[Bruce Nevin (2008.02.28.2005 EST)]
Anyone willing to discuss these problems of testing for “social” controlled variables.
/BNRe: Problems testing for “social” CVs
[was: Julian Day for
[From Martin Taylor 2008.02.28.23.00]
[Bruce Nevin
(2008.02.28.2005 EST)]Anyone willing to
discuss these problems of testing for “social” controlled
variables.
/BN
What aare you thinking of when you talk of a social controlled
variable? Are you thinking of the development of conventions such as
the ID stamp or the meanings of words and gestures, or something else.
If you are thinking of cultural conventions, I’m not sure they should
be called “controlled” variables, though they are the social
results of individual control (I talked about how they develop in my
first CSG meeting, Durango 93, which Dag may still have available on
tape).
Maybe you might define the topic of the conversation you are
interested in pursuing.
Martin
[From Bjorn Simonsen (2008.02.29, 12:15 EUST)]
Bruce Nevin (2008.02.28.2005 EST)
Anyone willing to discuss these problems of testing for "social"
controlled variables.
Reading this, I thought as Martin. "What aare you thinking of when you
talk of a social controlled variable?".
I think upon social controlled variables are nearly the same as
individual controlled variables. Or more exact, I don't think there is a
social brain that is able to controll "social" variables. In social
interactions, I think some people control different variables (one
person controls to hit the ball and another person controls to fetch the
ball). Sometimes two or more people agree that they will control the
same variable, and sometimes two or more people agrre to control
different variables.
We may discuss problems of testing "Which controlled variables do we
have when two or more control systems, or agents, as Kent McClelland
call them, seek to control a single variable in a tracking system".
Later we may go up some levels and discuss problems of testing such
variables.
bjorn
Anyone willing to discuss these
problems of testing for “social” controlled
variables.
[From Bill Powers (2008.02.29.0414 MST)]
Bruce
Nevin (2008.02.28.2005 EST) –
Just a general remark or two. Psychologists have developed a habit of
using surreptitious and covert methods of obtaining data about people,
possibly because they have been unable to discover very many real facts
about behavior and fear that any knowledge on the part of subject about
what is being studied will reduce their correlations to zero.
However, if they were to focus on more obvious aspects of behavior, they
might find that it’s really rather easy to find out about controlled
variables. You might have asked, for example, “How important is this
date-time stamp format to you?” I might have replied, “It’s a
bit of a nuisance to type it in by hand, but it’s worth it because I like
seeing who is writing and what is being answered, it seems to have
reduced the number of mis-attributions considerably, and it’s useful for
the purpose of searching for CSG-related posts – but only if everyone on
CSGnet uses it.” I would place its value to me about the same as the
value of spending a hour discussing it on the net once a year.
As to studying social controlled variables, I agree with Martin Taylor’s
observation (echoed by Bjorn Simonsen) to the effect that this can mean
only variables relating to social processes that individuals perceive and
control. It’s useful to know what other people think are important social
issues, because that tells us what they will seek and what disturbances
they will oppose. For example, we get to know which people like to
manipulate other people, so we can avoid them, and which people share
similar (or at least interesting) system concepts so we can spend more
time with them. Of course there is also the objective of simply coming to
understand human behavior better, which interests a lot of us
here.
Best,
Bill P.
/BN
No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.5.516 / Virus Database: 269.21.1/1303 - Release Date:
2/28/2008 12:14 PM
[From Bruce Nevin (2008.02.29.2005 EST) ]
Bill Powers (2008.02.29.0414 MST)–
Just a general remark or two. Psychologists have developed a habit of using surreptitious and covert methods of obtaining data about people, possibly because they have been unable to discover very many real facts about behavior and fear that any knowledge on the part of subject about what is being studied will reduce their correlations to zero.
However, if they were to focus on more obvious aspects of behavior, they might find that it’s really rather easy to find out about controlled variables. […]
Did my use of surreptitious and covert methods annoy you? Supposing it did annoy you, or someone, the feeling of annoyance would indicate that a controlled variable was disturbed – some expectation of appropriate behavior, perhaps. How would you suggest testing for such a variable? The only proposals I have heard here in the past for testing for control of social norms involve violating those norms and watching for resistance to the disturbance. With norms, the resistance may take the form of ignoring the transgression, which is a mild form of shunning. Your rather frosty “you can do as you like” responses indicate your readiness to do this.
(Norms: socially established reference values for variables that are controlled in order to participate in social processes. Perception and control of the variables themselves, as well as the reference values, are established and maintained in the course of participating in social processes.)
Of course we could ask people hypothetically “how would you feel if someone were to do X”. Among other reasons for indirect (“surreptitious and covert”) methods is that direct questions embody presuppositions that prejudice the answers, and more importantly they call attention to the norms that you are testing for, which may be what one is supposed to do, but not what the person actually does. In speech, for example, pronunciations when talking to one’s neighbors differ from pronunciations in more formal settings or when talking to an outsider such as an interviewer for sociolinguistic research. You hear a remarkable shift when the interviewee goes to answer a telephone call from a friend. So instead of asking “how do you pronounce ‘house’,” you ask questions whose answers are likely to include the word “house,” or set up conversations in which the words whose pronunciation concerns you are likely to occur, where the conversants are socially related in pertinent ways.
A more fundamental reason for indirect methods is that norms must be unconsciously controlled in order to be authentic. (I can revisit the basis of this assertion if you’ve forgotten. It has to do with manipulation and reliability.) The interviewer is participating in a social interaction, namely, the interview. By calling attention to the norm, he or she forces control according to the norm to be conscious rather than unconscious.
In the case of the CSG-L tag, direct questions would probably not have disclosed that the timestamp is too unreliable to be of literal value, and that it serves primarily to distinguish two posts on the same day from the same person. Control of “membership” and inside vs. outside had only begun to emerge when I blew the whistle on myself for the sake of not risking my bona fides any farther in this sometimes rather testy group.
Bjorn Simonsen (2008.02.29, 12:15 EUST) –
I think [that] social controlled variables are nearly the same
as individual controlled variables. Or more exact, I don’t
think there is a social brain that is able to control
“social” variables.
I agree with you. But there is more involved.
Bill Powers (2008.02.29.0414 MST)–
As to studying social controlled variables, I agree with Martin Taylor’s observation (echoed by Bjorn Simonsen) to the effect that this can mean only variables relating to social processes that individuals perceive and control.
The more important and more general case is variables which are controlled in order to participate in social processes (such as speaking and understanding, recognizing one’s respective social affiliations, etc.) and whose reference values are set in the course of controlling these variables in order to participate in those social processes.
/Bruce Nevin
Re: Problems testing for “social” CVs [was: Julian Day for
[From Bruce Nevin (2008.02.29.1229 EST)]
From Martin Taylor 2008.02.28.23.00 –
Sorry, Martin, didn’t see this right away, but I think my reply to Bill and Bjorn is responsive to your questions.
/BNRe: Problems testing for “social” CVs
[was: Julian Day for
[From Martin Taylor 2008.02.29.14.09]
[Bruce Nevin
(2008.02.29.1229 EST)]
From Martin Taylor
2008.02.28.23.00 –
Sorry, Martin,
didn’t see this right away, but I think my reply to Bill and Bjorn is
responsive to your questions.
As is appropriate to the date, I leap to the conclusion
that you mean this [Bruce Nevin (2008.02.29.2005
EST) ]:
The more important
and more general case is variables which are controlled in order to
participate in social processes (such as speaking and understanding,
recognizing one’s respective social affiliations, etc.) and whose
reference values are set in the course of controlling these variables
in order to participate in those social processes.
My understanding of this is that social control is not
necessarily involved, though it may be (the difference is roughly
between the evolution of English through mutually influenced
reorganization among the talkers, and the Academie Française
attempts to control the usages of French by declaring some usages
proper and some to be avoided).
For the best part of February, I’ve been composing a message on
that topic, in answer to [Bill Powers (2008.02.04.0820 MST)]. I don’t
know when it will be done; however, if you have the CSG Durango 93
tape and it hasn’t rotted away, you can find my discussion of the
matter. It followed on a CSGnet thread in which Rick had not
understood my point, but he said he did understand it after the talk.
So maybe that’s the best source, if it is available anywhere. I
thought I had a copy, but I can’t find it now.
Martin
Re: Problems testing for “social” CVs [was: Julian Day for
[From Bruce Nevin (2008.02.29.1520 EST)]
Martin Taylor 2008.02.29.14.09
···
–
As is appropriate to the date, I leap to the conclusion that you mean this [Bruce Nevin (2008.02.29.2005 EST) ]:
Yes. Emphasizing the utility of the datestamp, are we? I forgot to update the time when I copied it from an earlier one.
Bruce Nevin (2008.02.29.2005 EST) –
The more important and more general case is variables which are controlled in order to participate in social processes (such as speaking and understanding, recognizing one’s respective social affiliations, etc.) and whose reference values are set in the course of controlling these variables in order to participate in those social processes.
Martin Taylor 2008.02.29.14.09 –
My understanding of this is that social control is not necessarily involved, though it may be (the difference is roughly between the evolution of English through mutually influenced reorganization among the talkers, and the Academie Française attempts to control the usages of French by declaring some usages proper and some to be avoided).
I guess that what you mean by “social control” is overt assertion of preferred reference values for perceptual variables, such as the dictates of the Academie Française. I’m not particularly interested in such prescriptive efforts of people who identify with a dominant social class to enforce the shibboleths by which they guard their borders and denigrate people who by speaking differently advertise their identification with some other class of people.
