environmental gain

CSGnet Archive CSG1993
1

[Martin Taylor 931004 19:00]
(Bill Powers 931004.1000 to Hal Pepinski 931003c)

Bill, I admire your restraint and good nature. But I'd like to follow one
paragraph...

The organism-environment relationship would
be symmetrical if the amplification in the closed loop were
equally distributed around the loop. In fact, the amplification
involved is concentrated almost exclusively inside the organism.
The environment part of the loop almost always introduces
_attentuation_. This places the controlling agency inside the
organism.

This is true when you are dealing with a non-living part of the environment.
It is not true when the critical part of the environment is another control
system, and THAT is the condition for (almost?) all of Hal's discussions.

Hal is talking about interpersonal issues, and under those conditions the
gain in the loop may be equally distributed within and outside "the
organism" (meaning either one of the parties--"you" in what follows).
Now you get into a state in which control of your own perception can be
effected not only by applying substantial force to a dissipative world,
but also by "informing" the partner what perception you would like to achieve.
You can use the partner's power source. That's why, some many months ago,
we had an argument about language being developed precisely so that one
person COULD see what another was doing by seeing (hearing) what they were
doing, in contradiction to Marken's PCT aphorism number one.

In an interpersonal conflict, as opposed to an intrapersonal conflict,
you precisely DO NOT make that information available to the partner (enemy).
That means that you are likely to need physical force to control your
perceptions. It is only the fact that (often? always?) there are
ways of satisfying the higher-level references of both enemies that allows
them to terminate the conflict. But that is done by "informing" the parties
that such a possibility exists, or by one or both reorganizing or simply
changing mechanisms for satisfying the higher-level perception. If there
were not many ways of acting so as to control any perception, there could
be no resolution of interpersonal conflicts.

What's the point? That the analysis is different when the environment
has power gain, and that the environment always has power gain when it
involves another control system, as it usually does with social animals
like us.

Martin

2

From Tom Bourbon [931005.0906]

[Martin Taylor 931004 19:00]
(Bill Powers 931004.1000 to Hal Pepinski 931003c)

Bill said:

The organism-environment relationship would
be symmetrical if the amplification in the closed loop were
equally distributed around the loop. In fact, the amplification
involved is concentrated almost exclusively inside the organism.
The environment part of the loop almost always introduces
_attentuation_. This places the controlling agency inside the
organism.

Martin said:

This is true when you are dealing with a non-living part of the environment.
It is not true when the critical part of the environment is another control
system, and THAT is the condition for (almost?) all of Hal's discussions.

Hal is talking about interpersonal issues, and under those conditions the
gain in the loop may be equally distributed within and outside "the
organism" (meaning either one of the parties--"you" in what follows).
Now you get into a state in which control of your own perception can be
effected not only by applying substantial force to a dissipative world,
but also by "informing" the partner what perception you would like to achieve.
You can use the partner's power source.

. . .

What's the point? That the analysis is different when the environment
has power gain, and that the environment always has power gain when it
involves another control system, as it usually does with social animals
like us.

A point well taken, Martin. For person (control system) A, some of the
*sources* of disturbances to a controlled variable have a gain of their own,
when the disturbances come from another living control system. But the
*effects* of those other sources still add to all others, animate and
inaminate, to produce the net disturbance on the controlled variable. The
fact that some sources of disturbances are living control systems does not
*always* matter. Whether a cursor I try to control is disturbed by a
computer-generated random random function, or by another person using a
second control handle, often makes no difference: I eliminate the effect
of the net disturbance the same way in either case.

The high-gain properties of the environment *are important* when another
person and I agree to cooperate to affect the cursor, or if one of us decides
to help the other (with or without the knowledge of the one helped), or if
one of us attempts to control the actions of the other, or if I discover
that another person is the source of disturbances and I decide I want to
bring them to a halt, to name only a few possible interactions between two
people/control systems. In any of those cases, the success or falure of the
intended control is affected by the relative gains of the two systems. And
the possibility of conflict is always real.

When I started working on the adaptive control system we discussed recently,
I was attempting to model the fact that people often alter their gain factors
in order to increase or decrease their effect on a variable affected by two
people/control systems. They do that as a way of perceiving what they
intend to perceive in their interaction with another person. In those cases,
a person is deliberately altering what an observer might identify as the
degeree of symmetry in the interaction. The fact that two control systems
interact does not require that the interaction be permanently one of
symmetry, or of asymmetry, as judged by an observer, or that each party wil
always operate with the *same* gain. But however they might *adjust* their
own gains, unless one of them simply "lets go of the stick" and stops
affecting the shared variable, each person/control system remains a
high-gain controller of the kind Bill described.

To Hal, I think these simple models of interacting control systems, and
the tasks in which two people, or a person and a PCT model interact, all
represent small beginnings of PCT research on the questions that concern and
interest you.

Until later,
Tom Bourbon
Department of Neurosurgry
University of Texas Medical School-Houston Phone: 713-792-5760
6431 Fannin, Suite 7.138 Fax: 713-794-5084
Houston, TX 77030 USA tbourbon@heart.med.uth.tmc.edu