Open loops?

[Tom_BOURBON3]

Tom Bourbon [941004.1422]

Clark Mcphail [941003.1607] (I made a csg-l header for you, Clark.)

Tom Bourbon [941003.1032]

. . . As for the
social nature of psychophysics, I believe that is a topic ripe for the
plucking by a PCTish social scientist.

Could Muzafer Sherif's (1935) work contribute to this discussion? His
training in classical psychophysics was supplemented by extensive reading
in the social sciences which led him to the hypothesis that "social norms"
might have more to do with an individual's perception and behavior than the
physical properties of the world in which that perception and behavior took
place.

Could Sherif's work contribute to the discussion? Certainly, but with some
clear limits on its role. You described his classic work on the autokinetic
phenomenon. (I'll not quote all of your post here.) I think his idea of
studying perception during social interactions was elegant, but in that line
of research, Sherif's unavoidably traditional training shows through.

Both Rick [From Rick Marken (941003.2330)] and Bill
[From Bill Powers (941003.2025 MDT)] have commented on some limitations on
Sherif's measures of performance, which Sherif took to be measures of
perception. My first caveat concerning Sherif's work returns to Fechner's
warning that we never directly measure perception in a psychophysical study.
That sounds like one of the warnings that always prompts a reply of, "oh,
but we all know that," but if we do, we certainly don't act as though we do.
My impression is that Sherif assumed he had identified real changes in
perceptions, induced by social interactions.

Of course, had Sherif known of CST-PCT, he would have done what Bill Powers
suggested, and what I would suggest to any up and coming young PCT social
scientist -- include a condition in which the subject uses a control device
to hold the spot of light motionless. The person's adjustments of the
device would give a nice indication (but still not a direct measure) of how
much motion the person saw. If you were to try that procedure with two or
more people, you would have an interesting experiment in social perceptual
control. Would the subjects disturb one another's perceptions of
motionlessness so severely that no one could hold the spot still? Would you
see evidence of the raging conflict I predict, when each person tries to
control his or her unique (and not really shared at all) perceptions of
_where_ the spot should remain and of _how much_ motion counts as
"motionless?" What would the members of a group do when they discovered
they were disturbing one another, or they were locked in conflict? Would
their individual reference perceptions change and if so would they become
more, or less, similar? Would some stop using the control device altogether?
And on it goes.

Is this remotely related to which you had in mind Tom?

As you can see, it is, but I also had in mind the idea that even the
_simplest_ and most basic psychophysical measures -- the "purest" if you
like -- are created in a social setting where subjects control for
something (or some things) in addition to, or quite different from, what
the experimenter wants to control. Even the purest psychophysical measures
(absolute and differential thresholds), which people often assume tell us
something direct about isolated parts of the subject's sensory systems, are
open to interpretation as examples of perceptual control, by whole persons,
during a social interaction. Here are even more opportunities for a PCT
social scientist. You people may get to have _all_ the fun.

Later,

Tom

[Tom_BOURBON3]

Tom Bourbon [941003.1032]

(I believe there is something here for the sociologists and social
psychologists.)

[From Bill Powers (941002.0830 MDT)]

Bill Leach (941001.1851 EDT) reminds me of a comment I had meant to make
on

[Martin Taylor 940929 10:30]

Others have offered similar comments on this thread:
[From Rick Marken (941001.1030)] Loop decomposition
[Bill Leach 941001.18:51 EST(EDT)] Re: Help!

Martin:

When an electronic engineer wants to examine an unknown control system,
one way is to break the feedback connection and examine the S-R
behaviour of the broken loop.

Bill:

It's very difficult to break a control loop that a person normally uses,
without disrupting the control system. As soon as the loop is broken,
the person loses control and furthermore knows that control is lost, for
actions no longer affect perceptions. The normal consequence of breaking
the external part of a control loop is for the person to switch to some
alternate method of control as soon as loss of control is detected, or
to give up controlling completely until the outer part of the loop is
restored.

Yes. That is what the person will do. But most subjects in a classical
psychophysical experiment _do not_ give up; they keep going. I believe
they are controlling for something else, something very different from the
perceptions the experimenter wants to study. (More on that a few lines below.)

