fairness and meaning

[Bill_Powers]

[From Bill Powers (2003.10.24.1226 MDT)]

Bruce Nevin (2003.10.22 23:00 EDT --

First, I appreciate the careful and patient approach you're taking to
getting the "first-principle" stuff said correctly. It is slow and takes a
lot of thought, but that's how worthwhile conclusions are found. My
observations are not meant to replace yours, but are simply put on the
table for consideration.

Here's a scenario. I control a perception of certain words appearing on
the screen in front of me in a certain structured way. Let's say I write this:

    Linguistic structure is in our shared environment.

I'll assume you agree that I control perceptions of structure in this,
e.g. the words must come in that particular order to be an English
sentence, and this particular sentence. The input functions, references,
and output functions for my control of these perceptions accord with
conventions established as common knowledge among users of the English
language. In the course of learning those conventions I have become so
organized as to control these perceptions by means of various behavioral
outputs, in this case typing.

The perception in the sentence above, to the best of our knowledge, begins
with a pattern of light and dark on the viewer's retina. Several levels up,
these intensity signals give rise to configuration signals (letter and word
shapes). By convention for English, we read from left to right, which
establishes the sequence in which we will attend to the words or their
meanings. The writer, of course, produced the elements of the sentence with
the same ordering, or at least elements that ended up in that ordering.

You, then, having also learned the common-knowledge conventions of
English, have become so organized as to construct the same structure out
of your perceptual inputs from your environment (a shared environment
thanks to the Internet). Recognizing that structure (which you have
created, but which you perceive to be the words, etc. that I have put in
the environment), and remembering common knowledge conventions as to what
intentions one has when producing such a structure, you 'read' in it a
perception of my intentions; in particular, a perception of meanings that
you perceive that I intended you to understand from it. But you also infer
meanings that I did not intend. You say:

     You're asserting that you have direct knowledge of the
     environment, bypassing perception.

This is an inference that you have constructed. But you could not have
done so without first understanding what is directly said in my sentence:

Here is what I mean. When you see a configuration, "C", you assume that
there is something in the environment with the shape of a C. You do
something to the environment so that I, too, see a C, and can verify that I
do by making another one that you agree is the same as the one you showed
me. Does this indicate that there is something in the environment with the
shape of a C? No, it does not. All we can say is that there is something in
the environment that gives rise to a perception of a C shape when visual
inputs are processed by a human perceptual system.

The problem is that we can't independently observe the shape of something
in the environment. All we can do is try to test whether each person is
having the same reproducible perceptual experience. Suppose, for example,
that between our eyes and objects in the environment there is a mapping
that reverses everything left for right. If that is the only
transformation, then a reversed C-shape in the actual environment will give
rise to a correct-looking perceived C. Likewise, we will scan the written
sentence in a way that is actually right to left in the environment, but
which appears left to right in perception.

    Linguistic structure is in our shared environment.

I suppose that the meaning of this depends critically on what you mean by
"shared environment." Something is certainly in the environment, but is it
literally shared? To create structures that we experience as linquistic
structures, we must do things to the environment, but what is done to the
environment need only evoke the proper perceptions; it need not itself have
the same structure.

As you know, I am trying to learn how to pronounce Chinese sentences
written in Pinyin (romanized syllables). I have practiced pronouncing
individual syllables so they sound to me like what native speakers are
saying, and the program, which can record and analyze my recorded speech,
gives me ratings from barely passing to WOW. I have learned the four tones
and the particles, and how to produce them in a fairly authentic manner, by
initating native speakers.

But after that, I turned to the sections on sentences, in which whole
strings of syllables have to be produced. And suddenly the Chinese speakers
are no longer saying the same words correctly. They pronounce two words
said like "jian" and "er" as "jar". They no not reproduce rising and
falling tones as in the earlier practice. The rises and falls are all but
gone, and superimposed on them are much larger inflections that are
evidently the same sort of expressive tonal changes that we use in English.
Yet with practice, I began to hear the familiar words from the single-word
practice -- despite the fact that the speakers are saying them differently.
What they are doing, it seems, is adjusting the way they produce sounds so
that at some level, perhaps configurations or events, the listener
perceives the intended words. The reality is distorted to make the
perception come out right.

I guess that's the point I'm getting to so slowly. The reality is
manipulated so the word perceptions come out right; the word perceptions
are manipulated so the meaning perceptions come out right. The greatest
variation is at the _start_ of this chain, as is typical of all control
processes. If structure were truly in the environment, we would find the
greatest variability at the end of the chain, in the meanings. But we
don't. It's the other way around. Initially, the variations appear at the
end of the chain, but we adjust the beginning of the chain until the end,
where the meanings are, is properly controlled. Then we find that it is the
beginning of the chain where the most variations appear.

I trust that you will take these observations and dump them in the hopper,
where the mill will combine them with other ingredients to produce the next
iteration of this gradually growing analysis.

Best,

Bill P

[bnhpct]

[From Bruce Nevin (2003.10.22 23:00 EDT ]

Bill Powers (2003.10.24.1226 MDT)–

When you see a configuration, “C”, you assume that

there is something in the environment with the shape of a C. You do

something to the environment so that I, too, see a C, and [I] can verify [for you] that I

do by making another one that you agree is the same as the one you showed

me. Does this indicate that there is something in the environment with the

shape of a C? No, it does not. All we can say is that there is something in

the environment that gives rise to a perception of a C shape when visual

inputs are processed by a human perceptual system.

Right. Furthermore, for the point that I’m making, it doesn’t matter whether that something in the environment actually ‘has the shape of a C’, whatever that might mean apart from perception.
You mentioned the possibility of some apparatus that inverts the images, and made the point that if such an apparatus did the inverting for both of us (for me typing, for you reading), then we could no longer say that the physical transform in the environment had the same structure as our respective perceptions. But this doesn’t matter, you see. It is an additional layer of physical transformation, but it works the same way for both of us. Without that added layer, we have both come to be organized so as to construct those structure-perceptions from such inputs. This ancillary apparatus is the same for both of us, each cancelling the effect of the other, so they might as well not be there. Set up the inverting apparatus for just one of us, though, and we’re in trouble, because the additional transformation that it introduces is not included in the way that we have come to be organized, and is no longer cancelled out.
Limiting it to the configuration of a letter C introduces a similar problem. We are then no longer talking about linguistic structure at a level that is useful for this discussion. For the phonological contrasts between phonemes, there is appreciable structure (albeit still at the very lowest level), but the graphological contrasts between letters are a much simpler matter. They only enter into a discussion of linguistic structure in an indirect and, especially for English, a somewhat muddled way. They are an imperfect representation of the phonemic contrasts, and only make sense when we substitute phonemic word-shapes in place of graphic word-shapes. This makes my example of words typed and appearing on the screen an unhappy one. Too many graphological irrelevancies intrude, while the phonological essentials to which they imperfectly correspond get left out. There is an added layer of physical transformation, and while it is the same for both of us, and the effect on one end is more or less cancelled on the other end, we should not focus on the apparatus of this transformation.
I know that you want to make this as ‘concrete’ and specific as possible, but I think at this stage we need at least to assume the capacity to recognize words, and an association of them into word-classes such that (to a first approximation, to simplify the discussion) the classes, each represented by a particular member word, may occur in some sequences but not in others: John ate the cake or (in a dream, a joke, etc.) the cake ate John but normally not the ate John cake and for most people (other than folks like Pope, Dryden, and Browning) not ate John the cake. If the classes or variables are named N and V, with a constant the, then we will try to make sense of N V N, with the before either or both words from the N class, but V N N is not conventional English word order (though mind you it is the preferred word order in some other languages), and the V N N just doesn’t make it.
So if I say to you structure is in the environment, this fits a familiar sentence-form that we can represent here as N is P N. Many other sentences satisfy this form, such as The cake is on the table, Freddy is in trouble, The train is on time, and so on.
There are of course important differences between these sentences, and to a truly surprising extent control of these differences turn out to amount to controlling a refinement of these word classes, such that the satisfiers of some sentence forms have the same relative normalcy, limitation as to context, etc., as the satisfiers of certain other sentence forms, in extended derivation-like chains of sentence-differences. I can’t expect you to make the necessary investment to get into the really interesting aspects of formal structure (and of linguistic information). Nor do I expect you to take this on faith as a premise for any argument I am going to construct here. The point of mentioning this is only to assure you that there is more to it, and to suggest that if you do look into it you will find the science quite solid. And also to justify ignoring those important differences between the above sentences and, for present purposes, staying with the rather crude degree of structuring that is represented above by the simple sentence-forms N V N and N is P N.
The inference is only that there is some physical transform of that structure

More than that: If you can demonstrate to me some way that we can do without the hypothesis that there is a physical transform of linguistic structure in the environment, I will cheerfully abandon it.

