Beliefs reconciled

[Martin Taylor 990106 11:11]

[From Rick Marken (990106.0750)]

The disputation is about a different assertion--that _all_ beliefs are
reference providers and only reference providers, and that _no_ belief
is a perception, imaginary or otherwise.

Then it was an unnecessary dispute since no one on the "beliefs are
references" side -- not Bill Powers, not me -- said that beliefs are
_only_ reference perceptions.

I'm glad to hear you say that, because I've understood something very
different from some of your messages. I'll quote just one example of
several possibilities.

I think I agree with pretty well all of the message to which I am now
replying (but I reserve the right to change my mind later, in case I
have interpreted some of your words differently than you intended them).
Here's the example...

-------example 1--------

[From Rick Marken (990104.1020)]

Martin Taylor (990104 12:00) --

One can test whether belief B is a controlled perception--i.e.
that it has a reference value B', and actions are performed to
maintain the value of B near B'.

I think B' is what we would call the "belief"; if B is a
perceptual variable then it's what you actually perceive.
...
When I say "I believe that god created all living things in
two 24 hour days" I am not describing what I perceive but what
I _want_ to perceive (or imagine)

-------end of example------

On several occasions, I have presented what I thought was a simple example
of a belief that I was not controlling (an imagined perception of a
state of the world or of the way the work works) and in each case you
turned it around and said it was in fact a reference level. You did so
in spite of the fact that I presented the example with the assertion
that I really did not care whether the belief had to be changed
when contrary evidence presented itself--no resistance to disturbance of
the belief. So I had reason to hold the belief that you thought
beliefs were _only_ reference levels. I am glad now to alter that
belief.

When you say that Bill Powers did not say that beliefs are only reference
perceptions, I find it hard to reconcile with Bill's own statement

Bill Powers (990105.0903 MST)

I think it more likely that what we call a belief is a reference-perception
that specifies the "right" state for some variable perception which can be
in any state.

and (same message)

In fact, in all cases I can think of, when there is a deviation of the
perceived world from the state defined by a belief, it is the perception
that changes while the belief remains the same, implying that the belief is
the reference signal.

But Bill can speak for himself on this. I'm glad that you, at least, agree
that some beliefs are just perceptions, while other beliefs, being
controlled, have a reference value and outputs that set reference values
for other perceptions.

Martin

[From Rick Marken (990106.1010]

Martin Taylor (990106 11:11) --

I'm glad that you, at least, agree that some beliefs are just
perceptions

Imagined perceptions. I just don't ever use the word "belief"
to refer to something I am perceiving. I do use it to refer to
an _interpretation_ of what I am perceiving (for example, when
I intrepret my perception of the behavior of some congressman
as being that of a "hypocritical asshole"; I say "I believe Henry
Hyde is a hypocritical asshole", for example). The interpretation
is an imagined perception; the "yes" vote for impeachment is
an actual perception; no belief involved.

Best

Rick

···

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

[Martin Taylor 990106 14:00]

[From Rick Marken (990106.1010]

Martin Taylor (990106 11:11) --

I'm glad that you, at least, agree that some beliefs are just
perceptions

Imagined perceptions. I just don't ever use the word "belief"
to refer to something I am perceiving. I do use it to refer to
an _interpretation_ of what I am perceiving (for example, when
I intrepret my perception of the behavior of some congressman
as being that of a "hypocritical asshole"; I say "I believe Henry
Hyde is a hypocritical asshole", for example). The interpretation
is an imagined perception; the "yes" vote for impeachment is
an actual perception; no belief involved.

This distinction is a bit problematic. What perception is _not_ an
interpretation, in your view?

I do agree, though about the correction to "imagined" perceptions. It
is, after all, what I suggested in my initial foray into this area.
However, though this may be arguable, I doubt one can ever assert that
a particular perception is devoid of current external input, or that
any particular perception is devoid of any contribution from
imagination. It's more a matter of degree than a difference of kind, I
think (I believe?).

···

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

There is, however, a much more significant issue relating to the
structure of the perceptual control hierarchy, and this may be where
a lot of the earlier confusion arises. In the hierarchy considered
without imagination, every perceptual signal is the result of a
perceptual input function acting on a set of lower level perceptual
signals or sensory inputs. Every perceptual signal is sent to either
or both of two places--the comparator in an elementary control unit
that controls this particular perceptual signal, and to an input of a
higher-level perceptual input function (PIF).

