PCT study of culture

CSGnet Archive CSG2003
1

[From Bruce Nevin (2003.03.12 11:14 EST)]

I’ve changed the Subject field since the “Re: How culture colors the
way the mind works” thread has gone its own way.

Bill Powers (2003.03.09.0817 MST) –

Bruce Nevin (2003.03.08 15:13 EST) –

I don’t claim expertise or experience as an anthropologist or
ethnologist. I did live for two years in Greece 1964-65 and became fluent
in the language, and I did spend four years 1970-74 doing anthropological
linguistics (and some wildcat community development work) with a Native
American community in the Idaho corner of California. But that doesn’t
count for much.
The primary methodology for investigating cultures is participant
observation. Methods of participant observation have been developed,
refined, and disputed by anthropologists and ethnographers for a century
or so, with precursors of various less systematic kinds. Here’s one
informal statement of methodology:
Perhaps the most difficult work the anthropologist has to do is to
make his way into another codification of reality. It is difficult to
realize that what we know about human nature, about motivation, about
emotions, about satisfactions is not necessarily true of all human
beings, but may be true only of the human beings who have been brought up
within our own cultural framework. When we see the behavior of a people
of a different culture, it is difficult to refrain from making
automatically our usual assumptions as to their motives. When we read
about another culture, we have to remember to read every word; we have to
forget our habits of fast reading, because, when we begin a sentence, we
can never predict what its end will be.

Every one of the brief cultural descriptions I have given in this
volume represents the work of years. In reading accounts of a culture, I
examine each described act carefully and repeatedly, trying to eliminate
the bias and categories of the reporting ethnographer [who wrote the
account] until I can uncover each situation as it occurred. Only then can
I begin to search for the meaning of the act to the actor; and only then
can I try to discover how the situation is culturally shaped and
presented to the experiencing individual.

To make this inquiry, I have worked minutely with each culture, over
a period of years. I have paid attention even to apparently irrelevant
detail, until some item struck a resonant note. I have then gone back and
forth in my reading, seeing new relevance, noticing the obscure or
seemingly trite or obvious. I have done this until nothing remained queer
any more, and little seemed inexplicable. At this point I made tentative
predictions as to what I should find in the linguistic structure, and in
the mythology. If these predictions were confirmed, I went back to the
descriptive account, strengthened in my conclusions and with my
vision sharpened, and then I went to the language with the search for
further confirmation. This process occupied much time; but even apart
from the time-consuming work, I found that the process of understanding a
radically different codification of reality was one of long duration, one
of waiting until the acquaintance ripened and the bewildering and
contradictory became clear and simple.

Eventually, I plunged into the experience of another culture, until I
found myself liking what the members of the other society liked,
frightened by what gave them fear, grieved by their sorrows, delighted in
the situations which filled them with satisfaction and joy; and until
their categories seemed natural and right, and my own rigid and
misleading. In fact, I sometimes lost myself in the way of the life of
the society I was studying until my children complained that, after all,
they were not Trobrianders or Tikopia! What I had to say seemed so
obvious to me that, at this point, I could bring myself to write at all
only by reminding myself that I was writing for people who were ignorant
of this particular culture.
That’s Dorothy Lee, from the Introduction to Freedom and
Culture (Prentice-Hall 1959) which I recommend to you for a quick
overview of some of the more positive possibilities for human culture.
She omits the totalitarian, oppressive, and suppressive cultures, of
which there are also many, because her interest is, as I take it, in
fostering those forms of culture that are good for us, culture as support
and guide. The Jules Henry book that I mentioned earlier is a critique of
American culture as adversary to the autonomy of the individual.
There is nothing in this of population statistics. Instead, your own
person is the principle instrument of investigation. An analog of
learning a language with native fluency and then writing a grammar of it
in English, it is a method of becoming a person who perceives and
controls as members of this (initially) quite foreign community do -
gaining native fluency - and of then stepping apart from that and writing
it up as a description that natives of one’s own culture can understand
and evaluate. Having undertaken this truly vast task, however
incompletely, one comes back to one’s own native way of being human with
a kind of cognitive parallax. Without this dual perspective one’s own
culture remains as though invisible.
I have nothing decisive or definitive to say about American culture, just
a few perhaps suggestive comments. Culture at its best is a support and a
guide to conduct for the individual. Historically, cultures change
slowly. The present rate and complexity of change, especially
technological change, have famously speeded up, and the rate of contact
between people of different backgrounds. Barbara Tuchman, in her
marvellous A distant mirror: the calamitous 14th century,
explicitly draws parallels between that distraught “bad time for
humanity”, that “time of disarray”, and our own. Indeed,
it had antecedents and the repercussions persisted long after, so that
our European ancestors were damaged goods on arrival, raw from Germanic
migrations, Visigoth and Viking invasions, the Crusades, the Black
Plague, the Inquisition, the Thirty Years War, the Hundred Years War, and
so on. By contrast, peoples in the Americas and in Africa were remarkably
stable in their relations with their neighbors, warfare for the most part
almost a quasi-ceremonial sport with the function of maintaining mutual
boundaries, in which youths could test and prove valor. It is true that,
for example, ancestors of people speaking languages of the Hokan
affiliation were scattered to lands around the periphery of the
Sacramento Valley by invading Penutian speakers, but the
“invasion” took several thousand years, with no evidence of
settled warfare. (Among the descendants of the former are the Pit River
and Yana/Yahi people whose languages I studied, and among those of the
latter are their neighbors the Wintu people frequently cited by Dorothy
Lee.) In the Far East, the Han peoples rather absorbed even the Mongol
invaders, in their Aikido kind of way.
Culture in disruption and reorganization is a poor support and guide to
conduct for the individual. More improvisation is required -
improvisation upon existing cultural themes over gaps that have opened
between precept and actual conduct. Improvisation doesn’t suit everyone
equally well. Some demand that there be one right way and that any way of
bridging such gaps different from this right way (which one has of course
claimed for one’s own) is wrong.
It is not hard to see how the fragmentation and internal
contradictoriness of American culture, and its lack of support for the
autonomy of the individual even while espousing the independence of the
individual (example vs. precept), has emerged from cultural collisions in
the history of this country.
Your experience of alienation is itself entirely characteristic of
American culture today, and finds expression even in your seeing the
culture of science as vastly superior to the ethnocentric and prejudiced
views of those around us. But far from smiling dismissively, I agree with
you that the methodologies of science are for getting at what is
“really going on” without confusion from what we wish or fear
or neutrally expect. I am a great believer in the value of science for
making things better. And that is why I persist in trying to apply PCT to
what is really going on in language and culture.
The methodology of physics and the methodology of participant observation
are both ways of doing science. The methodologies of the physical
sciences are not particularly aimed at or even cognizant of culture as
such. They just aim at eliminating pre-judgement of any kind (but here
unfortunately blindness to one’s own culture - lacking that parallax -
undermines the integrity of science, most obviously, though only
scratching the surface, as handmaiden of industry - but that’s another
thread). The central difference is that in the physical sciences the
subjectivity of the investigator is inadmissible as an instrument of
investigation, and this is why the methodology of physics cannot inform
us about culture.
What of the methodology of PCT? The Test aims to get at the subjectivity
of the observed organism. The Test can identify some shared environmental
variable, a perception of which the subject is controlling. But it is not
immediately apparent for what higher-level purpose that relatively
low-level perception is controlled at the observed value(s). The
questions about higher-level purposes for observed resistance to observed
disturbance to observed environment variables that we hypothesize are
controlled - that’s where we begin to investigate culture. We want
to identify, and then disturb, and measure resistance to the disturbance
of, a social artifact, but we are arrested in our progress at the
very first step: the hypothesis that posits the controlled variable.
On this list over the years we can see many, many instances of swift,
sure assumptions made about high-level CVs where the only evidence at
hand is resistance to disturbance of some environment variable at a much
lower level. There is no safeguard against imposition of what is feared,
or hoped, or just plain neutrally expected to be so.
The arduous methodology of the ethnographer, in which the investigator’s
subjectivity is an instrument of investigation, appears to be a necessary
part of the process. For a foreign culture, that means going “out
there” or “in there” as a foreigner, becoming native
there, then writing up our experience for a reader with our original
native point of view - maintaining both modes of reality concurrently.
For our own culture, it means going out as a foreigner elsewhere,
becoming native there, then coming back with foreign eyes to where we
started so that we can see it anew - or in order to see it distinctly for
the first time. Because to distinguish what is we must be able to
perceive what it is not.
Perhaps there is some methodology other than painstaking immersion and
participant observation. So far, no one has thought if any, that I know
of.
In linguistics, the kind that I do at least, the appeals to the
investigator’s subjectivity are subject to verification by the pair test
and other substitution tests (for phonemic contrasts) and by the relative
acceptability-ranking of satisfiers of a sentence-form (preservation of
which is the criterion for mappings leading to specification of
operators, operands, and reductions within the set of sentences), but in
practice immersion and participant observation is the way to generate
hypotheses to be verified. In Generativist Linguistics ™ it’s all
about the “linguistic intuitions” of native speakers. And
language is by far the easiest aspect of culture to investigate
systematically.
A last note. The State (“government”) is a relatively recent
human invention, and the occasion of many of our discontents. Because
anthropology has articulated to us denizens of modern nation-states many
extant stateless societies it has particular relevance to anarchist and
“libertarian” interests. Anarchy is a range of views and
practices under the broad umbrella of the realization “that it is
possible and desirable for society to organize itself without
government” (Colin Ward (1973) Anarchy in Action, London:
Allen & Unwin). If we veer to politics we should at least be aware of
this.