···

At 01:38 PM 10/24/2003 -0600, Bill Powers wrote:

The problem is that we can’t independently observe the shape of something

in the environment. All we can do is try to test whether each person is

having the same reproducible perceptual experience. Suppose, for example,

that between our eyes and objects in the environment there is a mapping

that reverses everything left for right. If that is the only

transformation, then a reversed C-shape in the actual environment will give

rise to a correct-looking perceived C. Likewise, we will scan the written

sentence in a way that is actually right to left in the environment, but

which appears left to right in perception.

Linguistic structure is in our shared environment.

I suppose that the meaning of this depends critically on what you mean by

“shared environment.” Something is certainly in the environment, but is it

literally shared? To create structures that we experience as linquistic

structures, we must do things to the environment, but what is done to the

environment need only evoke the proper perceptions; it need not itself have

the same structure.

As you know, I am trying to learn how to pronounce Chinese sentences

written in Pinyin (romanized syllables). I have practiced pronouncing

individual syllables so they sound to me like what native speakers are

saying, and the program, which can record and analyze my recorded speech,

gives me ratings from barely passing to WOW. I have learned the four tones

and the particles, and how to produce them in a fairly authentic manner, by

initating native speakers.

But after that, I turned to the sections on sentences, in which whole

strings of syllables have to be produced. And suddenly the Chinese speakers

are no longer saying the same words correctly. They pronounce two words

said like “jian” and “er” as “jar”. They no not reproduce rising and

falling tones as in the earlier practice. The rises and falls are all but

gone, and superimposed on them are much larger inflections that are

evidently the same sort of expressive tonal changes that we use in English.

Yet with practice, I began to hear the familiar words from the single-word

practice – despite the fact that the speakers are saying them differently.

What they are doing, it seems, is adjusting the way they produce sounds so

that at some level, perhaps configurations or events, the listener

perceives the intended words. The reality is distorted to make the

perception come out right.

I guess that’s the point I’m getting to so slowly. The reality is

manipulated so the word perceptions come out right; the word perceptions

are manipulated so the meaning perceptions come out right. The greatest

variation is at the start of this chain, as is typical of all control

processes. If structure were truly in the environment, we would find the

greatest variability at the end of the chain, in the meanings. But we

don’t. It’s the other way around. Initially, the variations appear at the

end of the chain, but we adjust the beginning of the chain until the end,

where the meanings are, is properly controlled. Then we find that it is the

beginning of the chain where the most variations appear.

I trust that you will take these observations and dump them in the hopper,

where the mill will combine them with other ingredients to produce the next

iteration of this gradually growing analysis.

Best,

Bill P

[Richard_Marken2]

[From Rick Marken (2003.10.26.2130)]

Bruce Nevin (2003.10.23 14:59 EDT)--

My apology, Rick, for the very long delay in replying!

No problem. I can tolerate a pretty long transport lag for these higher level variables.

But another way of putting your (1), which is I believe what Bill had in mind, and is certainly what others over time have had in mind, is that for many words it is difficult to pin down any specific perceptions that are reliably associated with them as their referents. What are the referents of the verb get, for example, or the preposition up.

I agree that words can have several perceptual referents. I thought you were saying that a word like "fairness" doesn't have _any_ perceptual referent. If that's what you mean, then I think we are pretty far from agreement.

Me:

The final meaning I get -- and the one that led me to disagree with what you said
-- is: 4) people use words like "fair" and "unfair" to point to principle
perceptions but these words don't really point to these perceptions because the
perceptions themselves don't exist -- they are just "supposed".

Bruce:

Perhaps this is more clear now. The perceptions have not been demonstrated to exist on the Principle level of the hierarchy. So far, the only perceptions that have been discussed are at lower levels, primarily at the relationship level.

I don't see what demonstrating the existence of perceptions at _any_ level of the hierarchy has to do with the meaning of words. Indeed, I think it's impossible to "demonstrate" perception to anyone other than oneself. Maybe what you mean is that no one has tested to see if people control for "fairness". But I think they have. I gave an example of a little test for control of fairness. My kids thought it was unfair if one of them got to both divide a treat and also be the first pick their half. I could see that that was unfair myself. But the kids thought it was fair if one divided while the other picked first. I thought that was fair too. So what was a disturbance to my perception of fairness was also a disturbance to my kids' perception of fairness. I think my kids and I were (and still are) perceiving and controlling for a very similar perception when we control for fairness.

Best

Rick

···

---
Richard S. Marken
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400

[bnhpct]

[From Bruce Nevin (2003.10.27 17:15 EDT ]

Bill Powers (2003.10.24.1226 MDT)–

Something is certainly in the environment, but
is it

literally shared? To create structures that we experience as
linquistic

structures, we must do things to the environment, but what is done to
the

environment need only evoke the proper perceptions; it need not itself
have

the same structure.

The problem is that we can’t independently
observe the shape of something

in the environment.

Since we are only inferring that something is so about the environment,
we do not have to address this problem.

All we can do is try to test whether each
person is

having the same reproducible perceptual experience.

We need to test whether two people come to have the same
perceptual experience. We can’t do this for other perceptions, e.g.
qualia like the color green, but for language we can do this because the
other person can repeat the same words, and the first person will assent
that these are the same words. The other person can even paraphrase by a
purely formal rearrangement of the words, and the original speaker
assents that this second utterance “says the same thing” as the
first. (The availability of certain restructurings such as passive vs.
active, or extraction (N V N ↔ it is N that V N), is itself part
of the structure of language.)
Consider a simple lab scenario. A speaker writes something down and then
says it carefully and clearly. A hearer writes down what was said,
verbatim. An observer compares the two writings. The participants being
careful, conscientious, and native speakers of the same language, the
writings turn out always to be the same. Or, conversely, whenever the
writings are not the same, it turns out that one or both have been less
careful and conscientious.
The same linguistic structure is controlled by the speaker and by the
hearer.
Suppose there is no structure immanent in the speech signal. Zero
information. Then how does this remarkable coincidence come about? Just
as the speaker is saying something (such as the sentence “Linguistic
structure is in our shared environment”), the hearer just happens to
construct the same perceptions that the speaker was controlling. There is
no basis for doing so, since there is no information in the speech
signal. The hearer constructs different perceptions when the speaker says
something different, and those perceptions always just happen to
correspond to the perceptions that the speaker was controlling --always
with no information in the environment. Zero information in qo and qi
(proximal variables in the environment), but the full panoply of
linguistic structure constructed in perceptual signals p as these
input signals ascend the branches of the hierarchy that are used in
controlling language – perceptions constructed identically in the hearer
as in the speaker hearing herself.
Accepting this, to me, extremely implausible position, one must ask: how
did they come to be organized in the same way? By this hypothesis,
remember, there never was any information in the environment on
the basis of which either party could learn the structures of English.
That is a stupendous claim for reorganization. Enter Chomsky’s Innate
Ideas, stage left. At which point we must wonder by what miracle the
structures of language became innate in the human species, sans structure
in the speech signal. Exit innate ideas?

And as they exit, stage right, let us ask (as even Chomsky does) how much
is innate and how much is in the environment – what is the balance? This
question must be addressed, eventually, at three points: in the
execution, as speakers speak and hearers understand; in the learning, as
language is ‘acquired’; and in the evolutionary origination of language.
We are primarily concerned with the first and most immediate of these
now, but, as we have just seen, the other two become relevant from time
to time, and, critically, we must avoid conflating them.

The hearer constructs the same linguistic perceptions that the speaker
was controlling by speaking. How much information must be present in the
environment for the hearer to do that?

The reality is

manipulated so the word perceptions come out right; the word
perceptions

are manipulated so the meaning perceptions come out right. The
greatest

variation is at the start of this chain, as is typical of all
control

processes. If structure were truly in the environment, we would find
the

greatest variability at the end of the chain, in the meanings. But
we

don’t. It’s the other way around. Initially, the variations appear at
the

end of the chain, but we adjust the beginning of the chain until the
end,

where the meanings are, is properly controlled. Then we find that it is
the

beginning of the chain where the most variations
appear.

