[From Bill Powers (960909.1400 MDT)]
Bruce Abbott (960909.0855 EST)--
What I said was that Simon recognized that for purposive systems, variation
in behavior reflects mainly variation in the environment.
Now that I've read this statement three or four times, I'm beginning to
wonder what it means, or if it's even true. Is a piece of wood a purposive
system? Surely the variability in the behavior of a piece of driftwood
floating on the ocean is due ENTIRELY to the variability of its environment
-- wind, waves, and currents. On the other hand, the variability in the
behavior of a purposefully steered boat is FAR LESS than the variability of
its environment; wind, waves, and currents have almost no effect on the course.
Statements like yours are true or false depending on the specific meanings
that are plugged in as examples of the squishy words, like "environment,"
"variable," and "behavior." For example, if by "behavior" you mean the
observable motor actions of the organism, as opposed to their controlled
consequences, and if by "environment" you refer primarily to disturbances
rather than controlled consequences, then we can say that the behavior of a
purposive control system is AT LEAST AS VARIABLE as the variability of its
environment. It is normally MORE VARIABLE because not only do actions change
to oppose disturbances, they change as much more as necessary to make the
controlled input follow changes in the internal reference signal. Also,
whether you see behavior as variable or as constant depends on the level of
perception you apply to it: is the behavior of a marching band's bass
drummer variable because he repeatedly lifts the stick and bangs it against
the drum, or is it constant because he is maintaining the same pattern of
beats for a long time? [P.S. -- what DO you call that thing with a padded
ball on the end with which one hits a bass drum?]
The problem with abstract generalizations is that they seem general only as
long as you're thinking of the original context from which you drew them. As
soon as a new context arises, the meanings of the words shift and the
generalization might suddenly become false. Language enables us to generate
facile verbal generalizations, but it's far from easy to come up with a
generalization that's truly general -- one that remains true for all
specific instances of the general terms.
Best,
Bill P.
ยทยทยท
-------------------------------------------------------------------------