To focus consciously on norms (reference values) which normally are unconscious, which were unconsciously acquired and maintained in the course of interacting with one’s peers, and by conscious practice to change them, is a very uncomfortable thing for most people. It is artificial. It is not normal. It takes time to gain fluency controlling familiar variables according to the reset norms. You can’t practice with your friends and neighbors (unless some of them are doing it too), and you can’t practice with members of the social class for whom those norms are normal, that is, unconsciously acquired, maintained, and used, because they will detect any slip, any deviation as betraying what kind of person you really are, and (unless they are social change advocates sympathetic to the enterprise) will reject you for trying to “pass”.
For this reason, the efforts of the Academie Française are not serious efforts to reform the way hoi polloi talk, they are a vigorous policing of the boundaries between hoi polloi and the elite. My use of “hoi polloi” here is another example. In the 19th century only the educated knew Greek and Latin well enough to use the phrase comfortably with one another. It means “the many”. You will often find those without the requisite education saying or writing “the hoi polloi”, which is redundant in the way that El Camino Real Road is redundant. Similarly “matchmaking ala Pride and Prejudice,” “rebelling against the social roles of the day, (ala Austen)” where the French phrase with two words has been anglicized into a single English word. I found both those quotes on the web just now.
I remember your 1993 Durango talk (which I’ve only seen on the videotape), and I still agree with it. I think I’m talking about the same thing.
Now to update that timestamp, even though its accuracy to the hour and minute makes no real difference.
/Bruce
[From Bill Powers (2008.02.29.1348 MST)]
Bruce
Nevin (2008.02.29.2005 EST) –
Did my use of
surreptitious and covert methods annoy you? Supposing it did annoy you,
or someone, the feeling of annoyance would indicate that a controlled
variable was disturbed – some expectation of appropriate behavior,
perhaps. How would you suggest testing for such a
variable?
What I decided to object to was finding that you weren’t really
interested in the subject of the date-time stamp, so none of the
arguments against its use and none of the alternatives that you were
offering were genuine – you just wanted to see if I was controlling for
using it by pretending that you didn’t want to use it. Since I took your
comments at face value and didn’t want to urge you to use the convention
against your will, I indicated that I could live without it on your posts
and wouldn’t ignore your posts if you chose not to use it. The
“frostiness” you detected was probably due to the fact that I
replied more mildly than your conscience led you to expect, so you
interpreted this as
“a mild form
of shunning”. So am I to take it that if I had scolded
you for deviating from the norm, or had put pressure on you to conform to
the convention, you would have seen that as a mild form of acceptance? I
think your imagination is perhaps too active here, but then that’s what
makes soap operas, isn’t it?
A more
fundamental reason for indirect methods is that norms must be
unconsciously controlled in order to be authentic. (I can revisit the
basis of this assertion if you’ve forgotten. It has to do with
manipulation and reliability.) The interviewer is participating in a
social interaction, namely, the interview. By calling attention to the
norm, he or she forces control according to the norm to be conscious
rather than unconscious.
“Norms must be unconsciously controlled in order to be
authentic?” An interesting theory which I don’t recall ever having
believed, but I’m more interested in the “unconscious norm”
idea itself. If one adheres unconsciously to a norm, how does it change
when one’s consciousness is directed to it? I can understand how some
reference conditions are not ordinarily in awareness, and how awareness
can be directed through questioning so as to include what was formerly
unaware (why are you reading this?), so what was unconscious becomes
conscious. But why should that change the reference condition? This
touches on the method of levels, because we commonly find that a person
is aware of one side of a conflict but not the other side, until
attention is directed appropriately. Then the conflict is normally seen
and resolved, and there is only one reference condition left.
If calling attention to a norm that seems (reliably) to govern a person’s
actions results in a person’s disavowing it, the implication is that
there is a second norm in conflict with the first one. In MOL sessions
this is often found to be true, and the result in the end is that some
new norm, reference condition, evolves and replaces both of the others
that were in conflict. Then there is only one reference condition in
effect, whether the person continues to be aware of it or not. If you
then call the person’s attention to the reference condition, the person
will simply tell you what it is, having no reason to disavow it.
Therefore the only reason I would expect a person to tell two stories
about a social norm or reference condition (other than strategic reasons
for lying, which are not the issue here) would be that there is an
internal conflict about it. The person wants to adhere to the norm for
one set of good reasons, and also wants to deviate from it for a
different set of good reasons. Commonly, only one side of the conflict is
in awareness.
In the case of
the CSG-L tag, direct questions would probably not have disclosed that
the timestamp is too unreliable to be of literal value, and that it
serves primarily to distinguish two posts on the same day from the same
person.
Are you arguing now about deficiencies of the time stamp, or is this
another disturbance designed to see whether it’s a controlled variable? I
could argue that your statements about what the discussion
“disclosed” are a bit presumptuous or simply flat wrong, but
would I then be arguing against a real point, or just demonstrating
something of interest to you? I don’t know the terms of this discussion,
and now that I know that you have used, and may still be using,
deception, I have no trustworthy way of finding out what they really are.
So I won’t rise to the bait, in case that’ what it is. It’s just another
version of “The boy who cried ‘wolf!’”, part of every child’s
education about why it’s self-defeating to lie.
Oh, what a tangled web we weave, when first we practise to deceive, Sir
Walter Scott said (I just looked it up). Psychologists who use deception
have reaped their just rewards, including jokes which are quite popular:
Psychologist A: “Hi, Mike, how are you?” Psychologist B (to
himself): “What does he mean by that?”
If you want to know what a person’s reference conditions are, ask. If the
person is unsure or in conflict, the least you can then do is listen
while he works it out and resolves the problem. He’s not there just so
you can learn something.
Best,
Bill P.
[From Bruce Nevin (2008.02.29.2105 EST)]
I said (2008.02.29.2005 EST)
Supposing it did annoy you, or someone, the feeling of annoyance would indicate that a controlled variable was disturbed – some expectation of appropriate behavior, perhaps. How would you suggest testing for such a variable?"
You replied in a way that is consistent with your having read this as “I believe this annoyed you. This indicates …” etc. That is not what I said. I could be wrong, but this suggests to me that your annoyance influenced your perception of what I wrote.
The question still stands. How would you suggest testing for such a variable?
Bill Powers (2008.02.29.1348 MST) –
“Norms must be unconsciously controlled in order to be authentic?” An interesting theory which I don’t recall ever having believed, but I’m more interested in the “unconscious norm” idea itself.
Bruce Nevin (2008.02.29.2105 EST) - this message –
I’ll try to clarify what I mean.
There are prescriptive norms that are taught, or their violations taunted, such as “manners” that differ from your everyday unthinking behavior.That’s the French Academy sort of thing. I’m not talking about such norms.
Say “pod” as in “pea-pod”. Then say “Pa’d” as in “Pa’d go if he could.” Listen carefully to the vowel between the p and the d as you repeat these words. Are they the same or different? They are the same in my wife’s speech; she is from the Chicago area. I would expect that for you they are the same as well, since the merger of those two vowels is widespread through dialects in the midwestern US. In my dialect they are different. “Pod” has a vowel midway between the a of “father” and the o of “bode” (as in “it does not bode well”). Similarly, the first vowel in “windy” (as in “the windy city”) and “Wendy” (the girl who helped Peter Pan). The words “pan,” “pen,” and “pin” have distinct vowels in many dialects of English, including what I speak, but in these midwestern US dialects (and some others) the vowels of “pen” and “pin” have merged. These are norms of pronunciation that we acquired unconsciously in the course of using them in conversation with our peers. They are not easy to change after a person establishes his or her identity at puberty.
Such norms as these – for example whether your smile exposes upper teeth alone or both upper and lower – are not prescriptively taught, and deviation from them is not singled out and taunted unless the deviation becomes a shibboleth of us vs. them identification.
Bill Powers (2008.02.29.1348 MST) –
If one adheres unconsciously to a norm, how does it change when one’s consciousness is directed to it? I can understand how some reference conditions are not ordinarily in awareness, and how awareness can be directed through questioning so as to include what was formerly unaware (why are you reading this?), so what was unconscious becomes conscious. But why should that change the reference condition? This touches on the method of levels, because we commonly find that a person is aware of one side of a conflict but not the other side, until attention is directed appropriately. Then the conflict is normally seen and resolved, and there is only one reference condition left.
If calling attention to a norm that seems (reliably) to govern a person’s actions results in a person’s disavowing it, the implication is that there is a second norm in conflict with the first one. In MOL sessions this is often found to be true, and the result in the end is that some new norm, reference condition, evolves and replaces both of the others that were in conflict. Then there is only one reference condition in effect, whether the person continues to be aware of it or not. If you then call the person’s attention to the reference condition, the person will simply tell you what it is, having no reason to disavow it. Therefore the only reason I would expect a person to tell two stories about a social norm or reference condition (other than strategic reasons for lying, which are not the issue here) would be that there is an internal conflict about it. The person wants to adhere to the norm for one set of good reasons, and also wants to deviate from it for a different set of good reasons. Commonly, only one side of the conflict is in awareness.
Bruce Nevin (2008.02.29.2105 EST) - this message –
These are useful questions. I’m not quite sure how to engage them with the matter at hand. How does such a norm change when called to conscious attention? It depends.