It’s the other way around. There are degeneracies in the speech signal,
but the lost information (which was intended by the speaker) is
reconstructed by the hearer. You are going all the way to
“meanings”, which you imagine to be at the top of the hierarchy
for controlling language. Information from more than one level of
linguistic structure is present in the speech signal. For example,
phonemes at morpheme or word boundaries are often pronounced differently
than in the interior of a morpheme. A stock example is the difference
between night rate (for telephone or telegraph) and
nitrate. When morphemes are combined, their phonemic shape may
change. This often passes unnoticed in your native language:
skweet for “Let’s go eat.” It’s much more noticeable in
an unfamiliar language, e.g.

suddenly the Chinese speakers

are no longer saying the same words correctly. They pronounce two
words

said like “jian” and “er” as “jar”. They no
not reproduce rising and

falling tones as in the earlier practice. The rises and falls are all
but

gone, and superimposed on them are much larger inflections that are

evidently the same sort of expressive tonal changes that we use in
English.

Say these two sentences out loud and compare the stress pattern in the
word pigpen:
John was in the pigpen with the pigs.
It wasn’t John who was in the pigpen with the pigs, it was Frank.

Yet with practice, I began to hear the
familiar words from the single-word

practice – despite the fact that the speakers are saying them
differently.

What they are doing, it seems, is adjusting the way they produce sounds
so

that at some level, perhaps configurations or events, the listener

perceives the intended words. The reality is distorted to make the

perception come out right.

No, it’s a simple case of conflict. There is only one voice pitch
variable, but there’s more than one system controlling it. What is
happening is that structure at several levels, each of which uses voice
pitch, is being collapsed to a single output of voice pitch. In the
pigpen example, above, both pitch and (probably, but disputedly)
amplitude are involved in word stress, and although the pitch of the
relative clause is reduced to a flat, low line, a bit of the amplitude
increase on pig remains. I think Chinese only uses pitch within a
word, but uses both pitch contours and amplitude or other markers of
emphasis for syntactic intonation contours. I haven’t listened to your
Chinese examples, but I would guess that, for example, a relatively large
range of difference between a high and a low pitch becomes compressed to
a smaller range, but is still present within that smaller range as words
bearing those pitches occur within an intonation contour that requires
the voice to be on the whole higher at one point and lower at another.
The two levels of structure, intonation contour and lexical pitch, are
collapsed to a single voice pitch output. Both levels of structure are
present in the speech signal.

    /Bruce

···

At 01:38 PM 10/24/2003 -0600, Bill Powers wrote:

[bnhpct]

[From Bruce Nevin (2003.10.27 17:48 EDT)]

Rick Marken (2003.10.26.2130)–

Bruce Nevin (2003.10.23 14:59 EDT)–
[…] for many words it is difficult to pin down any specific perceptions
that are reliably associated with them as their referents. What are the
referents of the verb get, for example, or the preposition
up.

I agree that words can have several perceptual referents.

Sounds easy and obvious. For starters, what are the several different
specific perceptual referents for up in the examples I gave,
viz.
look
them up
get up in
the morning
up to that
time
it’s up to
you
use
up

I thought you were saying that a word like
“fairness” doesn’t have any perceptual referent. If
that’s what you mean, then I think we are pretty far from agreement.

It was Bill that said that some words have no meaning. This appeared to
me to be due the difficulty identifying any specific perceptions that
words like “patriot” (an example in the SF story) refer to. My
rejoinder was that meaning is not limited to reference.

The final meaning I
get – and the one that led me

to disagree with what you said – is: 4) people use

words like “fair” and “unfair” to point to principle

perceptions but these words don’t really point to

these perceptions because the perceptions themselves

don’t exist – they are just “supposed”.

Bruce:

Perhaps this is more clear now. The
perceptions have

not been demonstrated to exist on the Principle level

of the hierarchy. So far, the only perceptions that

have been discussed are at lower levels, primarily

at the relationship level.

I don’t see what demonstrating the existence of perceptions at any
level of the hierarchy has to do with the meaning of words.

You claimed that the word fairness refers to a Principle
perception “fairness”. That claim is state above and as
follows:

Rick Marken (2003.09.20.0940)–

Researchers at the Yerkes laboratory report
that monkeys

will protest getting unequal rewards for equal work.

[…] The researchers see this as evidence of the

monkey’s ability to detect “fairness”, which would mean

that the monkeys can perceive and control the world in

terms of principles. This would mean that the monkeys

are at about the same level of perceptual development

as most four year old children, who can can perceive

this principle […].

I questioned that claim, and still do. This is not a claim that the word
fairness has no meaning.

I gave an example of a little test for control
of fairness. My kids thought it was unfair if one of them got to both
divide a treat and also be the first pick their half. I could see that
that was unfair myself. But the kids thought it was fair if one
divided while the other picked first. I thought that was fair too. So
what was a disturbance to my perception of fairness was also a
disturbance to my kids’ perception of fairness. I think my kids and
I were (and still are) perceiving and controlling for a very similar
perception when we control for fairness.

Fair enough. But this seems to me to be only a sequence for controlling a
relationship of equality, not a principle. You could use those words to
talk about this with them, and with us. You could talk about a sequence
(one person cuts, the other person chooses) that results reliably in more
nearly equal division. When your child says that it is “unfair”
to get less than their sibling (a relationship perception), are they
reporting anything more than emotion associated with failure to control
that relationship? When “unfair” is also said about a sequence
arrangement whereby the other person can bring about the
“unfair” relation of inequality, evidently the word refers to
perceptions of more than one kind. The “unfairness” is
transitive to the means for bringing it about. Unfairness seems to be a
word for negative emotions. “Fairness” seems to be defined in
negative terms: talk about “fairness” here seems to be nothing
more than a story told about the absence of this “unfairness”.
By the way, what perception does “fair” refer to in the first
sentence of this paragraph?

    /Bruce

Nevin

···

At 09:31 PM 10/26/2003 -0800, Rick Marken wrote:

[Richard_Marken2]

[From Rick Marken (2003.10.27.2000)]

Bruce Nevin (2003.10.27 17:48 EDT)

Rick Marken (2003.10.26.2130)--

I agree that words can have several perceptual referents.

Bruce:

Sounds easy and obvious. For starters, what are the several different specific perceptual referents for up in the examples I gave, viz.

    look them up
    get up in the morning
    up to that time
    it's up to you
    use up

I'd have to use more words or, better, pictures, to show you what I think are the perceptual referents. But I don't think I need to. I bet you know as well as I do what the different referents of "up" are in these examples. But there examples are a great way of showing that words can have several perceptual referents. Good work.

It was Bill that said that some words have no meaning...My rejoinder was that meaning is not limited to reference.

Ah, yes. That other kind of "meaning" that has no meaning to me because I don't know to what it refers.

Rick:

I don't see what demonstrating the existence of perceptions at _any_ level of the hierarchy has to do with the meaning of words.

Bruce:

You claimed that the word fairness refers to a Principle perception "fairness"...

I questioned that claim, and still do. This is not a claim that the word fairness has no meaning.

Ah, so you think the perception to which the word "fairness" refers is not a Principle type of perception, a la the hypothetical HPCT hierarchy. Is that it? If so, that's fine, as long as we agree that there is a perception to which the word refers.

Rick:

I gave an example of a little test for control of fairness...

Bruce:

Fair enough.

Ah, so you do know what "fair" means.

But this seems to me to be only a sequence for controlling a relationship of equality, not a principle.

OK. Call it what you will. It seems like most people would describe "fairness" as a principle. For example, here's a little discussion of Shakespeare's _Measure for Measure_:

The Lord Chancellor, and later the Court of Chancery worked
on the principle that "No wrong should be without an adequate
remedy"—fairness, even if that meant bending the law a bit.

So this writer, at least, classifies the perception that corresponds to "fairness" (which is even described in more detail as the perception of "No wrong without adequate remedy") as a principle. But if you want to think of it as a sequence then be my guest.

You could use those words to talk about this with them, and with us. You could talk about a sequence (one person cuts, the other person chooses) that results reliably in more nearly equal division. When your child says that it is "unfair" to get less than their sibling (a relationship perception), are they reporting anything more than emotion associated with failure to control that relationship?

Of course. They don't say it's "unfair" when they are unable to control the relationship between the position of a cursor and the position of a target. They (and I and most everyone I talk with on a daily basis) use the word "unfair" to refer to a particular perception -- unless, of course, they are trying to be funny, in which case they might say that the world is being "unfair" to them when they are unable to keep the cursor on the target.

