Shared references

CSGnet Archive CSG2004
61

From[Bill Williams 30 June 2004 1:40 PM CST]

[From Bruce Gregory (2004.0630.0653)]

        >>Bill Willliams 30 June 2004 2:25 AM CST

        >>Bruce,

        >>You are making use of some terms such as "individual" and
        >>"public" in context in which there is massive equvocation.
        >>For you a "public" is not the sum of "individuals." Neither,
        >>it has to be said, is a "public" a magical super-organic
        >>creature.

Fair enough. What then is a "public"?

In John Dewey's terms a public consists of a group of people organized to accomplish some purpose none of them could
accomplish by themsselves.

It seems to me that a "market" can be modeled by a set of interacting individual agents. Does an economist disagree? If so, exactly how?

Ordinarily when economists model a market in terms of "free trade" between individuals, they make an implicit assumptions concerning law and order. With out law and some means to inforce it you would have the context that economists ordinarily assume. The foundation then of "a market" is a perception of what has been called "recognized interdependence." And, the law, a judicial, and police function can not function in terms of sale to the highest bidder. The creation of laws,
the making of judgements, and the exercise of the police power can be carried out in terms of the economists notion of maximization.

        In economics I can see that starting with what is ordinarily
       meant by an "individual" results in readily identifiable

        problems-- such as explaining how a transaction is possible,
        and what determins a transaction.

Really?

Yes.

How strange.

Indeed, it is. As Solo once said, if you make it fancy enough our graduate students will believe anything.

Of course the model must include, in addition to individual agents, objects that they can exchange and services that they can perform.

Right, and lots, lots more.

        The result in economics of
        an individualist approach was a failure to develop an understanding of the
        properties of a monetary system. This failure was particularly evident in
        the incomprehension of causal factors that genrated the great depression.
        The PCT sophistology and its individualism seems to have as one of its
        implications the conclusion that Keynes' work-- which assumes a community --
        is nonsense.

You haven't convinced me.

OK. Not too surprizing since what I have done is described the situation rather than reproduced the arguments.

I assume that markets are described by interacting agents

I don't believe that this is a warrented assumption. There is a lot more involved.

and I do not think that Keynes's work is nonsense.

Good. I wold note that neither does Milton Friedman. The income-expenditure indentity seems to have persuaded nearly all of the economists.
The problem now seems to be how to construct a viable micro-foundation for the Keynesian system-- and this effort isn't making much
progress.

>Perhaps I am the exception that tests the rule.

The Keynesian system is a large, complex and to some extent confused scheme. Different people find different aspects of it appealing.

Bill Williams

.

62

From[Bill Williams 30 June 2004 2:40 PM CST]

[From Rick Marken (2004.06.30.1000)]

Bill Williams (30 June 2004 1:15 AM PST]

Rick Marken (2004.06.29.1420)]

In the long run the "plausibility" of PCT will get sorted out on the basis
of scientific tests, all of which currently show PCT to be the best model of
behavior around.

You are assuming that the PCT sophistology is neccesary to support the
modeling of human behavior.

Not at all.

However, in your next sentence it seems you contradict yourself, by making a
false claim.

Any model of human behavior can be tried. PCT is the only model
right not that can account for things like control relative to a varying
reference for the state of a controlled variable.

One doesn't need Bill Powers' PCT to do this. And, PCT is not a model it is
an ideology.

I don't actually see myself as an opposed to approaching human behavior
from the standpoint of control theory.

Right. It's just PCT to which you are opposed. Which I think is really great
for PCT.

And, I find it convient as well since I don't have to defend stuff like "it won't cost anything to send people to Mars.'

Or, assertions that the Keynesian system is nonse

Bill Williams

63

[From Jim Beardsley (2004.06.30.1630 EDT)]

[Bill Powers (2004.06.30.0829 MDT)]

>Bruce Gregory (2004.0630.0834)--

As to pattern recognition, since all perceptions involve pattern
recognition in your terms, we still need terms to designate
different forms of pattern recognition, such as perception of
intensities, sensations, configurations, and so on to the
patterns we call system concepts. Each one, of course, would
require a different method of pattern recognition --
an input function that recognizes a set of sensations as the
sound of an oboe could not perform the function of recognizing a
succession of configurations as a particular kind of movement
like the rotation of a second hand on a clock.

Each different kind of pattern recognition would require a
different kind of computation. Is that a reasonable idea?

Is one of the points of contention at this moment about whether, hypothetically, as each kind of pattern-recognition ("PRecog"?) computation is successfully refined (a la Bill's bottom-up, best-next-action approach), additional patterns will likely &or presumably emerge between the otherwise disparate methods of computation themselves, making it theoretically possible, if not feasible or practical, that such computations could be further refined and consolidated into (partially?) 'universal' PRecog computations, at which time more patterns may likely emerge, etc etc..? (And doesn't this pattern of processes itself analog what evolution itself facilitates?)

From my experience and arm-chair observation, such symmetries (even complex ones) can occasionally be recognized in useful detail by gifted &or lucky individuals before or while trying to model or otherwise code them, through little more than private imagination or energetic dialog, without the effort and expense of scientific or otherwise robust modeling, coding, and testing. (I didn't intend this as one of my points, but with the challenges faced by other researchers of complex pattern recognition, I personally doubt anyone here [nothing personal, mind you] will find such shortcuts without successfully modeling somewhat complex perceptions, whether or not PCT or HPCT is employed.)

Even if some consensus emerges as to the likelihood of computational patterns also emerging between different types of pattern recognition, I am quite doubtful that any consensus will emerge, at least here, as to the values of &or necessities for actually launching and cooperating through certain efforts and their expenses to accomplish them.