    /Bruce

Nevin

···

At 12:41 PM 3/9/2003, Bill Powers wrote:

2

[From Bruce Nevin (2003.03.06 15:13 EST)]

Bill Powers (2003.03.09.0817 MST) –

Bruce Nevin (2003.03.08 15:13 EST) –

Far from smiling dismissively, I agree with you that the methodologies of science are for getting at what is “really going on” apart from what we might wish or fear or neutrally expect.

the interesting thing about this kind of

cultural influence is that it is designed to minimize cultural influences

True of participating in any culture.

It is designed to

increase one’s consciousness of such influences, and to teach one how to

resist them as far as possible – or at least to see the attempt to resist

as a virtue.

This, I think, is an overstatement. The methodologies of science are not particularly aimed at or even cognizant of culture as such. They just aim at eliminating pre-judgement of any kind.
Your experience of alienation is itself entirely characteristic of American culture today. Seeing the culture of science as vastly superior to the ethnocentric and prejudiced views of those around us is also perfectly consistent with this alienation.
It is not hard to see how the fragmentation and internal contradictoriness of American culture, and its lack of support for the autonomy of the individual even while espousing the independence of the individual, has emerged from cultural collisions in the history of this country. For starters, our European ancestors were damaged goods on arrival, raw from Visigoth and Viking invasions, the Crusades, the Black Plague, the Inquisition, the Thirty Years War, the Hundred Years War, and so on. Barbara Tuchman’s marvellous A distant mirror: the calamitous 14th century explicitly draws parallels between that “time of disarray” and our own.

If we do

something that disturbs others, we will see that they resist the

disturbance, and very often find that they will inform us of exactly what

we have disturbed and what will happen to us if we do it again, or too

often. If they don’t tell us, we can often figure it out. The nature of a

culture is not at all mysterious – it is, in fact, usually pretty blatant.

Ah yes, but the difficulty is that the way in which they resist (or evade) disturbance, the means by which you are informed and by whom, and just what you have to go on in order to figure it out, is often not at all obvious to a foreigner. Heck, just drive a few hundred miles south and west and walk in on a Pueblo just as friendly as you can be and see if you can figure out what you’re disturbing and what the consequences are.
When you disturb a variable in your shared environment that another person happens to be controlling they may resist the disturbance as you say (or not, if to do so conflicts with their control of some other variable). With further Testing and eliminating of alternatives this may tell you something about what that individual is controlling at that time in that place. To get from there to a generalization that this is a culture trait requires something more.
Intellectual analysis or argument about this is far from direct experience as a participant in a different culture. Americans are notoriously insulated from cultural difference, e.g. almost all of us are monolingual. The first generation of children from every successive wave of immigrants erases any trace of their overt differentness, beginning with a studied ignorance of their parents’ language. How then does one become aware of alternative ways of being human? Of how they exist and persist? Of how it is for an individual to be a human in a different way? It is very easy in this country to go through life unaware even that alternatives do exist in any real sense beyond some TV National Geographic diorama of distant exotica.
It is no easy thing to gain explicit knowledge of a culture - even one’s own! Methods of participant observation have been developed, refined, and disputed by anthropologists and ethnographers for a long time. Here’s one statement of methodology:
Perhaps the most difficult work the anthropologist has to do is to make his way into another codification of reality. It is difficult to realize that what we know about human nature, about motivation, about emotions, about satisfactions is not necessarily true of all human beings, but may be true only of the human beings who have been brought up within our own cultural framework. When we see the behavior of a people of a different culture, it is difficult to refrain from making automatically our usual assumptions as to their motives. When we read about another culture, we have to remember to read every word; we have to forget our habits of fast reading, because, when we begin a sentence, we can never predict what its end will be.

Every one of the brief cultural descriptions I have given in this volume represents the work of years. In reading accounts of a culture, I examine each described act carefully and repeatedly, trying to eliminate the bias and categories of the reporting ethnographer [who wrote the account] until I can uncover each situation as it occurred. Only then can I begin to search for the meaning of the act to the actor; and only then can I try to discover how the situation is culturally shaped and presented to the experiencing individual.

To make this inquiry, I have worked minutely with each culture, over a period of years. I have paid attention even to apparently irrelevant detail, until some item struck a resonant note. I have then gone back and forth in my reading, seeing new relevance, noticing the obscure or seemingly trite or obvious. I have done this until nothing remained queer any more, and little seemed inexplicable. At this point I made tentative predictions as to what I should find in the linguistic structure, and in the mythology. If these predictions were confirmed, I went back to the descriptive account, strengthened in my conclusions and with my vision sharpened, and then I went to the language with the search for further confirmation. This process occupied much time; but even apart from the time-consuming work, I found that the process of understanding a radically different codification of reality was one of long duration, one of waiting until the acquaintance ripened and the bewildering and contradictory became clear and simple.