By the way, what perception does "fair" refer to in the first sentence of this paragraph?

For me, it refers to the same perception it always refers to: a perception of appropriateness or justice. When you say "that's fair" or "fair enough" in reply to what I post I assume you mean that what I said was in some way appropriate or just -- le mot juste.

But perhaps you meant that I'm pretty enough. But I'll just assume that you meant something more like what I said above.

Best regards

Rick

···

--
Richard S. Marken
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400

[Richard_Marken2]

[From Rick Marken (2003.10.27.2100)]

Bruce Nevin (2003.10.27 17:15 EDT ) to Bill Powers (2003.10.24.1226 MDT)--

Excuse me for butting in but this is just too much fun.

We need to test whether two people come to have the same perceptual experience. We can't do this for other perceptions, e.g. qualia like the color green, but for language we can do this because the other person can repeat the same words, and the first person will assent that these are the same words.

But we still don't know whether the two people have the same perceptual _experience_ of the word. If I say "green" and you repeat it, all I know is that I heard you say "green" too. I don't know what you actually experienced when you said "green". You might have experienced "what I hear as "red"; same as is the case for the color itself -- the one that we both call "green", but could be experienced by you as what I see as red -- or for any perception, for that matter.

The other person can even paraphrase by a purely formal rearrangement of the words, and the original speaker assents that this second utterance "says the same thing" as the first.

This just assures you that are both referring to the same perception. It doesn't assure that you both experience the perception in the same way.

Consider a simple lab scenario. A speaker writes something down and then says it carefully and clearly. A hearer writes down what was said, verbatim. An observer compares the two writings. The participants being careful, conscientious, and native speakers of the same language, the writings turn out always to be the same. Or, conversely, whenever the writings are not the same, it turns out that one or both have been less careful and conscientious.

The same linguistic structure is controlled by the speaker and by the hearer.

Suppose there is no structure immanent in the speech signal. Zero information. Then how does this remarkable coincidence come about?

The same way it comes about with other perceptions, like color: the two people have the same perceptual functions that construct the word perception from sensory input. That's why I can paint a green patch and you can imitate it; indeed, you can imitate my green using pure (narrow band reflective) green or by combining yellow and blue. We can produce the same perceptions (or the color green, or the sound of the word "green") to the extent that we have the same perceptual functions that construct those perceptions as a function of sensory data.

Just as the speaker is saying something (such as the sentence "Linguistic structure is in our shared environment"), the hearer just happens to construct the same perceptions that the speaker was controlling. There is no basis for doing so, since there is no information in the speech signal.

The basis for doing so is the perceptual functions in both individuals.

The hearer constructs different perceptions when the speaker says something different, and those perceptions always just happen to correspond to the perceptions that the speaker was controlling --always with no information in the environment.

Amazing, isn't it? But not amazing once you understand how perception works.

Zero information in qo and qi (proximal variables in the environment), but the full panoply of linguistic structure constructed in perceptual signals p as these input signals ascend the branches of the hierarchy that are used in controlling language -- perceptions constructed identically in the hearer as in the speaker hearing herself.

Right. Because p = f(qi). It's the nature if f() -- and qi-- that determines p. But qi alone doesn't determine what p will be. The same qi (with the same presumed "information" about what it "is") will be perceived differently by people with different perceptual functions.

Accepting this, to me, extremely implausible position, one must ask: how did they come to be organized in the same way?

That is a good question. I think it was evolution. At least, I think it was evolution that determined the classes of perceptual function, f(), that can exist in the brain.

By this hypothesis, remember, there never was any information in the environment on the basis of which either party could learn the structures of English.

You betcha.

That is a stupendous claim for reorganization.

It is, indeed. Though it's reorganization that is constrained to come up with particular classes of perception.

Enter Chomsky's Innate Ideas, stage left. At which point we must wonder by what miracle the structures of language became innate in the human species, sans structure in the speech signal. Exit innate ideas?

I think the question is "what is innate?" I think organisms do have an innate ability to perceive the world in certain _ways_ -- to construct perceptions of particular types. But what specific perceptions of a general type are constructed is, I believe, largely learned (the result, that is, of reorganization).

Well, that's how I see it, anyway. The environment -- including the contrived environment, which includes words -- just provides the raw material, in the form of sensory data, from which perceptions -- including the perception of words -- are _constructed_ by perceptual _functions_.

Best

Rick

···

---
Richard S. Marken
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400

[Bill_Powers]

[From Bill Powers (2003.10.28.0718 MST)]

Bruce Nevin (2003.10.27 17:15 EDT ]

                               EST, I presume.

The problem is that we can't independently observe the shape of something
in the environment.

Since we are only inferring that something is so about the environment, we
do not have to address this problem.

For practical "engineering" purposes maybe not. But this is a theory about
the nature of the very mind we are using to think about these things, so at
some point we have to address it. You're making a claim about the
_objective_ structure of the environment. If you'd be just as happy to say
you're talking about the structure of human perceptual input functions, so
it really doesn't matter whether there's anything in correspondence to
those perceptions in the hypothetical "shared environment," we could go on
from there. But it does seem to matter to you that language have an
objective existence somewhere outside human beings, so we can't just let
this go.

I'm starting to get an impression that you don't think there are any
nonverbal perceptions above approximately what I call the "relationship"
level -- that above that level, it's all language. You've resisted my
proposal that there is a specific "category" level, and now it seems that
this extends to the proposed principle level as well, and perhaps all
higher levels, at least in terms of nonlinguistic perceptions.

All we can do is try to test whether each person is
having the same reproducible perceptual experience.

We need to test whether two people come to have the same perceptual
experience. We can't do this for other perceptions, e.g. qualia like the
color green, but for language we can do this because the other person can
repeat the same words, and the first person will assent that these are the
same words.

How is this different from one person presenting qualia like green to
another person and the other person assenting that this is the same color
he experienced? You've just shifted the whole problem up a level of
perception, without solving it. I write the string of configurations
<WORD>, and you assent that you see the same string of configurations, but
we have no way of knowing whether your experience of <WORD> is the same as
mine. We're talking about shapes rather than colors, but that's a trivial
difference. This same problem exists at every level. If I say that object A
is "on" object B, and show you what I mean by arranging the two objects
that I see, you can assent that this arrangement is what you, too, call
"on," but you have no way of knowing if your experience of "on-ness" is the
same as mine,

There is no place where we can draw a line and say, "above this level, we
simply know that we are having the same experience." The problem extends to
all levels of experience.

My point is not that we do NOT have similar experiences, but only that the
apparent reliability of linguistic conventions is illusory. If there is
ever to be a proof that we have similar experiences at _any_ level, it
can't come from the arguments you're presenting, which postulate their own
conclusions.

But back to what I see as the big point, exemplified in the discussion of
meaning and words like "fairness." Is there a level of perception at which
we sense the presence of a principle without using any words? I'm perfectly
willing to agree that using words makes it easier to speak and reason --
and communicate, of course -- about higher-level perceptions, but is
"covert speech" the _only_ form in which these perceptions exist? If we go
on talking only about words, written or spoken, it can easily seem that
way, since everything has to be passed back and forth between us
(especially over the internet!) as strings of words. But what about
apprehension of non-linguistic principles? I think we learn scads of
principles in utter silence, simply by observing how the world responds to
our actions and coming to appreciate generalizations that extend over many
different experiences -- generalizations that would be very hard to put
into words, like the best way to manipulate spaghetti with a fork, or how
to cancel factors in equations..

Is it really true that when you hear the word "fairness," the only thing it
means to you is a relationship of equality? Does that fit the observation
that to be fair, we should allow the rich as well as the poor to sleep
under bridges? Anyone can see that this is an invalid example of the
principle, but is it necessary to speak about it in the right way to see
that it's invalid?

The ultimate argument, I think, the one that will win the day by demanding
that the argument move up a level, will be constructed with linguistics as
its sole subject matter. I can't construct it yet, but I can see the outlines.

Linguistics, after all, exists in human perception, especially among
linguists. It should, therefore, exemplify all of the levels of perception
that human beings possess. For example, perceiving this pattern

  WORD

is an example of configuration perception.

For a higher-order example, in the string of configurations

APPLE PIE ORDER

we perceive something called "ordering" or "sequence" that makes it
different from

ORDER APPLE PIE

We also perceive a difference in the meaning of the two arrangements of the
configurations, but first we have to be able to see that the arrangements,
the sequences if we read left to right, are different. That perceived
difference shows that the two strings of words are being perceived in
_non-linguistic_ terms, the same non-linguistic terms that make us see

GOSTAK DISTIM DOSH

as different in some way INDEPENDENT OF MEANING from

DOSH DISTIM GOSTAK.