The cultural and social norms of the sort that I am discussing are controlled in more than one way simultaneously. There is no conflict between the two systems controlling it because one of them controls with lower gain and the other with higher gain. (But see the note following.) To use a word for communication it must only be recognizably distinct from other words that are similar to it. The pronunciation can vary over a considerable range, so long as the pronunciations of similar (“neighboring”) words co-varies so as to maintain their mutual distinctness. Take the word “house”. The sound between the h and the s sounds is a diphthong, two vowel sounds run together. The first part of the diphthong sounds like the vowel of “hut” in dialects common in Canada and the old-timers’ dialect where I live. In other dialects it sounds like the a in “father,” though shorter. In parts of Philadelphia and in many dialects through the south, it sounds like the a in “hat” or even the a in “late”.
Note: it may be that one control system is controlling distinctness of words from one another, with gain varying depending upon how predictable the word is from context, and the other is controlling the “target pronunciation” of the word in conformity with the norms of the speaker’s dialect.
Most people command more than one dialect. The differences between everyday talk with peers and formal talk with teachers, interviewers, and the like, are most marked in people who are not members of the most privileged class of people, and least evident in the latter, who often regard their dialect as the only proper way to speak (and similarly for other cultural norms, hence “manners,” etc. ). If you call an “everyday talk” norm to a person’s attention, certainly while in a formal setting but often more generally, they often in fact do disavow it, and are careful to demonstrate their command of the “official” norms. This is fundamentally a question of identity, who you are, what kind of person you are, who your people are, where you come from, where you belong, and the exceptions to disavowing “homeboy” talk are generally (and frequently belligerent) assertions of “homeboy” identity being of value, and you can take your fancy-pants presumption and stuff it.
It’s not immediately obvious how to engage the method of levels with the norms of pronunciation (or of smiling, or other body language, etc.). It might be possible to use MOL to get from pronunciations etc. to the control of identity and group affiliation. There very well might be a conflict there, but the resolution typically is a higher level of control switching from one dialect to another as appropriate, without feeling that one’s identity is at risk by doing so.
However that may be in a given individual, control of pronunciation (smiling style, body language, etc.) is normally and almost always out of awareness because you are focussing attention on other variables (meanings, quality of relationship, progress of social transaction, etc.) for which these are the well habituated means. Focusing attention on such details can interfere with performance, somewhat like paying attention to the details of balancing a bicycle, or trying to talk when there’s an echo on the line.
The norms for the lower-level control (the manner of doing the higher-level control) are unconsciously acquired in childhood. Martin’s 1993 presentation speaks to this. The effect is that the manner of doing the higher-level things (pronouncing words, smiling, etc.) comes into conformity across the members of a group of people. There seem to be a number of factors determining what constitutes such a group, and there almost always are several such groups living contiguously or overlapping in the same community. At puberty, children cement their identification with one group or another of those available in the community. (This has been well demonstrated by work done by Bill Labov and his students. I have reported previously on his seminal research into dialect change on Martha’s Vineyard, where I live, in his Master’s thesis, published in I think 1963 as “The Social Motivation of a Sound Change.” I think I’ve mentioned their work with Black English Vernacular.)
Consciously paying attention to these variables and trying to change them, as in Shaw’s “Pygmalion” or the musical “My Fair Lady” based on it, amounts to rejecting those whose manner you are changing from, and not being accepted, or at least being suspected, by those whose manner you are changing to. In the black community, it is (or used to be) called “passing”, i.e. passing for white, if the person’s skin is light enough, and is often derogated by terms like Uncle Tom, Oreo (black on the outside, white on the inside), etc. if it is not. An ample arena for conflict.
Bruce Nevin (2008.02.29.2005 EST) –
In the case of the CSG-L tag, direct questions would probably not have disclosed that the timestamp is too unreliable to be of literal value, and that it serves primarily to distinguish two posts on the same day from the same person.
Bill Powers (2008.02.29.1348 MST) –
Are you arguing now about deficiencies of the time stamp, or is this another disturbance designed to see whether it’s a controlled variable?
Bruce Nevin (2008.02.29.2105 EST) - this message –
No, I was saying that direct questions about CVs would probably not have disclosed this fact about what we are controlling with the CSG-L tag. I had previously noted reasons why the clock time component has little reliable information, and the only function identified for it is to distinguish different message sent by the same person on the same day.
Bill Powers (2008.02.29.1348 MST) –
I could argue that your statements about what the discussion “disclosed” are a bit presumptuous or simply flat wrong, but would I then be arguing against a real point, or just demonstrating something of interest to you? I don’t know the terms of this discussion, and now that I know that you have used, and may still be using, deception, I have no trustworthy way of finding out what they really are. So I won’t rise to the bait, in case that’ what it is. It’s just another version of “The boy who cried ‘wolf!’”, part of every child’s education about why it’s self-defeating to lie.
Bruce Nevin (2008.02.29.2105 EST) - this message –
Yes, my doing this has damaged my reliability (a perception that you and perhaps other members of the group construct and project onto me). Why would I take that risk? Is it because I don’t care? Or is it because I am taking stronger measures to communicate something? What is this perception called reliability? If you think about that, you are thinking about a social variable.
And yes I do understand that you have just publically labeled me a child who needs to be educated as to proper behavior. But you are not annoyed, or that’s what you said when I suggested that you might be.
Bill Powers (2008.02.29.1348 MST) –
If you want to know what a person’s reference conditions are, ask. If the person is unsure or in conflict, the least you can then do is listen while he works it out and resolves the problem. He’s not there just so you can learn something.
Bruce Nevin (2008.02.29.2105 EST) - this message –
I am still waiting for serious discussion of how we learn about control of variables of this sort, which are not amenable to direct questioning. They are not amenable to direct questioning because a person’s control of them often changes when they are brought to conscious attention. A person’s control of them, and a person’s reported perception of them, changes because of differences between what one does “naturally” and what one “should” do, to put it somewhat simplistically. Calling attention to them results in controlling what one “should” do, but the “should” is not always the official standard of what is “correct”, if the person rebelliously identifies with the norms of a less favored group.
For more complete disclosure, I brought up the CSG-L tag as an off-hand comment about its inconvenience and it’s rather precious uniqueness to this group. I immediately after (first response) acknowledged that I probably shouldn’t have brought it up, and was willing to drop it. However, it was defended with such vigor and complexity that it was clear a lot more was going on, so after that I opportunistically seized upon it to enquire into what was going on by disturbing what I surmised might be controlled variables. It seems that because of the nature of those variables (as discussed above) there are strong feelings that disturbing them to verify control is inappropriate. That seems to me to be what you are saying – you have plainly declared that disturbing certain variables to ascertain control is off limits, while the whole literature of PCT gives a pivotal role to the methodology of disturbing variables to identify the CV and its reference value. This robust denunciation I take as confirmation of what I have said above about the unusual status of such variables, and the need for them to remain at an unconscious level of control in order to ensure their authenticity and the reliability of those controlling them together. Maybe there is some other explanation, but that is what I offer, and those are the ample reasons that I offer it.
I hope that is plain enough speaking to get us above the emotional crosscurrents at least for a bit.
/Bruce Nevin
Supposing it did annoy you, or
someone, the feeling of annoyance would indicate that a controlled
variable was disturbed – some expectation of appropriate behavior,
perhaps. How would you suggest testing for such a variable?"
[From Bill Powers (2008)]
I’ll start with that – tell me when I’ve included enough information for
you to find useful. Next time I’ll add the month.
I said
(2008.02.29.2005 EST)
You
replied in a way that is consistent with your having read this as “I
believe this annoyed you. This indicates …” etc. That is not what
I said. I could be wrong, but this suggests to me that your annoyance
influenced your perception of what I
wrote.
I replied that I decided to object, rather than saying that what you
wrote annoyed me, which would not be PCT-correct, would it?. I get
annoyed at things that I want to change (with about medium effort). It’s
an action, not a reaction.
The question
still stands. How would you suggest testing for such a
variable?
Bill Powers (2008.02.29.1348
MST) –
If you’re in a research setting, and the people you’re studying have
given permission to be studied, then you do it as the Test would suggest.
I suspect that linguists already do it that way, by and large. If you’re
having a peer-to-peer conversation where it’s assumed that the
participants are trying to communicate about something, you don’t do it
at all except perhaps to clarify what is being communicated. Hidden
agendas are a breach of trust.
“Norms must be
unconsciously controlled in order to be authentic?” An interesting
theory which I don’t recall ever having believed, but I’m more interested
in the “unconscious norm” idea itself.
I’ll try to
clarify what I mean.
There are
prescriptive norms that are taught, or their violations taunted, such as
“manners” that differ from your everyday unthinking
behavior.That’s the French Academy sort of thing. I’m not talking about
such norms.
Agreed. These could be called coercive norms – the taunter is trying to
control the behavior of the tauntee by causing error.
The words
“pan,” “pen,” and “pin” have distinct
vowels in many dialects of English, including what I speak, but in these
midwestern US dialects (and some others) the vowels of “pen”
and “pin” have merged. These are norms of pronunciation that we
acquired unconsciously in the course of using them in conversation with
our peers. They are not easy to change after a person establishes his or
her identity at puberty.
The fact that you know about them says that they’re not unconscious in
the Freudian sense (permanently unavailable to conscious inspection, as
opposed to “preconscious”). It might better be said that such
modes of pronunciation are not normally in awareness, although closer
inspection can reveal them if one is interested in doing so.