Eventually, I plunged into the experience of another culture, until I found myself liking what the members of the other society liked, frightened by what gave them fear, grieved by their sorrows, delighted in the situations which filled them with satisfaction and joy; and until their categories seemed natural and right, and my own rigid and misleading. In fact, I sometimes lost myself in the way of the life of the society I was studying until my children complained that, after all, they were not Trobrianders or Tikopia! What I had to say seemed so obvious to me that, at this point, I could bring myself to write at all only by reminding myself that I was writing for people who were ignorant of this particular culture.That’s Dorothy Lee, from the Introduction to Freedom and Culture (Prentice-Hall 1959) which I recommend to you for a quick overview of some of the more positive possibilities for human culture. (She omits the totalitarian, oppressive, and suppressive cultures, of which there are also many, because her interest is, as I take it, in fostering those forms of culture that are good for us, culture as support and guide. The Jules Henry book that I mentioned earlier is a critique of American culture as adversary to the autonomy of the individual.)

I have not, however, found that any two people agree

as to exactly what “the social arrangement” is. Each person acts so as to

maintain what he or she thinks everyone else has agreed to or expects,

according to the way that individual perceives the social arrangement. This

alone is enough to prove that there is no objective social arrangement,

but only a sort of vector made of a lot of individual perceptions and

reference signals which lead in a variety of directions.

Well, of course, this convergence is how it comes about and is sustained. Notice how people get in line (or “on line” as they say in New York). Then spend some time in Israel. You’ll wait a very long time. When you have learned how to get on a bus there, catch a plane to London and see how the same means work out. There is a stable convergence in each place.

Culture at its best is a support and a guide to conduct for the individual. Cultural change is slow. The rate and complexity of technological change have famously speeded up, and of contact between people of different backgrounds. Culture in disruption and reorganization is a poor support and guide to conduct to the individual. More improvisation is required - improvisation upon existing cultural themes over gaps that have opened in actual conduct. Improvisation doesn’t suit everyone equally well. Some demand that there be one right way and that any way of bridging such gaps different from this right way (which one has of course made one’s own) is wrong.

This is why social

studies are always, and must be, population studies, which come up with

statements that are true of populations, but hardly ever of individuals.

I agree with the observation, but the causal conclusion is perhaps warranted only for populations with fragmented cultures in process of reorganization. It is a remark upon what is extant today but perhaps not upon what is possible or what has been ordinary for humans throughout most of our history.

Take another look at the paper on statistics I wrote for Wayne

Herschberger’s book, Volitional Action.

Unfortunately, I have never had a copy of the book or of either of your papers in it. I guess you mean the “Quantitative measurement” paper and not the “semi-scientific essay”. As I understand your description here, it is possible to get a statistically significant positive correlation which is true of a population while being false for each and every individual in the population. Several questions arise. How probable is this? Is there a higher degree of correlation factor such that at least one individual in the population must show a positive correlation, two if higher yet, and so on? Can the measure of statistical significance be reformulated in terms of the number of individuals who must match the correlation and would this be a useful measure? I know you’re not out to reform the field of statistics, but perhaps someone would find it an intriguing challenge. But even if that is never done, and accepting that all population studies are suspect and must be evaluated with care, it does not follow that generalizations never apply accurately to individuals.

So, the first question again: how do cultural controlled variables become established as controlled perceptions in an individual? How does an infant learn to control the perceptions of its process of urinating in culturally acceptable times, places, and ways? I quoted Lee’s description of how the process begins with a mother and baby in China. Perhaps you remember, or can be reminded, of how it went with your children. What was controversial about Ben Spock was his “permissiveness”. In a culture that honors autonomy of the individual more than ours does, one cannot give what the child already has. In a Chinese household a child may seem to us pretty much to rule the roost. Spoiled, we might say.

Then the second question: how do culturally normal reference values for these perceptions get established. You paint a picture of conflict, of compromise to avoid conflict, of generalization to be “friendly” to those who control those CVs at those values and to be “hostile” to those who don’t. But that this picture is a projection of expectations that themselves are American cultural values seems hardly possible to you.

I don’t think that cultural controlled variables are learned separately from the culturally normative reference values for them. Albert Schweizer spoke of the three ways to teach a child. To paraphrase, there are three ways to establish cultural controlled variables and their reference values: by example, by example, and by example. If one person’s example conflicts with that of another it adds to our knowledge of kinds of people, their roles, their personae. If precept and example conflict we consider the source of each. Unless some innate mechanisms are defective (as may be with autism) well before adulthood we have learned to control a repertoire of culturally recognized roles and personae as ways of being “ourselves”. In any example we perceive both the controlled variable and the values at which it is controlled simultaneously; we perceive them as a way to conduct oneself as that kind of person in that context. The selection, then, is what kind of person we are to enact when “the same” context arises for us. And that context is, above all, what kinds of people, in what roles, presenting what personae, whether they are physically present or represented by obligations and expectations that we perceive between them and other factors in the environment.

···

At 12:41 PM 3/9/2003, Bill Powers wrote:

3

[From Fred Nickols (2003.03.12.1255 ET)] --

[From Bruce Nevin (2003.03.12 11:14 EST)]

I've changed the Subject field since the "Re: How culture colors the way
the mind works" thread has gone its own way.

Here's a thought: One re-sequencing of the words in the new subject line
yields "study of PCT culture." I wonder what such a study might yield.

Just a thought.

Fred Nickols
nickols@safe-t.net
www.nickols.us

4

[From Bill Powers (2003.03.12.1041 MSST)]

Bruce Nevin (2003.03.12 11:14 EST)--

In case I haven't said it before, I truly admire the depth of your
knowledge and thinking about matters of language and society. When you go
beyond "Lite mode" arguments (and even sometimes when you don't), your
writings are a pleasure to read and think about. I hope that my objections
and alternatives have not hidden my basic respect for your capacities as a
thinker and analyist. I can only hope that my comments don't come across to
you as a strain on your patience.

Just a couple of remarks for now. You have said twice now that it would be
difficult to test for controlled social variables:

>We want to identify, and then disturb, and measure resistance to the
disturbance of, a >social artifact, but we are arrested in our progress at
the very first step: the >hypothesis that posits the controlled variable.

I can see that this would pose a problem if the social artifact had any
independent existence of its own: just where on the social CV would you
push to see if there is resistance? If you try to push on the distance
between two Arabs, your hand will encounter only thin air.

It is far easier to imagine disturbing a variable that _a person_ is
controlling, and this is how I would proceed in investigating social
reference levels. I would say to my Jewish friend, "Want half of my BLT?"
This, plus appropriate follow-up interactions, would tell me whether my
friend perceived being Jewish as requiring the avoidance of things like
bacon. By such means I could determine, within reasonable bounds, not only
how my friend perceives Jewishness, but what reference conditions he
maintains and what actions he finds permissible to carry out, at least with
a friend, to defend those variables against disturbances.

What I am describing is close to, but not the same as, what Lee says in the
cited segment. It helps a great deal to be able to perceive something like
what the other person perceives, in order to propose specific Tests, but
I'm not convinced that total immersion is essential. Maybe it is; I haven't
tried it.

