[From Bruce Abbott (941011.2050 EST)]
Tom Bourbon [941011.1416] --
Bruce Abbott (941007.1240 EST)
Ah, so you did see my post, after all. I was hoping to learn what you
thought of my remarks.
Sorry Tom, I just hadn't had time to frame a reply. Here it is:
The first major complaint we have with much of the traditional literature is
over things, not directly related to the phenomenon of control, that should
concern _all_ behavioral scientists. In the majority of traditional research
by psychologists (and by researchers who borrow their experimental methods
from psychologists), experimenters intend to learn about and talk about why
individuals behave as they do. The problem is that most experimental
methods and statistical analyses used in traditional research are suitable,
not for the study of individuals (idiographic research), but for determining
statistical features of numbers in distributions (nomothetic research) --
measures of central tendency and dispersion, counts of frequencies,
estimates of probabilities and proportions, and so on.
To some extent you are preaching to the choir here. My background is in
operant conditioning and, as you might expect from that fact, I would much
rather see data from single subjects than group averages. I do not agree,
however, that the use of statistical averaging necessarily invalidates a given
finding. As an example, Ebbinghaus's careful studies of his own capacities
for learning and retaining nonsense syllables have been replicated time and
again in group-based studies. In some cases averaging is the only practical
route available. Sternberg's (1966) investigations of search in short-term
memory come to mind, which used average reaction times while manipulating the
length of the list to be searched as evidence that subjects search the items
in STM in serial order.
. . . There is no assurance that a traditional
study, or a body of traditional literature, will reveal _anything_ about the
phenomenon of control, in general, or about the specific variables a
particular subject controlled, in particular. Even studies in which certain
stimulus-response associations repeat with very high reliability may tell us
about the reliability of irrelevant side effects of behavior, and not about
whether a subject is controlling, and if so, what is controlled. Those
questions can only be answered by research specifically designed to test for
controlled variables.
An here, as with Rick Marken's argument, we come to the crux of the matter...
my thesis has absolutely nothing to do with whether these other studies can
tell you anything about control per se. Rather, I'm arguing the inverse
proposition, that control theory should have a lot to say about SOME of well-
established, replicable phenomena to which it applies. I'm also arguing that
some facts established in non-PCT-oriented research should help to illuminate
such things as the processes by which perceptual signals are created and
transformed, or what determines WHICH perceptual control systems get access to
the final common behavioral pathway at any given moment.
Many of these findings will in fact have direct relevance to PCT
(perceptual studies, for example).
The truth of this claim will be determined empirically. I do not assume
that the results of _all_ perceptual studies will be relevant to PCT. In
fact, I believe many of them will be irrelevant.
I agree. I never made the claim that all would be relevant.
To gain acceptance of PCT as a general theory of behavior, we need to show,
wherever possible, how previous work "fits into" the PCT framework.
Wherever possible, yes; everywhere, no -- and the distinction must be drawn
rigorously. "Seeing relevance everywhere" is the path followed by some of
our more successful control theory "colleagues" -- the popularizers who have
bastardized the theory beyond all recognition and thereby achieved
publication, acceptance, and fame. Almost without exception, the titles of
their articles, chapters and books present control theory as a "framework,"
a "perspective," or a "viewpoint" on traditional topics. They have easily
made old ideas, even old fictions, "fit into" PCT.
I can empathize with you here, having experienced similar frustrations. I
don't believe that PCT should be called upon to explain everything, especially
not bad data from poorly conceived, qualitative studies, for the same reason
that I would not attempt to use Newtonian mechanics to explain the growth of
plants. One reason why I subscribed to CSG-L was to develop my understanding
of PCT well enough to discriminate those cases where PCT can and should be
applied from those where it should not. The fact that some individuals
misunderstand and misapply control theory is no argument against applying it
correctly where it makes sense to do so.
Tom Bourbon [941011.1658] -- Re: Rescorla-Wagner and PCT
I have read your reply but please bear with me--it will be a little while
before I can respond (I'm teaching a 12 hour load). For the time being, let's
assume that V and L represent perceptual quantities, that dV = aB(L - V) and
V = V + dV compute the output, that the environment is represented as Vin =
Vout, and that the input function is Vp = Vin (i.e., input gain = 1). These
assumptions make V and the external variable presumably being controlled
(quantity of salivation per CS presentation) identical for the simulation, but
this could easily be corrected. (The parameters a, B, and L are curve-fitting
parameters in the model, but the way the model has been used, it is said that
the important thing is that it predicts the correct functions and qualitative
relationships).
Oh, one more thing. I have a computer and I know how to program simulations.
Cheers,
Bruce