Fw: [COMPLEX-M] Engineering Perception

I note that Bill Powers' remarks on my posting to this list are less

acerbic

than Mary Powers. However, since I would imagine that I am the list member
referred to in his comments by stating that "at least one member of the
complexity list seems to find it boring to consider how the "bits and
pieces" of the brain relate to perception, yawn, yawn". I have to say that
my first reaction to this was to reply equally dismissively. However I will
just say that having spent a considerable portion of my life dealing with
people with defined lesions and consequent disabilities I have a great
interest in how these, what I unfortunately called "bits and pieces",
actually do distort behaviour. AND HENCE PERCEPTION.
I will merely reiterate to Powers that my view of his book Behaviour: The
control of Perception was favourable in the fact that it was a seminal

break

from previous arguments.
That Powers cannot understand that his approach should be rejected is his
problem not mine. He says that he takes the stand of a physical scientist
and engineer. That is a valid approach if one is trying to build a machine
which can simulate perception in that its behaviour modifies its internal
model/programmes, stores incoming data, corrects this data on further

inputs

and continually tests its models by behaving in the real world and again
updating and so on. This is the same model essentially as that of Richard
Gregory. There is no reason at all why such a machine should not work and
appear to be a perfect model of perception and behaviour in its operation

as

specified by an observer. It is just not how biological perception works.
Perception is being in the world. What the human perceives is the result of
evolution and is adequate and trustworthy for a specific animal in its

given

niche. Bees and dragonflies, toads and hedgehogs and horse perceive very
well in their niche. Perception is a resonance phenomenon of the total
organism. that there are processes involved in this and that there are
feedback mechanisms etc in the functioning of the system is not in doubt.
But the processes do not produce images in the human machine control system
inside the head. Perception is the behaving in the real world and that is
the only one we have. That there are aspects of reality which are beyond

our

organic perceptual systems is not in doubt. Reality is much, much, bigger.
What we have done as humans is to make machined which an produce images

for

us as if they constituted reality. We cannot see the far off galaxies and
structures which make up our universe. No astronomer has ever seen what
radio waves convey until they are made into images. We look with amazement
at pictures of far distant entities and believe we are seeing the thing
itself. Humanity has always investigated the invisible and made more and
more instruments to do so. Our niche expands as we do so. We have made
prostheses to enable us to expand our reality but this will never cease.

Our

hypotheses are tested in reality, Gregory is correct in this; but these are
the hypotheses of cognition and not of perception. I do not expect for one
moment that Powers will even contemplate this alternative view since to him
it would seem that engineering is the only answer. I would merely say that

I

···

-----Original Message-----
From: kim <kim@PSY.CO.UK>
To: COMPLEX-M@HOME.EASE.LSOFT.COM <COMPLEX-M@HOME.EASE.LSOFT.COM>
Date: 10 October 1999 16:48
Subject: [COMPLEX-M] Engineering Perception

am as scientific as he ever has been. Just that my starting point and end
point are different.

kim@psy.co.uk

http://www.psy.co.uk
Psi INTERNATIONAL web site

[Reminder: To alter your list options: go to
http://home.ease.lsoft.com/scripts/wa.exe?SUBED1=complex-m]

[From Rick Marken (991010.1640)]

Forwarded from COMPLEX-M by Paul Stokes:

That Powers cannot understand that his approach should be
rejected is his problem not mine.

I really like this kind of honesty. Could you tell this fellow
that I have spent over 20 years looking for a reason to reject
Powers' model and haven't found it yet. So I would _really_
like to know why Powers' approach should be rejected.

Best

Rick

···

--

Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken/

[From Bruce Nevin (991010.2130 EDT)]

···

At 07:31 PM 10/10/1999 +0100, Paul A. Stokes wrote:

-----Original Message-----
From: kim <kim@PSY.CO.UK>
To: COMPLEX-M@HOME.EASE.LSOFT.COM <COMPLEX-M@HOME.EASE.LSOFT.COM>
Date: 10 October 1999 16:48
Subject: [COMPLEX-M] Engineering Perception

[Powers] says that he takes the stand of a physical scientist
and engineer. That is a valid approach if one is trying to build a machine
which can simulate perception in that its behaviour modifies its internal
model/programmes, stores incoming data, corrects this data on further inputs
and continually tests its models by behaving in the real world and again
updating and so on. This is the same model essentially as that of Richard
Gregory. There is no reason at all why such a machine should not work and
appear to be a perfect model of perception and behaviour in its operation as
specified by an observer.

Is this intended to be a description of perceptual control theory? It
sounds more like some version of cognitive psychology.

