[From Rick Marken (930905.1000)]
I made the mistake of reading the Opinion section of the
paper this morning. There was a column by Jeanne Kirpatrick
in which she chides Clinton for not heeding the "lessons of
history" -- in this case the lessons of Munich and Vietnam
re: Bosnia. The article reminded me of one of the (many)
profound ways in which PCT flies in the face of some of
the comfortable bromides of common sense (making understanding
of PCT -- let alone its acceptance -- apparently prohibitively
difficult). I realized that the notion of "lessons of history"
is just good old fashioned S-R theory dressed up in Political
Science duds. It just means that if X led to Y in the past
then we should not do X is we want to avoid Y (or we should
do X if we want to produce Y) now. Of course, the relationship
between X and Y is only statistical -- an acceptable limitation
for Political Scientists who, after all, are just another flavor
(and not a very good one) of Behavioral Scientist.
PCT has a problem because it doesn't say that the lessons of
history are wrong (or right) -- something with which a Behavioral
Scientist can argue. PCT says that they are irrelevant; there
is simply no way to produce an intended result (not-Y or Y) in
the same way on each occasion. This is because Y is multiply
determined -- not just by one's own outputs (X) but, also, by
varying environmental disturbances, D. So Y = X + D. Since D
is always changing (in unpredictable and, usually, undetectable
ways) there is, in principle, no way to know what to do (X)
in order to produce the intended state of Y. The X that produced
a particular value of Y in the past (the lessons of history)
almost certainly WILL NOT produce the same result now -- due to
changes in environmental circumstances. Historical "relativity"
is apparently as frightening as moral relativity C'est la vie.
The "lessons of history" also reflects the lay reverance for
undertstanding the world in terms of observed regularities instead
of underlying models. People can point to Hitler's invasion of
Czechoslovakia and say "see what appeasement does". It's tougher
to look at the nature hierarchical control systems and figure
out how how best one can avoid conflicts between them -- in
principle -- keeping in mind one is one oneself.
Bill Powers -- Great post on Gibson. J.J was a nice guy and he made
some interesting observations but he sure didn't understand science
(modelling).
Gary Cziko -- Very interesting posts. I'll try to get to them
tmmorrow when I'm on vacation. (PS. I did the polarity reversal
study; Bill P. suggested it; Tom B. could have done it but I got
there first; actually, it was first done by sone manual control
theorists.There is a version of my conflict study where the
coordinates of the mouse shift relative to the X-Y movement of
the cursor; it doesn't shifty continuously but it definitely
requires some serious learning in order to gain control).
Best
Rick