If the "imagination loop" is included in a single elementary control
unit, it is as a connection between the output and the perceptual
signal. It is a transfer function that produces the expected effect
of the output on the perceptual signal (apart from time-compression).
This transfer function is, in some sense, a model of the environmental
feedback function that connects this same output to this same perceptual
signal through the outer environment. The output of this transfer
function acts like a perceptual signal, and presumably is an "imagined
perception."

The issue, in these terms, is that some of the "beliefs" that have been
bandied about are like perceptions of the form of the transfer function,
rather than like its output, and some are like perceptions of the states
of parameters of the transfer function. Where do these perceptions fit
in a diagram of the hierarchy? It looks to me as if one has to allow
perceptual signals to go to a third place in addition to their own
comparator and the next-level PIFs. They may also go to an input of
the imagination-loop transfer function.

Maybe I'm missing something important here, or ignoring something I ought
to have known. But I don't know where. If a belief is a (largely)
imagined perception, by what route is it instantiated? Whence come the
inputs to the (I)PIF of which it is the output? For it is those inputs
that must be controlled if the belief is to act as a source of reference
levels.

Martin

[From Rick Marken (990106.1520)]

Martin Taylor (990106 14:00) --

What perception is _not_ an interpretation, in your view?

All perception. The only reason we think perception is an
interpretation is because some smart philosophers noted long ago
that we don't have a direct line to reality; whatever is out there
(if anything -- yikes;-)) on the other side of our perceptual
systems, we will never have direct access to it, except as a model
developed by scientists who are even smarter than the philosophers.
Scientists develop models to explain the black box (see CSG Logo at
the CSG site) that is the "real world" on the other side of our
perceptual systems; and they test these models by seeing if the
predicted perceptions occur.

Before I understood all this high falutin' philosophy stuff I
didn't think of what I saw, felt and heard as "perception"; it
was reality, as far as I was concerned. And even now, when I
look at the world I deal with it -- the computer, the keyboard,
the people, etc-- as though it were reality. Perception is not
an "interpretation" until you've been through lots to schoolin'.
It's that pre-schoolin' perception I'm talking about. We don't
typically believe or disbelieve our perceptions; I don't ask
myself if that's really a Mac G3 on my desk or a cleverly
disguised armadillo. It just is what I see.

The rest of your questions are way too complicated for me. I'll
be happy if I can just figure out an answer to the question: how
does an outfielder catch fly a ball on a windy day?

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 Gregory (990106.1725 EST)]

Martin Taylor 990106 14:00

The issue, in these terms, is that some of the "beliefs" that
have been
bandied about are like perceptions of the form of the
transfer function,
rather than like its output, and some are like perceptions of
the states
of parameters of the transfer function. Where do these perceptions fit
in a diagram of the hierarchy? It looks to me as if one has to allow
perceptual signals to go to a third place in addition to their own
comparator and the next-level PIFs. They may also go to an input of
the imagination-loop transfer function.

Maybe I'm missing something important here, or ignoring
something I ought
to have known. But I don't know where. If a belief is a (largely)
imagined perception, by what route is it instantiated? Whence come the
inputs to the (I)PIF of which it is the output? For it is those inputs
that must be controlled if the belief is to act as a source
of reference
levels.

Very good questions. I'd like to understand this too.

Bruce Gregory

[From Bruce Gregory (990106.2000 EST)]

Rick Marken (990106.1520)

Before I understood all this high falutin' philosophy stuff I
didn't think of what I saw, felt and heard as "perception"; it
was reality, as far as I was concerned. And even now, when I
look at the world I deal with it -- the computer, the keyboard,
the people, etc-- as though it were reality. Perception is not
an "interpretation" until you've been through lots to schoolin'.
It's that pre-schoolin' perception I'm talking about. We don't
typically believe or disbelieve our perceptions; I don't ask
myself if that's really a Mac G3 on my desk or a cleverly
disguised armadillo. It just is what I see.

You really ought to read _Visual Intelligence_ by Donald Hoffman. Then
again, its likely to prove too high falutin' for your tastes.

The rest of your questions are way too complicated for me. I'll
be happy if I can just figure out an answer to the question: how
does an outfielder catch fly a ball on a windy day?

I hope Bill will not be content with quite so superficial a response. Your
answer implies a less than impressive grasp of the issues for one who speaks
so authoritatively. Might you be simply a _true believer_ after all? More's
the pity.