This analysis can be applied to any linguistic statement, proposition, law,
conclusion, or principle, quite aside from its meaning to linguists.

Consider these shapes: O C S

and these: W X Z

Do not these groups of shapes seem to be of different kinds, or categories?
To me, the difference appears long before I can formulate, in words, some
statement of what makes them different. Once I see the difference, I can
say that

   O C K

is wrong because K doesn't belong in that category.

Those "shapes" could, of course, be whole words, and we could quickly see
differences of other kinds:

set 1: an, at, to, id

set 2 mississippi, caterwauling, speciation, indeterminate

so the set

an, at, platitudinous, id

would contain a "wrong" element, quite independently of the meanings of the
elements. We can see what is wrong without stopping to formulate a sentence
that "explains" the wrongness.

What I'm getting at is that all the levels of human perception apply to
linguistic objects and manipulations just as they do to nonlinguistic ones.
The levels of the human hierarchy that I have proposed are "meta" to the
whole subject of linquistics, just as they are "meta" to subjects like
economics, fishing, basketball, and bank robbery. Language is just one of
the things we do by using the human hierarchy of control systems. It is a
product of the hierarchy, not the hierarchy itself.

I hope it seems to you that this gets us closer to the core of our divergences.

Best,

Bill P.

[Richard_Marken2]

[From Rick Marken (2003.10.28.1515)]

Bill Powers (2003.10.28.0718 MST)--

Bruce Nevin (2003.10.27 17:15 EDT )--

>We need to test whether two people come to have the same perceptual
>experience. We can't do this for other perceptions, e.g. qualia like the
>color green, but for language we can do this because the other person can
>repeat the same words, and the first person will assent that these are the
>same words.

How is this different from one person presenting qualia like green to
another person and the other person assenting that this is the same color
he experienced?

Exactly my point in Rick Marken (2003.10.27.2100) where I say, among other things:

But we still don't know whether the two people have the same
perceptual _experience_ of the word.

More evidence of off-line collusion;-)

But back to what I see as the big point, exemplified in the discussion of
meaning and words like "fairness." Is there a level of perception at which
we sense the presence of a principle without using any words?

This is a good way of putting it. I'd be interested in hearing what Bruce has to
say. My answer, based on my experience, is that there is, indeed, a a level of
perception (people might prefer to call perceptions at that level "cognitions"
rather than "perceptions" since they seem independent of sense modality -- you
can't hear, see, taste or smell them) at which we sense the presence of a
principle without using any words.

What I'm getting at is that all the levels of human perception apply to
linguistic objects and manipulations just as they do to nonlinguistic ones.
The levels of the human hierarchy that I have proposed are "meta" to the
whole subject of linquistics, just as they are "meta" to subjects like
economics, fishing, basketball, and bank robbery. Language is just one of
the things we do by using the human hierarchy of control systems. It is a
product of the hierarchy, not the hierarchy itself.

Yes!! Well put.

Best regards

Rick

···

--
Richard S. Marken, Ph.D.
Senior Behavioral Scientist
The RAND Corporation
PO Box 2138
1700 Main Street
Santa Monica, CA 90407-2138
Tel: 310-393-0411 x7971
Fax: 310-451-7018
E-mail: rmarken@rand.org

[bnhpct]

[From Bruce Nevin (2003.10.28 15:22 EST]

Bill Powers (2003.10.28.0718 MST)–

The
problem is that we can’t independently observe the shape of
something

in the environment.

this is a theory about
the nature of the very mind we
are using to think about these things, so at

some point we have to address it. You’re making a claim about the

objective structure of the environment.

I claim that it exists, not what it is. Our effects on it are our means
for controlling our perceptions of it. We can observe those effects by
means that are independent of our perceptual input functions for
language, which give us our subjective experience of language in use.
from observations in a variety of perspectives, extending our senses by
instrumental means, we can infer properties of the environment, without
ever having any claim to direct knowledge bypassing perception. This is
no different from any science.

If you’d be just as happy to say

you’re talking about the structure of human perceptual input functions,
so

it really doesn’t matter whether there’s anything in correspondence
to

those perceptions in the hypothetical “shared environment,” we
could go on

from there. But it does seem to matter to you that language have an

objective existence somewhere outside human beings, so we can’t just
let

this go.

I will be perfectly content to let it go, once there is an alternative
physical explanation for the ability of one person to repeat what another
has said (which is not the same as imitating the sounds they make).
Posing the nature of this problem has been the main thrust of what I have
written. I will lay it out more carefully so that perhaps it will be a
little harder to ignore this time.
As Rick has kindly reminded us
p =
f(qi)
You and he seem to be arguing that
p =
f()
But what is going on in qi is certainly relevant – the input functions f
could produce no perceptual input p without it! – and that which is
relevant in it is put there by the behavioral outputs qo of a speaker of
the language.
The behavioral outputs qo of the speaker have effects on the environment
that are perceived by the hearer
qi -
d
The hearer constructs
p =
f(qi)
The constructs quite a number of perceptions controlled in parallel
strata or layers of linguistic structure – phonemes, syllables,
morphemes, words, etc.
The speaker A and hearer B now reverse roles. B undertakes to repeat what
A said. The behavioral outputs qo of the speaker B have perceptible
effects on the environment
qi -
d
The hearer A constructs
p =
f(qi)
A constructs quite a number of perceptions controlled in parallel strata
or layers of linguistic structure – phonemes, syllables, morphemes,
words, etc.
A perceives a repetition of what he just said – the same phonemes,
syllables, morphemes, words, phrases, clauses, sentences, etc. All the
structure-perceptions of the utterance are absolutely identical to the
structure-perceptions that A was controlling by speaking to B just a
moment ago.
Now A repeats what B just said. And B perceives that all the structure of
A’s utterance is absolutely identical to what A had said before, which B
had just repeated.
To be sure, the outputs qo of B are not identical to qo of A. The aspects
that are different, such as voice pitch, precise phonetic shape of
phonemes, duration of syllables, amplitude, and so forth, are not
elements of the structure of the utterance. It’s not that they are
disturbances, they just don’t make a difference in the input functions f
of the hearer.
Nor of course are the same disturbances d present in the environment.
These are additional non-structural variations that the hearer ignores.
Assume a quiet, acoustically protected environment for the present. We
can talk about how this ignoring is done later, but it is a
separate discussion.

Now,

    p  =

f(qi)

In the usual situation that we diagram, the CV is some independent
variable in the environment. However, in this case the CVs in the
environment are the speech outputs of the other person

    qi = qo +

d

Therefore, by substitution

    p  =

f(qo + d)

and since we have abstracted d for the moment,

    p =

f(qo)

For clarity I should distinguish qiA from qoB and the converse, but I
have put it this way because this is the form of the environmental
feedback by which the speaker monitors her own speech. Substituting, as
above, qo for qi, the speaker, A, is monitoring

    pA =

fA(qoA)

even as the hearer, B, is also listening to qiB = qoA, so that

    pB =

fB(qoA)

Then they reverse roles. Since B intends to repeat what A said, and since
all parties agree that B is successful in doing so, the references for
the CVs to be repeated are set the same as the perceptual inputs that B
just heard from A:

    rB = pB

By substitution of pB = fB(qoA), then

    rB=

fB(qoA)

As speaker, B monitors

    pB =

fB(qoB)

even as the hearer, A, is attending to

    pA =

fA(qoB)

They reverse roles again. Since A intends to repeat what B said, and
since all parties agree that A is successful in doing so, the references
for the CVs to be repeated are set the same as the perceptual inputs that
A just heard from B:

    rA = pA

And, as above

    rA =

fA(qoB)

A has assented that B was successful in the repetition. This means
that

    rA =

fA(qoB) = fB(qoB) = rB

    rA =

rB

We have already agreed that A and B have come to be organized in the same
way in respect to this control of language, so

    fA =

fB

Now, closing the loop through the environment we have

    qo = qi -

d

or, abstracting d under laboratory conditions

    qo =

qi

These physical effects in the environment are the means by which both the
speaker and the hearer control their perceptual input p.

Trivially, if

    qi = 0

(no information in the environment), then

f(qi) gives the same value for p no matter what effect the speaker has on
the environment

    qo =

qi

Or even

    qo =

qi-d

(For surely unpredictable disturbances do not enable f to construct
p.)