Despite Bruce Nevin's position, and my temptation to believe he is honing into something important, and also despite some remarkable accomplishments by immense human efforts, the human condition, forget about CSGnet, seems (to me, at least for now) plagued worst of all by contentions about whether and what immediate tangible actions, let alone the "shared" references behind them, toward which any 2 or more people will manage to cooperate with success, even when a particular success can be defined in such a way that its definition doesn't change over time.

However, if I could choose to witness only one cooperative success, Bill Power's suggestion is one of the strongest contenders in my mind..

[Btw, my imagination compels me to apologize for mentioning "expenses" and "efforts" since I'll be surpised if someone else will resist his obsessive compulsion to parrot his opinions about all-too-well-known expensive efforts..]

Enjoying, and especially appreciating, these threads..
Jim

64

[From Bill Powers (2004.06.30.1516 MDT)]

Bruce Gregory (2004.0630.1414) --

Bill Powers (2004.06.30.0829 MDT)

OK, the picture I get is of the head of the MGM lion opening his mouth and
roaring (head 3/4 view oriented with the nose to my right). The triangle
has the point at the top, and is about twice as large as the lion's head.
Its color is red, and the line thickness is appreciable. Is that what you
imagined?

I don't see the point of this exercise. Suppose I said, "Yes! You got it
exactly right." What next? In my view we are imagining similar things. Is
this what is meant by constructing a perception?

Don't make it more complicated that it was intended to be. The point of the
experiment is to see what happens when you try to transfer a nonverbal
perception (the picture you imagined) into my perceptions, using words. I
gave you a description of the picture that came into my awareness as I read
your description. The next step I had in mind (where we are now) would be
for you to take the words I used in describing what was in my mind, and see
if what they bring to your mind is like what you started with. If not, you
could tell me how my picture needs to be changed to be more like yours, and
I could make the changes according to how I understand your instructions,
and then describe the result back to you. And so on. I'm not trying to pull
any tricks on you -- there's no "right answer" that I know but you don't. I
think this will illustrate something, but to avoid putting words in your
mouth I'd like to see how this plays out first. I should have said all this
first, but I just dived into it.

I think it's best to use real examples. Here is a computer program (not
really, but you can read it):

1. Morse writes the time in the margin of the crossword puzzle.

2. Morse studies the next unsolved definition until he sees the answer, and
writes the answer into the puzzle.

3. Morse looks looks to see if there is another unsolved definition. If so,
he returns to step 2. Otherwise, he proceeds to step 4.

4. Morse writes down the time next to the first time, subtracts the earlier
from the later time, and writes down the result.

5. Program ends.

So can you perceive this process and imagine how it goes? Can you perceive
the purpose of this program? I'm sure you can, there's no trick question
here.

I think I see the problem. No, I cannot perceive the process, and no, I
cannot perceive the purpose of the program.

I've repeated essentially what Colin Dexter said as he described what Morse
did at breakfast one day. I think you're still crediting me with more
sublety than intended. Can you imagine Morse carrying out the steps above
until they stop? Can you see that the process _would_ come to an end? Have
I left out some crucial information without realizing it?

Could it be that the word "perceive" is causing the problem here? In PCT, a
"perception" is simply information of any kind entering the brain and being
received at various levels. Experientially, this just means you apprehend
something, know that it is going on or exists. I could have asked, do you
know what the process and its purpose are? You could substitute "recognize
a pattern in" for "perceive" in the above, if you want it in your terms..

This is probably enough to talk about now.

Best,

Bill P.

65

From [Marc Abrams (2004.06.30.1911)]

[From Bill Powers (2004.06.30.1516 MDT)]

Could it be that the word "perceive" is causing the problem here? In

PCT, a >"perception" is simply information of any kind entering the
brain and being >received at various levels.

Nice tight definition. No wonder the neuroscientists have it all wrong.
They have been studying the wrong things. I bet the second edition of
B:CP will get the neuroscientific community straightened out once and
for all.

But I think I'm going to wait for B:CP the movie.

Marc

Considering how often throughout history even intelligent people have
been proved to be wrong, it is amazing that there are still people who
are convinced that the only reason anyone could possibly say something
different from what they believe is stupidity or dishonesty.

Being smart is what keeps some people from being intelligent.

Thomas Sowell

Don't argue with an idiot; people watching may not be able to tell the
difference.

Anon

I don't approve of political jokes. I've seen too many of them get
elected

Anon

···
66

[From Dick Robertson,2004.06.30.2050CDT]

···

----- Original Message -----
From: Bruce Gregory <bruce_gregory@CHARTER.NET>
Date: Wednesday, June 30, 2004 6:05 am
Subject: Re: Shared references

[From Bruce Gregory 92004.0630.0704)]

> Dick Robertson,2004.05.29.1530CDT
>
> Good question. Something that we shrinks believe we need to
work with
> all the time. "Construct" needs a careful definition the way I
> use it, if I would at all. We need to use inference and
> to 'intuit/infer" what CV the other person is
controlling, and also
> make inferences about how well,he/she is controlling it (them), and
> futher try to infer the degree of pain or pleasure the other
person is
> trying to tell us about . In this process we put a lot of stock on
> that not-well-definable concept "empathy" to infer that "that
way I'm
> feeling while I'm hearing/observing you talk about this probably
> me a powerful hint about how you are feeling as you do so."

Yes, I agree.

>
> We recognize that we can never "see" into the other person's
mind, but
> when we reflect the feeling+concept that we infer and he/she
> out audibly and says, "Yeah, that's it," then we assume
that we are
> sharing the "experience" as much as any two humans can.