Bruce Gregory

[From Rick Marken (990106.2020 PST)]

Me:

Before I understood all this high falutin' philosophy stuff I
didn't think of what I saw, felt and heard as "perception"; it
was reality, as far as I was concerned...Perception is not an
"interpretation" until you've been through lots to schoolin'.

Bruce Gregory (990106.2000 EST) --

You really ought to read _Visual Intelligence_ by Donald Hoffman.
Then again, its likely to prove too high falutin' for your tastes.

I looked it up at amazon.com. It looks like a pretty nice, standard
book on visual perception. What does Hoffman say that I ought to
know? I believe in the physics and neural models so I believe
that perception is a function ("interpretation") of an external
reality that doesn't "look like" what we see. All I was saying is
that the notion that perception is an _interpretation_ is an
intellectual notion. We don't experience our perception as an
"interpretation", even when we are experiencing perceptions that
we know, based on reasoning, could not "really" be happening (like
the movement of the cliffs in the waterfall illusion). At least,
that's the way I experience my perception; it's reality to me.

Visual illusions are fun because we know, intellectually, that what
seems to be happening couldn't really be happening; illusions make
us aware, through reasoning, of the fact that what we perceive is the
result of neural computation. But the perceptions themselves don't
seem like an interpretation; they seem like what's going on in
reality. That's why books like Hoffman's are so much fun; they call
attention to the fact that what we take for granted as "reality" is,
in fact, just a function (interpretation) of something we will never
experience directly.

Best

Rick

···

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

[Martin Taylor 990107 0:10]

[From Rick Marken (990106.1520)]

Martin Taylor (990106 14:00) --

What perception is _not_ an interpretation, in your view?

All perception. The only reason we think perception is an
interpretation is because some smart philosophers noted long ago
that we don't have a direct line to reality;

This is a simple cop-out, isn't it? I asked you to distinguish between
what you say is "belief"--an interpretation--and "perception", when
we just finished agreeing that belief was an "imagined perception."

This was the context...

[From Rick Marken (990106.1010]

Martin Taylor (990106 11:11) --

I'm glad that you, at least, agree that some beliefs are just
perceptions

Imagined perceptions. I just don't ever use the word "belief"
to refer to something I am perceiving. I do use it to refer to
an _interpretation_ of what I am perceiving ...

Does this mean that to you an "imagined perception" is not a perception,
since no perception is an interpretation, a belief is an interpretation,
and a belief is an imagined perception?

If an "imagined perception" is not a perception, your use of the word
is a bit misleading.

···

---------------
However...

You didn't comment on my observation that no perception is purely imagined,
and probably no perception is devoid of imagination. So I ask again how you
distinguish between perceptions that are interpretations--beliefs--and
perceptions that are not. (To put my view on the table, any perceptual
input function constitutes an interpretation of its inputs, so clearly
Rick's meaning of "interpretation" is different from mine).

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

The rest of your questions are way too complicated for me.

But they go to the heart of the question of what constitutes an "imagined
perception." I think that when one uses "imagined perception" as a basis
for discussing some properties of PCT, one ought at least to have some
notion of where these imagined perceptions might come from, and where
they appear in the hierarchy. If you don't have such an idea, all you
are doing is word-spinning. I had the impression that your position on
the matter was very clear to you. Mine is only as clear to me as I stated
in Martin Taylor (990106 14:00), though a good diagram might make what
I said a little more intelligible.

Martin

[From Bill Powers (990107.0000 MST)]

Martin Taylor 990106 14:00 --

I can agree that some of the things we call beliefs are imagined
perceptions, if we're talking about imagining something rather than
perceiving it in the normal way. I don't agree that ALL the things we call
beliefs are imagined perceptions. And maybe none of them is.

One of the problems is where imagined perceptions come from. According to
hypothesis, they are short-circuited copies of the reference signals that a
system would otherwise send to lower systems. They become perceptual
signals in the higher system, just as if the lower systems had succeeded
instantly and perfectly in making the environment and the lower systems
create that perception in the higher system.

This process is therefore driven, as usual, by the reference signal
entering the higher system. The imagined perception is made to match that
reference signal perfectly, without the usual constraints that exist when
the action passes through the environment. In imagination, I can lift my
car with one hand to see if my wallet skidded underneath it when I dropped
the wallet. That is, of course, impossible to do in the normal mode.

