[From Bruce Abbott (950914.1840 EST)]
Not long ago Bill Powers and I examined some data extracted from Collier,
Hirsch, and Hamlin (1972) concerning the regulation of food intake and body
weight. I've been looking into the literature of which this study was a
part, and came across a short article by Bob Bolles (originator of the
"species-specific defense reaction," or "SSDR" notion) in a 1980 anthology
edited by Fred Toates and Tim Halliday. What follows is an excerpt from
Bolles's paper, which was entitled "Some Functionalistic Thoughts About
Regulation."
Consider again the marvellous kidney. Perhaps it is thought to be an
exceptional regulatory device because it is so efficient and so automatic.
I would argue that, on the contrary, the ADH control of kidney excretion is
a typical control device. Basically, blood plasma Na+ concentration is free
to vary as it will when Na+ intake and loss and water intake and loss vary.
But as the concentration increases this increase is detected, it triggers
ADH release, which causes the kidney to retain water. There is no mandate
to control tonicity; there is no proper value that it is obliged to take.
It is just that as Na+ concentration increases a new mechanism comes into
play that puts a limit upon its increase.
In some animals, presumably those that have evolved under conditions where
it was important to have a strict limit, the correction mechanism is very
efficient so that it provides a flat, sharp-cornered ceiling on the rise of
Na+ concentration. This is not some special feature of water balance, I
would argue, but a characteristic of all systems that are thought of as
regulatory. If there is some adaptive cost associated with too high a value
of this parameter, or too low a vlaue of that one, then we can expect there
to be some device that comes into action to limit variation. The more
efficient mechanisms, like the adjustment of urine Na concentration, are
simply those that provide a sharp corner and a flat ceiling for some
parameter. Those where variation poses less challenge or implies less cost
may provide a more sloping function. But the principle is the same in
either case. Some physiological parameter varies freely with the factors
that govern it, but when it reaches some critical limit it begins to
activate a new factor that imposes a limit. The floor or ceiling is usually
not encountered all at once, more likely the control device was there all
along and just growing in importance as the controlled parameter varied.
That is the way, I suggest, that all regulation works. It is instructive in
this connection that none of the physiological parameters that we think of
as well regulated are limited on both the high side and the low side by the
same mechanism. In each case one thing keeps it from rising too high and a
different device keeps it from falling too low. That is the essence of
real-life control systems. In no case do we find a real control device that
measures a descrepancy from some preordained value and applies a correction
that is proportional to the discrepancy. There are, as far as I know, no
systems of that sort, there are only floors and ceilings, typically
unrealted to each other, to limit the variation in the system. Often, as we
saw with the fly, there is only a ceiling and nature provides the floor.
[Pp. 65-66]
Bolles, R. C. (1980). Some functionalistic thoughts about regulation. In
Toates, F. M., and Halliday, T. R. (Eds.), _Analysis of Motivational
Processes_. New York: Academic Press, P.p. 63-75.
Bolles is talking specifically about physiological regulation, in particular
the regulation involved in a "feeding control system," and not about
servomechanisms such as is exemplified by the system controlling muscle
length and tension. In this context he is arguing that control over some
physiological variable is acheived by mechanisms which act to limit, either
on the high side or on the low side (or both) the excursion of the variable,
as opposed to a classical control system in which there is a physical set
point, comparator, error signal, and continuous opposition to deviations via
a high-gain loop. Such limits could take simple forms, as when (to use one
of Bolles's examples) a cat grows heavier until its excessive weight begins
to limit the number of mice it can catch.
I don't wish to quote Bolles more extensively than I already have, but the
thrust of the position he outlines is that body weight, fat storage, etc.
may not be regulated around some set point at all, but may result instead
from a kind of dynamic equilibrium at which the animal settles under given
conditions. Thus, reduced body weight follows when food is adulterated with
quinine, because palatibility is one of the factors that determines the
equilibrium value. In such a system, the search for a comparator and a
physical reference signal would be futile.
Comments? (;->
Regards,
Bruce