If we put that aside, I think that part of my thesis would be upheld by
Dorothy Lee's experiences. I am, for all practical purposes, as totally
immersed in my society as possible for me, yet I don't think I have ever
met any person who could be considered representative of everyone else but
me. Furthermore, I am quite sure that there is no such person. I have
formed some generalizations about others in my society, but it has always
been quite clear to me that they did not apply to everyone, and in fact the
older I get, the less general any of my generalizations seem. If Dorothy
Lee immersed herself in a foriegn society and came to know it as a native,
then she, too, would have formed generalizations about the other people,
and they would have been no more true about everyone in that society than
my generalizations about my own society are. Well, they probably would have
been more true than mine from the standpoint of being consistent and
systematic, but that would not make them any less statistical in nature.
Generalizations, as understood in the social sciences, by their very nature
are not true of everyone.

You haven't yet budged me from my position that social variables exist only
in individuals; I now add, ... and can be detected only through
interactions with individuals. I will include linguistic phenomena in that
assertion. The way we find out about cultures is through interacting with
individuals and then forming generalizations -- which are, of course,
perceptions in ourselves of the same type as the social perceptions we
investigate. Generalizations which refer to all individuals in a culture
are _prima facie_ false. Generalizations which refer to a culture treated
as a population are, if correctly drawn, true but not necessarily
applicable to any individual in that culture ("The Achumawi male is 5 feet
5.173 inches tall"). Those that are true of individuals are true only of
_specific_ individuals -- the subset of those tested who behaved exactly
according to hypothesis. And as I'm sure you will agree, that subset is
never as large as the set of individuals tested. Sometimes it is not even
half as large as the tested set (for example, it could be a a little over a
third of the population if three alternatives were tested).

Best,

Bill P.

5

[From Bruce Nevin (2003.03.12 17:28 EST)]

Bill Powers (2003.03.12.1041 MSST)–

this is how I would proceed in investigating
social

reference levels. I would say to my Jewish friend, “Want half of my
BLT?”

This, plus appropriate follow-up interactions, would tell me whether
my

friend perceived being Jewish as requiring the avoidance of things
like

bacon. By such means I could determine, within reasonable bounds, not
only

how my friend perceives Jewishness, but what reference conditions he

maintains and what actions he finds permissible to carry out, at least
with

a friend, to defend those variables against
disturbances.

How do you know that offering to share your sandwich should or could be a
disturbance to a perception that you call Jewishness?
You’re presuming a lot of prior knowledge - cultural knowledge. Having
attained knowledge of some controlled cultural variables, you can disturb
one and find out something about the other person’s reference value for
it. The question was how is such knowledge attained in the first place.
If you disturb this cultural variable, you do so by disturbing some
lower-level variables in your shared physical environment (real or
imagined) - a sandwich, meat from a pig in the sandwich, processes of
eating, etc. Your change of a variable in the physical environment is
significant only because his control of it is means for controlling some
higher-level variables that are not present in the physical environment -
what you called Jewishness. (He might call it keeping kosher or something
else, but what you call it is immaterial, the control of it is not by
means of words.)

Cultural variables are not generalizations about people. They are
perceptions that people control. They are not perceptions of physical
phenomena (so “being 5 feet 5.173 inches tall” has no
relevance).

If you walk up to me and bend your elbow so that your right hand with
fingers extended is directed more or less toward my right arm, it would
be quite ordinary for me to do likewise and for each of us to grasp the
other’s hand, and it would be somewhat unusual for me not to, that is,
for me not to would ordinarily call for some explanation (is he holding
something, does he not see me, is he angry with me, etc.). Culturally,
this is “shaking hands”. Not all peoples do this. Not all
Americans do it in just the same way, notably African Americans. Even in
the Euro-American version there are differences in detail, such as
grasping the upper arm with the other hand, clasping the other hand over
as well, and so on, all of which have some nonverbal communicative
significance. There are variables involved here that are strictly in the
shared physical environment, but that which is cultural here is not in
the physical environment.

The generalization is not that every Euro-American makes the reciprocal
gesture on perceiving you walking up and making that gesture, but that
doing so (or not doing so) has certain significance about that person and
about your relationship, and such significances are themselves
perceptions that you and he will continue to control in your further
interactions.

When we gesture “no” we most often do so by rotating the head
about a vertical axis so that the nose moves side to side. When a Greek
gestures “no” they do so by rotating the head slightly back on
a horizontal axis side to side through the occiput, sharply lifting the
chin, perhaps lifting the eyebrows a bit, and making an ingressive
apico-alveolar click (what we rather unrepresentatively write as
“tsk!”). A similar, but slower, gesture prevails among the
Turks. Between Rome and Naples, I am told, are a number of small enclaves
such that as you travel down you pass from communities with one style of
gesturing “no” and communities with the other. Those with the
Greek style are in locations where Greek colonies were established
2500-3000 years ago. In India at least some of the (many) peoples there
waggle their head gently about a horizontal axis fore and aft through the
nose, one ear up and then the other, usually smiling the while, and I
think (not sure) this is an affirmative gesture.

Until now perhaps you took the business of shaking your head and nodding
your head pretty much for granted. Or maybe you’re shaking your head. Or
maybe nodding it.

    /Bruce

Nevin

···

At 02:08 PM 3/12/2003, Bill Powers wrote:

6

[From Bill Powers (2003.03.12.1836 MST)]

Bruce Nevin (2003.03.12 17:28 EST)--

How do you know that offering to share your sandwich should or could be a
disturbance to a perception that you call Jewishness? If you disturb this
cultural variable, you do so by disturbing some lower-level variables in
your shared physical environment (real or imagined) - a sandwich, meat
from a pig in the sandwich, processes of eating, etc. Your change of a
variable in the physical environment is significant only because his
control of it is means for controlling some higher-level variables that
are not present in the physical environment - what you called Jewishness.
(He might call it keeping kosher or something else, but what you call it
is immaterial, the control of it is not by means of words.)

But all of the variables we control at higher levels are constructed on a
world of lower-level variables, some of which we can correlate with
environmental variables. We can disturb any higher-level variable by
arbitrarily altering one of the lower-order variables of which it is a
function (assuming that we know how an individual has created such
constructions). In my understanding of HPCT, the only way to disturb any
perception higher than first order is to cause a lower-level perception, or
the proper combination of such perceptions, to change.

Could it be that you're thinking of a social or cultural variable as
something the higher levels of the brain can detect directly, bypassing all
the lower levels? I really don't believe that's what you think, but perhaps
my mentioning this might jog something.

Cultural variables are not generalizations about people. They are
perceptions that people control.

To identify a cultural variable that more than one person controls -- to
say "people" control variables of that kind -- is to generalize in the
statistical sense. I am speaking of the observer's generalizations, not
those of the person being observed. However, now that you mention it, any
person who perceives and controls a cultural variable, I would maintain,
does so on the basis of generalizing from observations of and interactions
with many individuals.

They are not perceptions of physical phenomena (so "being 5 feet 5.173
inches tall" has no relevance)

No, but saying "Jews keep Kosher" is a generalization of the same kind (not
quite as restrictive); it is an average over all people who are
identifiable as Jews but who most certainly do not _all_ keep Kosher.

