Living an Illusion, Complex Controlled Variables

[From Rick Marken (951129.1000)]

Bruce Abbott (951129.1040 EST)--

In focusing on this supposed defect of the study, you have managed to
deflect attention from the issue this example was intended to address. It
was claimed that all the observations from operant conditioning studies are
useless because they are merely a product of the "behavioral illusion." I
offered this study as a counterexample, and so far the only replies I have
received are (a) Rick's assertion (without supporting argument) that his
claim is still true and (b) quibbles about Herrnstein's interpretation.

Did you read my post "On pigeon's and thermostats" before writing this? If
so, don't you count my example of experimentation on a thermostat as an
argument supporting the possibility that the results of the pigeon experiment
reflect a behavioral illusion?

Your statements above suggest that you still don't understand what the
"behavioral illusion" is. "Observations from operant conditioning studies"
are not "a product of the 'behavioral illusion.' " The observations, like the
observation that one line appears longer than the other in the Muller-Lyer,
are correct; it is the interpretation of those observations that is wrong
(illusory). In the Muller-Lyer you are correct to say that one line looks
longer than the other; the mistake is to say that one line is actually longer
than another. In the "behavioral illusion" you are correct to say that the
stimulus appears to cause a response; the mistake is to say that the stimulus
actually causes the response.

The "behavioral illusion" refers to the fact that all apparent _causal
relationships_ observed in conventional behavioral research (including all
operant conditioning studies) are illusory if the system you are studying is
a control system. The behavioral illusion works as follows:

1) You observe that some behavior, o, (such as pecking at key A or B) is a
function of some stimulus situation, s, (such as pictures). That is, you
observe:

(1) o = f(s)

2) You conclude that the relationship between o and s (f()) reflects a
characteristic of the organism. For example, you conclude that differential
responding to the pictures reflects the organism's ability to discriminate
the pictures; different o's result from different values of s because the
bird can disciminate different values of s.

3) The behavioral illusion means that conclusion 2 is wrong. The relationship
between s and o does NOT reflect a characteristic of the organism; it
reflects a characteristic of the environment. This is because, in a control
system, the relationship between s (disturbance) and o (output) reflects the
inverse of the feedback function (g()) that relates o to the controlled
variable, q. That is,

(2) o = 1/g(s)

where

q = g(o) + s.

Equations (1) and (2) describe the same observed relationship between s and
o. Equation (1) attributes this relationship to the organism (f()); equation
(2) attributes the same relationship to the envrionment (g()).

So while there IS a relationship between o and s, this relationship tells
you, not about the nature of the organism (f()), but about the nature of the
organism's relationship to the variables it controls (g()). In the pigeon
example, it means that the observed relationship between pictures and pecking
reflects the inverse of the relationship between pecking and the variable
influenced by the pictures (if there is such a variable; I presume there is
but you have to do The Test to determine what it is).

With Herrnestein's experiment, I've offered the exception that tests the
rule.

You can't show that there is NO behavioral illusion by showing that there IS
a relationship between variables; that's like showing that there is no
Muller-Lyer illusion by showing that one line really looks longer than the
other. The only way to show that there IS an a behavioral illusion is to show
that the organism is controlling q; the illusion is then exposed by showing
how the obsered relationship between s and o depends on g(). Conversely, the
only way to show that there is NOT a behavioral illusion is to show that
the organism is not controlling any variable, q; this will show that there
is, indeed, a lineal causal path connecting s to o via the organism
function(f()).

In other words, the only way to "prove" that the results of conventional
operant conditioning experiments are NOT useless is by TESTING to see if the
organisms in these experiments are controlling perceptual variables (q). Of
course, this is the step that you (and the rest of conventional
psychologists) understandably consider unimportant and uninteresting.

Chris Cherpas (951129.0912 PT)--

What, for example, has PCT experimental work contributed to our
understanding of "concepts?" In short, is there anything beyond "thermostat
research" here?

I describe some research that deals with control of sequences and programs in
a paper called "The hierarchical behavior of perception". It was never
accepted by a real journal. It is available at:

http://www.ed.uiuc.edu/csg/documents/behave_p.erc.html

I am trying to continue my research in this area; but my motivation is low
because there are only four people in the world who are interested in it or
understand it (and one is me).

Best

Rick

from Phil Runkel on 1 Dec 95.

        Dear Rick Marken: You recently said that your motivation to do
research is low because there were only four people in the world who
understood your research or cared about it.

        I beseech you to continue your research.

        I suppose you are telling us that only three people have given
you clear cause that they understand what you have done and care about
it. I CARE ABOUT IT. Of all investigations in psychology or other
social science, indeed of all investigations in any branch of science
that I know about, the work you and those other three (and maybe three
more you forgot to think about) are doing is the work I care about the
most. The very most. Whether I have given you evidence that I
understand it, or maybe some of it, I must of course leave to your judgment.

        I have not been joining in the arguments in the net, because I
must discipline myself if I want to get some pages written for my book.
But I will push aside the book, any time, to tell you how much I prize your
work.
        Please continue your research.

        Signed, a friend.