(re) organization

[Peter Burke 9951026.1335]

[From Bill Powers (951026.1205 MDT)]
With this picture in mind, how would we answer your question above? The
answer becomes exceedingly simple. If one ECU controls just a single
dimension of perception, the only way the perceptual signal needs to
vary is in _amount_ (in the nervous system, frequency of firing). The
perceptual signal does not look like the thing perceived; it is simply a
signal that can be stronger or weaker as the set of inputs to the
perceptual function comes closer to or gets farther from the pattern
that generates maximum perceptual signal. It is a scalar signal -- that
is, a signal whose state is represented by one number rather than a set
of numbers (a vector).

I had understood this, so my question was not clear. Consider for an ECU a
number of input signals that get combined to form a perception. They are
combined in a particular way, and of all of the things that might be
combined, some are excluded. So my question really is how is it that the
input signals that are included in the combination, and the particular form
of the combination is the correct one for the particular reference signal of
that ECU? I think that this may really be a question about how
(re)organization takes place. This, of course, can be handled by a
neural-net algorithm if one has already set up the condition to be matched,
maximized or minimized, and is also easier to do when the system is
"trained" by an experimenter. However, I believe we are talking about
self-organizaing or self-training systems, and that make the proposition
much more difficult.
The inverse question also holds, that is, (re)organization from the top down
(how refernce signals get determined as (weighted) combinations of outputs
from higher up. And, of course, the real problem is that all of this is
going on all of the time (to varying extents). It is just difficult for me
to see the algorithms, mechanisms, or whatever that structure or guide this
process.
Peter

ยทยทยท

------------------------------------------------------------------------
Peter J. Burke Phone: 509/335-3249
Sociology Fax: 509/335-6419
Washington State University
Pullman, WA 99164-4020 E-mail: burkep@unicorn.it.wsu.edu
------------------------------------------------------------------------