Am I right in thinking that “norm” is being used here more in
the descriptive than the prescriptive sense? People interacting with each
other come to control in ways that involve similar actions. This doesn’t
necessarily mean that they adopt (either consciously or unconsciously)
similar reference signals for those actions; it just means that in the
course of learning to control what they perceive, they reorganize until
their actions succeed in producing the desired results. I think it’s very
likely that no matter what we are learning in a social situation, we will
tend to converge to some ways of speaking, standing, gesturing, drinking,
and so on that achieve control of the relevant perceptions. It seems
reasonable to me that this would involve acquiring details of behavior
without being aware of doing so (although we can become aware of at least
some of the details).
Such norms as
these – for example whether your smile exposes upper teeth alone or both
upper and lower – are not prescriptively taught, and deviation from them
is not singled out and taunted unless the deviation becomes a shibboleth
of us vs. them identification.
Yes, I see and agree. No artificial social pressure is needed to achieve
the convergence (which Martin also talked about in that Durango
presentation). Don Campbell was delighted to learn that I had actually
hand-ground a telescope mirror, and he pressed me for details while he
was checking to see if his “lens-grinding” metaphor for social
adaptation was technically valid. We’re talking about phenomena more like
the natural tendency of the glass workpieces to arrive at a spherical
form simply because that form is the one that eliminates all the points
that stick out (though it doesn’t help much with pits from oversized
abrasive grains).
I think convergence may be a better term than norm, just because norm is
so often used in the prescriptive or coercive sense. The uynconscious
social norms you speak of don’t have the same role as reference signals
in PCT, which are individual rather than collective. Individuals are
relatively unaware of the outputs they are producing that control the
social variables they perceive, though by examining lower-level
perceptions they can get some inkling of what they are doing. The
inklings are not easy to see, however, because the connections between
muscle contractions and their perceptual results are not directly
visible. But physical realities can still see to it that the muscle
contractions are just the ones needed to adjust the details of action,
such as articulation, so as to create the intended proprioceptive and
auditory consequences. E. coli reorganization can explain how the
convergence is achieved even without anyone’s knowing exactly how it was
achieved.
If one adheres unconsciously
to a norm, how does it change when one’s consciousness is directed to it?
I
I asked that question before realizing the two senses of norm that come
up. One doesn’t “adhere” either consciously or unconsciously to
the kind of norm that is simply a result of reorganizations under similar
circumstances. If you’re from Boston and everyone refers to the structure
where you leave your vehicle as a caah paahk, then if you want them to
know where your car is, that’s how you tell them. You’ll reorganize until
they understand what you’re saying and act as if they do, which is one of
the layered protocols. If someone from Manhattan then asks where the caw
pawk is, you’ll need a second to realize what they’re trying to say, and
ask them if they’re trying to make fun of you. So adherence does get into
the act, but through other social
interactions, as you point out. A different kind of norm.
These are useful
questions. I’m not quite sure how to engage them with the matter at hand.
How does such a norm change when called to conscious attention? It
depends.
The cultural and social norms of
the sort that I am discussing are controlled in more than one way
simultaneously. There is no conflict between the two systems controlling
it because one of them controls with lower gain and the other with higher
gain.
When I spoke of calling a norm to conscious attention, I was using the
term strictly in the sense of a reference signal, an internal intention.
Calling them to attention involves little more than mentioning them.
Reference signals change when there is reorgnization at a higher level,
and the hypothesis is that reorganization follows awareness. But norms in
the other lens-grinding sense don’t depend on awareness; they’re just the
way things settle down to a state of minimum error after many
reorganizations and interactions with the outside world. To call
attention to them in the field of language, one needs to be a linguist
with more attention to detail than most people can manage.
That doesn’t eliminate conflict; it just says that the lower-gain system
will be the one with the biggest errors.
(But see
the note following.) To use a word for communication it must only be
recognizably distinct from other words that are similar to it. The
pronunciation can vary over a considerable range, so long as the
pronunciations of similar (“neighboring”) words co-varies so as
to maintain their mutual distinctness. Take the word “house”.
The sound between the h and the s sounds is a diphthong, two vowel sounds
run together. The first part of the diphthong sounds like the vowel of
“hut” in dialects common in Canada and the old-timers’ dialect
where I live. In other dialects it sounds like the a in
“father,” though shorter. In parts of Philadelphia and in many
dialects through the south, it sounds like the a in “hat” or
even the a in “late”.
Notice that these criteria are higher-level goals
that are independent of the particular lower-level articulations and
sounds by which they are achieved. In one geographic place they are
achieved by saying things differently from the way the same results are
achieved in other places. If the focus of awareness is on distinction
between words and recognizeability of meaning, reorganization can bring
that about by varying the way lower-level components are controlled until
communication is satisfactory. The latter is the lens-grinding aspect,
the non-prescriptive aspect of “norm.”.
Note: it may be
that one control system is controlling distinctness of words from one
another, with gain varying depending upon how predictable the word is
from context, and the other is controlling the “target
pronunciation” of the word in conformity with the norms of the
speaker’s dialect.
Most people command more than
one dialect. The differences between everyday talk with peers and formal
talk with teachers, interviewers, and the like, are most marked in people
who are not members of the most privileged class of people, and least
evident in the latter, who often regard their dialect as the only proper
way to speak (and similarly for other cultural norms, hence
“manners,” etc. ).
If one is consciously controlling for a target pronunciation because it’s
the social norm, then we’re back with the prescriptive meaning again – a
higher-order reference signal. That’s a higher-level goal than just
wanting to understand and be understood. Just wanting to reproduce a
specific kind of sound, however, is not a social norm.
This gets us definitely into the prescriptive meaning of norm.
It’s not
immediately obvious how to engage the method of levels with the norms of
pronunciation (or of smiling, or other body language, etc.). It might be
possible to use MOL to get from pronunciations etc. to the control of
identity and group affiliation. There very well might be a conflict
there, but the resolution typically is a higher level of control
switching from one dialect to another as appropriate, without feeling
that one’s identity is at risk by doing so.
However that may be in a given
individual, control of pronunciation (smiling style, body language, etc.)
is normally and almost always out of awareness because you are focussing
attention on other variables (meanings, quality of relationship, progress
of social transaction, etc.) for which these are the well habituated
means. Focusing attention on such details can interfere with performance,
somewhat like paying attention to the details of balancing a bicycle, or
trying to talk when there’s an echo on the
line.
The norms for the lower-level
control (the manner of doing the higher-level control) are unconsciously
acquired in childhood. Martin’s 1993 presentation speaks to this. The
effect is that the manner of doing the higher-level things (pronouncing
words, smiling, etc.) comes into conformity across the members of a group
of people.
I agree, the MOL application is to the prescriptive side, the side where
something social is being controlled and one can relatively easily become
aware of doing so. And that brings in psychological problems of all
kinds, but mainly conflicts both internal and external.
Yes, awareness goes to where the immediate problems are. This is really
why MOL therapy exists, because if the reason for a problem is that a
higher-level system is setting incompatible goals, reorganizing where the
problem seems to be will be ineffective – it will just change how one
has the same problem. In that case, of course, interfering with what is
going on is exactly what is needed. But what you are saying is also the
main reason that we don’t go around doing MOL with people without at
least some warning – conversations can get very strange when one person
stops answering questions and starts asking how it feels to wait for an
answer.
I think I can agree with all this now, recognizing that you’re not
describing prescriptive social norms but norms in the sense of
convergences to least-effort control, or something like that. I think
Martin would put that in thermodynamic terms.
Consciously
paying attention to these variables and trying to change them, as in
Shaw’s “Pygmalion” or the musical “My Fair Lady”
based on it, amounts to rejecting those whose manner you are changing
from, and not being accepted, or at least being suspected, by those whose
manner you are changing to. In the black community, it is (or used to be)
called “passing”, i.e. passing for white, if the person’s skin
is light enough, and is often derogated by terms like Uncle Tom, Oreo
(black on the outside, white on the inside), etc. if it is not. An ample
arena for conflict.
Yes, and largely because there is so little control over HOW we control
things – until we think to look at a lower level. The above is all about
the prescriptive meaning of social norms, which does bring in conflicting
reference conditions.
I am still
waiting for serious discussion of how we learn about control of variables
of this sort, which are not amenable to direct questioning. They are not
amenable to direct questioning because a person’s control of them often
changes when they are brought to conscious attention. A person’s control
of them, and a person’s reported perception of them, changes because of
differences between what one does “naturally” and what
one “should” do, to put it somewhat simplistically. Calling
attention to them results in controlling what one “should” do,
but the “should” is not always the official standard of what is
“correct”, if the person rebelliously identifies with the norms
of a less favored group.
I hope the distinction between prescriptive norms and convergent norms
qualifies as serious discussion. Concepts such as “correctness”
involve individuals’ reference signals, not just natural
convergences.
Direct questioning is fine. Just say, “I’ll like to ask some
questions about how you say things.” You can at least get at the
reference-signal kind of norms that way, which puts the others that you
observe in the other category. And you might even help someone resolve a
conflict or two of the precriptive sort.
Best.
Bill P.
Re: Problems testing for “social” CVs
[was: Julian Day for
[Martin Taylor 2008.03.01.09.47]
[From Bill Powers (2008)]
I’ll start with that – tell me when I’ve included enough information
for you to find useful. Next time I’ll add the month.
