated in order to keep certain sensory input variables
at (or at least as near as possible to) preset values (set-points).
By George, I think he's got it!!
(NO MENTION IS MADE OF OPTIMAL CONTROL
SYSTEMS, ADAPTIVE CONTROL, MULTIVARIATE CONTROL, AND SO ON WHICH MAKE
UP THE TOOLBOX OF THE MODERN CONTROL THEORIST).
Nor was any mention made of ANY kind of control system!! But forgetting
all that trendy stuff was really a big mistake -- how could I do that?
In my opinion, the manuscript's central thesis is a rather technical
way of expressing the "fact" that behaviour is (typically)
goal-directed
By George, I think he missed it! This is not even close to the
central thesis -- reading the paper might have helped this reviewer.
The central thesis was:
"If organisms are in a negative feedback situation with respect
to the environment, then their behavior will APPEAR to be SR,
reinforcement and cognitve when it is actually NOT -- it is
CONTROL OF PERCEPTION".
I think I said that in the paper. My goal was to encourage the opposition
to explain why they think that organisms are actually NOT in such a
negative feedback situation -- so that what they think they are
dealing with -- SR, reinforcement and cognitve behavior -- is what they
are actually dealing with. The paper showed that if organisms are in
a negative feedback situation with respect to the environment, then
behavior is NOT what it APPEARS to be (that's the whole point of the
title, fer chrissakes -- an elephant is NOT a snake, wall or rope --
though it might appear to be to a person who cannot see the whole
phenomenon -- elephant).
Consequently, I'm unconvinced that discovering what an
animal's sensory set-points are is any different from discovering what
its goals are.
I AGREE -- sensory set points ARE an animal's goals and discovering
what these set point are set for (what sensory variables are
controlled) is what PCT is all about. That is the methodology
I suggested in the paper (if organisms are in a negative
feedback situation this is the only kind of research that makes
sense) -- the test for controlled variables.
The latter is none other than a functional analysis of
behaviour which people have been trying to do for a long time.
Oh, so I should look to the resarch on "functional analysis of
behavior" for examples of research based on the test for controlled
variables? Thanks for the reference. How could I have missed all
that in my training as a psychologist; Gee, and I studied the
"functional analysis of behavior" with David Premack and he never
told us about testing for controlled variables. Must not have been
listening that day -- got an A in the course anyway, though.
For example, equation
(1) strongly suggests that the actions which tend to reduce the
difference between the desired state and the current state are of a
single type which differ only in magnitude depending on the size of the
error signal.
I can't parse this too well but apparently this reviewer disdains
simplifying assumptions. I guess he thinks that I'm only interested
in organisms with a single sensor and a single output variable.
No wonder non-linear systems crap is the hot topic -- the goal in
behavioral science seems to be to move as far from the simple case
as possible -- sort of like Galileo starting his work in physics
with a cyclotron.
Note further that the
assumption that the functions in equations (1) and (2) are linear is
ENORMOUSLY RESTRICTIVE; it allows the author to treat k.e, k.f and k.o
as numbers permitting the derivation of all the other equations which
appear in the manuscript.
More "simple is bad" stuff. But HOW might that simplifying assumption
influence the premise of the paper -- that negative feedback control
will LOOK LIKE SR, reinforcement and cognitive behavior? Not a peep.
Just too much simplifying. Geez -- don't people take elementary
science classes anymore?
For example, equation (5) cannot be said to establish
the "behavioural illusion" since it relies on the linearity
assumption.
Maybe this is something that could have been discussed if the paper
were accepted?? In fact, the "behavioral illusion" does not depend
on linearity -- but it certainly works in the linear case. These are
the kinds of things that should be debated AFTER the paper appears.
I failed to understand what the author was getting at in section 3.3.1
the "stimulus-response" view of control.
Quite true -- as revealed in the next sentences:
The pupillary reflex, for
example, was one of the first types of behaviour to be subjected to a
control theoretic treatment that the author is advocating and this
treatment is widely accepted - the pupillary reflex is a
servo-mechanism and similar treatments were offered a very long time
ago for other reflexes e.g., the muscle stretch reflex - the idea that
active muscle force, F, is related to the difference between its
stretched length, x, and its unstretched length x* is described by an
equation of the form, F = k(x*-x) is basic stuff. The discussion of
Warren et al. (1986) is very curious - their treatment of the control
of running is basically control theoretic: they argue that the optic
variable tau determines the input to the muscles necessary to achieve
the goal of placing the feet correctly.