I don't know that I would call this sharing an experience. But I agree
with what you are saying.

> Note that I didn't say sharing the CV. When that is the case I
> the easiest cases are objectively observable. That is,
when two of us
> are moving a burden from here to there that would be too heavy
for one
> alone, then it seems to me safe to infer that each party is
> controlling a very analagous variable.

I think it is safer to say that each party is controlling closely
coupled variables. The world looks slightly different from different
ends of the couch.

OK. I can live with that.

> Whether it is logical or illogical to say we are sharing the
_same_CV> I think resolves into a matter of sematics, calling for
coming to
> agreement (if possible) on the definition of "same."

We are each controlling our own perceptions of what we take to be a
common task. If you try to bring your end of the couch into one room
and I try to bring my end into another room, we will quickly discover
that we have a problem.

Check

Bruce Gregory

Dick R

There seems to be no shortage of people who value certainty above
truth.

67

From{Bill Williams 1 July 2004 1:35 AM CST]

[From Richard Kennaway (2004.06.30.1307 BST)]

In another message in this thread Bill Williams compared the
perception of the reasons for the constancy of a pendulum's period
with the perception of someone's reasons for driving to McDonalds.
This only serves to show that there are difficult and easy
perceptions in both categories. Picking a single example from each,
chosen to make a point, fails to make that point.

What the comparison succesfully illustrates is as Richard says, is
that there are "diffiucult and easy perceptions in both categories.
There are enough "easy" perceptions in the social category to sustain
a somewhat precarious civilization. Adopting an indivualist sophistology
contributes to making civilization more precarious. From the standpoint
of control theory it isn't neccesasry.

Bill Williams

68

[From Bruce Gregory (2004.0701.1002)]

Bill Powers (2004.06.30.1516 MDT)

Could it be that the word "perceive" is causing the problem here? In PCT, a
"perception" is simply information of any kind entering the brain and being
received at various levels. Experientially, this just means you apprehend
something, know that it is going on or exists. I could have asked, do you
know what the process and its purpose are? You could substitute "recognize
a pattern in" for "perceive" in the above, if you want it in your terms..

That is clearly the problem. In PCT, I construct a perception by thinking about something. I can certainly think about all the levels in the hierarchy, therefore I am constructing perceptions of them. I hope I am not the only person who failed to grasp this, but thanks for the clarification. Apparently very few PCT perceptions are perceptions in everyday language. That may be why PCT seems so difficult for many people to understand, not to say accept. It may be that what seems to be an unwillingness to accept the conclusions of PCT is based to some degree on this misunderstanding.

I recall Rick telling me at one point that he could perceive justice. I never understood what he was talking about until now.

Bruce Gregory

There seems to be no shortage of people who value certainty above truth.

69

[From Bill Powers (2004.07.01.0810 MDT)]

Jim Beardsley (2004.06.30.1630 EDT) --

> Each different kind of pattern recognition would require a
> different kind of computation. Is that a reasonable idea?

Is one of the points of contention at this moment about whether,
hypothetically, as each kind of pattern-recognition ...computation is
successfully refined ... additional patterns will ... emerge between the
... methods of computation themselves, making it theoretically possible
...that such computations could be further refined and consolidated into
... 'universal' PRecog computations, at which time more patterns may
likely emerge ...?

I'm not sure the contention is about anything more than terminology. I've
always thought of a "pattern" as something having at least some internal
structure (a checkerboard pattern) in which we can see the elements (the
dark and light squares) as well as their arrangement (which is the
pattern). The squareness of the dark and light squares could also be called
a pattern, since it's an arrangement of edges and corners, but I don't see
how the darkness or lightness of the squares could be called a "pattern."
I'd use the term "configuration" for the squareness and checkerboard
perceptions, but perhaps someone else would like to propose a separate sort
of pattern for the repetitiveness of the checkerboard.

It seems to me that each of the 11 levels that now make up the HPCT model
would require a quite distinct sort of computation to recognize the type of
perception it contains. The idea that the nervous system might exemplify
some "universal" pattern recognizer would imply that even a tadpole would
possess all the levels of perception that human beings have -- a
proposition I doubt. I imagine the 11 levels, or whatever the levels really
are, did evolve as you say, but one level at a time, building on the
pre-existing levels. Our lowest levels of control resemble those in all
other vertebrates, and even have striking resemblances to what we find in
cockroaches. So the lower levels probably don't change their main
principles of operation as higher levels are added.

All the pattern-recognizers I have seen in AI and AL are focused on
configurations the third of the eleven levels I have proposed. That
leaves a long way to go.

...with the challenges faced by other researchers of complex pattern
recognition, I personally doubt anyone here [nothing personal, mind you]
will find such shortcuts without successfully modeling somewhat complex
perceptions, whether or not PCT or HPCT is employed.)

I'm not really interested in modeling _complex_ perceptions per se -- only
in exposing the need to model perceivers that create aspects of the world
that seem to be common in human experience. The first step is to recognize
that there _are_ different types of aspects of the experienced world such
as intensity, sensation, and configuration, and the second is to recognize
these as products of human perceptual input functions rather than taking
them for granted as the givens of reality. That's a big enough step for
starters, considering that most people think of these things as existing
Out There.

Best,

Bill P.

70

[From Bill Powers (2004.07.01.0859 MDT)]

Bruce Gregory (2004.0701.1002) --

Could it be that the word "perceive" is causing the problem here?

That is clearly the problem. In PCT, I construct a perception by thinking
about something. I can certainly think about all the levels in the
hierarchy, therefore I am constructing perceptions of them.