[From Bill Powers (991011.0016 MDT)]

That Powers cannot understand that his approach should be rejected is his
problem not mine. He says that he takes the stand of a physical scientist
and engineer. That is a valid approach if one is trying to build a machine
which can simulate perception in that its behaviour modifies its internal
model/programmes, stores incoming data, corrects this data on further

inputs

and continually tests its models by behaving in the real world and again
updating and so on. This is the same model essentially as that of Richard
Gregory. There is no reason at all why such a machine should not work and
appear to be a perfect model of perception and behaviour in its operation

as

specified by an observer. It is just not how biological perception works.
Perception is being in the world.

That's not a very illuminating idea (perception is "being in the world.").
"Biological perception" is just perception, isn't it? Well, I should know
better than to ask, given the above paragraph. I can only say that its
author remains unfamiliar with PCT and apparently thinks it is something
else. Either that, or my seminal (but otherwise useless) book was not about
what I think it was about.

I retire from the field defeated by superior reasoning, or something, and
will leave Dr. James to his convictions.

Best,

Bill P.

[Martin Taylor 991011 14:43]

···

-----Original Message-----
From: kim <kim@PSY.CO.UK>
To: COMPLEX-M@HOME.EASE.LSOFT.COM <COMPLEX-M@HOME.EASE.LSOFT.COM>
Date: 10 October 1999 16:48
Subject: [COMPLEX-M] Engineering Perception

------------------------------
Back from Copenhagen to find another 150 messages waiting to add to the
2200 already in my in-box:-(
------------------------------

I find it hard to reconcile Kim's comments with my understanding of PCT.
For example, as an argument against PCT, Kim says:

But the processes do not produce images in the human machine control system
inside the head.

as if there was somewhere in PCT a construct that said there were images
inside the head!!! Where in Powers' writings could that notion have come
from?

What the human perceives is the result of
evolution and is adequate and trustworthy for a specific animal in its

given

niche.

Of course, reorganization works within an organism that has evolved to
be predisposed to perceive variations in the environment that have proved
useful to the survival of its ancestors. But evolution cannot be
everything. If it were, we would all perceive the same things,
and we don't. PCT says that we all reorganize (learn) to perceive those
things that our behaviour permits and encourages--those things that are
stably affected by our actions, and what those things are is going to
depend on the environment in which we learn to perceive.

Bees and dragonflies, toads and hedgehogs and horse perceive very
well in their niche. Perception is a resonance phenomenon of the total
organism.

Perception is the behaving in the real world and that is
the only one we have.

Isn't that the fundamental PCT notion? That the perceptual signals that
exist are generated by perceptual functions that _depend_ on the stability
of behaviour in the (assumed to exist) real world? It's how the world
reacts when an organism acts on it that determines what kinds of
perception will continue to exist as the organism matures.

PCT reorganization theory says that perceptions that fail to be stably
affected by behaviour are likely to alter over time. As far as I can
understand PCT, these statements of Kim's describe one aspect of PCT,
rather than criticizing it. All perception is in and of the world
outside the perceiving apparatus, but totally dependent on the way the
world changes when acted upon. One cannot separate the perception in the
long term (reorganization time-scale) from the behaviour that is the means
of controlling it, though in the short term (control loop time-scale) the
perceptual function is independent of the behaviour.

[Parenthetically, it is this aspect of reorganization that argues most
strongly for the general veridicality of perception. If most perceptions
were unrelated to stable properties of the real world, actions on that
world would not have allowed us or our ancestors to stabilize our/their
internal states against events impinging from that real world, and we
would not be here. -- this in response both to Kim and to [Bill Powers
(991007.1446 MDT)] "It's such a simple problem that the answer to it
must be obvious; however, I haven't seen it. It's just this: how can we
demonstrate or prove that perceptions tell us about the real world?"
On the assumption that there is a real world, the argument is that on
balance, most perceptions cannot be too far from telling us correctly
what is in it, despite the obvious likelihood that some are wildly
wrong and none are _exactly_ right.]

I do not expect for one
moment that Powers will even contemplate this alternative view since to him
it would seem that engineering is the only answer.

What "alternative view"? What alternative does Kim propose to the idea
that biological processes conform to the same physical laws as do devices
constructed by humans? After all, that's the "engineering" claim, isn't
it?

If Kim wishes to be scientific, he might contemplate the thermodynamic
argument for PCT, which is that there are only two ways to ensure that
one's internal structure is sustained in an environment violent enough
to dissociate one's molecules at a microscopic level and one's flesh and
bones at a macroscopic level. Nature uses both means: (1) build a strong
shell that isolates the internal structure from external buffeting, and
(2) apply forces that oppose the buffeting forces.

Failing either of those (the second of which implies PCT), an organism
must (3) rebuild what structure is broken by external forces, which implies
that there must be a reference state for the structure and mechanisms to
bring the actual state toward the reference--but this again is PCT. The
argument almost becomes a _reductio ad absurdum_, in that if the organism
is not _totally_ isolated thermodynamically from its environment (which
even a black hole is not), then if it does not use PCT mechanisms
to protect its internal states, it has to use PCT mechanisms to
restore those states. If not PCT, then PCT;-)

I would be very interested if Kim could propose an alternative method of
sustaining an organism's internal structure against external buffeting,
while neither invoking PCT nor describing a system that is formally
identical to PCT.