This is suitable for framing. The reviewer says that the pupillary reflex
and running are understood as control processes. Then, he explains both
in SR TERMS!!! F = k(x*-x) is an SR equation (the x* is unstretched
length, not a reference force) and tau determins the input to the
muscles necessary to achieve running-- S (tau) causes (R) muscle outputs.
Gee, he forget that R causes S also --AT THE SAME TIME!!! Control appears
(to this reviewer) as an SR process -- the very point I was making in
the section he was reviewing. Of course, the point that the reviewer
doesn't "get" is that SR IS AN ILLUSION (when there is control). Of
course, getting that would mean getting the idea that psychology
has got it COMPLETELY wrong about how control (goal oriented behavior)
works.
I think that the three types of "view" that the author discusses
represent ways of attacking the problem of understanding goal directed
behaviour
BZZZZT! Completely wrong (and completely consistent with conventional
psychological thinking). The point of the paper was that these are
APPEARANCES -- noticeable side effects of the process of controlling
sensory input. Focusing on these appearances is precisely the way
to miss the point (and not do what is necessary -- research aimed
at discovering WHAT organisms control and HOW they do it). I said
all this in the paper --quite clearly I thought -- maybe a bit
too clearly?
I FAIL TO SEE WHAT THE AUTHOR'S ANALYSIS ADDS IN THE WAY OF
CLARIFICATION OR METHODOLOGY AND DO NOT SEE IN WHAT WAY IT REPRESENTS
AN ORIGINAL CONTRIBUTION.
Indeed, he did fail to see. If he had seen, he would have understood
that I was suggesting a completely new approach to methodology -- one
aimed at the discovery of controlled input variables -- and one that
is completely incosistent with current approaches to the study of behavior.
Good job of failing to see -- control in action -- by this reviewer.
REPORT #3 Eliot Shimoff (shimoff@umbc4.umbc.edu)
Marken's "The Blind Men and the Elephant" seems like an attempt at a
grand unification theory, showing that S-R relations, reinforcement
theory, and cognitive psychology are all aspects a single process (a la
Powers [1978]).
Yep! See, Gary, it looks like your approach would not have worked either.
A serious attempt at unification must (in my opinion) (a) lead to
interesting NEW EXPERIMENTS, or (b) force CLARIFICATION of some muddy
concepts, or (c) make a SURPRISING PREDICTION (e.g., "If this theory is
correct, you should observe phenomenon X which is not predicted by any
other system").
We've a) got the new experiments. I b) tried to clarify what
might be considered a muddy concept -- at least in conventional
psychology -- the concept of control and I c) think I made
some surprising predictions -- in a control loop, S doesn't cause
R, though it appears to; reinforcement doesn't select behavior,
though it appears to and thoughts don't control responses,
though they appear to.
Marken suggests (I think) that the proper task of psychology is to
determine k.f, k.o, and k.e.
BZZT! Nope, the goal is to discover s* -- I said it in the paper --
but it was buried there in all the english sentences.
REPORT #4: Ed Fasse (edfasse@Athena.MIT.EDU)
Marken is suggesting that behavior is control.
CORRECT!
The idea that certain aspects of behavior
can be understood in terms of control is thus NOT NEW.
This is one of PCT's main problems -- not trendy. The new way to
do science seems to be to get the newest trendy methodology
(non-linear attractors, neural networks, you know the litany)
and apply it -- then it's worth considering -- never mind if
if actually accurately accounts for some data; tools are in,
observation of phenomena is out.
In any case the following question
is interesting: To what extent is behavior control.
Yes, indeed. That would have been something worth discussing in the
NOW IMPOSSIBLE discussions in Psycholoquy!
Those are not
Marken's words, but that is the question he is asking as I understand
it.
Not really. The question is -- what if organisms are in a negative
feedback situation with respect to their environment -- what would
we see? Well, I'll be darned, we'll see what looks like SR,
reinforcement and cognitive behavior -- BUT ISN'T.
1.1 I think that it is valid to say that the goal of control is to
produce consistent results, but I DISAGREE that this always takes
place in the context of an unpredictably changing environment. It is my
understanding that most computer hard disk drives use stepping motors
which are well enough behaved that feedback control is unnecessary.
Then it's not control. What a non-sequiter!!
2.1 The first sentence, "The basic requirement for control is that an
organism exist in a negative feedback situation with respect to its
environment" is FALSE. Marken later implies that stability requires
negative feedback. This is not true in general.