I see. This meaning of "perceive" doesn't really fit PCT, since it requires
consciousness, and also says that only those mental activities we call
"thinking" involve perceptions. In HPCT, thinking would involve
manipulating imagined perceptions, but mostly at the category level and up.
The lower level perceptions are what we think _about_, but they would still
be classed as perceptions in PCT. See the definition of perception in B:CP
(p. 286). I think I've been pretty consistent in sticking to it.

I hope I am not the only person who failed to grasp this, but thanks for
the clarification. Apparently very few PCT perceptions are perceptions in
everyday language.

It's more the other way around. Everything commonly called perception
(like thoughts in your mind) is perception in PCT. But in PCT, perception
can also be the things you think about, like intensities, sensations,
configurations, events, relationships, and categories. In common usage,
these other kinds of perceptions are taken for granted as aspects of the
external world; they aren't even recognized as depending on sensory systems
and neural processing.

Everyday language is pretty limited in providing ways to distinguish levels
of perception. We have "sense data" and "impression" and "perception" and
"concept", and that's about it, with no indication of the relationships
between these types of experience, or their relationship to hierarchical
control. I decided to discard all those old terms and try to describe
specific types that seem common in human experiences of the world, and then
to use the term "perception" to refer to them all, as a generic term.

This is no more presumptious that is for a physicist to use words like work
and energy and equilibrium in ways that are radically different from
common-language uses of the same terms.

That may be why PCT seems so difficult for many people to understand,
not to say accept. It may be that what seems to be an unwillingness to
accept the conclusions of PCT is based to some degree on this
misunderstanding.

True, but who has to change? Shall we require physicists to redefine "work"
as the dictionary defines it? Or as "anything that makes you tired?" A
technical subject needs a technical language, and anyone who wants to learn
the subject needs to learn the language. Seems to me someone I know wrote a
book about the language of physics. Well, there is a language of PCT, too,
defined just as rigorously and used just as consistently.

I recall Rick telling me at one point that he could perceive justice. I
never understood what he was talking about until now.

There may be a little more to that claim than meets the eye. At the lower
levels it's not hard to understand that behind the words we use there are
experiences: behind the word "green" is the experience of a color; behind
the word "motion" there is the experience of something moving. We have to
perceive something before we can decide what to call it, so perceptions are
not the words we used to describe them. The word is not the object, as they
say. But words and objects are both perceptions.

Higher-level perceptions -- concepts -- are not quite so easy to separate
from the words we use to describe them. They aren't as vivid or
tangible-seeming as lower-level perceptions. Yet if you ask yourself, for
example, how you know whether to call a given interaction between people
"just" or "unjust", it's clear that you have to judge the interaction
before you can put a word to it. And that means you have to perceive
whether it fits the principle of justice before you label it or start
thinking or talking about it. In ordinary circumstances, the word comes
hard on the heels of the perception and seems to be part of it. But it's
not: it's only a word.

Anyway, back to the example if that's OK with you. I gave you a description
of what came into my mind as you described the lion's head in the triangle.
I imagined the MGM lion roaring, his head in a 3/4 view, nose to my right.
It was a male, by the way, with a mane, and tawny. I imagined a triangle
with a thick border, its base horizontal and one point up, colored red,
around the lion's head. Is that the image you wanted to describe to me, or
does it need corrections to be the same as yours?

Best,

Bill P.

71

[From Dick Robertson,2004.07.01.1230CDT]

···

----- Original Message -----
From: Bill Powers <powers_w@FRONTIER.NET>
Date: Thursday, July 1, 2004 10:34 am
Subject: Re: Shared references

[From Bill Powers (2004.07.01.0859 MDT)]

>>Bruce Gregory (2004.0701.1002) --
>>
>>Could it be that the word "perceive" is causing the problem here?

>That is clearly the problem. In PCT, I construct a perception by
>about something. I can certainly think about all the
levels in the
>hierarchy, therefore I am constructing perceptions of them.

I see. This meaning of "perceive" doesn't really fit PCT, since it
requiresconsciousness, and also says that only those mental
activities we call
"thinking" involve perceptions. In HPCT, thinking would involve
manipulating imagined perceptions, but mostly at the category
level and up.
The lower level perceptions are what we think _about_, but they
would still
be classed as perceptions in PCT. See the definition of perception
in B:CP
(p. 286). I think I've been pretty consistent in sticking to it.

> I hope I am not the only person who failed to grasp this, but
thanks for
> the clarification. Apparently very few PCT perceptions are
perceptions in
> everyday language.

It's more the other way around. Everything commonly called perception
(like thoughts in your mind) is perception in PCT. But in PCT,
perceptioncan also be the things you think about,

Or don't think about, but just have under control in your behavior without conscious knowledge of, or might not have under control but just experience as your view of the sky, a sunset, etc (although I've never been totally sure that even then Some variable wasn't under control.

like

intensities, sensations,
configurations, events, relationships, and categories. In common
usage,these other kinds of perceptions are taken for granted as
aspects of the
external world; they aren't even recognized as depending on
sensory systems
and neural processing.

Everyday language is pretty limited in providing ways to
distinguish levels
of perception. We have "sense data" and "impression" and
"perception" and
"concept", and that's about it, with no indication of the
relationshipsbetween these types of experience, or their
relationship to hierarchical
control. I decided to discard all those old terms and try to describe
specific types that seem common in human experiences of the world,
and then
to use the term "perception" to refer to them all, as a generic term.

This is no more presumptious that is for a physicist to use words
like work
and energy and equilibrium in ways that are radically different from
common-language uses of the same terms.

> That may be why PCT seems so difficult for many people to
understand,> not to say accept. It may be that what seems to be an
unwillingness to
> accept the conclusions of PCT is based to some degree on this
> misunderstanding.

