[From Rick Marken (01.04.05.2200)]
Bruce Gregory (2001.0403.1555)--
Could you provide a functional diagram showing how a
"perceptual function" is constructed?
Sure:
er
> A REORGANIZING SYSTEM
v
--------|c|<-------
> >
> ---| c|--e--
v | |
o>--or->| p | |o| A CONTROL SYSTEM
^ |
__________|____________v_________________
qi ----------
REAL WORLD
The output of the reorganizing system (or) acts on the perceptual
function (p) of a control system. The reorganizing system output (or)
varies as a function of the difference between the actual (e) and
reference (er) for the error level in the control system; er is an
"intrinsic reference", probably fixed at zero.
Are perceptual constructions arbitrary or are they constrained?
They are arbitrary in the sense that they can be _any_ function
of real world inputs. They are "constrained" by the operation of
the reorgnanizing system to produce a control organization that
controls well (keeps error continuously near zero).
In neural networks, the constraints are called "the real world".
I don't believe "the real world" is a constraint in neural networks
any more than it is a constraint in a control model. Perceptions
in both are a _function of_ variables in the real world. The
perceptual function in PCT _is_ a neural network.
The real world is not constructed, it is assumed to exist.
Of course. I assume that as well. What PCT says is that it is
the perceptions that are a _function of_ this real world that
are the constructions.
PCT doesn't really have much to say about the workings of
the brain, does it?
Yes, of course it does. Read B:CP, for example.
When your wife walks into the room and you recognize her,
are you constructing a pattern or recognizing one?
I am constructing a pattern perception. When the value of
that (analog) variable matches my reference for "Linda"
(one of the configurations in the category "babe") I have
recognized her.
Is the process a control process?
Yes.
What perceptions are you controlling?
I could be controlling for "looking for my wife" or I could
be controlling for categorizing the person coming towards me
as "familiar" or "unfamiliar" or I could be controlling for
answering the question ""who is that"?
Was the process constructed as the result of random
reorganization of some control system?
As I recall, studies of infacts suggest that the face perception
functions are constructed by a reorganization process (possibly
random) that is apparently completed at quite a young age (before
the child is 6 months old, I think).
Is there a separate control system for each of the objects
you recognize? How is the appropriate control system activated?
Good questions. How about suggesting the research to test them.
Best regards
Rick
···
--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: marken@mindreadings.com
mindreadings.com