In other words, if there is no information in the environment, then p is
imaginary.

    p =

f()

Suppose by some actions I control some independent CV in the environment,
meaning, one that is not dependent upon this sort of reciprocal
correspondence that is a necessary and distinguishing characteristic of
language. Let’s say I paint something green. The perception that it is
green is a perception that is constructed by a perceptual input function
f:

    f(qi) =

p

But we do not doubt for a moment that there is something really in the
environment to be perceived. We can even determine, by instrumental
extensions of our senses, how that something is structured so that most
light energy is absorbed and only that energy is reflected to the eye
which falls within a range that we perceive as green. We’re still talking
about perceptions, to be sure, however extended. But it is by such means
that physics, chemistry, and the other sciences claim to infer (and test)
objective knowledge of reality.

In the same way, we can instrumentally identify variables in the acoustic
properties of the speech signal as A speaks and as B speaks. We can
identify changes in amplitude, including silences, concentrations of
periodic energy at different levels of the acoustic spectrum, changes in
these, different spectral distributions of aperiodic noise (of nasality,
sibilants, fricatives, affricates, stop releases, etc.) and so on. As
observers with our instruments, we see that in the speech of B these
variables are articulated the same as in the speech of A, and vice versa,
as each repeats what the other has said. This is quite apart from
anyone’s input functions for speech, pre-set in speaker and hearer (and
observer) by evolution and by epigenetic reorganization. Again, we’re
still talking about perceptions, however extended. But this kind of
concordance of disparate instrumental and perceptual means is one basis
for the claims of science to construct and verify objective knowledge of
reality.

The acoustic signal is the result of the speaker controlling all of these
layers of structure simultaneously. The speech signal is like a merger of
overlays. A very rough analogy: several sine waves can be combined
forming a single complex wave. By fourier analysis we can recover the
constituent simple sine waves. All are present in the complex wave.

As you work to develop your perceptual input functions for Chinese, you
find that the input functions that you established for words in isolation
are not recognizing the same words in running speech. This is because in
running speech the speaker is simultaneously controlling other structures
in addition to syllables, morphemes, and words. You might have already
noticed the same sort of thing even within words (as is the fluent
listener). The same phoneme is pronounced differently in different
syllables, the same syllable is pronounced differently when stressed or
unstressed, and so on, and the difference in each case is predictable.
The predictability is somehow built into the input functions as
differences that don’t make any difference.

If the intonation contour for the sentence calls for voice pitch to go
up, then the entire range from high tone to low tone is constrained to a
narrower band of pitch frequencies, because the lower frequencies are not
available within that portion of the intonation contour. At a part of the
utterance where the intonation contour is high, low tone may possibly be
higher in frequency than high tone at a part of the utterance where the
intonation contour is low.

The first language structures that a baby learns are the intonation
contours. Charles Ferguson, at UCLA, used to recommend that language
learners spend a long time (a month or two, maybe) just listening, not
attempting to speak. Then another period babbling, just producing the
acoustic outlines of assertions, questions, side comments and
subordinated clauses within these, and so on. These intonation contours
almost always remain foreign in a foreign accent, no matter how
proficient an adult learner has become. Then begin putting syllables and
some of the simpler words into the babbling.

The acoustic signal qo = qi -d (or in our laboratory qo = qi) is the
result of controlling all of these layers of structure simultaneously.
For example, there are physically identifiable traces in the speech
signal of the boundaries between syllables. These may be identified quite
independently of the perceptual input functions f that have been
established (by whatever means) in the two participants A and B in our
laboratory experiment. This is a physical transform of perceptual
variables controlled by A and by B.

I’m starting to get an impression that you
don’t think there are any

nonverbal perceptions above approximately what I call the
“relationship”

level – that above that level, it’s all language.

This is a mistaken impression. It is just that verbal reports are not
trustworthy by themselves. Even verbal rejoinders are not sufficient
evidence of resistance to disturbance for purposes identifying controlled
variables. For one thing, as we see demonstrated here from time to time,
control of the structures of language and of argumentation can be a quite
beguiling alternative to control of the variables to which they may
refer. For another, there is the little matter of awareness. In the old
distinction between precept and example, I’ll take example any day as
evidence of what someone is really controlling, as opposed to what they
tell you they are controlling, or even what they tell themselves
(convincingly!) that they are controlling.

How is this different from one person
presenting qualia like green to

another person and the other person assenting that this is the same
color

he experienced?

The word qualia is in contrast to quanta, and refers to the subjective
experience of one’s perceptions. Your subjective experience is as
inaccessible to me as is the Real “thing in itself” Ding an
sich Reality in the environment, and for the very same reason.

On the other hand, we could set up an exchange between A and B, analogous
to the language demonstration outlined above, in which each had a set of
color chips. A picks up a green chip and shows it to B. B picks up a chip
and shows it to A. A, comparing the two chips, either nods yes that it is
the same color or no that it is not. (We need to exclude use of language
for obvious reasons.) If the color differentiation among the chips is
great – just the six primary and secondary colors, say – then they
would reach agreement. But this is because we have put the work of input
functions and output functions into the environment and done it for them.
This gives the appearance of repetition vs. imitation. If they have a
full range of color chips, then it depends on which chip they pick up,
and how close is close enough, and that kind of discrimination and
imitation will differ from one individual to another and from one
occasion to another, idiosyncratically. This is because none of this is
pre-set by lifelong practice according to pre-established social
convention.

In all of this, as for language, the qualia of the subjective experience
has nothing to do with it.

There is no place where we can draw a line and
say, "above this level, we

simply know that we are having the same experience." The problem
extends to

all levels of experience.

For subjective experience, yes. The process outlined above establishes
that two individuals are controlling the same CVs in the same way,
whatever their subjective experience of their perceptions may be.

But back to what I see as the big point,
exemplified in the discussion of

meaning and words like “fairness.” Is there a level of
perception at which

we sense the presence of a principle without using any words?

I think yes, there is, and further that the stories we tell about those
perceptions are unreliable guides as to what they are.

Is it really true that when you hear the word
“fairness,” the only thing it

means to you is a relationship of equality?

Not at all. But in the examples given I saw no evidence of control at any
higher levels, and that is what I said and why I said it.

    /Bruce

···

At 09:05 AM 10/28/2003 -0700, Bill Powers wrote:

[Richard_Marken2]

[From Rick Marken (2003.10.29.1500)]

Bruce Nevin (2003.10.28 15:22 EST) –
As Rick has kindly reminded us

p = f(qi)

You and he seem to be arguing that

p = f()

Not at all. When I say p=f(qi) I mean p = f(qi). My point was that,
if the perceptual function, f(), is the same in two people then both people
will perceive variations in qi in the same way. The equivalence of perceptions
of person 1 and 2 comes from the fact that both are perceiving via the
same perceptual function, f(), not because qi is objectively the same for
both persons. For example, suppose that qi is actually a vector of two
variables, x and y, and that f(x,y) =2x+y. Now if x = 3 and y = 2 for both
people, then both will perceive 8. But suppose x = 3 and y = 2 for person
1 and that x = 4 and y = 0 for person 2. Now qi is different for both people
but both will perceive the same thing because both perceive qi in terms
of the same perceptual function, f(), of qi. Of course, this works for
both people; both will perceive 3,2 and 4,0 as the same. This is what actually
happens with color perception. We both see the same color (green, say)
if I see the proper combination of wavelengths (blue, yellow) and you see
only a single wavelength (green). The same could also happen with
words. We both could perceive the same word even if I say it by saying
the phonetic equivalent of 3,2 and you say the phonetic equivalent of 4,0.

Bill said:

But back to what I see as the big point, exemplified
in the discussion of

meaning and words like “fairness.” Is there a level of perception at
which

we sense the presence of a principle without using any words?
You say:
I think yes, there is, and further that the stories
we tell about those perceptions are unreliable guides as to what they are.
So you agree that there is a level of perception at which we sense a
principle, like fairness, without using words. That is, we do perceive
principles. You just believe that what we say about those principles are
unreliable guides to what we are actually perceiving. Is this right? If
so, I agree. I know “fairness” when I see it but I’m not sure I can
describe it perfectly accurately. Just pretty darn accurately.
Bill said:

Is it really true that when you hear the word
“fairness,” the only thing it

means to you is a relationship of equality?
You say:
Not at all. But in the examples given I saw no evidence
of control at any higher levels, and that is what I said and why I said
it.
I presume you are referring to the study of monkey’s refusing the lesser
reward after seeing their peers get better ones. What was the evidence
you saw that indicated to you that these monkeys were not controlling
for fairness? It seemed to me that you were saying that the researchers
were confused because they were trying to study the control of fairness
in monkeys; they were confused because fairness is just a word that “reifies”
something – the principle of fairness – that doesn’t really exist.
Best

Rick

···

–

Richard S. Marken, Ph.D.