In the behavioral-psychophysical-psychophysiological literatures authors
often say they have studied some part of a person-system as an input-output
device. Sometimes they even speak of "opening up a loop." Words do not make
it so. In classical psychophysics, the closest anyone comes to studying
perception as part of a control loop is in the "method of adjustment," where
the subject adjusts some parameter of a "stimulus" until it
looks-sounds-feels-etc "just right" -- just different, just not different,
barely there, barely not there, etc. All of the other methods treat
people as though they were lineal systems -- a fact, not merely my opinion.

Investigations of open-loop passive perception are peculiar in that they
make the test subject a partner, and even a substitute, for the nominal
experimenter. . . .

Yes. The experimenter relies on the subject slipping into a very tightly
constrained social role. The only other place in life, outside a
psychophysics lab, where you will find two people interacting as they do in
a classical psychophysical experiment is in a clinical test of hearing or
vision -- "Hold your finger up when you hear the sound; put it down when you
don't;" "Does it look clearer with this lens . . . or this one?" And so on.

What is generally done, as I understand it, is for the experimenter to
tell the test subject to make some standard indication when some feature
of the perceptual field is experienced -- but there is no way to know
what that feature is, in the test subject's world. Thus perceptual
experiments tend to be cast as S-R experiments, with occurrance of a
response being assumed to indicate occurance of the perceptual situation
that the experimenter has in mind. In very, very simple (low-order)
situations this may not be a bad guess, but it is always a guess, and it
will become progressively worse as more complex perceptions are
investigated.

Yes. That is exactly the point I have tried to make each time we come back
to the topic of psychophysics. The clasical methods are, by design, S-R
methods (not a pejorative, but a statement of fact). Good old Gustav
(Fechner) knew that was the case and cautioned that the only things we learn
as facts from such methods are some associations between stimulus (as we
define and manipulate it) and response (the reports from our subjects,
reports on which we as experimenters impose severe constraints). All else
-- ALL ELSE -- is inference (educated guess, but guess nonetheless).
Nothing in the data tells us anything direct about the subject's
perceptions. And, as Bill says, that is true for even the simplest, lowest
levels of perception -- is the light there, or not; is the sound there, or
not; and so on.
. . .

In summary, "breaking the loop" to observe open-loop responses to
arbitrary changes in the input is not a practical method for
characterizing human control systems. Nor are tests of passive
perception that do not involve control likely to tell us about the
characteristics of control systems.

Yes. Psychophysical studies _do not_ isolate individual parts of the
control loop. It is common practice for psychologists to refer to a
psychophysical measure (an absolute threshold, a difference threshold,
a psychophysical function relating magnitude of perception to magnitude of
stimulation, etc) as though it is a measure of the sensitivity of one
part of the person. They say they have measured "the sensitivity of the
eye," "the absolute threshold of the basilar membrane," or "the growth of
sensation as a function of rate of discharge in the ____ nerve," and so on.
Not so. In every case, they have measured an association between stimulation
and a severely socially constrained "response" of the whole person. No
exceptions.

When I say the subject's role is "socially constrained" in psychophysical
studies, I do not mean there is an immaterial "social influence" at work;
the experimenter asks the participant to assume a particular constrained
role and the participant agrees. The participant adopts an intention (more
or less and to the degree he or she understands it) to observe certain
"stimuli" and to "respond to them" in specific ways. The subject is
controlling for something in the social domain -- approval, course credit,
pay, time out from doing lousy work, a way to avoid or get into military
service, etc; the experimenter is controlling for seeing certain
relationships between "stimuili" and "responses" and then drawing or
supporting certain inferences about perception. Nobody in those experiments
is measuring the sensitivity of eyes, retinas, rods and cones, visual
pigments, or anything else more "basic" than that. Everything is a matter
of whole, social persons.

What I say is not intended as a rejection or discrediting of psychophysics,
but as a reminder of possible limitations on psychophysical data as
sources of insight into the phenomenon of perceptual control. As for the
social nature of psychophysics, I believe that is a topic ripe for the
plucking by a PCTish social scientist.