>If you walk up to me and bend your elbow so that your right hand with
fingers extended >is directed more or less toward my right arm, it would be
quite ordinary for me to do >likewise and for each of us to grasp the
other's hand, and it would be somewhat unusual >for me not to, that is, for
me not to would ordinarily call for some explanation (is he >holding
something, does he not see me, is he angry with me, etc.). Culturally, this
is >"shaking hands". Not all peoples do this. Not all Americans do it in
just the same way, notably African Americans. Even in the Euro-American
version there are differences in detail, such as grasping the upper arm
with the other hand, clasping the other hand over as well, and so on, all
of which have some nonverbal communicative significance. There are
variables involved here that are strictly in the shared physical
environment, but that which is cultural here is not in the physical
environment.

Nope, can't let you get away with that. Americans who do not customarily
shake hands are still Americans even if they are women. And this cultural
trait is not observable in everyone, for the simple reason that not
everyone does it. Your argument that "that which is cultural is not in the
physical environment" doesn't cut it: nothing we perceive is in the
physical environment, shared or not. Every perception is an abstraction and
a generalization carried out by human perceptual functions. Every
perception is a function of lower-level perceptions and can be disturbed by
altering lower-level perceptions in the right way.

>Until now perhaps you took the business of shaking your head and nodding
your head >pretty much for granted. Or maybe you're shaking your head. Or
maybe nodding it.

I was aware of such differences. The concept of "no" is certainly not a
part of the shared environment, but I can easily disturb your expectation
of a "no" meaning by shaking my head from side to side when you expect me
to nod (or whatever the yes gesture is).If I have arranged the experiment
properly, you will then do or say something intended to change my signal
for "yes" into the expected signal for "no", not because you want to see
that in the shared environment, but because you want to perceive that I
agree with you -- a social perception that is derived, in part, from my
signals. It is also derived in part from what I do, so if you ask me to
pass the salt, and I nod (signifying "no") and immediately pass it to you,
you will get conflicting signals and your perception of my degree of
acquiescence will lessen. From such interactions we become able to deduce
what other people mean by their words and gestures.

I agree with you that cultural perceptions are not part of the shared
environment. But like all other higher-level perceptions, they are
functions of lower-level perceptions and ultimately of environmental
variables. Everything you are saying about cultural perceptions is also
true of relationships, events, and configurations. They do not exist
outside us, either, even though our perceptions of them stem from variables
in the environment that interact with our sensory receptors. Any
higher-order perception can be disturbed by altering the right combination
of lower-order variables. So it's perfectly possible to do the Test with
controlled cultural perceptions -- one person at a time.

Your turn.

Best,

Bill P.

7

[From Bruce Nevin (2003.03.12 17:28 EST)]

Bill Powers (2003.03.12.1836 MST)–

Bruce Nevin (2003.03.12 17:28 EST)–

all of the variables we control at higher levels are constructed on a

world of lower-level variables, some of which we can correlate with

environmental variables. We can disturb any higher-level variable by

arbitrarily altering one of the lower-order variables of which it is a

function (assuming that we know how an individual has created such

constructions). In my understanding of HPCT, the only way to disturb any

perception higher than first order is to cause a lower-level perception, or

the proper combination of such perceptions, to change.

I agree. But the key phrase here is “assuming that we know how an individual has created such constructions”. We cannot just assume such knowledge. How do we obtain it? The observation is that people “have created such constructions” in a convergent way. How have they done so? These are the questions.

Could it be that you’re thinking of a social or cultural variable as

something the higher levels of the brain can detect directly, bypassing all

the lower levels? I really don’t believe that’s what you think, but perhaps

my mentioning this might jog something.

Of course not. I don’t think I’ve said anything to suggest that. Certainly didn’t intend to.

Cultural variables are not generalizations about people. They are

perceptions that people control.

To identify a cultural variable that more than one person controls – to

say “people” control variables of that kind – is to generalize in the

statistical sense.

  1. Observing a conversation, Fred recognizes that A has said “no”. The other participant in the conversation, B, also recognizes that A has said “no”. A intended to say “no”. All three people recognize a slight, quick hand gesture that is used by the Watusi people to signify “no”. All three people are controlling their separate perceptions of the same controlled variable. I’m standing there and I ask Fred “Why is B walking away with that chicken?” Fred says “A told him he didn’t want it.” Then Fred has to explain the gesture to me, which I didn’t notice among the various actions that I could see - that is, I didn’t single it out as a single discrete element. I needed to learn to control that perception as means of controlling a perception of its communicative significance. (I’m making up the attribution to the Watusi people.)

I am speaking of the observer’s generalizations, not

those of the person being observed. However, now that you mention it, any

person who perceives and controls a cultural variable, I would maintain,

does so on the basis of generalizing from observations of and interactions

with many individuals.

There are many redundancies in communication. As a child, B perceived that hand gesture together with words of denial and together with actions of denial so that its significance of denial became evident by example and possibly also by precept and is now evident to him even when the gesture stands alone without redundancy. Not much statistical generalization is required here. But there is much evidence in fact for the efficacy and actuality of statistical learning in humans and in primates. (I can send you Lila Gleitman’s chapter of my book, and her student Elissa Newport http://www.bcs.rochester.edu/people/newport/newport.html has done a lot of very interesting work in this area.)

They are not perceptions of physical phenomena (so "being 5 feet 5.173

inches tall" has no relevance)

No, but saying “Jews keep Kosher” is a generalization of the same kind (not

quite as restrictive); it is an average over all people who are

identifiable as Jews but who most certainly do not all keep Kosher.

Whew. Fast and loose here. First, I said no such thing. I asked “how do you know your offer of a sandwich will be a disturbance?” And then the perception that this phrase refers to is hardly an average. A stereotype, maybe?

Nope, can’t let you get away with that. Americans who do not customarily

shake hands are still Americans even if they are women. And this cultural

trait is not observable in everyone, for the simple reason that not

everyone does it.

I make no claim that all Americans do this, I mentioned one sort of exception, and specifically the convention to shake hands with a woman only if she offers to first is yet another cultural convention (or aspect of this one). I claim only that this is a convention recognized as such by any American when it occurs just as any Watusi recognizes that hand gesture of denial. How a given person controls this perception on a given occasion or even always does not vitiate that point.

Your argument that "that which is cultural is not in the

physical environment" doesn’t cut it: nothing we perceive is in the

physical environment, shared or not. Every perception is an abstraction and

a generalization carried out by human perceptual functions. Every

perception is a function of lower-level perceptions and can be disturbed by

altering lower-level perceptions in the right way.

You are being obtuse, Bill, and I think you know it. For the perception of someone else’s hand we reasonably presume that there is a “real” object in the physical environment that you share with that person, an object that could be investigated microscopically, surgically, chemically, etc. For the perception of a handshake it’s more complicated (a series of coordinated actions by two people involving their arms and hands, etc.) but still reasonably attributed directly to phenomena in the shared physical environment. The complication is because it involves control actions by two people, but those actions can still be regarded as physical phenomena without considering the control loops in each person. For the perception of an agreement (a perception which each of the people controls by controlling the handshake, in turn a perception which each of the people controls by means of their arms and hands) - for the perception of an agreement we have only something going on inside the head of person A and something going on inside the head of person B, and possibly going on inside the heads of witnesses to the handshake (a perception which they also control as means of controlling a perception of an agreement). The agreement cannot reasonably be attributed to any phenomenon that is present in the physical environment that is shared by A and B and any witnesses. Similarly for a denial, or the perception that a deer scapula is a hide scraper, or the perception that a check is means for purchasing things, such as a DVD player or a bag of groceries. The distinction corresponds, I think, to Korzybski’s levels of abstraction. And you agree with this distinction immediately below, regarding denial.