This particular case, as described, is NOT "model-based control." The link
between action and perception is NOT a model of the lower environmental
feedback functions. It is simply a direct connection that ignores all
practical limitations and constraints. That is one valuable aspect of
imagination: one can solve the higher-level control problem without having
to deal with the lower-level details. Lifting the car to look underneath it
would in fact solve the problem of looking for my wallet, if I could do it.
When the time comes to consider the details, of course, I would have to
reject that solution. But other solutions found in this way would turn out
to be practicable.

Notice that in order to imagine a perception under the present model, one
must first set a reference level for it, specifying the desired degree of
that perception. The perception doesn't "just happen." In the imagination
mode, there is never any significant difference between perception and
reference signal, because no disturbances can affect the perception. So it
makes no difference whether we say that kind of belief is a reference
signal or a perceptual signal: they will be essentially identical in
magnitude.

However, when such a belief is subject to a reality test -- when the
perception comes from lower systems rather than imagination -- it is
possible that the perceptual signal will no longer match the reference
signal. I may believe I can lift the car, but when I try it, I do not
perceive the car being lifted. This, if I choose to pay attention, is the
evidence that says my belief is wrong.

But my perception is NOT wrong: it reports that the car is still on the
ground, even though my reference level for it is still "off the ground."
The perception now reports the actual state of affairs. In the period just
before the belief is changed, then, which signal represents the (mistaken)
belief? Is it the perceptual signal reporting the car on the ground, or the
reference signal specifying that it is off the ground? I claim it is the
latter.

Best,

Bill P.

[From Oded Maler (990107)]

Rick Marken (990106.1520)]

Martin Taylor (990106 14:00) --

What perception is _not_ an interpretation, in your view?

All perception.

In the n-th time I recommend you read:

Quantum Psychology : How Brain Software Programs You and Your World
by Robert Anton Wilson
New Falcon Publications; ISBN: 1561840718

--Oded

[From Bruce Gregory (990107.0935 EDT)]

Rick Marken (990106.2020 PST)

First, let me apologize for my testy tone. I had a tooth extracted
yesterday afternoon, and it must have affected my otherwise sunny
disposition.

I looked it up at amazon.com. It looks like a pretty nice, standard
book on visual perception. What does Hoffman say that I ought to
know? I believe in the physics and neural models so I believe
that perception is a function ("interpretation") of an external
reality that doesn't "look like" what we see. All I was saying is
that the notion that perception is an _interpretation_ is an
intellectual notion. We don't experience our perception as an
"interpretation", even when we are experiencing perceptions that
we know, based on reasoning, could not "really" be happening (like
the movement of the cliffs in the waterfall illusion). At least,
that's the way I experience my perception; it's reality to me.

If you are at all like me, your beliefs differ considerably from you
see. Yet according to PCT, both are perceptions. Granted that this is an
interpretation, an intellectual notion if you like, but it still raises
questions that seem to me, at least, to be legitimate.

Visual illusions are fun because we know, intellectually, that what
seems to be happening couldn't really be happening; illusions make
us aware, through reasoning, of the fact that what we perceive is the
result of neural computation. But the perceptions themselves don't
seem like an interpretation; they seem like what's going on in
reality.

So do our beliefs, but they "look" very different. We attribute the
difference to the different sources of the perceptions. So while they
may all be perceptions, the perceptions are very different.

Bruce Gregory

[From Rick Marken (990107.0820)]

Martin Taylor (990107 0:10) --

This is a simple cop-out, isn't it? I asked you to distinguish
between what you say is "belief"--an interpretation--and
"perception", when we just finished agreeing that belief was
an "imagined perception."

Since, as Bill Powers (990107.0000 MST) notes, imagination (according
to the PCT model) is driven by the reference signal, I would say
that even imagined perceptions are beliefs that are, ultimately,
reference perceptions.

Does this mean that to you an "imagined perception" is not a
perception, since no perception is an interpretation

Perceptions, including imagined perceptions, can be fairly called
"interpretations", according to the PCT model. I'm just saying that
we don't _experience_ them (perceptions, anyway) as such.
Interpretation is a complex control process; we have learned (from
science) to "interpret" perception as an interpretation; but we (me,
anyway) don't _experience_ our perceptions as interpretations.