True, but who has to change? Shall we require physicists to
redefine "work"
as the dictionary defines it? Or as "anything that makes you
tired?" A
technical subject needs a technical language, and anyone who wants
to learn
the subject needs to learn the language. Seems to me someone I
know wrote a
book about the language of physics. Well, there is a language of
PCT, too,
defined just as rigorously and used just as consistently.

>I recall Rick telling me at one point that he could perceive
justice. I
>never understood what he was talking about until now.

There may be a little more to that claim than meets the eye. At
the lower
levels it's not hard to understand that behind the words we use
there are
experiences: behind the word "green" is the experience of a color;
behindthe word "motion" there is the experience of something
moving. We have to
perceive something before we can decide what to call it, so
perceptions are
not the words we used to describe them. The word is not the
object, as they
say. But words and objects are both perceptions.

Higher-level perceptions -- concepts -- are not quite so easy to
separatefrom the words we use to describe them. They aren't as
vivid or
tangible-seeming as lower-level perceptions. Yet if you ask
yourself, for
example, how you know whether to call a given interaction between
people"just" or "unjust", it's clear that you have to judge the
interactionbefore you can put a word to it. And that means you
have to perceive
whether it fits the principle of justice before you label it or start
thinking or talking about it. In ordinary circumstances, the word
comeshard on the heels of the perception and seems to be part of
it. But it's
not: it's only a word.

Anyway, back to the example if that's OK with you. I gave you a
descriptionof what came into my mind as you described the lion's
head in the triangle.
I imagined the MGM lion roaring, his head in a 3/4 view, nose to
my right.
It was a male, by the way, with a mane, and tawny. I imagined a
trianglewith a thick border, its base horizontal and one point up,
colored red,
around the lion's head. Is that the image you wanted to describe
to me, or
does it need corrections to be the same as yours?

Best,

Bill P.

72

[From Rick Marken (2004.07.01.1100)]

Bruce Gregory (2004.0701.1002)--

I recall Rick telling me at one point that he could perceive justice. I
never understood what he was talking about until now.

I found the notion of higher level perceptions, like perceptions of programs
and principles, to be one of the most difficult things to understand about
PCT. In the conventional psychology curriculum the term "perception" refers
to a specific set of phenomena that occupy no more than the first 5 or so
levels of the proposed PCT hierarchy of perception. My own studies of
perception focused on perceptions that I would now refer to as intensities,
sensations, transitions (apparent motion), configuration, relationship, and,
perhaps, events. These were perceptions of things that clearly were "out
there" in reality: tones, lights, pictures, moving figures, relative
position of things (eg. the moon and the horizon), etc. Higher level
perceptions -- programs (like the one described in an earlier post by Bill),
principles (like justice) and system concepts (being a Dodger fan) -- were
part of the study of cognition, and cognition was thought to be based on
perception.

I can now see that programs, principles and system concepts are perceptions
in the same sense that the perception of the intensity of a tone is a
perception. For example, a principle, like corruption, is a perception like
tone intensity in the sense that we can experience different levels of the
perception of corruption, from very corrupt to not corrupt at all, just as
we perceive different levels of tone intensity, from very loud to soft. The
difference between the perception of corruption and the perception of tone
intensity is in terms of what I would call "subjectivity". Tone intensity
seems very objective; like a part of external reality. A tone, especially a
very loud one, is experienced as an objective reality that is "out there" in
the world. Level of corruption, on the other hand, seems very subjective;
like a part of one's internal reality. Corruption, whether flagrant or
minor, is experienced as a subjective judgment that is "in here" -- in one's
mind.

The fact that corruption is a perception, functionally equivalent to the
perception of tone intensity, is most clearly demonstrated, I believe, by
the fact that people can _control_ this perception. That's why I posted that
New Yorker cartoon. The cartoon recognizes that people can control how
corruptly they behave. In order to be able to do this, people have to be
able to control a _perception_ of corruption, one that varies in magnitude
as a function of things these people do, like taking or not taking bribes,
soliciting or not soliciting single source contracts from people in high
office, etc.

I understand that it is hard to think of programs, principles and system
concepts as perceptions, when you are used to thinking of perception (as I
was) only as the sounds, colors, tastes, patterns, movements and events that
one sees (hears or tastes) "out there". But I think demonstrations of the
fact that people can _control_ things like programs, principles and system
concepts would show that these things _are_ perceptions that are
_functionally_ , if not experientially, equivalent to perceptions of tone
intensity. Therefore, I am currently working on developing a demonstration
of program control. The idea would be to have a program running off -- such
as the one Bill described in an earlier post -- and then, at some point,
have the program change to a different one. The subject would be able to
press a button to restore the original program, this keeping the program
perception under control (always the same program). In order to be able to
do this the subject would have to be able to _perceive_ which program is
occurring. The methodological problem (which I have still not solved to my
satisfaction) is to provide no lower level perceptual clues that the program
has changed. I'll try to have a program control demo worked up this weekend.
I think the best way to help people understand higher level perception is
let them see themselves controlling such perceptions.

Regards

Rick

···

--
Richard S. Marken
MindReadings.com
Home: 310 474 0313
Cell: 310 729 1400

73

[From Bruce Gregory (2004.0701.1423)]

Bill Powers (2004.07.01.0859 MDT)

I see. This meaning of "perceive" doesn't really fit PCT, since it requires
consciousness, and also says that only those mental activities we call
"thinking" involve perceptions.

That was not my intention. I only meant to say that thinking is a form of perception, not that all perceptions involve thinking.