Until now perhaps you took the business of shaking your head and nodding

your head pretty much for granted. Or maybe you’re shaking your head. Or

maybe nodding it.

I was aware of such differences. The concept of “no” is certainly not a

part of the shared environment, but I can easily disturb your expectation

of a “no” meaning by shaking my head from side to side when you expect me

to nod (or whatever the yes gesture is). If I have arranged the experiment

properly, you will then do or say something intended to change my signal

for “yes” into the expected signal for “no”,

vice versa, you mean?

···

At 09:26 PM 3/12/2003, Bill Powers wrote:

not because you want to see

that in the shared environment, but because you want to perceive that I

agree with you – a social perception that is derived, in part, from my

signals. It is also derived in part from what I do, so if you ask me to

pass the salt, and I nod (signifying “no”) and immediately pass it to you,

you will get conflicting signals and your perception of my degree of

acquiescence will lessen. From such interactions we become able to deduce

what other people mean by their words and gestures.

I agree with you that cultural perceptions are not part of the shared

environment. But like all other higher-level perceptions, they are

functions of lower-level perceptions and ultimately of environmental

variables. Everything you are saying about cultural perceptions is also

true of relationships, events, and configurations. They do not exist

outside us, either, even though our perceptions of them stem from variables

in the environment that interact with our sensory receptors. Any

higher-order perception can be disturbed by altering the right combination

of lower-order variables. So it’s perfectly possible to do the Test with

controlled cultural perceptions – one person at a time.

Your turn.

Best,

Bill P.

8

[From Bruce Nevin (2003.03.12 23:53 EST)]

Bill Powers (2003.03.12.1836 MST)--

I agree with you that cultural perceptions are not part of the shared
environment. But like all other higher-level perceptions, they are
functions of lower-level perceptions and ultimately of environmental
variables. Everything you are saying about cultural perceptions is also
true of relationships, events, and configurations. They do not exist
outside us, either, even though our perceptions of them stem from variables
in the environment that interact with our sensory receptors. Any
higher-order perception can be disturbed by altering the right combination
of lower-order variables.

I'm glad that we're in complete agreement this far.

So it's perfectly possible to do the Test with
controlled cultural perceptions -- one person at a time.

Sure. But first we have to hypothesize what the controlled variable is. To do that, we have to have some knowledge of the culture. Otherwise, we risk simply projecting the perceptions proper to our own culture onto the lower-level variables. This happens all the time. It's called ethnocentrism. A familiar analogy in language is the Japanese hearing "his laughter" when you say "his rafter". This is the reason for the methodology of participant observation. Do you have an alternative methodology?

all of the variables we control at higher levels are constructed on a
world of lower-level variables, some of which we can correlate with
environmental variables. We can disturb any higher-level variable by
arbitrarily altering one of the lower-order variables of which it is a
function (assuming that we know how an individual has created such
constructions). In my understanding of HPCT, the only way to disturb any
perception higher than first order is to cause a lower-level perception, or
the proper combination of such perceptions, to change.

The key phrase here is "assuming that we know how an individual has created such constructions". We cannot just assume such knowledge. How do we obtain it? The observation is that people "have created such constructions" in a convergent way. How have they done so? Martin Taylor had some suggestions. I don't know if he succeeded in implementing his recognizer. There was a fair amount of hand-waving at that stage it was at in 1993. The principle of convergence that he articulated I think is correct. A model demonstrating it is needed. The Crowd or Gather program is not such a model because the rings and arcs there are not and do not become controlled perceptions.

         /Bruce Nevin

···

At 09:26 PM 3/12/2003, Bill Powers wrote:

9

[From Rick Marken (2003.03.13.2215)]

Bruce Nevin (2003.03.12 23:53 EST)

>So it's perfectly possible to do the Test with
>controlled cultural perceptions -- one person at a time.

Sure. But first we have to hypothesize what the controlled variable is. To
do that, we have to have some knowledge of the culture. Otherwise, we risk
simply projecting the perceptions proper to our own culture onto the
lower-level variables.

Even if we do initially selected hypotheses about controlled variables from our
own controlled perceptions, the test will quickly reveal that they are wrong,
if they are wrong. The test isn't a one shot affair. You test, changing you
hypotheses about the variable controlled, until you have identified a variable
that is protected from all variations that would affect the state of that
variable if it were not controlled.

Knowing a culture can certainly be a good source of initial hypotheses about
what variables a presumed member of that culture might be controlling. But like
the example of the Jew and the BLT, it might turn out that your hypothesis,
based on your knowledge of culture, is wrong. I, for example, who was raised in
a Jewish culture, would accept Bill's offer of a BLT in an instant (especially
if he adds some sliced avocado). And how about a lobster for dinner!

This happens all the time. It's called ethnocentrism.

The test for controlled variables is ethnocentrism-proof, of done properly.

A familiar analogy in language is the Japanese hearing "his
laughter" when you say "his rafter". This is the reason for the methodology
of participant observation. Do you have an alternative methodology?

I suggest the test for the controlled variable. My wife used it with a Japanese
friend of ours to determine that she was warning us against eating oleander
rather than Chinese food. At first my wife thought our Fried Michiko was
telling her not to eat "oriental". But we ate chinese the next night so we knew
that couldn't be what she was controlling for (or Michiko would have tried to
protect Linda and me from the Pork Lo Mein). When Linda and Michiko were
walking in our garden a few days later Michiko saw an oleander plant, pointed
at the leaf and cried out "Oriental, Linda. Don't eat oriental". Voila. The
controlled variable: poisonous oleander.

Best regards

Rick

···

--
Richard S. Marken
MindReadings.com
marken@mindreadings.com
310 474-0313

10

From [ Marc Abrams (2003.03.13.0808)

···

At 05:28 PM 3/12/2003 -0500, you wrote:

[From Bruce Nevin (2003.03.12 17:28 EST)]

Bill Powers (2003.03.12.1041 MSST)--
At 02:08 PM 3/12/2003, Bill Powers wrote:

Bruce, are you trying to see if there are "intrinsic" social variables,
and/or how cultural variables or behaviors develop over time?

Marc

11

[From Rick Marken (2003.02.13.0840)]

Bruce Nevin (2003.03.13 10:33 EST)

for now I'd just like to figure
out how to model the convergence of control systems on higher-level
variables. Communication or signalling between autonomous agents would have
to be part of the model as means of controlling some variables that the
autonomous agents cannot control (or cannot control as well) independently
of one another.

I think this capability is already part of PCT. It's called _reorganization_. I
think people learn to control for "cultural" perceptual variables in the same way
they learn to control all other perceptual variables; though the E. coli like
control process of reorganization, which is aimed at keeping intrinsic error close
to zero. Bill Powers (2003.03.13.0719 MST) alludes to this process in his recent
post when he points out that:

The child learns the
dos and don'ts of its culture by (intentionally or otherwise) disturbing
the adults and other children and experiencing their efforts to correct the
errors they feel, even if the adults can't articulate and are only vaguely
aware of why these errors feel like errors.