You didn't comment on my observation that no perception is purely
imagined

That's a theoretical guess. It may be the case but I doubt it.
Imagination has not been an important consideration in modeling
behavior so far. We are able to model some rather complex controlling
without assuming that any aspect of the controlled perception is
imagined. So I am inclined to believe that perception does _not_,
in general, have a large component of imagination. The evidence,
so far, is that perception (p) is a function of environmental (e)
variables (p = f(e)) and imagination (i) is a function of higher
level reference (r) signals serving as inputs to perceptual
functions (i = f(r)).

So I ask again how you distinguish between perceptions that are
interpretations--beliefs--and perceptions that are not.

We know from science that what we perceive is an interpretation.
In PCT p = f (e); the perceptual signal is a function of
environmental variables; f() is the _interpretation_ of the
environment. But this is _theory_. I don't experience my
perceptions as interpretations.

Bruce Gregory (990107.0935 EDT)--

If you are at all like me, your beliefs differ considerably from
you see. Yet according to PCT, both are perceptions.

No, they are _not_ both perceptions. I think we are using the work
"belief" to refer to _reference signals_ (what you _want_ to perceive).
What you perceive is what you perceive; if it matches what you believe
(your reference for the state of that perceptual variable) then the
perception is in the reference (belief) state. If the perception is not
in the reference (belief) state then, presumably, you will act to
bring that perception to the reference (belief) state.

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 Gregory (990107.1055 EST)]

Bill Powers (990107.0000 MST)

One of the problems is where imagined perceptions come from.
According to
hypothesis, they are short-circuited copies of the reference
signals that a
system would otherwise send to lower systems. They become perceptual
signals in the higher system, just as if the lower systems
had succeeded
instantly and perfectly in making the environment and the
lower systems
create that perception in the higher system.

How do we know that what we are perceiving is the result of a
short-circuited signal rather than a signal originating at lower levels?

This process is therefore driven, as usual, by the reference signal
entering the higher system. The imagined perception is made
to match that
reference signal perfectly, without the usual constraints
that exist when
the action passes through the environment. In imagination, I
can lift my
car with one hand to see if my wallet skidded underneath it
when I dropped
the wallet. That is, of course, impossible to do in the normal mode.

How are we able to distinguish such an "impossible" perception from a
possible perception?

This particular case, as described, is NOT "model-based
control." The link
between action and perception is NOT a model of the lower
environmental
feedback functions.

If we lack a model of the lower environmental feedback functions, how do
we know that lifting the car does not convert it into a chocolate
eclair, for example?

It is simply a direct connection that ignores all
practical limitations and constraints. That is one valuable aspect of
imagination: one can solve the higher-level control problem
without having
to deal with the lower-level details. Lifting the car to look
underneath it
would in fact solve the problem of looking for my wallet, if
I could do it.
When the time comes to consider the details, of course, I
would have to
reject that solution.

Don't you really reject this solution without considering details? In
fact, do you ever really entertain it as a solution?

Notice that in order to imagine a perception under the
present model, one
must first set a reference level for it, specifying the
desired degree of
that perception. The perception doesn't "just happen."

More accurately a higher level system must set a reference for it, no?

In the
imagination
mode, there is never any significant difference between perception and
reference signal, because no disturbances can affect the
perception. So it
makes no difference whether we say that kind of belief is a reference
signal or a perceptual signal: they will be essentially identical in
magnitude.

This seems to be an interpretation rather than experience. Perhaps a
belief is neither a perception nor a reference level.

But my perception is NOT wrong: it reports that the car is
still on the
ground, even though my reference level for it is still "off
the ground."
The perception now reports the actual state of affairs. In
the period just
before the belief is changed, then, which signal represents
the (mistaken)
belief? Is it the perceptual signal reporting the car on the
ground, or the
reference signal specifying that it is off the ground? I
claim it is the
latter.

This seems to suggest that any time we fail to match a reference level
it is a mistaken belief. Is this what you had in mind? If I swing my
racket and miss the ball, is this a demonstration of a mistaken belief?

Bruce Gregory

[From Bill Powers (990107.1040 MST)]

Bruce Gregory (990107.1055 EST)--

How do we know that what we are perceiving is the result of a
short-circuited signal rather than a signal originating at lower levels?

Because the imagined signal instantly follows changes in the reference
signal and is unaffected by external disturbances. I'm speaking, of course,
both of subjective impressions when I imagine something, and of predictions
from the model of imagination.

How are we able to distinguish such an "impossible" perception from a
possible perception?

Partly through logic (the world never behaved that way before) and partly
through incompleteness of imagined perceptions. A truly complete job of
imagining something would be like an hallucination. Sometimes people do go
that far, and have real (and scarey) problems telling imagination from
reality.