It's more the other way around. Everything commonly called perception
(like thoughts in your mind) is perception in PCT. But in PCT, perception
can also be the things you think about, like intensities, sensations,
configurations, events, relationships, and categories. In common usage,
these other kinds of perceptions are taken for granted as aspects of the
external world; they aren't even recognized as depending on sensory systems
and neural processing.

O.K. I think I understood that.

Everyday language is pretty limited in providing ways to distinguish levels
of perception. We have "sense data" and "impression" and "perception" and
"concept", and that's about it, with no indication of the relationships
between these types of experience, or their relationship to hierarchical
control. I decided to discard all those old terms and try to describe
specific types that seem common in human experiences of the world, and then
to use the term "perception" to refer to them all, as a generic term.

I did not mean to voice an objection to that step.

This is no more presumptious that is for a physicist to use words like work
and energy and equilibrium in ways that are radically different from
common-language uses of the same terms.

I did not mean to imply that you were being presumptuous. The fact that terms used technically in physics are often used in different ways in every day life leads to unending confusion on the part of many students. That is all I was trying to say.

That may be why PCT seems so difficult for many people to understand,
not to say accept. It may be that what seems to be an unwillingness to
accept the conclusions of PCT is based to some degree on this
misunderstanding.

True, but who has to change?

No change is called for. It might help to remind people that perceptions in PCT often seem to bear little relationship to perceptions in every day life, that's all.

Shall we require physicists to redefine "work"
as the dictionary defines it? Or as "anything that makes you tired?" A
technical subject needs a technical language, and anyone who wants to learn
the subject needs to learn the language. Seems to me someone I know wrote a
book about the language of physics. Well, there is a language of PCT, too,
defined just as rigorously and used just as consistently.

I did not intend to accuse you or anyone else of inconsistency.

I recall Rick telling me at one point that he could perceive justice. I
never understood what he was talking about until now.

There may be a little more to that claim than meets the eye. At the lower
levels it's not hard to understand that behind the words we use there are
experiences: behind the word "green" is the experience of a color; behind
the word "motion" there is the experience of something moving. We have to
perceive something before we can decide what to call it, so perceptions are
not the words we used to describe them. The word is not the object, as they
say. But words and objects are both perceptions.

Higher-level perceptions -- concepts -- are not quite so easy to separate
from the words we use to describe them. They aren't as vivid or
tangible-seeming as lower-level perceptions. Yet if you ask yourself, for
example, how you know whether to call a given interaction between people
"just" or "unjust", it's clear that you have to judge the interaction
before you can put a word to it. And that means you have to perceive
whether it fits the principle of justice before you label it or start
thinking or talking about it. In ordinary circumstances, the word comes
hard on the heels of the perception and seems to be part of it. But it's
not: it's only a word.

Anyway, back to the example if that's OK with you. I gave you a description
of what came into my mind as you described the lion's head in the triangle.
I imagined the MGM lion roaring, his head in a 3/4 view, nose to my right.
It was a male, by the way, with a mane, and tawny. I imagined a triangle
with a thick border, its base horizontal and one point up, colored red,
around the lion's head. Is that the image you wanted to describe to me, or
does it need corrections to be the same as yours?

Your imagination is much more vivid than mine. I have a lot of trouble imagining things in the detail you describe. I also rarely have vivid dreams, or even dreams in which people I know play important roles.

Bruce Gregory

There seems to be no shortage of people who value certainty above truth.

74

[From Bill Powers (2004.07.01.1424 MDT)]

Dick Robertson,2004.07.01.1230CDT--

But in PCT,
> perception can also be the things you think about,

Or don't think about, but just have under control in your behavior
without conscious knowledge of, or might not have under control but just
experience as your view of the sky, a sunset, etc (although I've never
been totally sure that even then Some variable wasn't under control.

Agreed. Perception is the world coming in to us, or us converting the world
into representations we can grasp and understand. But I guess
"representation" is a bad word, because it implies that there are copies of
the external objects in the brain. The trouble is that this scene has been
reported and analysed by generations of visitors and there aren't any
unused terms lying around to distinguish a new point of view. If it is new
-- doesn't matter, really.

Best,

Bill P.

75

[From Bill Powers (2004.07.01.1442MDT)]

Bruce Gregory (2004.0701.1423) --

Your imagination is much more vivid than mine. I have a lot of trouble
imagining things in the detail you describe.

Yet I suspect you have your own ways of finding out if your meaning has
survived the passage into someone else's mind and back again, and your own
ways of both detecting and correcting transmission errors. That was the
point of the experiment, not to show you had any "deficiency."

Bruce Nevin is saying that we can reach understanding of each other that is
more perfect than other impersonal understandings because there is someone
on both ends trying to help in reaching it. I can see the logic of that,
though not in detail. It is probably true that in transmitting words --
particularly written words in a standard font -- we can fairly easily agree
on what word we are seeing. But that is at a pretty low level of
perception. What we use language _for_, on the other hand, is to transmit
meanings (As Bruce N. said), and the meanings of words come from our own
experiences. So the task isn't quite as simple at that level as Bruce implies.

Even our short-lived experiment demonstrates that point. What you meant by
your words about lions and triangles conveyed something to me that I pulled
out of my own experiences, which proved to be considerably more elaborate
and detailed than appropriate. If I had understood your description
correctly, I would have imagined -- what? And in what degree of detail or
vividness? This isn't a question of whose imagination is better than whose;
it's a question of reaching a correct understanding of an experience that
someone else had and is describing. It's not that easy.

