Feedback Functions

[Martin Taylor 2007.09.08]

[From Rick Marken (2007.09.08.1520)]

> Martin Taylor (2007.09.08.12.51)
>

To this point, I see no difference between a PCT approach and the
conventional approach.

The difference is the assumption that it is the value on the response
dimension -- N in this case -- that is perceived as matching the
magnitude of the stimulus. What we are saying is that it is the
_perceived_ value of N -- Pn -- that is perceived as matching the
perceived value of the stimulus. The difference between the PCT and
conventional approach is that, in PCT, it is assumed that the subject
is controlling for Pn = Pi; in the conventional approach it is assumed
that the subject responds with N = Pi.

I guess you have been skimming, rather than reading, my posts. The whole point of the summer student work that I discussed, and of my thesis, was that the matching was of perceptions -- perceived value of N, of grey value, of time difference, of hue, or orientation ... -- that mattered. The notion was even then that a dimension used as response was perceptually different from the same dimension used as stimulus. That would have made no sense if the response represented a phsyical rather than a perceptual magnitude. To understand this has nothing whatever to do with control.

> All that is asserted is that somewhere inside the

subject there is the possibility of comparing the magnitudes of two
perceptions of different kinds. Again, that's conventional
psychophysics.

I don't believe that conventional psychophysics treats the subjects'
"responses" as a perception, let alone one that is part of a
controlled perception.

I think you go overboard in your denigration of conventional psychologists and what they think. Yes, most don't think of behaviour as the control of perception, but in the sort of experiment we are talking about here, I don't think they are as naive as you suggest. I mean, even in 1923, Thurstone could write a paper titled "The Stimulus-Response Fallacy in Psychology" <http://www.brocku.ca/MeadProject/Thurstone/Thurstone_1923.html> in which he says:

···

------------------------
I am calling in question the stimulus-response formula which is explicit or implied in much of our psychological writing

and

Let us start the causal sequence with the person himself. Who and what is he? What is he trying to do? What kind of satisfaction is he trying to attain? What are the large types of self expression that are especially characteristic of him? What are the drives in him that are expressing themselves in his present conduct? Let us consider the stimuli as merely the environmental facts in terms of which he expresses himself.

and

I have listed a number of stages in the development of psychological acts in which they may be identical or similar. If we start with action at its source and follow the stages through which it becomes formulated into conduct, and the consequent satisfactions, we shall have a table as follows:

THE STAGES OF THE PSYCHOLOGICAL ACT

(I) Energy source
(2) Reduced threshold for stimuli
(3) Deliberate ideation
(4) The internal stimulus
(5) Imaginal hunt for external stimuli
(6) Overt hunt for external stimuli
(7) The external stimulus
(8) The consummatory overt act
(9) Overt consequences of the act
(10) Satisfaction to the actor, and quiescence at the energy source

(I) The energy source is the dissatisfaction in the physiological, mental, and social conditions which provokes action. These physical and mental conditions cover such wants as the satisfaction of hunger, bodily comfort, sex, social approval, social power. A state of dissatisfaction in any one of the instinct conditions is the starting point for action which is maintained until satisfaction is attained. Two actions belong in the same instinct category if they can by conditioning be readily substituted for each other.

--------------------------
(Gary...Maybe this article might qualify for "almost PCT"?)

I got the impression in graduate school that Thurstone was something of a hero to at least some of the faculty.

Back to the main theme. A response may not itself be considered to be a perception, but the assumption always is that it corresponds to a perception of some kind. Is this assumption chanbged by rercognizing that the response is an act that serves to control a perception of the experimenter being pleased that the subject is performing the experiment as intended? The act of responding cannot, in itself, influence the perception of the stimulus the experimenter presents, but if it turns out to be logically inconsistent with other responses, affect subject's perception of the experimenter's satisfaction. The easiest way for the subject to be consistent is to report something which (s)he perceives as bearing the correct relation to the stimulus. That's just a long-winded way of saying that from a PCT point of view the subject is doing what most psychophysicists would expect.

I don't think it's necessary to assert all the (imagined or real) failings of "conventioal psychologists" in order to advance the study of Perceptual Control Theory. It can stand on it's own feet, and from time to time can usefully take note of results obtained by "conventional psychologists". The data are there, and if the experiemnt is accurately described and is appropriate, why not use it? That the experimenter didn't know PCT is irrelevant.

Martin

[Martin Taylor 2007.09.09.00.35]

[Martin Taylor 2007.09.08.22.13]

Another little quote from Thurstone's 1923 article:

"Behavior starts in dissatisfaction and it terminates in satisfaction. It is in this sense that we can speak of a reflex circuit rather than the reflex arc."

It's a pity he didn't know control theory, and therefore couldn't develop the concept.

Martin

The Deja Vu still is there,
though. Unfortunately, Ina and I have discarded most of the books we had
in graduate school, so I can’t check whether it was in them. I do know
that the logarithm to power law relation was part of my undergraduate
engineering education, if not before. Maybe the Deja Vu feeling comes
from there rather than from a direct application to the two
laws.
[From Bill Powers (2007.09.09.0255 MDT)]

Martin Taylor 2007.09.08.22.13 –

I think that same engineering background is what got me thinking about
that article by Stevens in the 60s (I had seen the same “power”
idea before in his Handbook). I think it’s clear that Stevens took the
number generated by the subject as an indication of the magnitude of the
subjective response, the perception. It didn’t occur to him that the same
principle would apply to the perception of number.

I like your quote from Thurston – his work would definitely quality as
“Not Quite PCT” for Gary’s project. He came awfully close, as
did a number of others.