I hope you are going to post more than one message per year – or
per month, come to that. But I wouldn’t object if you were to make
your ID labels in the form “[From Bill Powers (2008) a.1.1.2]”
or whatever unique vaue you might come up with.] As I mentioned
somewhere, I’m trying to write a now rather long message in response
to [Bill Powers (2008.02.04.0820 MST)], to which a reader may wish to
refer when reading my forhcoming response (soon, I hope). That would
be rather less convenient if the reader had to search using as a
search string text quoted from your message.
On detecting pronunciation variants by observation as opposed to
by questioning:
Direct questioning
is fine. Just say, “I’ll like to ask some questions about how you
say things.” You can at least get at the reference-signal kind of
norms that way, which puts the others that you observe in the other
category. And you might even help someone resolve a conflict or two of
the precriptive sort.
For myself, I grew up in three different areas with distinct
pronunciation norms – Cheshire (England), East Lothian (Scotland),
and Toronto (Canada). I am normally not conscious of how I say
“house”, but if I listen to a tape of myself talking, I hear
something quite different from what I imagine I would hear. If I’m
asked how I say “house”, I consciously imagine the different
ways I might say it, and try to select one “haahs”
“hoos” or “huis” or something like that. A
conflict arises that is not there in any way I can detect when I am
just talking.
I do know that when I am with my brother, who grew up in these
same communities, I hear his pronunciation change depending on where
we are and to whom we are talking (though of course, mine does not). I
don’t think I have the facility of an actor to “talk English”
or “talk Boston” at will, at least not during conversation
at normal speed and when concentrating on the meaning rather than on
the sound.
I think it very UNlikely that “Direct questioning is just
fine” in this and in many other matters relating to conformity
with cultural norms.
Martin
PS. I repeat the question of why you write in blue rather than
the more easily read black? By “you” I don’t mean just Bill
P, by the way.
[From Bill Powers
(2008)]I’ll start with that – tell me when I’ve included enough information for
you to find useful. Next time I’ll add the month.I hope you are going to post more than one message per year – or per
month, come to that. But I wouldn’t object if you were to make your ID
labels in the form “[From Bill Powers (2008) a.1.1.2]” or
whatever unique vaue you might come up with.]
[From Bill Powers (2008.03.01.0832 MST)]
Martin Taylor 2008.03.01.09.47]
Well, suggest a format and I’ll use it. The one I’ve been using
is
[Name (yyyy.mm.dd.tttt zzz)]
where z is for time zone and t is for 24-hour time. The brackets are
intended to distinguish between the “[From” in the message body
and the “From” in the header. My main concern isn’t with
searching anyway, but just with keeping attributions straight, and not
having to look all over the place to find the information needed to do
so. The exact time doesn’t matter that much, though I don’t see any
reason to lie about it or round to the nearest hour to save two
keystrokes. Or put in a British dot where Americans use a colon and the
military and scientists (astronomers anyhow) don’t put anything.
On detecting
pronunciation variants by observation as opposed to by questioning:For myself, I grew up in three different areas with distinct
pronunciation norms – Cheshire (England), East Lothian (Scotland), and
Toronto (Canada). I am normally not conscious of how I say
“house”, but if I listen to a tape of myself talking, I hear
something quite different from what I imagine I would hear. If I’m asked
how I say “house”, I consciously imagine the different ways I
might say it, and try to select one “haahs” “hoos” or
“huis” or something like that. A conflict arises that is not
there in any way I can detect when I am just talking.I do know that when I am with my brother, who grew up in these same
communities, I hear his pronunciation change depending on where we are
and to whom we are talking (though of course, mine does not). I don’t
think I have the facility of an actor to “talk English” or
“talk Boston” at will, at least not during conversation at
normal speed and when concentrating on the meaning rather than on the
sound.I think it very UNlikely that “Direct questioning is just fine”
in this and in many other matters relating to conformity with cultural
norms.
I don’t see why not. The things you said about your own pronunciations
were interesting and sounded authentic, and moreover they call into
question the idea that because you’re aware of them, they aren’t
authentic. Most of us don’t have the ability (or the desire) to examine
and arbitrarily vary the mechanics of our behavior when our attention is
on higher-order controlled variables. Doing so would conflict with the
higher-order control systems anyway unless we could somehow turn them
off. But under the right circumstances, we can do some of that and with
the help of a neutral questioner we can probably do it better.
As to questioning others, the first consideration for me is, “First,
do no harm.” Harm includes withholding help when it’s possible.
Questioning for the sake of pure science, I think, does more harm than
good, so that isn’t what I recommend. Just wanting to find out how people
recover from hypothermia doesn’t justify the way the Nazis did it, and
the same goes for ignoring human relations when exploring human behavior.
But the real problem I see with not “just asking” is that the
alternative is to use guesses based on misleading information and
informal deductions from insufficient data, which leads to superstitions
and systematic delusions. You really can’t tell much about what’s going
on in another person just by making inferences from what is externally
visible (or audible). Some people think they can do that, but that’s
really just wishful thinking or egotism. Bad data are bad data. The best
observer of what goes on in another person is that other person. There
can still be communication difficulties, but even so asking is better
than guessing. And asking with awareness of levels, control processes,
and conflicts can be even more productive. This isn’t to say that sound
spectrograms and informed listening are useless, but their usefulness is
confined to low orders of control and perception. Uncertainties grow
rapidly as one tries to guess what’s happening at higher levels. Bruce N.
had to guess whether I was feeling annoyed. I didn’t have to
guess.
PS. I repeat the
question of why you write in blue rather than the more easily read black?
By “you” I don’t mean just Bill P, by the
way.
I don’t. When the text leaves my screen, it’s black, and when I
re-read it, it’s still black. What I see right here is black Courier
12-point or so text. Bruce N. appears to write in blue, and when I reply
to selected quotes, my text often gets contaminated and turns blue, too,
so I have to highlight my text and remove all formatting with
control-space (that’s in Eudora). Which I always do as far as I know,
though of course I could miss some. Or perhaps the removal of formatting
doesn’t stick when the message is transmitted. The more I find out about
the programs we use, particularly Windows but also the applications that
have to use it, the bigger the mess looks and the more numerous the loose
ends become. Here’s one. When Delphi is used to create two identical
bitmaps and to save the contents of one into the other for later
comparison, Windows decides that Delphi really doesn’t want another
bitmap and just hands back a pointer to the first one. You see, Delphi
doesn’t “own” the bitmap and what it wants doesn’t matter. So
when you read a new frame into the first bitmap and look at the saved
bitmap, you find the saved one gone and the new one in its place. Windows
is a maze of self-inflicted complexities that nobody will ever be able to
map.
In Eudora, the following font, with capitals a centimeter high, is called
Courier Humungous. And this is
“medium.” This is “large” and
what I guess is 12-point. It’s all black, though this is bold.
What do you see? Yours comes through as black “medium” size.
I’m sending “styled text only.”
Best,
Bill P.
Re: Problems testing for “social” CVs
[was: Julian Day for
[From Martin Taylor 2008.03.01.11.51]
Going back to the original question
[Bruce Nevin
(2008.02.28.2005 EST)]Anyone willing to
discuss these problems of testing for “social” controlled
variables.
followed by [Bruce Nevin (2008.02.29.2005
EST) ]:
The more important
and more general case is variables which are controlled in order to
participate in social processes (such as speaking and understanding,
recognizing one’s respective social affiliations, etc.) and whose
reference values are set in the course of controlling these variables
in order to participate in those social processes.
You raise a rather difficult issue, which has been sidestepped or
wrongly answered (in my view) by Bill P (labelled 2008, but apparently
Sat, 1 Mar 2008 01:37:54 -0700).
Direct questioning
is fine. Just say, “I’ll like to ask some questions about how you
say things.” You can at least get at the reference-signal kind of
norms that way, which puts the others that you observe in the other
category.
Bill’s answer is not unlike finding, out how well someone draws a
freehand circle by asking, and getting the answer “by making the
curvature the same all the way round” or “so that it looks
like a circle”. Even asking them to actually draw a circle won’t
tell you how they normally draw circles, because they are quite likely
to be running the lower-level control systems at higher gain or with
lower tolerance (a consequence of paying more attention to it, I
believe), so as to make a more precise circle than they would have
done had they merely wanted to draw a circle as an element in some
other figure.
You can’t easily apply “The Test” to find what
perceptions of a person’s conformance to social constructs are being
controlled, at east you can’t when the person is also controlling
perceptions relating to interactions with the observer. Disturbances
that affect the person’s perceptions of the observer will be opposed
in any control systems that relate to those perceptions. If the person
really and truly doesn’t care whether they are being observed, then
and only then will control actions be derived from systems that
control perceptions disturbed by the observer.
But that’s not as easy as, say, determining whether someone is
controlling x+y, x^2+y^2 or xy when asked to keep the area of a
rectangle constant. Rick can do that by disturbing x and y and seeing
what function of them remains most stable. That’s harder when you are
trying to disturb the person’s perception of their own speech or
gestural patterns while they interact in their normal social
environment.
One may be able to test higher-level social controlled variables,
but perhaps not ethically with any ease. One could offer a public
official a bribe to determine whether he is controlling some
high-level perception that would allow him to accept it, but one would
be unlikely to find a useful answer by asking “would you take a
bribe”.