In my experience, people hardly ever actually check to see if their words
have evoked in another person the meaning that was intended. I don't do
that very often myself, particularly in this medium where there is such a
time lag, but even in ordinary face-to-face conversation, you don't often
hear people pausing to see if the other has taken every word as it was
meant. In a linguistics laboratory, of course, it is possible to do this
back-and-forth checking with great thoroughness, and to show that
communication can be pretty good when it is done. But in the wild, the kind
of checking that is done is mostly spot-checking, and inference from
observing the general trend of replies or comments, or from observing
whether the actions that follow seem to indicate understanding. And in my
experience, the result is as often misunderstanding or false agreement as
it is understanding (more so with more complicated ideas, of course).

Bruce N. wondered how you could have appeared to agree with him some time
ago and now apparently be rejecting what you agreed to. This illustrates my
point. When we agree with people, we agree with what we perceive their
meaning to be, which of course indicates that we have found within
ourselves (at the time of agreement) meanings for the words that we find
acceptable. But on reflection, I think we often wonder just what we agreed
to, and see alternative meanings that are not so agreeable. Also, when we
observe the other person behaving and talking in ways that seem
inconsistent with what we thought we were agreeing to, we may realize that
we didn't understand what the other meant, and thus agreed to something the
other person didn't mean. So it is very understandable that I could
enthusiastically agree with you on the basis of what I thought you meant,
and a while later strongly disagree with the same words, having decided
that I had misunderstood your meaning.

The experiment did have a point, but maybe it has been made now by other means.

Best,

Bill P.

76

[From Bruce Gregory (2004.0702.1043)]

Rick Marken (2004.07.01.1100)

I think your project is very important and wish you well. I will point out one potential problem. I agree that there are degrees of corruption. But it is not all all clear to me that there are degrees of being a program. It seems to me more like an all or nothing situation and in that sense that something either does or does not conform to the pattern "program." That said, a working demonstration of the sort you envision would demonstrate that my view may be too simplistic.

Bruce Gregory

There seems to be no shortage of people who value certainty above truth.

77

[From Bill Powers (2004.07.02.0849 MDT)]

From Bruce Gregory (2004.0702.1043) --

... it is not all all clear to me that there are degrees of being a
program. It seems to me more like an all or nothing situation and in that
sense that something either does or does not conform to the pattern
"program." That said, a working demonstration of the sort you envision
would demonstrate that my view may be too simplistic.

There are very fews things that exist only in the "yes" and "no" states,
with nothing between and no alternative forms of either. I can tell you
from immediate experience that there is such a thing as "almost the right
program" and "hardly the right program at all." I've been writing a few of
the latter recently, but have made some progress on demos for the CSG
meeting. Why does it take so long to write an acceptable program? Because
programs take form gradually. You don't have "no program" at one instant
and "finished program" in the next instant.

The reason that this is so is that all perceptions are functions of many
perceptions of lower order. A program consists of many sequences of
operations connected by logical tests -- what are termed "flow control"
constructions such as if, for, while, and repeat ... until. Two programs
can differ from each other only in one or two details, or almost
completely. What they accomplish may be exactly the same, or totally different.

So a program is just as multidimensional as a face and can vary over a wide
range of forms just as faces can. Consider tennis. John McEnroe is a very
interesting commentator because he's been there and understands tennis
programs -- what are called strategies or game plans. One player will stand
back of the baseline and volley deep to keep the other player running
around until she makes a mistake. Another player will see the other player
tiring and rush the net after a return, but not so close as to invite a
passing shot or a lob.There's lots of "If he does this, do that, otherwise
do the other" in tennis, and that's a program. It's enough of a
recognizeable perception for McEnroe to identify a player's typical
strategy and say that if he doesn't modify it he'll be in trouble.

If you think of all the lower-order perceptions on which any higher-order
perception depends, I think you'll begin to see how even discrete-seeming
perceptions have an underlying continuum of variation. Is a rose either a
rose or not a rose? Ask any judge at a horticultural exhibition.

Best,

Bill P.

78

[From Bruce Gregory (2004.0702.1130)]

Bill Powers (2004.07.02.0849 MDT)

From Bruce Gregory (2004.0702.1043) --

... it is not all all clear to me that there are degrees of being a
program. It seems to me more like an all or nothing situation and in that
sense that something either does or does not conform to the pattern
"program." That said, a working demonstration of the sort you envision
would demonstrate that my view may be too simplistic.

There are very fews things that exist only in the "yes" and "no" states,
with nothing between and no alternative forms of either. I can tell you
from immediate experience that there is such a thing as "almost the right
program" and "hardly the right program at all." I've been writing a few of
the latter recently, but have made some progress on demos for the CSG
meeting. Why does it take so long to write an acceptable program? Because
programs take form gradually. You don't have "no program" at one instant
and "finished program" in the next instant.

While i certainly agree with what you say, I interpret it somewhat differently. No matter what state a program is in, from just beginning, to abandoned, to almost finished, to be being debugged, it seems to me that it is always a program (or not). A failed program can be fixed, but I am not sure I would want to locate the failed version and successful version on a continuum. If Rick's project is successful, it will presumably demonstrate the merit of such a continuous view.

So a program is just as multidimensional as a face and can vary over a wide
range of forms just as faces can.

Faces differ, but not in degree of "faceness" as far as I can tell. I would be reluctant to say that Rick's face scores a .867 while Bill's scores only .643. The same is true of tennis styles.

If you think of all the lower-order perceptions on which any higher-order
perception depends, I think you'll begin to see how even discrete-seeming
perceptions have an underlying continuum of variation. Is a rose either a
rose or not a rose? Ask any judge at a horticultural exhibition.

Roses do seem to form a continuum as a result of breeding programs. But I am still not sure I would opt for degree of roseness as a linear scale in the way that the distance between a cursor and a target vary in a linear fashion.