I think Rick’s reason, and mine, for coming down rather hard on
conventional psychologists is they they have been so damned haughty about
the “right way to think” that they have a bit of that coming.
Literature referees in psychology act more like goalies than
facilitators.

Best,

Bill P.

[From Fred Nickols (2007.09.09.0814 ET)]

To the end of Martin's post (see below)

[Martin Taylor 2007.09.08]

<snip a lot>

I don't think it's necessary to assert all the (imagined or real)
failings of "conventioal psychologists" in order to advance the study
of Perceptual Control Theory. It can stand on it's own feet, and from
time to time can usefully take note of results obtained by
"conventional psychologists". The data are there, and if the
experiemnt is accurately described and is appropriate, why not use
it? That the experimenter didn't know PCT is irrelevant.

Martin

Bravo, Martin! Bravo!

I can't think of a single thing that gets more in the way of advancing PCT than the seemingly fanatic dedication some have toward proving conventional psychology wrong or, conversely, to prove PCT right. If, as some say, nothing can be proven, both approaches are a waste of time and energy. It would seem to me to be a far better use of time and energy to establish the value of PCT via what Rick and Bill typically call for: modeling and experiments.

···

--
Regards,

Fred Nickols
Managing Principal
Distance Consulting
nickols@att.net
www.nickols.us

"Assistance at A Distance"

[From Rick Marken (2007.09.09.1210)]

Martin Taylor (2007.09.080 --

The whole point ..was that the matching was of perceptions...
mattered. The notion was even then that a dimension used as
response was perceptually different from the same dimension used as
stimulus.

The response is understood to be based on a perception in cross model
matching as well. My point was that responses are typically taken to
be outputs-- the last step in a causal chain that begins with the
stimulus. There is no concept of a controlled variable.

> Rick:
>I don't believe that conventional psychophysics treats the subjects'
>"responses" as a perception, let alone one that is part of a
>controlled perception.

Martin:
I think you go overboard in your denigration of conventional
psychologists and what they think.

Denigration? It looks more like an observation to me.

Yes, most don't think of behaviour
as the control of perception, but in the sort of experiment we are
talking about here, I don't think they are as naive as you suggest.

I'm not suggesting that conventional psychologists are naive. I'm
saying that they don't know how to study control. And just "thinking
of behavior as control of perception" does not necessarily get you
there. Just look at the work of Carver and Scheier. They "believe in"
control of perception (their theory is explicitly based on PCT) yet
they study control as though it were a cause-effect process, using
conventional methodology.

even in 1923, Thurstone could write a paper titled "The
Stimulus-Response Fallacy in Psychology"

I think it's interesting when an early thinker came close but did not
quite get the PCT view of behavior. But my experience is that these
"not quite" PCT views can be as big an obstacle -- possibly bigger --
to people getting PCT as can the most conventional, non-PCT view.
That's why aan article on "Non quite PCT" might be helpful. Not quite
means _not_. People who see in quotes from Dewey, Mead, James,
Thurstone, McDougall, Tolman, Skinner, Kantor, etc a harbinger of PCT
tend to be reluctant to give up the idea that these thinkers did not
actually "get there".

I don't think it's necessary to assert all the (imagined or real)
failings of "conventional psychologists" in order to advance the study
of Perceptual Control Theory.

I think it is essential to describe the possible "failings" of
conventional psychologists (which exist only if organisms are closed
loop systems) because they are what make it impossible for
conventional psychologists to see why PCT is needed. One failing of
conventional psychologists is that they study organisms under the
assumption that they are cause effect systems. This is a failing
because, if organisms are control systems, the results that are
observed will not be what they seem (they are a behavioral illusion,
as seems to be the case with the Power Law). A second failing of
conventional psychologists is that they study groups rather than
individuals under the assumption that each individual follows the same
cause effect law. This is a failing if organisms are control systems
_or_ if each individual follows different laws even if they are cause
effect systems.

It can stand on it's own feet,

Of course. And it does.

and from time to time can usefully take note of results obtained
by "conventional psychologists". The data are there, and if the
experiment is accurately described and is appropriate, why not use
it? That the experimenter didn't know PCT is irrelevant.

This is a bit more problematic. If organisms are closed loop systems
then the results obtained by conventional psychologists can be quite
difficult to interpret. Add to the the fact that most of the results
are from group studies and it's very hard to use the data from
conventional experiments. Sometimes data from non-PCT experiments can
be useful; I've used data from several experiments as a basis for
building control models. I did it for the Grey Lag goose egg rolling
observation; a two handed coordination task and, of course, for some
ball catching data. My current project, aimed at finding the purpose
subjects carry out in various conventionally done experiments, will
also be making used of conventional data.

But I do think it is important to forcefully explain and demonstrate
the fundamental difference between the conventional and the control
theory based approach to studying behavior. From a scientific
perspective, I think the methodological distinction between
conventional and PCT psychology is very important. It is also, I
believe, the reason why conventional scientific psychologists ignore
(or misunderstand) PCT. It's not so much because they don't like the
theory; the theory itself sounds just fine to most conventional
psychologists. Indeed, most such psychologists would probably say they
agree with PCT. But what they won't agreed to do is change the way
they go about the business of studying behavior (or mind). They
might, like Carver and Scheier, sign up for the theory but they won't
sign up for the methodology, and particularly the goals of the
methodology. The goal of conventional psychologists is to understand
S- R relationships; their interest in purpose (understanding of which
is the goal of PCT psychologists) ends when the participants in the
experiment do what the experimenter wants them to do so that what
really matters -- the S-R relationship -- can be observed.

Best

Rick

···

--
Richard S. Marken PhD
rsmarken@gmail.com