[From Gary Cziko 921204.2345 GMT]
Bill Powers (921204.1430)
I was hoping that my ruminations would disturb you in such a way as to
evoke some more thoughts on a CT view of evolution. I will need time to
digest them. Thanks.
In the meantime, I don't need much time to digest:
If I add in a disturbance to the error so that, e.g.,
error = reference signal - perception + 5, then error will still be
kept near zero.
But the perceptual signal will now be controlled about a level of
reference signal + 5.
But doesn't this then prove Audra's point showing that the perception is
not controlled (since it has changed as a result of the disturbance added
to the error signal) but rather the error is controlled (it has not changed
since it is still near zero)?
Nuts--I wanted Audra to be wrong about this and you and Marken right! It's
very disturbing to see someone with a background in STATISTICS do this to
us.--Gary
···
------------------------------------------------------------------
Gary A. Cziko Telephone: 217.333.8527
Educational Psychology FAX: 217.244.7620
University of Illinois E-mail: g-cziko@uiuc.edu
1310 S. Sixth Street Radio: N9MJZ
210 Education Building
Champaign, Illinois 61820-6990
USA
------------------------------------------------------------------
[From Rick Marken (921204.1800)]
Gary Cziko (921204.2345 GMT)--
But doesn't this then prove Audra's point showing that the perception is
not controlled (since it has changed as a result of the disturbance added
to the error signal) but rather the error is controlled (it has not changed
since it is still near zero)?
Read my algebra. The perception is no more disturbed by the addition of
the disturbance than it is when the reference level varies. The error
signal, however, is disturbed -- it DOES NOT go to zero -- it goes to
e + de. The error signal varies (mainly) as de. Adding dp to the perception
in the same way DOES NOT result in p equalling p + dp. So p is controlled
relative to r (or r + de) but e is NOT CONTROLLED-- it is not maintained
against disturbance at some reference level -- the purported reference
being zero.
Nuts--I wanted Audra to be wrong about this and you and Marken right! It's
very disturbing to see someone with a background in STATISTICS do this to
us.--Gary
Well, I don't want to be competitive about this. Audra is trying to
learn this stuff and, I think, asking the right kinds of questions.
Her intuition about controlling error, though wrong, at least does
make us look more closely at this phenomenon (control); and I think
she led us to some interesting (and surprising) discoveries about the
behavior of control systems.
Best
Rick
···
**************************************************************
Richard S. Marken USMail: 10459 Holman Ave
The Aerospace Corporation Los Angeles, CA 90024
E-mail: marken@aero.org
(310) 336-6215 (day)
(310) 474-0313 (evening)