The kid being reared in a Kosher family will find life a lot more pleasant if he
gets used to eating his matzo brie without a side of bacon. People "converge" on
cultural conventions by doing things that, in their family or social group, allow
them to feel more knachus than tsuris.

I think the idea that culture shapes what we control is an illusion similar to the
illusion that reinforcement shapes behavior. In both cases, the agent's main
concern is to control intrinsic variables (like blood sugar level, etc). The world
in which the finds himself constrains _how_ he can do this. The child raised in a
Kosher homes learns to do this by eating only food that is Parve. The rat in the
Skinner box learns to do this by pressing a lever. The agent chooses to control
certain perceptions in order to keep intrinsic error at zero. The environment or
culture doesn't choose which variables the person will control (although members
of a culture often _try_ to get kids to control certain perceptions if the kid is
being recalcitrant), the agent choose them.

When the agent learning to control cultural variables is a child, the choice of
which variables to control is probably made unconsciously; the child ends up
choosing to control perceptions that result in their experiencing the least amount
of tsuris from a parent. But I have had the rare opportunity to watch an adult (my
Mom) consciously choose to control for a particular set of cultural perceptions in
order to be part of that culture.

My mother was raised in a completely secular Jewish household. She married a
completely secular Jewish man (my Dad) and I was raised knowing little more about
Jewish culture than a few great Yiddish expressions (most of which are sprinkled
through this post). My Dad turned out to be kind of a Lothario so my mother
divorced him and decided that Dad was such a jerk because he had no religion (in
fact, my Dad found religion toward the end of his life but remained a jerk, though
a new kind of jerk). So my Mom decided to become Jewish. She started controlling
for things that she had never controlled for when I was a kid. For example, when I
was a kid bacon was mandatory at breakfast (a source of iron, she said). Now it
was forbidden. She married an Orthodox man (who has been a great husband for her)
and tries to keep Kosher , but only to the extent that it doesn't get in the way
of her other life style goals. So I have watched my mother consciously start
controlling for variables that would let her see herself as more "Jewish". It was
clearly not the culture that made her do this; it was an attempt to overcome
heartbreak (intrinsic error).

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

12

[From Bill Powers (2003.03.13.0719 MST)]

Bruce Nevin (2003.03.12 23:53 EST)--

>I'm glad that we're in complete agreement this far.

That was a surprise! I had thought you were saying that social perceptions
were an exception to the way other perceptions work, so I expected some
part of what I said to constitute a disturbance for you, which you would
correct. From the reply you did make, I now think that the problem is
completely different from what I thought it was. As you can see, I am Testing.

So it's perfectly possible to do the Test with
controlled cultural perceptions -- one person at a time.

Sure. But first we have to hypothesize what the controlled variable is. To
do that, we have to have some knowledge of the culture.

There it is! "To do that, we have to have some knowledge of the culture."
While some knowledge of the culture might speed things up a bit, it's not
at all necessary. Think about the E. coli way of approaching a source of
food. What's required is for
E. coli to make some move and then evaluate whether the situation is better
or worse. You might think, "In order to make a move relative to the target
position, E. coli must have some knowledge of the target position." But it
doesn't need that. If the result is worse, it tumbles again right away. If
better, it keeps going in the same direction until things get worse again.
It doesn't even know it's getting closer to the source of the attractant.
All it can know is the local gradient.

In doing the Test for social variables we can operate essentially the same
way. When we interact with another person, sooner or later there will be a
moment when something we start to do in relation to the other meets with
resistance. Alternatively, we will find that while we're simply going about
our business, the other person suddenly says, "Don't do that!" or exhibits
some behaviour to the same effect. For example, back away from the Arab,
who then thwarts the attempt to increase the distance between the two of
you by moving toward you. Actually, you might conclude that the Arab wants
to keep the distance between himself and you small, when in fact the Arab
is controlling for staying "within your breath," as they put it. It's a
smell or a level of humidity that's being controlled for (possibly), not a
distance, but the two variables are strongly linked and thus difficult to
separate by the Test, expecially if you're convinced you've found the real
controlled variable and stop Testing.

Actually it's quite easy to sense the presence of a cultural variable in
interactions with other people, though once you realise that something is
going on, it may take a lot of effort to narrow it down.If nothing you did
disturbed anything the other person was controlling, you would never
experience any opposition from the other person, nor would the other person
gratuitously start trying to make you behave differently..I think this must
be exactly how ethologists go about their business, with an efficiency that
is greatly improved by their sensitivity to disturbances. It's similar to
the total-immersion process any child goes through. The child learns the
dos and don'ts of its culture by (intentionally or otherwise) disturbing
the adults and other children and experiencing their efforts to correct the
errors they feel, even if the adults can't articulate and are only vaguely
aware of why these errors feel like errors.

> ... we risk simply projecting the perceptions proper to our own culture onto
>the lower-level variables. This happens all the time. It's called
ethnocentrism.

Of course. That's what happens when you make a guess about the other
culture and then defend it instead of Testing it. If you're Testing, the
next thing to do is to make a prediction: If this person is really refusing
my sandwich because of a dislike for whole wheat bread as I am guessing,
then if I offer the same sandwich but with white bread, that person will
accept it. So I then, at the next opportunity, offer half of my BLT on
white bread -- only to have it refused, too. Scratch that hypothesis. Maybe
it's the mayo? And so forth, until finally, months later. the bacon is
revealed as the culprit, however unlikely it is that any human being would
not like bacon. A lot more work might be required to arrive at the
principle of keeping Kosher, but if I kept asking "why?" eventually I'd get
there, especially if I were aware of principles as a universal type of
human perception and were looking for a principle.

> A familiar analogy in language is the Japanese hearing "his laughter"
when you say >"his rafter". This is the reason for the methodology of
participant observation. Do you >have an alternative methodology?

Absolutely not. That's what I'm describing when I describe how the Test is
done. You do something and check for resistance, then keep changing what
you do to Test more refined hypotheses, until every time you expect
resistance you get it, and every time you expect no resistance there is
none. You follow up by saying to the Japanese, "He drove a nail into his
rafter." The puzzlement that follows such a nonsense statement (drive a
nail into laughter?) is a sign that something was disturbed that would not
have been disturbed if the Japanese meant what you meant.

The key phrase here is "assuming that we know how an individual has created
such constructions". We cannot just assume such knowledge. How do we obtain
it? The observation is that people "have created such constructions" in a
convergent way. How have they done so?

I think I have to retract my qualifier about knowing how a person has
created a particular perception. The test will reveal that, as closely as
we are ever going to know by any means other than neural mapping. When no
systematic way of producing hypotheses works, we try a variation at random
and observe the results. The key to success is to keep trying new
hypotheses until every prediction of resistance or no resistance is correct.

It's _essential_, for doing the Test, to give the other person a means for
making a correction when you disturb something. In fact, no actual change
of the variable being disturbed should occur: your disturbance should be
small enough for the other person to cancel its effect completely, and slow
enough so no significant error ever develops. As you move the
experimenter's end of the rubber bands around, you want the knot to stay
over the dot. If the knot simply moves as you disturb it, it's not under
control, not in its spatial dimensions. If you seize the knot and move it
off the dot, you can't see whether it's being controlled or not.