If we lack a model of the lower environmental feedback functions, how do
we know that lifting the car does not convert it into a chocolate
eclair, for example?

When you're imagining, you're NOT basing higher-level perceptual signals on
perceptions in the systems below the level where the imagination connection
is in effect. We aren't actually lifting the car, so what is "really"
happening is irrelevant. Actually, nothing is happening because the
reference signals are not reaching the lower systems.

Don't you really reject this solution without considering details? In
fact, do you ever really entertain it as a solution?

Of course not -- this was an example to get the point across. In other
cases, those where we're not so sure about what is possible, exactly this
sort of situation occurs. We hold an imaginary conversation with our boss,
persuading him to give us a raise, but when we actually try to have the
conversation, the boss doesn't do anything we imagined him to do. Or in
planning a job, we imagine all kinds of difficulties we must overcome --
but when we actually do the job, they don't turn up, and it's easy (one
origin of procrastination).

Notice that in order to imagine a perception under the
present model, one
must first set a reference level for it, specifying the
desired degree of
that perception. The perception doesn't "just happen."

More accurately a higher level system must set a reference for it, no?

Yes. I include "higher systems" in the less-precise term "one".

This seems to be an interpretation rather than experience. Perhaps a
belief is neither a perception nor a reference level.

One can only judge from the properties of beliefs. I'm presenting my
proposal. Perhaps a belief is chocolate syrup, but that's not what I'm
proposing.

This seems to suggest that any time we fail to match a reference level
it is a mistaken belief. Is this what you had in mind? If I swing my
racket and miss the ball, is this a demonstration of a mistaken belief?

Which kind of belief are you talking about? You could say that you swing
the racket to sustain the belief that you can hit the ball. When you miss,
you find this belief to be mistaken, as you suggest. You may then modify
your belief about your ability to hit the ball. But if you're imagining the
ball, your swinging is based on the belief that you're not hallucinating a
ball -- that there is a real ball there. That kind of belief is more like a
mistaken perception, a confusion between imagination and reality.

It's also possible that you're not imagining anything and are only trying
to hit the ball. In that case I wouldn't think that "belief" would be an
appropriate term.

Best,

Bill P.

[From Bruce Gregory (990107.1335 EST)]

Rick Marken (990107.0820)

No, they are _not_ both perceptions. I think we are using the work
"belief" to refer to _reference signals_ (what you _want_ to
perceive).
What you perceive is what you perceive; if it matches what you believe
(your reference for the state of that perceptual variable) then the
perception is in the reference (belief) state. If the
perception is not
in the reference (belief) state then, presumably, you will act to
bring that perception to the reference (belief) state.

O.K. Let me see if I have this right. A belief is the output of higher
level perceptual system that serves as a reference level for a lower
level system. There is no difference between a belief and any other
reference level for a perception. The only way you can identify your
beliefs is by short circuiting the higher level system so that it will
perceive its output which will be equivalent to the reference level for
the lower system. There is, therefore, a higher level perception
associated with every statement we consider to be a "fact". This higher
level system establishes the reference level for a lower level
perception. One higher level system presumably can establish a
perceptual reference for many lower level systems (I believe an apple
has a shape that falls within a certain range, a color that falls with a
different range, a taste that falls within yet another range...) If I
taste what appears to be an apple and the taste falls outside the range
(it tastes like a peach, for example), I alter the perceptual signal to
bring it back into the range established by the higher level system. To
accomplish this, there must be an environmental feedback loop that
allows me to alter the taste of something or I will short circuit the
taste control loop and experience an imaginary taste that matches the
output of the higher level system.

Is this correct?

Bruce Gregory

[From Bruce Gregory (990107.1415 EST)]

Bill Powers (990107.1040 MST)

Bruce Gregory (990107.1055 EST)--

>How do we know that what we are perceiving is the result of a
>short-circuited signal rather than a signal originating at
lower levels?

Because the imagined signal instantly follows changes in the reference
signal and is unaffected by external disturbances. I'm
speaking, of course,
both of subjective impressions when I imagine something, and
of predictions
from the model of imagination.

O.K. So it is the existence of disturbances that tells me that I am
perceiving rather than imagining something.

One can only judge from the properties of beliefs. I'm presenting my
proposal. Perhaps a belief is chocolate syrup, but that's not what I'm
proposing.

I like the chocolate syrup model. I'll keep it in reserve.