A low gain positive feedback system can be stabilized -- but the
loop gain must be so low that there really is no "control" being
exerted. I never implied that stability required negative feedback.
Nevertheless, both of these comments are irrelevant to the point
of the paper; I mentioned stability briefly just to satisfy those
who knew that there were dynnamic considerations that must be
satisfied for control. I just meant to say; don't worry; the
solutions work dynamically -- so the algebra is a good
representation of the functional relationships that actually occur
in a negative feedback situation.
< I will now try to skip a lot of stuff which is basically
irrelevant to the point of the paper>
I am not happy with section 2. Control theory has been applied to
understanding behavior before, notably in motor control. I do not
feel that this presentation is particularly adept. It did not teach me
anything I didn't know before.
Obviously!!
Again, what I see as interesting is not
"Can aspects of behavior be understood in terms of control?". Yes, they
can. It's been done before.
Well, gee, if it's been done before, why isn't anyone worried about
the possibility that they are using TONS of grant money, resources,
time and energy studying AN ILLUSION!! and wasting testbook space
explaining these illusions as saying something important about the
nature of behavior?
What I see as interesting is "Can ALL behavior be understood in terms of
control?".
Great. We could have discussed that (and formulated research strategies
for approaching an answer) if this paper were a target article in
Psycholoquy. Now you'll just have to wonder about it all be yourself.
What does Marken mean to say that the environmental
variable is seen as a stimulus. How is the human aware of the
environment except via the sensory information s? Or is d sensory
information too?
The OBSERVER of the behaving system sees it has a stimulus -- it's
the DISTAL STIMULUS to psychologists. That was obvious from
context.
How can we model the process that generates s*?
Good question -- and we have a model that does just that; and it's
a working model called HPCT. But the paper was not about modelling.
It was about what behavior LOOKS LIKE if in FACT organisms are
in a negative feedback SITUATION with respect to their environment;
there was no theory in the paper AT ALL. Apparently when people see
the word "control" they just assume you are dealing with control
theory. Well, this little paper was about the phenomenon of control;
this is what psychologists should learn about first -- well, that
is, if they want to understand what they are doing; if they just want
tenure and fame maybe they'd just better treat it as a theory -- and
an old, out of date one at that.
One cannot say control explains behavior until one explains the generation
of s*, which affects behavior.
When did I ever claim that control (I assume theory) explains behavior?
I said behavior MIGHT be control -- and if it is then one can derive
from equation 4 what the relationship between observable variables will
look like -- SR, reinforcement and cognitive -- when an organism is
controlling.
"Organisms behave in order
to keep sensory inputs at these reference values (Powers, 1989)." That
is debatable, but that's the kind of statement that would be good to
discuss by the community at large.
Too bad -- there will now be no discussion.
I feel the theoretical, mathematical analysis of the paper
is WEAK, and NOT PARTICULARLY ORIGINAL.
Feelings. Ooooooh. Feeeeelings. Actually, he's right. Bill P. did basically
this analysis in his Psych Review articles. So maybe it's true -- the
reviewer might already not understand this weak analysis sufficiently.
REPORT #5: ANONYMOUS
It reminds me of the sort of off-the-wall discussing we used to do late in
the evening after studying for comps. There's been so much specific
research since then, I would have expected such speculations to take a
new tack.
That's what I get for not going to MIT or Harvard. I missed the discussion
where they explained that behavior is control so studies of the
external and internal causes of dependent variables reveal little more
than well-known laws of the environment.
REPORT #6: Vasant Honavar (honavar@iastate.edu)
I recommend that the paper be resubmitted for publication in a revised
form.
Well, thanks. I suppose I could change all that stuff about SR,
reinforcement and cognitive behavior being APPEARANCES; maybe if
I explained how important it is to study SR, reinforcement and
cognitive behavior as a way to understand control-- bet that would
help. I could show how all those great studies I referred to in the
paper have contributed to our understanding of control; like that
wonderful Warren et al materpiece that shows conclusively that a
visual variable (tau) can tell the muscles precisely waht to do
to keep people walking upright -- what an incredible variable to
discover.
------
OK. I feel better now. Fear not, I will NOT send this back to
Harnad (or the reviewers). I just can't repond to this crap
non-sarcastically anymore.
Back to the lab.
Best regards
Rick
**************************************************************
Richard S. Marken USMail: 10459 Holman Ave
The Aerospace Corporation Los Angeles, CA 90024
E-mail: marken@aero.org
(310) 336-6215 (day)
(310) 474-0313 (evening)