Martin

[Martin Taylor 991014 9:46]

On Monday I posted the following message, but it has not arrived in my
CSG postbox. My ISP seemed to be having troubles with the mail system at
that time, because at the next attempt to get my CSG mail, I got mail I
had already read, but no new stuff. So I apologise if this is a duplicate
posting.

Martin

···

----------------------------------------------------------------------

[Martin Taylor 991011 14:43]

-----Original Message-----
From: kim <kim@PSY.CO.UK>
To: COMPLEX-M@HOME.EASE.LSOFT.COM <COMPLEX-M@HOME.EASE.LSOFT.COM>
Date: 10 October 1999 16:48
Subject: [COMPLEX-M] Engineering Perception

------------------------------
Back from Copenhagen to find another 150 messages waiting to add to the
2200 already in my in-box:-(
------------------------------

I find it hard to reconcile Kim's comments with my understanding of PCT.
For example, as an argument against PCT, Kim says:

But the processes do not produce images in the human machine control system
inside the head.

as if there was somewhere in PCT a construct that said there were images
inside the head!!! Where in Powers' writings could that notion have come
from?

What the human perceives is the result of
evolution and is adequate and trustworthy for a specific animal in its

given

niche.

Of course, reorganization works within an organism that has evolved to
be predisposed to perceive variations in the environment that have proved
useful to the survival of its ancestors. But evolution cannot be
everything. If it were, we would all perceive the same things,
and we don't. PCT says that we all reorganize (learn) to perceive those
things that our behaviour permits and encourages--those things that are
stably affected by our actions, and what those things are is going to
depend on the environment in which we learn to perceive.

Bees and dragonflies, toads and hedgehogs and horse perceive very
well in their niche. Perception is a resonance phenomenon of the total
organism.

Perception is the behaving in the real world and that is
the only one we have.

Isn't that the fundamental PCT notion? That the perceptual signals that
exist are generated by perceptual functions that _depend_ on the stability
of behaviour in the (assumed to exist) real world? It's how the world
reacts when an organism acts on it that determines what kinds of
perception will continue to exist as the organism matures.

PCT reorganization theory says that perceptions that fail to be stably
affected by behaviour are likely to alter over time. As far as I can
understand PCT, these statements of Kim's describe one aspect of PCT,
rather than criticizing it. All perception is in and of the world
outside the perceiving apparatus, but totally dependent on the way the
world changes when acted upon. One cannot separate the perception in the
long term (reorganization time-scale) from the behaviour that is the means
of controlling it, though in the short term (control loop time-scale) the
perceptual function is independent of the behaviour.

[Parenthetically, it is this aspect of reorganization that argues most
strongly for the general veridicality of perception. If most perceptions
were unrelated to stable properties of the real world, actions on that
world would not have allowed us or our ancestors to stabilize our/their
internal states against events impinging from that real world, and we
would not be here. -- this in response both to Kim and to [Bill Powers
(991007.1446 MDT)] "It's such a simple problem that the answer to it
must be obvious; however, I haven't seen it. It's just this: how can we
demonstrate or prove that perceptions tell us about the real world?"
On the assumption that there is a real world, the argument is that on
balance, most perceptions cannot be too far from telling us correctly
what is in it, despite the obvious likelihood that some are wildly
wrong and none are _exactly_ right.]

I do not expect for one
moment that Powers will even contemplate this alternative view since to him
it would seem that engineering is the only answer.

What "alternative view"? What alternative does Kim propose to the idea
that biological processes conform to the same physical laws as do devices
constructed by humans? After all, that's the "engineering" claim, isn't
it?

If Kim wishes to be scientific, he might contemplate the thermodynamic
argument for PCT, which is that there are only two ways to ensure that
one's internal structure is sustained in an environment violent enough
to dissociate one's molecules at a microscopic level and one's flesh and
bones at a macroscopic level. Nature uses both means: (1) build a strong
shell that isolates the internal structure from external buffeting, and
(2) apply forces that oppose the buffeting forces.

Failing either of those (the second of which implies PCT), an organism
must (3) rebuild what structure is broken by external forces, which implies
that there must be a reference state for the structure and mechanisms to
bring the actual state toward the reference--but this again is PCT. The
argument almost becomes a _reductio ad absurdum_, in that if the organism
is not _totally_ isolated thermodynamically from its environment (which
even a black hole is not), then if it does not use PCT mechanisms
to protect its internal states, it has to use PCT mechanisms to
restore those states. If not PCT, then PCT;-)

I would be very interested if Kim could propose an alternative method of
sustaining an organism's internal structure against external buffeting,
while neither invoking PCT nor describing a system that is formally
identical to PCT.

Martin