It's also possible that you're not imagining anything and are
only trying
to hit the ball. In that case I wouldn't think that "belief"
would be an
appropriate term.

So belief only applies when we are operating in imagination mode?

Bruce Gregory

[From Rick Marken (990107.1300)]

Bruce Gregory (990107.1335 EST)--

O.K. Let me see if I have this right.

I think you have it right up to here:

If I taste what appears to be an apple and the taste falls outside
the range (it tastes like a peach, for example), I alter the
perceptual signal to bring it back into the range established
by the higher level system. To accomplish this, there must be
an environmental feedback loop that allows me to alter the
taste of something or I will short circuit the taste control
loop and experience an imaginary taste that matches the
output of the higher level system.

Imagination is certainly one way with the problem. I, however,
would probably deal with it by throwing away the peach- tasting
apple and getting a new one that (hopefully) tastes like an apple.
If I can't find such an apple I'll probably just give up and have
a nectarine.

Bruce Gregory (990107.1415 EST) --

So belief only applies when we are operating in imagination
mode?

We (Bill and I) are trying to see how some phenomena described
by the somewhat vague term "belief" can be understaood in terms
of PCT. Apparently, you (and some others) don't like what we
are saying. That's fine. But why keep your point of view secret?
How about describing the phenomenon of belief that is interesting
to you and give us the explanation that makes sense to you. I'm
having trouble understanding what it is about what we are saying
that is causing you so much apparent grief.

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 Gregory (990107.1710 EST)]

Rick Marken (990107.1300)]

Bruce Gregory (990107.1335 EST)--

> O.K. Let me see if I have this right.

I think you have it right up to here:

> If I taste what appears to be an apple and the taste falls outside
> the range (it tastes like a peach, for example), I alter the
> perceptual signal to bring it back into the range established
> by the higher level system. To accomplish this, there must be
> an environmental feedback loop that allows me to alter the
> taste of something or I will short circuit the taste control
> loop and experience an imaginary taste that matches the
> output of the higher level system.

Imagination is certainly one way with the problem. I, however,
would probably deal with it by throwing away the peach- tasting
apple and getting a new one that (hopefully) tastes like an apple.
If I can't find such an apple I'll probably just give up and have
a nectarine.

But if you believed that all Jews were stingy and you encountered a
generous Jew, would you simply go looking for a stingy Jew to match your
desired perception? And if you couldn't find one, would you settle for a
generous Gentile? I'm trying to reconcile what you are telling me about
belief with the notion "I'll see that because I believe it." I don't
seem to be having much luck, however.

Bruce Gregory (990107.1415 EST) --

> So belief only applies when we are operating in imagination
> mode?

We (Bill and I) are trying to see how some phenomena described
by the somewhat vague term "belief" can be understood in terms
of PCT. Apparently, you (and some others) don't like what we
are saying. That's fine. But why keep your point of view secret?
How about describing the phenomenon of belief that is interesting
to you and give us the explanation that makes sense to you. I'm
having trouble understanding what it is about what we are saying
that is causing you so much apparent grief.

You misunderstood me. Having my tooth pulled caused me grief, this
discussion only causes me slight confusion. It seems to me you are
telling me that if I believe the Earth goes around the Sun I will
perceive the Earth going around the Sun. Well, I do believe the former,
but I do not seem to perceive the latter. I am also trying to understand
why some reference signals seem to differ from others. I pick up a glass
of water by making what I perceive match my reference level, but I do no
such thing when I believe the Earth orbits the Sun. I control a
perception in the first case, but I don't seem to control a perception
in the second case. Beliefs seem to me to be statements that I affirm,
while purposeful behavior seems to involve intended outcomes I achieve.
I am having trouble seeing that they are "really" the same thing.

Bruce Gregory

[From Bill Powers (990107.1618 MST)]

Bruce Gregory (990107.1415 EST)--

So belief only applies when we are operating in imagination mode?

According to my usage of the term, that would make sense. When I say I
"believe" something, to me that means specifically that I don't "know" it
-- that is, I have either no evidence or insufficient evidence for it. If I
say I believe it is night-time in Australia right now, I mean that I have
no way to verify that -- it's reasonable, but there's no easy way to prove
it, and I'm speaking from expectations or logic, not knowledge.

As to "knowing," my usage is like Rick's. If I can see it, or direct
evidence of it, I say I know it. By that, I mean that its existence or
behavior seems independent of my belief or disbelief in it; it seems to be
a given of experience.