Perhaps this is simply terminology.

Bruce Gregory

There seems to be no shortage of people who value certainty above truth.

79

[From Bill Powers (2004.07.02.1012 MDT)]

Bruce Gregory (2004.0702.1130) --

While i certainly agree with what you say, I interpret it somewhat
differently. No matter what state a program is in, from just beginning, to
abandoned, to almost finished, to be being debugged, it seems to me that
it is always a program (or not). A failed program can be fixed, but I am
not sure I would want to locate the failed version and successful version
on a continuum. If Rick's project is successful, it will presumably
demonstrate the merit of such a continuous view.

You're talking only about categorizing a program as belonging to the
general class called "a program". But the levels in HPCT don't work like
that. We don't usually sense simply "sensation or no sensation" or
"configuration or no configuration" or "program or no program," though
that's possible. We sense _this particular kind_ of sensation as present to
a greater or lesser degree, _this particular kind_ of configuration as
present to a greater or lesser degree. Is that color green? Well, sort of.
Maybe a little bluish, too. Is that a picture of an elephant? Yes, a very
good one, for a three-year-old, though I would have guessed it might be a
donkey or a banana.

So we perceive a square-root-taking program, or a serve-and-volley program,
or a build-strength-in-the-center program. Is this guy telling me about a
legitimate chance to make some money, or is this the Pigeon Drop scam? When
we first observe a program we may take a while to decide whether it's this
program or that program, because both program-perceivers are responding
somewhat, at the same time (remember that in HPCT we're using the
Pandemonium model). We may seldom experience a particular program in its
purest maximum-perceptual-signal form. A perfect finesse in bridge; a
totally successful bluff in poker, a serve-and-volley strategy that never
results in a return winner.

So a program is just as multidimensional as a face and can vary over a wide
range of forms just as faces can.

Faces differ, but not in degree of "faceness" as far as I can tell. I
would be reluctant to say that Rick's face scores a .867 while Bill's
scores only .643. The same is true of tennis styles.

Precisely the distinction I want to make. It is not "faceness" that we
usually perceive when we recognized a face (though that is a possibility in
a specific case), but "Bill's face" and "Rick's face." The variability in
the configuration perception is in how much signal we get from the
Bill's-face perceiver and from the Rick's-face perceiver. If you think in
terms of feature-detection types of pattern recognition, the variability is
in how much of each feature, or what state of each feature, is being
perceived. This is the variation in lower-order perceptions that results in
variation at the higher order. The face is finally classified only when the
continuously-varying numbers pass whatever thresholds have been selected.
Before that final high-level classification, many possibilities are being
entertained and we perceive more than one at a time: "I think that could be
Rick, but he's too short, more like Julius."

If you think of all the lower-order perceptions on which any higher-order
perception depends, I think you'll begin to see how even discrete-seeming
perceptions have an underlying continuum of variation. Is a rose either a
rose or not a rose? Ask any judge at a horticultural exhibition.

Roses do seem to form a continuum as a result of breeding programs. But I
am still not sure I would opt for degree of roseness as a linear scale in
the way that the distance between a cursor and a target vary in a linear
fashion.

Again, it's not degree of roseness that we perceive unless its only the
general class we're concerned with (is it a rose or a peony?). What the
judge is concerned with is how close this particular rose comes to being an
ideal exemplar of an American Beauty. Linearity of the scale is not
required, unless you mean only an ordered arrangement along a line rather
than a fit to a first-degree equation.

Perhaps this is simply terminology.

In this case I think it's meaning that matters. The general principle here
is that a perception at any level (other than the category level itself) is
not simply an either-or classification. If that were so, there would be
only 11 different perceptions possible in HPCT, one at each level. The
levels are defined in terms of _types_ of perceptions, with multiple
(hundreds to thousands of) perception-recognizers of the same type at each
level. Many of them, I don't know how many, would be active at the same
time, operating in parallel, putting out perceptual signals indicating how
well the incoming signals from lower levels exemplify whatever combination
of values a given input function is organized to report as a perceptual
signal -- large signal for a perfect match, small signal for hardly any
match. Or, in other cases, large signal for high end of the scale, small
signal for low end of the scale.

There may be other models that would work better (I don't know), but this
is the one I use. I tried to make this clear in B:CP, but that's hard to
achieve.

Best,

Bill P.

80

[From Bruce Nevin (2004.07.02 14:00 EDT)]

Bruce Gregory (2004.0629.1606)--

Bruce Nevin (2004.06.29 15:29 EDT)

Bill Powers (2004.06.29.0934 MDT)--

I make no conclusion that his perceptions are organized as mine are. I only
construct a perception of his perception of a certain CV of mine.

I'm sure this is obvious to people who understand PCT, but exactly how
does one "construct a perception" of someone else's perception? I thought
a perception was input to a comparator in a control loop. Do you have
access to other people's neural signals? Is this another perceptual
ability that only I lack?

I have described how conventionalized CVs can provide a scaffolding for
constructing perceptions of other people's intentions. Those CVs are all
constructed of perceptual inputs to control loops. I have not said that
this is error free. Language enables error-free transmission of linguistic
information, but the associations and interpolations of meanings upon that
information by the recipient is variously reliable, sometimes quite
unreliable. (Perceptions of what I mean by these three sentences vary, but
among English-speaking readers perceptions of what words and sentences I
have typed do not.)

It is also possible to ignore much of the linguistic information that is
transmitted to you, and to associate and interpolate meanings upon
fragments of it. This is what you have done above.

         /Bruce Nevin

···

At 04:08 PM 6/29/2004 -0400, Bruce Gregory wrote: