[From Bruce Abbott (960531.13:00 EST)]
Brian D'Augostino (960531.1022?) --
Brian, the top part of your message was missing/garbled; I've attempted to
reconstruct the header. The message STARTS with the following:
yWPC< they termed a "set
If there was anything prior to this I didn't receive it.
Bruce, you seem to be saying that the control systems model does
_not_ require the assumption of actual control system components at
the level of physiology. This is news to me.
Perhaps I wasn't as clear as I could have been on this point. Booth (and
many others) apparently thought they were going to find _neural_ components
to produce the reference signals, error signals, and so on, and when they
failed to identify such structures mediating the control of body weight,
food intake, water balance, and so on, they concluded that in most cases,
physiological systems simply are not organized this way. This is a bit like
concluding that aerodynamic principles cannot apply to birds because a
detailed examination of bird construction shows that birds do not have
propellers or jet engines.
If control systems exist in a living organism, the physiological components
must be there (of course!), and they must be connected in the right way so
as to produce and transmit the appropriate signals. But there are many
physical arrangements that will produce the same _functional_ unit. For
example, a neuron whose rate of firing begins to increase only when its
inputs exceed a certain threshold may function as a comparator whose
reference level input is represented by the threshold rather than another
input neural signal. If that threshold is altered by variations in the
level of a certain hormone, then in effect the reference level is determined
by the hormone level. The mistake Booth and others seem to have made was in
taking the control-system diagram too literally, as if each part of the
system had to be physically represented exactly as in the diagram, with
discrete "devices" performing each separate function: perception, set-point
representation, comparison, output. The functions are there, but typically
the structures that provide them are not so cleanly differentiated as to
function, and the signals may be represented in a variety of ways in
addition to neural impulses.
On two occasions on this net I asked what has been done to
substantiate the kind of speculation about physiological mechanisms
in B:CP. I got no answer. Tom Bourbon is currently working in
this area, but when I asked him about empirical research, he said
that control theory was not nearly as well established in
physiology as Bill Powers seems to think. In fact, Bruce's
quotation from D. A. Booth in the same post I quoted above
indicates the opposite. Bill, since you recently said on this net
that you were looking for collaborators on PCT research, why don't
you work with Tom on these sorts of questions? Do you not agree
that empirical work of this sort is foundational to PCT?
Bill and I are currently working together on exactly this sort of research
program, although we are still in the early stages. At present we are
examining food consumption and body weight in rats that are able to access
food both in their home cages and in an experimental chamber, while
manipulating the amount of food available for consumption in the home cage.
Through these manipulations we hope to determine what variables are being
controlled, how the various control systems are tied together, and what the
parameters of control are.
I know that Bill _has_ collaborated with Tom Bourbon and would still be
doing so if Tom's funding (and job!) had not been cut, and would certainly
be more than happy to continue if Tom were again in a position to gather
data. But there is much more to be done, and Bill is looking for yet others
who might be ready, willing, and able to take the plunge. Perhaps someone
like you, Brian? (:->
Regards,
Bruce