Later,

Tom

[Clark_Mcphail1]

Tom Bourbon [941003.1032]

What I say is not intended as a rejection or discrediting of psychophysics,
but as a reminder of possible limitations on psychophysical data as
sources of insight into the phenomenon of perceptual control. As for the
social nature of psychophysics, I believe that is a topic ripe for the
plucking by a PCTish social scientist.

Could Muzafer Sherif's (1935) work contribute to this discussion? His
training in classical psychophysics was supplemented by extensive reading
in the social sciences which led him to the hypothesis that "social norms"
might have more to do with an individual's perception and behavior than the
physical properties of the world in which that perception and behavior took
place.

His problem was to find 'a fluid and ambiguous situation for which
individuals did not have previously established standards" (1936:ix) and
therefore one in which he could study the development of individual
standards and then social standards and their consequences for perception
and behavior. He concluded that the autokinetic phenomenon - the apparent
movement of a pinpoint of light against a totally darkened field - might
provide such a situation. He brought individual subjects into a room,
seated them at a table containing a signal key, and explained:
        "When the room is complete dark, I shall give you the signal READY and
        then show you a point of light. After a short time the light will
start
        to move [sic]. As soon as you see it move, press the key. A few
        seconds later the light will disappear. Then tell me the distance it
        moved. Try to make your estimates as accurate as possible. (1936:35)

On the first day each S made 100 judgments. The iniital judgments of
distance varied considerably. Subsequent judgments varied less and less,
the majority converging around a narrower range, unique to each subject.
Sherif computed the median of each individual's distance estimates and
termed this the individual's standard. (Parenthetically, for now at least,
a PCT researcher could replicate this and then by some means challenge or
disturb the judgments and thereby do "the test" to verify that the
individual had indeed established his or her unique "standard" or reference
signal from the multiple perceptions which had been judged and stored away
in memory.)

On a second day, three individuals, who had gone through the procedure just
described, returned to Sherif's laboratory where each made and witnessed
one another making another distance judgments of another set 100
autokinetic exposures. Initially each individual's judgments clustered
around his or her prior individual standard. But subsequent judgments
gradually shifted in the direction of the judgments given by the other
individuals. Sherif reported that no individual returned to his or her
initial range or median nor did any individual establish a new or unique
range or median standard. Instead, the initially divergent individual
standards converged and a new, Sherif termed it "emergent", standard was
established which was unique to this group.

Sherif subsequently did a number of variations on this psychophysical
experiment. One variation reversed the order described above; that is,
three individuals' initial exposure to the autokinetic phenomenon was in a
situation of copresence in which each individual in turn gave his or her
estimate of the distance the light appeared to move. Again there was
initial variation in their judgments, although less so than when
individuals estimate autokinetic distance, but those judgments quickly
converged in the direction of a consensus "standard" for the remainder of
the 100 trials. After three such copresence sessions each individual
returned for a solitary judgment session. At issue was whether or not each
person would establish a new range of judgments (and median therein) or
would give judgments corresponding to the standard establish in the
copresence judgment situation. The latter was the case and Sherif thereby
claimed support for his contention that in a new and unfamiliar situation
people can "interact" and establish a new or "emergent" standard for
behavior in that situation.

When individuals who had participated in any of the copresence experiments
(before or after participating alone) were asked to return one month, six
months, even 12 months later and against asked to judge another 100
autokinetic exposures, their judgments corresponded very closely with this
standard which had been established in the previous "social interaction"
situation.

I think these classic experiments in social psychophysics should be of
interest to people on the net because it addresses a situation of how
individuals may establish a reference or perceptual standard in a new
situation as well as how two or more individuals may do so in a situation
of copresence and minimal interaction as well as the apparent "enduring"
nature of such standards. With replication and the introduction of "the
test" social scientists might learn more about both the production,
maintenance and alternation of perceptual standards.

Is this remotely related to which you had in mind Tom?

Clark McPhail
Professor of Sociology
326 Lincoln Hall
University of Illinois
702 S. Wright
Urbana, IL 61801 USA
off/voice mail: 217-333-2528 dept/secretary: 217-333-1950
fax: 217-333-5225 home: 217-367-6058
e-mail: cmcphail@ux1.cso.uiuc.edu