All this is quite subjective.

Best,

Bill P.

[From Chris Cherpas (990108.0200 PT)]

Bill Powers (990107.1618 MST)--

All this is quite subjective.

Probabilists going back to Bruno de Finetti, and
even Rev. Thomas Bayes, I imagine would agree.

A few verbose points that occur to me about belief:

1. Belief is a pre-PCT term, so I imagine it does
    not have a unique translation into PCT; more
    like a (later) Wittgenstein-like "definition."
    In the typical language game, we don't generally
    bother invoking "belief in X" if the lower-level
    perceptions upon which X is based are accepted as
    givens, just as PCTers don't bother qualifying
    every X by saying "the perception of X" unless one
    is advertising that more will be said about X
    to oppose disturbances to the perception of X.

2. Even though beliefs and knowns are both algebraically
    encompassed in the vanilla flavors of probability,
    in which the knowns of certain and impossible
    are 1 and 0, respectively, these are conveniences
    or trivial cases of belief; if a discontinuity
    between these boundaries and what lay between
    were not recognized as a qualification to the
    system, something like a parallel, (albeit) useless,
    modal logic might even arise in which every
    probability would have its own modal operator.
    In PCT terms, known cases might be those in which
    one is unaware of any conflicts that cannot be
    resolved via an existing higher order control
    system. This awareness would involve imagination,
    I imagine.

3. The "put up or shut up" criterion of subjectivist
    probabilists, in which the expected value of a
    random variable -- estimated by what the believer
    is willing to gamble, or accept as a pay-off to
    not play -- seems related to the combined maximum
    outputs, or gains, of the set of control systems
    that would contribute to opposing the perception
    of values in the distribution that are different
    than the expected value.

4. The "expectation" would correspond to the reference
    value for the perception "being bet on," which itself
    would be perceived in the imagination mode. The
    mathematical expectation is the deterministic part,
    so to speak, but that doesn't mean necessarily that
    the believer also believes s/he is being prophetic.
    If my hypothesis is X=666, then I can revise it on the
    basis of an experiment; in fact, from an axiomatic
    perspective, whether I learn anything about X is
    questionable if I don't bother to adequately model it
    (i.e., at least represent the expected value via
    imagination mode, or say what imaginary distribution
    from the perception of X arises).

5. In any case, learning becomes irrelevant if belief
    is the outcome of a committed optimist or pessimist
    higher-order position, just as invoking belief
    contributes more incoherence than information
    to a discussion if one has not taken some minimal
    effort to consider what set of possible outcomes
    to which one is assigning relative degrees of belief.
    Of course, there is always the surprise, which we
    can take or leave when it happens, but cannot
    literally anticipate. A reflective believer can,
    however, globally model his or her believing process
    as, say, a maximizer of gain/benefit versus, say, a
    maximizer of minimum loss; these idealized strategies
    may even say something about one's intinsic variables,
    given that people are born with so-called temperaments,
    based on (just guessing) different functions relating
    error to rate of reorganization, overall sensitivities
    (or gain levels) on the input or output sides of the
    control systems they develop, concentration levels
    of particular chemicals, rates of metabolism, etc.,
    that may be identified one as, say, relatively "risk
    avoidant." They can also model their believing
    in terms of their organization of relatively high-level
    control systems, using the method of levels.

6. Incoherence is as inevitable as uncontrolled perceptions,
    so I accept statements such as, "I don't always believe
    in my beliefs," as revealing psychological reality.
    However, not learning about models of probability, at
    the extreme end of ignorance is a akin to never
    bothering to see that one side of a coin is heads
    and the other is tails, and still being willing to bet
    on heads for the outcome of a coin toss. The studies
    that show people making incoherent probability judgments,
    or lumping the full probability range into certain, impossible,
    and "I'm a helpless victim of chance," reveal something
    of the problem, especially since specific training
    appears to make a difference. The same studies (e.g.,
    Tversky's, Piaget's) also reveal the ignorance of the
    researchers, since their results appear to be
    dependent on specific wordings of questions
    or other contrived situational factors used in the
    studies. PCT might guide some experimentation that
    avoids these biases and may have practical value for
    educating the world's future decision-makers (i.e., kids).

N. The first time I met my wife, she said, "I don't believe
    in anything," to which I had to honestly reply, "What a
    coincidence: I believe in everything."

Atheist, praying in a foxhole,
cc