Best,

Bill P.

13

[From Bruce Nevin (2003.03.13 10:33 EST)]

Marc Abrams (2003.03.13.0808)--

Bruce, are you trying to see if there are "intrinsic" social variables,
and/or how cultural variables or behaviors develop over time?

Important questions when we get there, but for now I'd just like to figure out how to model the convergence of control systems on higher-level variables. Communication or signalling between autonomous agents would have to be part of the model as means of controlling some variables that the autonomous agents cannot control (or cannot control as well) independently of one another.

The further questions you raise are implicit in this. In evolutionary terms any "signals" and their significance must themselves be a product of convergence such as we seek to model, but in terms of a new individual learning they may pre-exist as a product of such convergence among ancestor individuals, some of whom are currently demonstrating their use for cooperative control. Processes of learning contribute new disturbances to continuing processes of convergence. It is an open question exactly what is intrinsic. I would assume for starters that what is biologically intrinsic is also not specialized for communication, convergence, and cooperation, but some non-biological constraints are in what is informationally required for reliable communication.

         /Bruce Nevin

···

At 08:11 AM 3/13/2003, Marc Abrams wrote:

14

From [ Marc Abrams (2003.03.13.1103) ]

[From Bruce Nevin (2003.03.13 10:33 EST)]

Important questions when we get there, but for now I'd just like to figure
out how to model the convergence of control systems on higher-level
variables.

Interesting. This is one of the things I hope to see in my research. I am
currently of the opinion that some or all of the higher level
variables are actually a "network" (i.e., an interdependent rather then in
a pure hierarchical structure, which indicates a dependency on the level
above _only_ ) I believe I can show this. ( i.e. a network of higher level
variables that might exist on one or more of the higher levels )

You might want to look at the AHP software for help in this regard.
http://www.expertchoice.com/ This company uses a hierarchy only. I have
software that allows the use of a network as well. It's currently being
sold and used as a prioritization tool for decision making but is really
quite useful in helping to structure thinking and comparing variables. A
free demo software is available at the site.

Communication or signalling between autonomous agents would have
to be part of the model as means of controlling some variables that the
autonomous agents cannot control (or cannot control as well) independently
of one another.

This gets to be kind of tricky. :slight_smile: Let me try an example on for size. What
people often say to reduce error ( and fulfill lower level reference
conditions ) are often "understood" differently then what was intended.
This of course increases rather then decreases the error associated with
the higher level goals. An autonomous agent can only control what comes out
of his mouth. He cannot control how someone else interprets it. This of
course could create error in and a disturbance outside the control system.
A common "understanding" ( from Imagination/memory ) becomes _part_ of the
lower level input to an upper level perception. Creating a new variable and
around the loop we go again :-). The input function has to be some kind of
an integration ( a conjecture on my part ). With all the different signals
coming into it I can't see haw a simple addition or multiplication would
suffice.

"Conflict" in HPCT can take on two different states, either between to or
more agents or internally in one. The use of the term in PCT suggests a
common goal ( common reference condition ) and differing perceptions. Since
most reference conditions are defined by the error from the variable above,
It would be quite unique for 2 or more people to have the _same_ exact
reference condition, ever. It seems reasonable to think that "in the
ballpark" would be more accurate to describe how two or more people might
agree on a common reference condition, and I would think you would find
totally different variables effecting and causing that structure in each
individual.

Again, something I hope to see in my research

Marc

···

At 10:33 AM 3/13/2003 -0500, you wrote:

15

[From Bill Powers (2003.03.13.1133 MST)]

Marc Abrams (2003.03.13.1103)--

>The input function has to be some kind of
>an integration ( a conjecture on my part ). With all the different signals
>coming into it I can't see haw a simple addition or multiplication would
>suffice.

Addition (with sign) and multiplication are simply two operations that a
synapse can imitate, A neural network that constitutes an input function
can carry out dozens of such computations in series and in parallel, and so
perform functions that we would describe by polynomials. Include
time-integration, and the computation is expanded to include the behaviors
we represent with differential equations, and just about every other
possible mathematical operation.

The mention of addition and multiplication was intended only to give
examples of the kinds of computations that neurons can do. With such
building blocks, very complex input functions can be constructed -- or
could be, if we only knew what functions to put in them.

In fact, students of perception have very little understanding of the kinds
of computations done even at the third level, perception of configurations.
Perceptrons, single and multilayer, are just barely able to recognize a few
simple configurations. They still can't report a rotating cube as a cube
independently of its orientation, or pick out the angle of rotation
regardless of the appearance of the cube. Transitions, events,
relationships, and so on are still far out of reach in terms of modeling
how they are perceived.

In our models we have to rely on subjective experience to discover what
classes of perceptions exist. We have to assume that if a controlled
variable is found, there is a perceptual signal corresponding to it. As
modelers, we can program a display to look like a figure changing shape,
and we can posit that there is a perceptual signal corresponding to the
momentary orientation of the figure. In a model, when the computer makes
the figure rotate, it also makes a simulated perceptual signal's magnitude
vary with the angle. That's how we bypass the fact that we can't make a
model that will, all by itself, generate a perceptual signal corresponding
to the angle of rotation. We can't just present a picture to the program
and have it recognize the shape it's looking at, the orientation, and the
rates of change.

Subjective experience, comparing experiences with others, and testing to
see what variables people can actually control, are the main tools we have
right now for modeling behavior. That will be true for a very long time,
unless someone comes up with a real breakthrough in artificial perception.

Best,

Bill P.

16

From [ Marc Abrams (2003.03.13.1633) ]

[From Bill Powers (2003.03.13.1133 MST)]

Include
time-integration, and the computation is expanded to include the behaviors
we represent with differential equations, and just about every other
possible mathematical operation.

That was exactly my point.

The mention of addition and multiplication was intended only to give
examples of the kinds of computations that neurons can do. With such
building blocks, very complex input functions can be constructed -- or
could be, if we only knew what functions to put in them.

Yes, I know this.

Transitions, events,
relationships, and so on are still far out of reach in terms of modeling
how they are perceived.

Yes, I know and understand this

In our models we have to rely on subjective experience to discover what
classes of perceptions exist. We have to assume that if a controlled
variable is found, there is a perceptual signal corresponding to it.

Don't know how it could be any other way. :slight_smile:

As
modelers, we can program a display to look like a figure changing shape,
and we can posit that there is a perceptual signal corresponding to the
momentary orientation of the figure. In a model, when the computer makes
the figure rotate, it also makes a simulated perceptual signal's magnitude
vary with the angle. That's how we bypass the fact that we can't make a
model that will, all by itself, generate a perceptual signal corresponding
to the angle of rotation. We can't just present a picture to the program
and have it recognize the shape it's looking at, the orientation, and the
rates of change.

I understand this.

Subjective experience, comparing experiences with others, and testing to
see what variables people can actually control, are the main tools we have
right now for modeling behavior. That will be true for a very long time,
unless someone comes up with a real breakthrough in artificial perception.

I _fully_ understand this. :slight_smile:

Thanks Bill.

Marc

···

At 12:00 PM 3/13/2003 -0700, you wrote: