social control

CSGnet Archive CSG1993
1

[From Rick Marken (930519.1400)]

Bruce Nevin (Wed 930519 12:26:11) --

Think of a child learning to throw and catch a ball. Reference
perceptions are set based upon experience of trial and error in
throwing and catching a ball.

Again, I don't know what you mean by " reference perceptions".
Really -- I'm not trying to be difficult. I presume you are using the
term "reference perception" to refer to reference signals or to the
reference state of a controlled perception. In either case, these
are not set "based upon experience of trial and error"; reference
signals are set by disturbances to controlled variables (since
they are set by the outputs of higher level systems in order
to compensate for disturbances to the variables controlled
by these systems); reference states of a controlled perceptions
are set by the reference signals.

The trial and error involved in learning to throw and catch is a
process of learning to control; that is, learning what to perceive,
how to vary the output so that the perception remains under
control, how that perception can be varied to control higher
order perceptions, etc. Your description above suggests that
there is a setting for each "reference perception" which, when
learned by trial and error, is the the "right" setting to select
in order to throw or catch. I know you know that this is not
the way control systems work -- yet I can't seem to derive
any other meaning from your words. You can see my problem,
I hope.

But these properties of the environment are in the environment
as a byproduct of perceptual control by other people--specifically,
as a byproduct of their control of perceptions that the child is
learning to control in a way that they can recognize. In this
respect, the properties of the child's environment that give rise
to words and phonemic distinctions (in that order, and then to
more words once the system of distinctions is controlled) is unlike
the physical properties of the environment that imputedly give rise
to such things as parabolic trajectories.

I'll buy that. But they are still just the child's perceptions. What you
describe is exactly what happens in Tom's cooperation experiment;
each subject is able to control ONLY when the other subject is also
controlling. But that is just a property of the subject's connection
to their own perceptions -- there is a "social" aspect to the feedback
function that connects the actions of the subject to the perceptions
he or she controls. But no new properties of the PCT model are required
to explain control in this situation.

In other words, the perceptions
used for communicating between A and B must be pre-set in A and in B
such that each may produce repetitions recognizable by the other and
such that each associates imagined perceptions with them for which
they can substitute actual perceptions that they can agree are
"the same thing" in their shared environment, etc.

I don't get what you mean by "perceptions must be pre-set". All that
needs to happen, it seems to me, is that A be able to select words
(actions) that will result in the desired perception of B . For example,
if I want B to lift the other end of a table and I might be able to produce
this perception by saying "lift". But if that fails I will vary my actions
(utterances) and say "lift please" or "lift or else" or "yoo hooo" or
whatever works. There may be considerable variation in the "preset"
utterance I mak in order to get the desired perception (B helping)
on each occasion.

Reading your post over again, I get the impression that your main
problem is with the current random reorganization model of PCT;
is that it? You seem to be saying that some built in "knowledge" is
needed by a system that will allow people to learn to communicate.
Is that right? Suppose that a random reorganization model of subjects
A and B in Tom's experiment WAS able to reliably discover the
cooperative solution ? Would that convince you that maybe
it's a little early to start worrying about whether we need to imagine
"intelligent" reorganization processes? I bet (and I have done some
simulation work on this -- I think Tom has too) that two interacting
random reorganization systems can learn to control cooperatively when
this is required (by the "social environment") in order to achieve their
goals.

Can you now see why I believe this question of how we acquire
reference perceptions is important?

I think so. Is it because you think there is something special about
learning to control variables when the feedback function is "social"
(depends on the controlling of other control systems) rather than
"physical"?

Can you see that HPCT does
not give a coherent account of the skillful use of language
with reference to perceptions, the unskillful use of language
with reference to actions, the pointing and gestures and
jumping up and down, and the other means of accomplishing this end
that the human participants demonstrated?

Well, not really. But, I'm willing to listen and try to understand
your position.

Best

Rick

2

[Hans Blom, 930615]

(Rick Marken (930614.1200))

This is impossible to resist. The discussion of social control systems
reminds me of some of the statements that my biology and physiology
colleagues often try on me. Let me try the following paraphrases on you,
where the 'individual' is a cell and 'social' refers to an organized
collection of cells, i.e. a (human) body:

Tom Bourbon (930614.0840)

"They" (entities called body control systems -- body systems that
control) are actually individual cells, each of which has goals
or intentions that bring them into common action around a set of
shared variables in their environments.

This sounds all right, doesn't it?

Bill Powers (930613.2200 MDT)] --

Body cells often try
to use the concept of body rules and body demands as a way of
avoiding responsibility for the things they do with their own
actuator mechanisms. They like to pretend that they are
governed by something outside themselves; in that way they can do
what they really want to do, but when problems or objections
arise they can pass the buck. "I'm only following orders."

But this sounds funny, isn't it? Responsibility isn't a term we use
for cells. Could it be that notions like responsibility are orthogonal
to our field of study? Or is a cancer cell worthy of our reproach?

The body's systems are real perceptions -- they are an important and
interesting byproduct of many cells simulaneously controlling
their own perceptions. But the body's system's (and what they do) are
just a side effect of the operation of individual control systems.
These side effects can be quite useful to many of the individuals
involved in creating the side effects of their mutual efforts-- for
example, the food that is obtained, the move to a better feeding place,
etc. But the explanation of the body phenomena exist in each of
the individual control systems and their (mutual) relationship to the
variables they are controlling.

Let's try not to forget that when we study PEOPLE, we focus on THEM as
our natural units. We do not regard their components as units of discourse,
nor the multi-unit assemblies that they might form in order to make life
easier for each. In my opinion, there is no reason NOT to talk about
social control systems; 'higher level' concepts often emerge when many
units interact; in statistical mechanics you can talk about concepts like
temperature and pressure, concepts which do not apply to single molecules.

Greetings,

Hans Blom

3

[From Oded Maler 931615]

ยทยทยท

*
* [Hans Blom, 930615]
*
* (Rick Marken (930614.1200))
*
*
* Bill Powers (930613.2200 MDT)] --
*
* >Body cells often try
* >to use the concept of body rules and body demands as a way of
* >avoiding responsibility for the things they do with their own
* >actuator mechanisms. They like to pretend that they are
* >governed by something outside themselves; in that way they can do
* >what they really want to do, but when problems or objections
* >arise they can pass the buck. "I'm only following orders."

Etc.

This is exactly what I was going to say, but it is useless. It turns
out that there is something in those cell assemblies (aka BP and RM)
which resists the idea that their level of organization is not the
ultimate one. While they can look *down* and see how their own
selves emerge from autonomous cells, who control some biochemical
individual perceptions, they cannot look *up* and see the emergent social,
historical entities, for which "autonomous" human beings are nothing
but sensors, actuators and comparators.

--Oded

--

Oded Maler, LGI-IMAG, Bat D, B.P. 53x, 38041 Grenoble, France
Phone: 76635846 Fax: 76446675 e-mail: maler@imag.fr

4

[Hans Blom, 930617]

Let me try to make clear what I mean when I talk about social
control emerging from individual control systems. I will use a
constructive (modelling) approach that anyone can easily test.

Take a number of control systems. In this example I will use
'robots' that move about. Give each of them some control laws:
don't get too close to others, don't get too far away from
others, but otherwise you can move freely (use random numbers
to create a random walk for each individual, for instance). The
details of the laws don't matter much: under the most general
conditions a flock, herd or some such will be formed. Do NOT
introduce non-local control laws; each individual is supposed
to use only its own sensors and actuators to establish its
position.

Each individual is a separate and autonomous control system,
controlling only for its own goals. But the flock will prove to
be a 'higher level' control system, although no higher level
rules are programmed in in any way. Try to use The Test, in
whatever way you want, to disturb 'flockness'. Pick up an
individual and drop it again some way off. If the other
individuals can still perceive the lost one, the flock will
reform. Limit movements by introducing obstacles or corridors:
the flock may temporarily separate while moving around an
obstacle, but it will reform again.

You can call the flock a social control system. No single
individual has a notion of what a flock is: it just controls
its own movements. Yet a (moving but stable) flock emerges.

Many more examples of emergent behavior can be found in: From
animals to animats, Proceedings of the First International Con-
ference on Simulation of Adaptive Behavior. Eds: Jean-Arcady
Meyer and Stewart W. Wilson. MIT Press, Cambridge, Mass., 1991.

Greetings,

Hans Blom