My answer to Bruce’s question is pretty much: "At this
point, I don’t know how you would test for what social variables are
being controlled (direct questioning might be a suitable approach for
this, but even that is doubtful), and still less do I know hiw you
woudl go about disturbing then in a way that would allow you to
perform “The Test”.
Martin
[From Bill Powers (2008.03.01.1322 MST)]
Martin Taylor 2008.03.01.11.51 --
Bill's answer is not unlike finding, out how well someone draws a freehand circle by asking, and getting the answer "by making the curvature the same all the way round" or "so that it looks like a circle". Even asking them to actually draw a circle won't tell you how they normally draw circles, because they are quite likely to be running the lower-level control systems at higher gain or with lower tolerance (a consequence of paying more attention to it, I believe), so as to make a more precise circle than they would have done had they merely wanted to draw a circle as an element in some other figure.
You won't know if any of that is true until you try it. If you're anything like me, you find many more difficulties when imagining doing something than you encounter when actually doing it. I can tell you several methods I use in trying to draw a good circle when it matters, which I also use less carefully when it doesn't.
You can't easily apply "The Test" to find what perceptions of a person's conformance to social constructs are being controlled, at east you can't when the person is also controlling perceptions relating to interactions with the observer. Disturbances that affect the person's perceptions of the observer will be opposed in any control systems that relate to those perceptions. If the person really and truly doesn't care whether they are being observed, then and only then will control actions be derived from systems that control perceptions disturbed by the observer.
The observer gets to observe the person under many conditions, some directly relating to the controlled variables and others at both higher and lower levels. The self-consciousness and potential embarrassment you allude to do happen, but when they do they can become the focus of attention for a while, and when they abate, as they most likely will do, the focus can return to the questions of interest, and will probably do so spontaneously. If you're somewhat experienced with the method of levels, you encounter these sorts of problems repeatedly, and you also see how they melt away when approached properly. Ninety per cent of the time, I'd say. It's just not as hard to do these explorations as you're imagining it to be.
But that's not as easy as, say, determining whether someone is controlling x+y, x^2+y^2 or xy when asked to keep the area of a rectangle constant. Rick can do that by disturbing x and y and seeing what function of them remains most stable. That's harder when you are trying to disturb the person's perception of their own speech or gestural patterns while they interact in their normal social environment.
Remember the Dick Robertson and David Goldstein experiment with self-concept? They had subjects rate themselves on several dimensions, and then while discussing the result with the subject they simply inserted comments such as "I don't think you're that way at all." They got clearly identifiable opposing protests from 24 (as I recall) of the 26 subjects, the other two reactions being unclassifiable. That's a pretty good hit rate for verifying control of a very high-order perception. When you actually try doing these things they turn out to be easier, and work better, than one might imagine. And don't forget that you can always ask the subject to help out in looking for the information. If you can prove that an aware subject reacts differently from an unaware one, then obviously you have a way to measure the true reaction already and don't need to ask (otherwise you wouldn't know there was a difference). If you don't know what the true reaction is, there are no provable grounds for claiming that they are different from the aware reaction. What you have to watch out for is drawing conclusions about the aware statements just on the basis of which you think must logically be true.
One may be able to test higher-level social controlled variables, but perhaps not ethically with any ease. One could offer a public official a bribe to determine whether he is controlling some high-level perception that would allow him to accept it, but one would be unlikely to find a useful answer by asking "would you take a bribe".
I maintain that a neutral questioner who earns the confidence of the public official could find a way to get a truthful answer without having to offer a bribe. But this takes a direct person-to-person interaction, not just handing a person a standardized questionnaire to fill out. If a person trusts you to remain neutral, it's amazing what the person will be willing to tell you.
This is approaching a different subject I think is important -- the relationship of psychological testing to ethics. I feel rather strongly that just testing a person to find out things about him must take the back seat the inquiry verges on a therapy session. A public official who takes bribes is quite probably not in the best psychological shape. So what does the investigator do? Call the cops? Explore with the official what underlies this behavior? Or just say "Here, fill this out and I'll be back for it in a few minutes", and try not to look anyone in the face?
Best,
Bill P.
···
--
No virus found in this outgoing message.
Checked by AVG Free Edition. Version: 7.5.516 / Virus Database: 269.21.2/1305 - Release Date: 2/29/2008 6:32 PM
[From Bruce Nevin (2008.03.01.1601)]
Bill Powers (2008.03.01.0832 MST)--
The things you said about your own pronunciations were interesting and
sounded authentic, and moreover they call into question the idea that
because you're aware of them, they aren't authentic.
That wasn't what was claimed. What is claimed is "if you're manipulating
the references consciously, they're not authentic." In other words, if
you're trying to sound like people who talk a certain way, your
conscious manipulation of the kinesthetic and auditory references is
inauthentic. As you said, most of us are not actors who are capable of
that.
It's a different matter if you unconsciously shift dialect or social
register according to the company you find yourself in. That's what
Martin was describing (2008.03.01.09.47). You're shifting from one to
another manner in which you are fluent. How do you come to command more
than one manner? People unconsciously pick up mannerisms that are new to
them, but that is a slower process, and only after protracted contact
does it amount to acquisition of a new dialect or register. It is faster
in children. Some adults are more sponges than others in this regard.
The difference probably has to do with gain on control of your identity
distinct from those around you. Adult immigrants can and usually do
retain their Italian or Vietnamese or Iranian "accent" long after
gaining fluency in English. Those who gain a more native manner of
speaking resist disturbance to those aspects of their identity less
strongly. No one is comfortable going back to being a small child, and
to learn to speak a language in the manner of a native speaker may
require that.
Martin Taylor 2008.03.01.09.47--
I think it very UNlikely that "Direct questioning is just fine" in
this
and in many other matters relating to conformity with cultural norms.
Bill Powers (2008.03.01.0832 MST)--
I don't see why not. The things you said about your own pronunciations
were interesting and sounded authentic, and moreover they call into
question
the idea that because you're aware of them, they aren't authentic.
Most of
us don't have the ability (or the desire) to examine and arbitrarily
vary
the mechanics of our behavior when our attention is on higher-order
controlled variables. Doing so would conflict with the higher-order
control systems anyway unless we could somehow turn them off. But
under
the right circumstances, we can do some of that and with the help of a
neutral questioner we can probably do it better.
It may be a mistake to think of these as higher and lower levels in the
perceptual control hierarchy. Much that is going on in language is going
on through systems operating in parallel. I think that is the case here.
One system keeps words distinct. The gain varies according to how
predictable the word is, i.e. how much information it contributes at
that point. Observe the different pronunciations of "the," "have,"
"not," for example. In "I haven't got it" the likelihood of any word but
"not" coming between "have" and "got" is so low that the latter is
reduced to a syllabic n followed by t or glottal stop (no vowel). This
system doesn't care how the words are pronounced, within a very wide
tolerance, so long as each is distinct from the others that could be
there, with the caveat that, in reduced pronunciation, a word frequently
becomes auditorially indistinguishable from another word or words, and
this system does not care so long as only one of them is possible (or
highly likely) to occur in that syntactic and semantic context. This
system must involve inputs at a fairly high level of meanings and
syntactic combinabilities. It is not concerned with how different kinds
of people talk.
Another system has a very wide range within which to specify the target
pronunciation of each word (the kinesthetic and auditory CVs for
pronunciation) according to the norms of different kinds of people. This
system must have among its inputs perceptions of how different kinds of
people talk, what it sounds like, and among its references what it takes
to sound like that. It is not concerned with meanings or syntactic
construction at all.
Bill Powers (2008.03.01.0832 MST)--
As to questioning others, the first consideration for me is,
"First, do no harm." Harm includes withholding help when it's
possible. Questioning for the sake of pure science, I think,
does more harm than good, so that isn't what I recommend. Just
wanting to find out how people recover from hypothermia doesn't
justify the way the Nazis did it, and the same goes for ignoring
human relations when exploring human behavior.
An example may help. Labov and his students wanted to test hypotheses
about certain pronunciations in NYC differing by social class. He and
his students, with concealed tape recorder, took turns standing near the
escalator on busy days and asking anonymous passers-by where a certain
department was. They would presumably walk away periodically. I think
the pronunciation difference was evident in the phrase "fourth floor."
It doesn't matter here, so suppose that was it. The department they
asked for was on the fourth floor. They got a lot of examples of people
saying "fourth floor." They did this in a lower-middle-class department
store, and they did it in an upscale department store. The stores were
in close proximity in Manhattan. The results clearly confirmed the
hypothesis. No one was harmed or seriously inconvenienced.
(By the way, your response above is a stunningly rapic confirmation of
Godwin's Law
http://en.wikipedia.org/wiki/Godwin's_law.)
Bill Powers (2008.03.01.0832 MST)--
But the real problem I see with not "just asking" is that the
alternative
is to use guesses based on misleading information and informal
deductions
from insufficient data, which leads to superstitions and systematic
delusions.
The cultural norms that we are talking about are very distinct as
observable descriptive data. People consistently make them distinct.
That's how culture works. What is being teased out here is what is being
controlled by making these distinctions of pronunciation, etc. As with
the example above, you make hypotheses based upon informal observation,
and you test them in the field. Cultural differences are consistent
across groups of people who thereby mutually distinguish themselves from
one another. If you don't get consistent results (the same pronunciation
of "fourth floor") for all the people who otherwise have traits that are
also associated with the hypothesized group (shopping in a working class
store because they can't afford the upscale store; shopping in the
upscale store and wouldn't be caught dead in that dump over there, and
there's nothing anyone would want there anyway) then the hypothesis is
wrong.
