[from Mary Powers 9212.05]
Mark, Audra:
Bill said something like this the other day, but I think it bears
repeating/rephrasing. When you try to pick one part of a control
system as being the most important - control of error - you are
missing the point that all "parts" are equally important.
Basically you are thinking lineally - here's this part and then
this one and this one and AHA! this one, which is the "really
important" one. When all the parts are lying around, sensors and
comparators and so on, you don't have a control system. When they
are linked together, you do. All are essential. Asserting that
one part is more important than the others is only to say that we
CAN perceive and categorize control systems as having parts, but
that doesn't mean that it's useful to do so. The interesting
thing about control systems is what emerges from the linking
together of their components. Evaluating the worth of one part
over others is reductionism - good for intellectual arguments but
not likely to be very productive.