Senior Behavioral Scientist

The RAND Corporation

PO Box 2138

1700 Main Street

Santa Monica, CA 90407-2138

Tel: 310-393-0411 x7971

Fax: 310-451-7018

E-mail: rmarken@rand.org

[Bill_Powers]

[From Bill Powers (20903.10.30.0851 MST)]

Bruce Nevin (2003.10.28 15:22 EST])--

Rick has said what I was preparing to say, but I'll give it a go, too.
Start with your rendition of the input function (a very good idea to
approach it this way!):

As Rick has kindly reminded us

        p = f(qi)
You and he seem to be arguing that
        p = f()

Not quite, but close. Recall the demo I put on at the Los Angeles meeting
-- the 150 control systems operating simultaneously and independently
through a shared environment containing 150 variables. In any one of these
control systems, the input function would be written as

p = a[1]*v[1] + a[2]*v[2] + ... + a[150]*v[150]

There are 150 environmental variables (common to all the control systems)
designated v[1] to v[150]. For any one control system, the magnitudes of
these variables are summed with weights a[1] to a[150]. the sum being the
value of the perceptual signal. The demo was initialized with 150 sets of
150 weights, the weights being randomly chosen in the range from -1 to
1. With the output weighting matrix being the transpose of the 150 x 150
input weighting matrix (how it comes to be that is a matter for
reorganization to explain), the demonstration showed that each of the 150
control systems could bring its own perceptual signal very close to its
given reference signal, independently of the reference signals entering the
other 149 control systems.

In this demo, we can see the 150 perceptual signals. In a hierarchical
system, they would become the inputs to the perceptual input functions of
the next level of control systems. But going the other way, what and where
is the qi corresponding to each of the 150 perceptual signals?

The answer to "what" is "nothing" and and to "where" is "nowhere." There is
no single qi in the environment that corresponds to any of the perceptual
signals. Instead, each perceptual signal stands for a unique (most likely)
weighted sum of _all 150_ environmental variables. Change the weights, and
the corresponding perceptual signal stands for a different function of the
same 150 variables. By picking the input weights at random, I demonstrated
that control is possible for _any random way_ of creating qi's.

For that is what these weights do: they _create_ input quantities, qi's.
And this shows that it is not necessary for input quantities to make sense
to an outside observer, or even to be physically meaningful. Each time I
restarted that demo program, a new set of random weights was created, yet
(you will remember) each time all the control systems were able to bring
their respective perceptual signals close to the values of their (randomly
chosen) reference signals.

Now try to imagine a hierarchy built up in this way. At higher levels, the
input signals for _each_ control system are the perceptual signals from (in
principle) _all_ of the lower-order systems. If the weights are chosen
randomly, and if the output weighting matrix is the transpose of the input
weighting matrix at each level, each control system will be able to control
its own perceptual signal relative to the net reference signal being given
to it by the next higher level of system. So the whole hierarchy, with its
input weightings being set at random in every system at every level, will
create for itself a whole series of universes of input quantities, with all
systems at all levels controlling successfully and independently.
Controlling what? A universe of "input quantities" it has created for
itself -- not out of nothing, but nevertheless without regard to the actual
structure of Reality.

Now the next step of this somewhat involved argument. We have a random
hierarchy that controls its random perceptions by means of systematic
actions on the real reality, out there. Nothing in that real reality,
however, has any direct connection to the perceptions being controlled.
Every perception, we might say poetically, is a function of every quark in
the universe that affects our sensory endings.

BUT:

The demo in LA assumed that the _only_ effects on the environment variables
were the output actions of the 150 control systems. There were 150 random
disturbances, but they were constant. And the environmental variables
didn't have any effect on _each other_. The environment was essentially
unstructured. What happens if we put a system like this into a structured
environment?

Now, I think, we will begin to see that conflicts will develop in the
randomly-organized hierarchy. Presumably, reorganization will therefore
commence, due to effects of the enviroment on the physical structure of the
controlling system. And reorganization will not cease until the control
systems are once again all keeping their perceptions at their respective
reference levels of the moment. The perceptions they are controlling will
now be different functions of the environmental variables than they were
before: the weights will have changed. Each perceptual signal will still be
a function of many or all of the lower-order perceptual signals (sensor
signals at the first level), but the functions will now reflect the
structure of the environment, or enough of it to allow survival to learn
again another day. We will still not be able to find a single qi to go with
each perceptual signal, but the structure of reality will nevertheless be a
factor in the organization of the control systems.

Among the structured elements of reality are other people. This gets us
back to the question at hand. When people operate on their environments to
evoke perceptions in themselves and in other people -- the process we call
"communicating" -- the transfer of information is mediated by the external
reality. But does this mean that somewhere between the perception in one
person and the perception in another person, there is a "qi" that is the
same for both of them? This is a possibility, or would be but for one
aspect of reality I have left out so far: the degrees of freedom of the
real world.

In the LA demo, there were 150 variables and 150 perceptual signals under
control. But I could just as easily have used 1500 environmental
variables, with 1500 weights that would have been 90% redundant. Since
there are only 150 perceptual signals, it's clear that the actual state of
the environmental variables could not possibly be represented by the
available perceptual signals. And more directly to the point, it's clear
that there would be infinitely many different combinations of values of the
environmental variables that would produce exactly the same set of 150
perceptual signal values.

So what would it mean, in terms of the state of the environment, if we
found that the 150 control systems continuously maintained their 150
perceptual signals in a match with the 150 steady or varying reference
signals? Would it mean that the environment was being maintained in a
constant state? Far from it! In fact, the probability would be that
bringing the perceptions to the same set of reference levels a thousand
times would result in 1000 different states of the environment. Of course
those differences would be invisible to anyone (or any higher-level system)
looking just at the perceptual signals. That Observer would think that the
environment had come back to exactly the same state all 1000 times.

I think we can be pretty confident that this is the actual case. There may
be 10^10 neurons and 5*10^10 connections in the brain, but there are
10^30th or more quarks in the universe, even the local parts of the
universe, maybe even in just my little finger, I should think. So it is
literally impossible to step in the same river twice, or even to repeat the
same step twice. Our universe consists of all the perceptual aspects of it
that matter to us, leaving by far, vastly by far, most of its aspects
unknown and uncontrolled.

If you look at sonograms of spoken language, it may seem that people repeat
the waveforms when they repeat the same words. In fact, they never repeat
the same waveforms, as close inspection will always show. It's just that
our visual perceptions of waveforms depend on only a few of the degrees of
freedom that are actually there, even at the gross level of where the
darkened marks are located, and how dark they are. Since our perceptions
repeat, we assume that the reality must have repeated.

That is clearly a wrong assumption.

Best,

Bill P.

[Philip_Runkel]

Phil Runkel replying to Bill P's of 2003.10.30.0851 MST:

Bill: Your essay on the relation between me and the universe (via the
150-and-more variables) slips into my mind with less friction than
anything I have ever read before. And offers a picture of how the
correspondences and no-correspondences can be there together.
Thanks. --Phil R.

[Bill_Powers]

[From Bill Powers (20903.10.30.0851 MST)]
[Correction to original post with this date-time stamp]

There may be 10^10 neurons and 5*10^10 connections in the brain, but there
are 10^30th or more quarks in the universe, even the local parts of the
universe, maybe even in just my little finger, I should think.

The number of connections was supposed to be 5000 per neuron, so the total
should have been 5*10^13. Even at 50,000 per neuron, the total would be
only 5*10^14, which is 100,000 billion times smaller than that
modestly-estimated number of quarks.

Best,

Bill P.

[bnhpct]

[From Bruce Nevin (2003.10.31 12:44 EST])]

Bill Powers (20903.10.30.0851 MST)–

Rick has said what I was preparing to say,

He was talking about the subjective experience of
one’s perceptions. I don’t see any connection.

Recall the demo I put on at the Los
Angeles meeting

– the 150 control systems operating simultaneously and
independently

through a shared environment containing 150 variables.

