[from Jeff Vancouver 960606.1345 EST]
[From Rick Marken (960604.0800)]
Jeff Vancouver (960603.1430 EST) --
>We define goals as internally represented desired statesDesired states of what? Perceptual variables, I hope;-)
Yes, the role of the input function is very much in the paper. We call it
the perception of the current or anticipated state. Although it strikes
me as inconsistent to use the term perception to mean, generally, the
output of an input function but, specifically to refer to the output of
input functions in the non-intrinsic hierarchy (i.e., perceptual
hierarchy). We try to be clear that a perception need not refer to the
system's translation of the environment into a mental representation, it
could also refer to translations of a state of the system into a metric
that can be used to compared the current state with the desired state.
>we were both surprised by the reception that the paper got from the
>reviewers.Maybe the psychological _zeitgeist_ is now right for PCT.
This is what I am thinking, but one must be very careful in making the
argument. Remember Anthony began by condemning Caesar. (not that I
condemn PCT).
>When I talk to colleagues in these areas they also agree that the control
>process is an accepted concept.Control theory (the control _process_) has always had some acceptance in
psychology. What has been missing from psychology is 1) an understanbding of
the nature of control as a phenomenon and 2) an understanding of the fact
that behavior IS control -- in fact, not in theory.
I would not question that control is a fact, but I would question the
premise that is it the fact that explains all of behavior. It may, I am
just questioning it. Can you allow us that?
Psychologists (other than PCTers) who accept or use control theory as a model
of behavior do so in the context of a cause-effect model of behavior; they
accept the theory, but not the fact, of control. This means that they don't
really understand the appropriate application of control theory to behavior.
That is perhaps true of some, even many psychologists. But I have met
others, particular in the cognitive areas that have grave doubts about
the simple cause effects model.
I find that the best way to determine whether a psycholgist understands
control is to watch what he or she does, not what he or she says. The
psychologist who says "control theory is the greatest thing since sliced
bread" while manipulating independent variables to determine their effect on
behavioral responses, does not understand control.
This is where we begin to part company and where those who might have
been interested turn away because it is so antithetical to previous beliefs.
But we part company because I think you are wrong. The TEST does it
(actually would be a little cleaner if it did). To quote RUnkel (1990):
"3. apply various amounts and directions of disturbance directly to the
variable
4. Measure the actual effects of the disturbance" (p. 14)
This is a maniputate the independent variable, measure the dependent
variable experiment. Only, rather than measuring disturbances, one might
measure behavior (actions) instead or as well (it would be cleaner as the
disturbances might be coming from more than the focal system).
Avery Andrews (960604) gives some great hints about where to look for
articles that have contributed to the idea that control theory is not up to
the task of explaining behavior. Surely the "best" article on this topic --
the one that tried to reject PCT in one swell foop -- is one by Fowler and
Turvey (1978) called "Skill acquisition"; it's in G. Stelmach (Ed.)
Information processing in motor control and learning. New Yory; Academic
Press.
Thanks, this will be helpful. I have found the general criticism to be
of the nature Avery mentions. That is over-simplifying the model. THis
is a major aspect of my current arguments. More examples will be good.
BTW, Bill P. notes that the motivation (e.g., hunger) psychologists have
been looking a control systems for some time. THis is true. Cannon at
the beginning of the century coined the term homeostasis. Homeostasis was
also supposedly trashed long ago. But, my reading of Mook is that the
counter-argument was something like this: Cannon proposed that rats
control x (x _means_ hunger). Studies of input functions (experimental,
cause-effect studies) reveal that x is not being controlled. In fact
(a little while later) y and z are proposed as the inputs that are
controlled. More studies and y and z (well almost z) are supported.
THerefore, Cannon was wrong. Follow the argument? I don't.
Finally, of the issue of the term goal. First, let me say that I am
simpathetic to using the term reference signal. Second, we were clear
that goal need not be an achievement but may very well be, if not most
often is a maintenance goal. Third, the reason we went with goal boils
down to the Anthony argument above. It would be very hard to sell a
paper titled "The reference signal in psychology...". THe goal construct
in psychology is much more likely to hook psychologists in. I am trying
to convince these people that they have really already largely adopted a
PCT approach to looking at psychology. We try to show them the advantage
of pushing a little bit more forward in that direction. Of course, one
reason for this is that my coauthor has no particular enthusasiam for
PCT. It is not that he does not like it, but he is very aware that
others do not.
Bill's (969694.0800 MDT) problem with the goal term is one of degree, not
kind. It is not that we cannot hold desired _future_ states, it is just
that they might play a lesser role than psychologist give credit.
Whatever, I think the difference (if that is the correct word) is the lag
in the system. Goals with long lags look like what Rick and Bill think
of when they think of goals. Goals with short lags are the kind where
much of your research focuses. It kind of makes one think about how the
practicalities of doing research influence the nature of the theories
used the think about the phenomenon.
One more thing, Bill P. I realized that I might have sounded a little
testy in my response to your test of my understanding of PCT. After
reading another of your posts to me I understand your position better and
in fact am amazed at your restraint. A reasonable interpretation of my
June 3 post is that another yahoo has bastardized PCT to the psychological
community. I respectively hope that after reading it, Rick will attest to
it (I'll send you a copy if you would like). I am not saying he or you
(particularly he) will like it, but he might find (by using words like
goal) that I have not misrepresented PCT. I have my own ideas about this
stuff, many others have ideas about this stuff, and they are reflected in
that paper.
Later
Jeff
and indicate that it would be handled in the higher levels or