Bill Powers (2008.03.01.0832 MST)--
Bruce N. had to guess whether I was feeling annoyed. I didn't have to
guess.
I ASKED if you were annoyed. (Your preferred methodology.) I then said
IF you or someone else were annoyed, that would indicate a disturbance
to a controlled variable. I don't know why you persist in misreading
what I wrote. Since it is behavior, it must be maintaining some variable
in a desired state. But you've made it very clear that annoyance is not
any part of it. That's very clear because simply asking always gets a
correct and complete answer.
In case it's not clear, I'm still working on Bill's challenge some years
ago to write something that might be titled "Language and the Control of
Perception". There's a lot of spadework that has to be done first, and
teasing out parallel as well as hierarchical systems of control is not
simple.
Bruce Nevin
[Martin Taylor 2008/03.01.17.43]
[From Bruce Nevin (2008.03.01.1601)]
teasing out parallel as well as hierarchical systems of control is not
simple.
They aren't mutually exclusive, as I believe you known. Classical HPCT is massively parallel at all levels.
Martin
[From Bruce Nevin (2008.03.01.2322)]
Bruce Nevin (2008.03.01.1601)--
>teasing out parallel as well as hierarchical systems
>of control is not simple.
Martin Taylor (2008/03.01.17.43)--
They aren't mutually exclusive, as I believe you know.
Classical HPCT is massively parallel at all levels.
Yes. It's the interrelation of parallel systems that gets more
complicated. Is it just that the outcome is a combination of their
several contributions, or are there inputs from one system across to
another? I assume the former without specific evidence for the latter.
There is a third system controlling syllable prosody, and a fourth
controlling intonation prosody, which is more of a gestural aspect of
language.
To clarify about syllable prosody (I know you're familiar with this,
Martin), the same word (or morpheme) can lose or gain vowels or
consonants to satisfy syllable constraints. These are not the same in
every language. In English, for example, we have the alternation of "a"
and "an", and the two main pronunciations of "the", with a vowel like
the second vowel in "holdup" when it comes before a consonant, and
pronounced like "thee" before a vowel, with a "y" sound making a
syllable onset.
A field of research called optimality theory has been very successful
over a great variety of language, and is I think quite compatible with
PCT. For any given word or morpheme there is a base form from which all
the others can be predicted (setting aside paradigms where actually
different words have merged, as in go-went, where went is historically
from wend). The deviations from the base form is accounted for by a kind
of sieve of different constraints. The connection with PCT is to see
that these constraints are CVs. The constraints are preferentially
ordered, and if an earlier one is applied it may change the shape of the
word so that it no longer is a disturbance to a later one. Many of these
constraints seem to be universal, but their preferential ordering
differs from one language to another. Of course the optimality theorists
do not think in terms of CVs and control of perceptual input, because
they do not know about PCT. I think there is great potential here. The
process is very neo-Darwinian, which should interest Gary Cziko,
generating a variety of candidate pronunciations and eliminating them
until only the optimal pronunciation -- the best fit to the constraints
in that language -- remains. English places the syllable onset
constraint (a syllable must begin with a consonant) above faithfulness
constraints (produce the base form faithfully), so when speaking
carefully we insert a glottal stop between a word that ends in a vowel
and one that begins with a vowel ("Oprah answered the question.") This
is of course the condition for saying "a" or "an", etc.
It applies well also, I think, to a pandaemonium treatment of syntactic
dependencies. As an utterance begins, there are many alternative
expectations as to what can co-occur with each word as it appears, but
as more words come the candidate set is reduced to one, or a few in the
case of ambiguity. But ambiguity in a sentence usually passes unnoticed
because of other contextual information, either in the more extended
discourse, or extralinguistic, what they call "pragmatics".
We're talking about parallel CVs controlled at different relative gain,
where the relative gain varies from one language to another.
There are other constraints which are not due to the inter-influence of
parallel control systems. Acoustical properties determine that in
certain regions in the vocal tract, shifting the location at which you
narrow or close the air passage with the tongue forward or back a bit
makes no discernible acoustic difference, but at other points a slight
shift makes a much more perceptible acoustic difference. This is why
there are no languages with contrastive articulations at the position of
j without also having contrastive articulations at the positions of
b-d-g, and there are none with articulations at the position of q (which
is behind g), or at positions between d and j or between j and k (which
do occur in various languages), without also having contrastive
articulations at the positions of b-d-j-g. Other constraints are
informational, certain mathematically characterizable properties that a
language must have in order for people to be able to use it for the
purposes for which we use language. I can enumerate these if you like.
Such "external" constraints limit what is possible for language rather
as properties of the environment constrain what it is possible to
control by, say, the observable actions of lifting and moving things.
I am not able to work on this very much or very consistently, having
many other obligations, but it is much on my mind.
/BN
[From Bill Powers (2008.03)]
Bruce Nevin (2008.03.01.1601)]
It's a different matter if you unconsciously shift dialect or social
register according to the company you find yourself in. That's what
Martin was describing (2008.03.01.09.47). You're shifting from one to
another manner in which you are fluent. How do you come to command more
than one manner?
I suppose the same way you learn to speak French when you're in France (if you can).
People unconsciously pick up mannerisms that are new to
them, but that is a slower process, and only after protracted contact
does it amount to acquisition of a new dialect or register.
I think the emphasis on "unconscious" is unnecessary. I see the point about people lapsing into their natural modes of speech unconsciously, but we do that with everything, don't we? You can breathe unconsciously, and usually do, but you can also do it consciously, perhaps not quite as tuned in to the body's real needs but workably. We walk unconsciously, type unconsciously, react to jokes unconsciously -- until something calls attention to these things, and then they go on working, but consciously. Since reorganization follows awareness, the behavior might change a bit when consciousness is present, if there's sufficient error somewhere. If picking up mannerisms is unconscious and slower, that's because attention is elsewhere and whatever reorganization is going on is slower because of not being focused. But athletes practice new mannerisms very consciously, hoping to master them well enough to rely on them to work automatically, the preferred mode since they will not change so fast when not in awareness. Look at how Tiger Woods has completely revised his stroke several times very systematically and very quickly. Yet he still speaks of his putting as if someone else is doing it.
It may be a mistake to think of these as higher and lower levels in the
perceptual control hierarchy. Much that is going on in language is going
on through systems operating in parallel. I think that is the case here.
The higher-lower dimension is suggested by which aspects of speech are varied as a means of controlling higher-order aspects of experience. One can use various learned systems for producing specific speech patterns, but the choice of which learned pattern to use when interacting with different people would be a higher-order matter.
One system keeps words distinct. The gain varies according to how
predictable the word is, i.e. how much information it contributes at
that point.
Fine, and then we have to ask, why do we use those systems to keep words distinct? The answer would be the higher-order goal served by doing so. If communication worsens (as in a noisy place), we tell the lower systems to make the contrasts clearer. That looks like a hierarchical arrangement to me.
Observe the different pronunciations of "the," "have,"
"not," for example. In "I haven't got it" the likelihood of any word but
"not" coming between "have" and "got" is so low that the latter is
reduced to a syllabic n followed by t or glottal stop (no vowel). This
system doesn't care how the words are pronounced, within a very wide
tolerance, so long as each is distinct from the others that could be
there, with the caveat that, in reduced pronunciation, a word frequently
becomes auditorially indistinguishable from another word or words, and
this system does not care so long as only one of them is possible (or
highly likely) to occur in that syntactic and semantic context. This
system must involve inputs at a fairly high level of meanings and
syntactic combinabilities. It is not concerned with how different kinds
of people talk.
That make sense -- what is controlled is not the words so much as the meanings we are try to grasp or convey by using them. Once that goal is achieved, there is no further motivation to reorganize. We stop reorganizing when higher-order understanding is achieved, not when pronunciation is perfect (unless we're singers or orators). Again, Martin Taylor's layered protocol idea naturally arises.
An example may help. Labov and his students wanted to test hypotheses
about certain pronunciations in NYC differing by social class. He and
his students, with concealed tape recorder, took turns standing near the
escalator on busy days and asking anonymous passers-by where a certain
department was.
That sounds innocuous enough, and harmless. If what you're doing isn't causing any problems, there's no reason to stop. It's not behavior I object to, but its results when they are bad. The people weren't asked to reveal anything they'd rather not see made public. As you say, no one was harmed or seriously inconvenienced. No problem that I can see. They weren't led falsely to believe that the questioner was interested in something and drawn into a serious discussion of it under false pretenses.
(By the way, your response above is a stunningly rapic confirmation of
Godwin's Law
Godwin's law - Wikipedia
Yes, I plead guilty. The Nazis did so many bad things that they're a cheap source of examples. I wish I knew what rapic means, though -- the web was no use even with Webster's unabridged aboard, and the various acronyms didn't seem to fit. What does your Oxford say? I don't have one.
Bill Powers (2008.03.01.0832 MST)--
> But the real problem I see with not "just asking" is that the
alternative
> is to use guesses based on misleading information and informal
deductions
> from insufficient data, which leads to superstitions and systematic
delusions.The cultural norms that we are talking about are very distinct as
observable descriptive data. People consistently make them distinct.
That's how culture works.
Well, I'd say that's how reorganization works, but I agree that at these lower levels, observations can be more reliable. I think I said that.