Try as I might, I have not been able to see how this explains how A and B
repeat each other’s linguistic structures (phonemes, syllables,
morphemes, words, etc.), and confirm that they have done so.
One person can repeat exactly what the other person said, and the other
can confirm this and repeat it back, whereupon the first person can
confirm both that last repetition and (by means of it) their own prior
one. And an observer fluent in the same language can confirm that all the
things said are repetitions of one another. (Bear in mind the difference
between imitation and repetition.)
The two (or three) people are structured in the same way with respect to
speaking and hearing their language – but not with respect to
hearing these particular utterances on this occasion! Indeed, it can be
an entirely novel utterance that probably no one has ever said before, as
Chomsky’s famous “colorless green ideas sleep furiously” was
novel in 1957.

Whether you imagine with naive literalness that the shapes of syllables
and words float somehow in the vibrating air between them, or at the
other extreme splinter the origins of perceptual inputs over a vast field
of quarks or even of strings and the supernumary dimensions that
differentiate them, that problem is unchanged. And so far unanswered. In
fact, you must answer it to avoid solipsism.

We could discuss this very interesting proposal at great length, but
until its relevance is shown, my further comment on it has no place in
this thread.

What’s missing is qi = qo. That requires the environment.

If you look at sonograms of spoken language,
it may seem that people repeat [you mean imitate]

the waveforms when they repeat the same words. In fact, they never repeat
[you mean imitate]

the same waveforms, as close inspection will always show.

Of course not. Nor was that ever claimed, and in fact I have emphasized
the opposite many times.
To say that qo is a physical transform of the linguistic structure
perceptions controlled by the speaker does not require that there be only
one possible such transform, and a one-many transform from p = r to qo
does not preclude a many-one transform from qo = qi to p = r.
Around the loop, continuously, a difference in one loop segment is
transformed to another difference in the next loop segment, and the
loop is closed through the environment. The transform from neural
signal to physical effect is not exceptional, it is essential.
Furthermore, in this case the CV in the environment is the other’s qo, a
physical transform of perceptual variables that the other is controlling
– that other who you know (or assume) is controlling perceptions of
linguistic structure that you also control, and just as the physical
effects of your own speaking (i.e. your perceptual inputs of the physical
transforms qo = qi) are not a constant one-one transform but variable
over a one-many range, yet you perceive them as the syllables, words,
etc. that you intend, so also his. This is not a problem, because you are
not imitating; you are repeating. The ability to repeat rather than
imitate is possible with language. You must understand the difference and
bear it in mind if you want to concern yourself with language and
perceptual control. In the above, you have not.

    /Bruce

Nevin

···

At 10:21 AM 10/30/2003 -0700, Bill Powers wrote:

[bnhpct]

[From Bruce Nevin (2003.10.31 13:50 EST)]

(Bill, I see that this was the post by Rick that you were referring
to.)

Rick Marken (2003.10.29.1500)–

if the perceptual function, f(), is the same
in two people then both people will perceive variations in qi in the same
way. The equivalence of perceptions of person 1 and 2 comes from the fact
that both are perceiving via the same perceptual function, f(), not
because qi is objectively the same for both persons.

In the situation I described, both people are perceiving the same qi at
the same time: the hearer to control recognizing it, and the speaker to
control speaking it (more exactly, to control having spoken it). There
can be no question that it is objectively the same qi – it is the very
same qi.

It is because of properties of qo imparted by the speaker’s control
actions that the hearer can recognize in perceptions constructed from (qi
= qo) the perceptions intended by the speaker. Those properties are a
physical transform of the perceptions controlled by the speaker and
recognized by the hearer. That physical transform, qo = qi, in the
environment, is the means by which the perceptions in the hearer come to
be the same as the perceptions in the speaker. (I am talking about
perceptions of linguistic forms, here, don’t go off on a wooly chase
about perceptions of meaning.)

For example, suppose that qi is actually a
vector of two variables, x and y, and that f(x,y) =2x+y. Now if x = 3 and
y = 2 for both people, then both will perceive 8. But suppose x = 3 and y
= 2 for person 1 and that x = 4 and y = 0 for person 2. Now qi is
different for both people but both will perceive the same thing because
both perceive qi in terms of the same perceptual function, f(), of qi.

This is not relevant to the situation that I described, in which qo = qi
is the same for both people.

I presume you are referring to the study of
monkey’s refusing the lesser reward after seeing their peers get better
ones.

Also your examples of cut/choose, and other examples subsequently
given.

What was the evidence you saw that indicated
to you that these monkeys were not controlling for fairness?

This is putting it backward. It was you who said that the scientists’ use
of the word fairness means that they must be claiming that the
monkeys are controlling on the Principle level. I need more evidence than
the scientists’ use of the word fairness, whatever that actually
means to them, to conclude that the monkeys are controlling a perception
on the Principle level. As I said, all I saw was evidence of control on
the Relationship level. (And that’s all I’ve seen so far in subsequent
examples of humans said to be controlling “fairness”.)
In order for you to conclude from the scientists’ use of the word
fairness that they must be claiming that the monkeys are
controlling on the Principle level, you must have believed that the word
fairness necessarily refers to control of a perception on the
Principle level. It seems to me that the word fairness is used to
refer to a variety of things, kind of like up and get. (And
of course those scientists probably haven’t thought much about the
perceptual hierarchy, but we’ll let that pass.)

Now, what does it take to demonstrate that a given perception is on a
certain level? Some descriptive attributes, loop delay greater at higher
levels than lower, and I/O relations to other levels – those are three
kinds of criteria that I can think of off the top of my head. If it looks
like a relationship, and doesn’t have I/O connections with Programs, it’s
probably not a Principle. But the socalled “fairness doctrine”
is a bit of legal code from the FCC that reads rather like a program.
Maybe a Principle of “fairness” is implemented there by means
of Programs. It may be currently in serious danger of disappearing. That
would be unfair, don’t you think?

    /Bruce

Nevin

···

At 03:01 PM 10/29/2003 -0800, Richard Marken wrote:

[Bill_Powers]

[|From Bill Powers (2003.10.31.1203 MST)]

Bruce Nevin (2003.10.31 12:44 EST] --

I think there is an answer to your proposals, but I will have to think
about this for a while to find a demonstration. Briefly, qi is NOT in
general equal to qo, and your qi may be a transform of my qi (and vice
versa). The question that must be answered is how two people can go through
the verification process you describe and still be perceiving differently.
I believe this is not only possible but the most likely case. But I can't
prove it yet.

Best,

Bill P.

[bnhpct]

[From Bruce Nevin (2003.10.31 14:42 EST)]

Bill Powers (2003.10.31.1203 MST)--

···

At 12:07 PM 10/31/2003 -0700, Bill Powers wrote:

I believe this is not only possible but the most likely case. But I can't
prove it yet.

That will be very interesting indeed. I'm open to the possibility.

         /Bruce

[Bruce_Gregory4]

[From Bruce Gregory (2003.10.31.1448)]

Bill Powers (20903.10.30.0851 MST)

Not quite, but close. Recall the demo I put on at the Los Angeles
meeting
-- the 150 control systems operating simultaneously and independently
through a shared environment containing 150 variables. In any one of
these
control systems, the input function would be written as

p = a[1]*v[1] + a[2]*v[2] + ... + a[150]*v[150]

Very impressive! Am I correct that this is not a hierarchical model?

[bnhpct]

[From Bruce Nevin (2003.10.31 15:02 EST)]

Bill Powers (2003.10.31.1203 MST)--

Briefly, qi is NOT in general equal to qo,

Disturbances are not a problem for my proposal that there must be in the environment a transform of the structure that holds among the perceptions of the speaker in order for the same structure to hold among those of the hearer. And indeed, for speaker and hearer to have by epigenetic reorganization come to be organized alike for control of speaking English. And behind that for whatever genetic specializations for language as may exist in humans to have arisen in the evolution of the species.

Or are you questioning that they are organized alike? After all, they are not exempt from this absence from the environment of anything but strings and quarks, all else being perceptions. Are the means of perceiving and controlling (receptors, neurons, etc.) themselves only perceptions, and, if so, perceptions constructed where and by what? There are shoals to navigate beyond this immediate issue of language and repetition.

and your qi may be a transform of my qi (and vice versa).

It may be that this is not a problem either, but I'll have to wait until you elaborate on it.

         /Bruce Nevin

···

At 12:07 PM 10/31/2003 -0700, Bill Powers wrote: