Sameness

[From Rick Marken 920623.0945)]

Martin Taylor (920622) says:

As I see it, the key problem is more like: third, how does the brain identify
that there exists N objects of a certain kind rather than one strong exemplar
of the kind.

Very interesting question.

Let me make sure I've got it. Is the question something like " How
can I recognize a repetition of the same pattern - like AAAAA? Doesn't
the "A" detector get "used up" with just one?"

One obvious answer is that there must be several detectors for the
same sensation, configuration, transition, etc. How this is architected
is the big question. How does a model like "pandemomnium" solve this
problem? Or does it?

If there is an ECS controlling for the perception of an X, it
should satisfy its reference if an X is in its input. What distinguishes
thes existence of exactly three Xs in its input from the existence of one X.

Again. I'd guess that "three X's" requires three X detectors firing
maximally and simultaneously. The "threeness" must be detected by a
count detector that takes the X detector outputs as inputs. How that's
done, I don't know. Looks like a research/modelling problem to me.
Get those grad students on it, Martin.

But problems such as this one are
fairly central, I think, and must be solved within the "natural" hierarchic
structure if HPCT is to be taken seriously as the instantiation of PCT that
corresponds to real living beings.

How are these "central" problems solved by existing cognitive models? These
models might give some hints about possible mechanisms. But I think detailed
research might be the best first step. For example, how about a study
in which a person controls N of the same objects simultaneously? See how
control varies as a function of N. The problem would be to keep the number
of output degrees of freedom constant while the perceptual degrees of
freedom increase.

     The basic statement of PCT, that behaviour
is all and only the control of perception, seems incontrovertible. How that
fact is developed into structure is not.

Right! But the idea of behavior as controlled perception is THE important
concept -- the resulting shape of the PCT structure will start to emerge
once there are more than three people in the world who are doing research
and modeling based on an understanding of the fact that behavior IS control.

HPCT seems a very sensible proposal,
but there may well be other equally sensible instantiations, and not all
living things necessarily use the same instantiation.

No doubt.

Simplicity is to be preserved where possible in science, and HPCT is a simple
structure. If it can solve problems as basic to perception as "there's an
X and there's another" or "I see a lot of Xs among the Ys" without the
introduction of new structures or concepts, so much the better. At present
I don't see the answer.

There are TONS of unanswered questions -- fortunately for those of use who
don't have billions of dollars to entertain ourselves with and, thus, are
forced to get our kicks from by trying to understand nature. I don't see
the immediate lack of a definitive answer to the question "how do you
see a lot of Xs among the Ys" as posing any more fundemental challenge to
PCT than it does to S-R or cognitive models. It's just another interesting
question -- and one that is far more interesting when asked in the context
of an organizational model of behavior that actually works (PCT).

But I think there should be something we could learn from existing models,
like neural networks. After all, such models are just models of pieces of
a control organization -- posing as a complete model of behavior.

By the way, Martin (and other speech/language afficionados) -- any comments
about the sound adaptation study that I mentioned in an earlier post?

Regards

Rick

ยทยทยท

**************************************************************

Richard S. Marken USMail: 10459 Holman Ave
The Aerospace Corporation Los Angeles, CA 90024
E-mail: marken@aero.org
(310) 336-6214 (day)
(310) 474-0313 (evening)

[Martin Taylor 920623 13:30]
(Rick Marken 920623.0945)

As I see it, the key problem is more like: third, how does the brain identify
that there exists N objects of a certain kind rather than one strong exemplar
of the kind.

Let me make sure I've got it. Is the question something like " How
can I recognize a repetition of the same pattern - like AAAAA? Doesn't
the "A" detector get "used up" with just one?"

I'm no longer trying to restate what I thought Bruce was getting at, but
speaking for myself alone...No, that's not really the question. More like
how do I know there are three A's in BAbNvXCA3kiAlaft without explicitly
counting them in serial order, bringing each into central attention separately.

How does a model like "pandemomnium" solve this problem? Or does it?

I know of no satisfactory neural net proposal that solves this problem.

Again. I'd guess that "three X's" requires three X detectors firing
maximally and simultaneously. The "threeness" must be detected by a
count detector that takes the X detector outputs as inputs. How that's
done, I don't know. Looks like a research/modelling problem to me.
Get those grad students on it, Martin.

I don't think this solves the problem. How do these "X" detectors know that
they are each looking at a different examplar?

I don't have students. I work in a government lab, and can engage contractors
on research if the authorities agree that the problem is worth spending
money on.

But problems such as this one are
fairly central, I think, and must be solved within the "natural" hierarchic
structure if HPCT is to be taken seriously as the instantiation of PCT that
corresponds to real living beings.

How are these "central" problems solved by existing cognitive models? These
models might give some hints about possible mechanisms. But I think detailed
research might be the best first step. For example, how about a study
in which a person controls N of the same objects simultaneously? See how
control varies as a function of N. The problem would be to keep the number
of output degrees of freedom constant while the perceptual degrees of
freedom increase.

I'm not trying to compare HPCT with any other model. I take it to be a claim
of HPCT that it is the "theory of everything" for psychology, and indeed, I
have been trying to turn people on to it by saying that HPCT is to psychology
what Newton's laws were to physics 300 years ago. What I am trying to do is
to justify that claim, because I at present believe it to be true. Hence...

I don't see
the immediate lack of a definitive answer to the question "how do you
see a lot of Xs among the Ys" as posing any more fundemental challenge to
PCT than it does to S-R or cognitive models. It's just another interesting
question -- and one that is far more interesting when asked in the context
of an organizational model of behavior that actually works (PCT).

is perhaps true, but somewhat off the point. Seeing multiple things is an
obvious and low-level ability that sufficiently complex living things have.
It is at the level where HPCT should be able to provide a natural answer.
Sure, there are tons of questions. It might be a good idea to develop an
inventory of ones that seem simple to pose and that should be answerable
either by analysis or through experiment within the (H)PCT paradigm.

Martin