What is being teased out here is what is being
controlled by making these distinctions of pronunciation, etc.
That's more difficult and not as clear-cut, especially if you don't know of any way to ask the person that will yield good information. We could easily agree on the distinctions that are apparently being made, and think up tests to check our perceptions. But when it comes to why they are made, the guesswork gets fuzzier, particularly when the psychologizing starts.
As with the example above, you make hypotheses based upon informal observation, and you test them in the field. Cultural differences are consistent across groups of people who thereby mutually distinguish themselves from one another. If you don't get consistent results (the same pronunciation of "fourth floor") for all the people who otherwise have traits that are also associated with the hypothesized group (shopping in a working class store because they can't afford the upscale store; shopping in the
upscale store and wouldn't be caught dead in that dump over there, and
there's nothing anyone would want there anyway) then the hypothesis is
wrong.
Here the problem is that of sampling a few people and then assuming that everyone who shares a few superficial characteristics with them has the same (hypothesized) reasons for doing what they do. The so-called consistencies in this sort of data are pretty miserable. Remember that study that Gary Csiko found, showing that published sociological research articles based their conclusions on corrrelations that averaged, if I remember right, 0.26. In my book, that's "no relationship." But the addiction to statistics is pretty hard to kick.
Bill Powers (2008.03.01.0832 MST)--
> Bruce N. had to guess whether I was feeling annoyed. I didn't have to
guess.I ASKED if you were annoyed. (Your preferred methodology.) I then said
IF you or someone else were annoyed, that would indicate a disturbance
to a controlled variable. I don't know why you persist in misreading
what I wrote.
Well, you also mentioned that my reply sounded "frosty", and I did notice that you didn't ask whether I was delighted. I sort of deduced that you suspected that I was a bit annoyed, but if you say you proposed that attitude at random, I'll take your word for it.
Since it is behavior, it must be maintaining some variable
in a desired state. But you've made it very clear that annoyance is not
any part of it.
No, I didn't. I said I was, indeed, annoyed, but I did not put that in a way that made it seem that you were making me be annoyed; I took responsibility
for deciding to object, as you will see if you reread the post. You don't make me annoyed, I make me annoyed. I get into the state I call annoyance when I'm using more effort than I want to waste in trying to correct a persistent error of moderate importance to me.
That's very clear because simply asking always gets acorrect and complete answer.
God, that sounds like the ghost of Bill Williams. When did I ever say such a silly thing? Answer: never. Kindly get your words out of my mouth.
In case it's not clear, I'm still working on Bill's challenge some years
ago to write something that might be titled "Language and the Control of
Perception". There's a lot of spadework that has to be done first, and
teasing out parallel as well as hierarchical systems of control is not
simple.
That is going to be a great event, Bruce. Annoying or not, you are still an outstanding investigator and expositor and when you put your mind to it, you write some of the most interesting stuff in this discussion list. That's why I keep forgiving you.
Best,
Bill P.
···
--
No virus found in this outgoing message.
Checked by AVG Free Edition. Version: 7.5.516 / Virus Database: 269.21.2/1305 - Release Date: 2/29/2008 6:32 PM
[From Bruce Nevin (2008.03.05 2141 EST)]
Bill Powers (2008.03) --
Bruce Nevin (2008.03.01.1601) --
Bruce Nevin (2008.03.01.1601) --
How do you come to command more
than one manner?
Bill Powers (2008.03) --
> I suppose the same way you learn to speak French
> when you're in France (if you can).
I think what you mean by that is a concerted effort to learn French
while you're in France. Your coda "if you can" has no other
interpretation that I can think of. But what I am talking about is what
linguists (presumptuously, I think) called "code-switching" within the
same language. It may be that you always speak in the same manner no
matter who you are talking with, but it is very normal, ordinary,
commonplace, and expectable that a given person speaks in one manner in
a formal situation (to an interviewer, a judge in a courtroom, when
introduced to a respected VIP), and in a different manner when speaking
with a family member, childhood friend, neighbor of many years'
standing, etc. These are differences that one does not typically mount a
concerted effort to learn the way one sets out to learn a foreign
language.
Bruce Nevin (2008.03.01.1601) --
People unconsciously pick up mannerisms that are new to them, but
that is a slower process, and only after protracted contact does it
amount to acquisition of a new dialect or register.
Bill Powers (2008.03) --
> I think the emphasis on "unconscious" is unnecessary.
When learning this socalled "code switching" from one "register" to
another the attention is on higher levels of control of at least two
sorts of perceptions. One is meanings and confirmation that you have
replicated the intended information among your own CVs. Another is a
perception that we are related as peers. If one person persists in
holding to a formal and "correct" way of talking while the other is
speaking in "homebody" way that is an assertion that they are not peers.
I believe it is error in control of peer relationships and "belonging"
that drives reorganization of the manner of speaking, "body language,"
and other variables of personal style. Attention is focused on the
higher levels where error occurs, and the lower levels where
reorganization is taking place are out of awareness. There may be a
generalization here about attention drawn to error at a higher level
triggering reorganization of lower-level means of control to which
attention is not directed because it is directed to the higher level
where the error occurs.
Bill Powers (2008.03) --
> ... athletes
> practice new mannerisms very consciously, hoping to master
> them well enough to rely on them to work automatically, the
> preferred mode since they will not change so fast when not
> in awareness.
Yes, it is possible to make a concerted effort to observe the relevant
differences and practice them, but this is not what people ordinarily
do. The more profound differences e.g. between Black English Vernacular
and formal English the way white folks talk may well require a concerted
effort; I can't speak to that, and am not. In some communities more than
one distinct language is involved in code switching.
Bruce Nevin (2008.03.01.1601) --
It may be a mistake to think of these as higher and lower
levels in the perceptual control hierarchy. Much that is
going on in language is going on through systems
operating in parallel. I think that is the case here.
Bill Powers (2008.03) --
> The higher-lower dimension is suggested by which aspects of
> speech are varied as a means of controlling higher-order
> aspects of experience. One can use various learned systems
> for producing specific speech patterns, but the choice of
> which learned pattern to use when interacting with different
> people would be a higher-order matter.
A bit of a misunderstanding here, and entirely my fault, because I did
not include enough in that statement. The word "these" refers to systems
controlling perceptions like distinctness, syllable prosody, intonation
prosody, and manner of speaking in parallel, each as means for
controlling higher-level perceptions including information replication
in the recipient ("meanings"), peer/non-peer relationships, and attitude
toward the information (including emotions).
Bruce Nevin (2008.03.01.1601) --
This
system doesn't care how the words are pronounced, within a very wide
tolerance, so long as each is distinct from the others that could be
there, with the caveat that, in reduced pronunciation, a word
frequently becomes auditorially indistinguishable from another word or
words, and this system does not care so long as only one of them is
possible (or highly likely) to occur in that syntactic and semantic
context. This system must involve inputs at a fairly high level of
meanings and syntactic combinabilities. It is not concerned with how
different kinds of people talk.
Bill Powers (2008.03) --
> That makes sense -- what is controlled is not the words so
> much as the meanings we are try to grasp or convey by using
> them. Once that goal is achieved, there is no further
> motivation to reorganize. We stop reorganizing when
> higher-order understanding is achieved, not when
> pronunciation is perfect (unless we're singers or orators).
In parallel, however, another system does care how the words are
pronunced, within a narrower range, so as to speak like this kind of
person or that kind of person, perhaps imperfectly to the ear of an
outside judge, but in a way that in the speaker is set as a reference
for e.g. talking to a judge or to Josie down the block.
Bill Powers (2008.03.01.0832 MST)--
> > But the real problem I see with not "just asking"
> > is that thealternative is to use guesses based on
> > misleading information and informaldeductions from
> > insufficient data, which leads to superstitions
> > and systematic delusions.
Bill Powers (2008.03) --
>The cultural norms that we are talking about are very distinct as
>observable descriptive data. People consistently make them distinct.
>That's how culture works.
Bruce Nevin (2008.03.01.1601) --
Well, I'd say that's how reorganization works, but I agree
that at these lower levels, observations can be more
reliable. I think I said that.
I'm glad you now agree that at these lower levels which I am talking
about (and whose references are set in processes of reorganization due
to error at higher levels, and therefore not in awareness) observation
is acceptably reliable.
I would press this farther and ask your agreement that observation is
more reliable than "just asking" and that "just asking" directly about
such variables is unreliable. It is unreliable because the asking calls
attention to these lower-level variables are the means of controlling
higher-level perceptions of social status, peer relationship in
conversation, etc.
Try it some time. Notice some peculiarity of pronunciation, and ask the
person why they pronounce "house" (or whatever) that way. Imitate their
pronunciation to make your point. Unless they have made a study of the
matter they are very unlikely to be able to give an informative answer.
Very likely your conversation with them will become awkward and
uncomfortable. Why are you shifting the topic from the interesting
information we were talking about to the way I pronounce "house"? This
doesn't feel like a peer relationship. How do you pronounce that word,
yourself? What does the difference mean, and in particular what are you
inferring about me and my kind of people vs. you and your kind of people
from this difference in pronunciation? And so on. Not a simple question
at all!
Some adaptation of MOL process (with prior agreement, of course) might
confirm this in a way that you would regard as more reliable -- When I
asked you about how you pronounce "house", what occurred for you? How
did you feel? What's behind that?
Sorry to be so slow responding. Healing this shoulder surgery slows down
work and interferes with sleep.
/BN