facilitation and interference (was cooperation and competition)

CSGnet Archive CSG2000
1

[Martin Taylor 2000.07.03 09:45]

[From Bill Powers (2000.07.02.20953 MDT)]

>Anyway, "facilitation" it is, as far as I am concerned, just to ease
>communication. "Facilitation" occurs when the effects of A's actions
>(intended or side-) make it easier for B to control.

Right, and there are so many ways to do this that the category is really
pretty arbitrary
.... However,
since you find "facilitation" to be a usable term, I could suggest
"interference" for an equally broad antonym referring to an objective
relationship (by which I mean a relationship that can be observed whether
intended by someone or not).

I could live with "interference" as the opposite of "facilitation,"
though I don't think its connotations are much better. It does have
the advantage of not implying reciprocity.

I don't think either category is arbitrary at all. There aren't "many
ways to do this". There are three for "facilitation", four for
"interference". "Interference" has one possibility that "facilitation
does not:

   --A disturbs B (if intentional, this may become "conflict").

Here's what I think is the _complete_ list--add more if there are more:

   -- A's actions influence the way in which B's output can affect B's CEV.
   -- A's actions influence the way B's sensors react to B's CEV
   -- A's actions influence B's power supply.

Here's a diagram (???) indicates places where A might have an
influence, and shows why I think the list above is complete.

                 > / Power Supply (from the outer world)
                 > /
                 > ???
      -----------o------- /
     > > /
perceptual output
   input function
  function |
     > >
====^===================V====
     > >
    ??? |
     > >
     +--------???--------
     >
    ???
     >

There's one other apparent possibility that some might think belongs
here, but it really doesn't:

   * A's actions influence B's reference signal value

(I use the * in the way writers on linguistics do, to indicate a
wrong or improper example).

The only way A can affect B's reference signal value is by affecting
some higher level output in B that affects the reference value in
question. A can do this _only_ by one of the four routes above,
typically by disturbing some perception whose value is part of a
higher-level controlled perception in B. So it is not an extra way
for A to facilitate or interfere with B's control. In fact it doesn't
even affect B's ability to control. Moreover, it requires a hierarchy
in B for this to have an effect, and we are talking only about
elementary control units.

The main thing I'm interested in here is to distinguish intentional from
unintentional consequences. I think that intentional consequences are far
more likely to occur than unintentional ones.

I think you need to re-think that one!

We act in a world in which many living control systems (not only
human LCSs) sense and could act on the same entities in the
environment. In the absence of facilitation, we would be endlessly
disturbing one another, and in many cases conflicting with one
another. We don't, not to nearly the extent that a randomly organized
set of LCSs would do in that environment.

But neither do we _intentionally_ refrain from actions whose main
effects would control our perceptions but whose side effects would
disturb other LCSs perceptions. We have reorganized so that not only
do we refrain from many of the possible disturbing actions, we act so
that the side effects of our actions are often (unintentionally)
facilitatory. Our reorganizations have created social structures in
which unintended facilitation is the norm.

I may work for a salary that gets us food, the side effect of which
is that someone (say Dora) quite unknown to me has an environmental
feedback path available that would not otherwise be there--Dora might
find a car available to buy, if my work is on an assembly line. She
might be able to fly across an ocean, if my work is as an airline
pilot (or plane designer). Or she might be able safely to store her
salary in a bank, if my work is as a bank auditor.

Not knowing of Dora's existence, I have no "intention" to facilitate
her perceptual control, but the side effects of my own perceptual
controls do exactly that. What Dora does to facilitate my perceptual
control, I don't know. Perhaps she pays some money to my employer,
which allows him to pay me, which allows me to get food. Perhaps she
is an engineer who designs better computers that allow me to produce
the graphic images I enjoy. I don't know what she does, and she
doesn't know I exist, but unless the car she buys (partly facilitated
by my acts on the assembly line) collides with mine, she is more
likely to facilitate my control than to interfere with it.

Of course, she may (unintentionally) interfere with my perceptual
control, too. She may buy the last available kiwi fruit that I wanted
for a party. Her car does emit CO2 that influences the weather in
which I must operate (so does she).

If people intend to
cooperate, cooperation is more likely than if they don't know how to
control for cooperation, or control for the opposite. They can still
accidentally facilitate each others' control processes (or do it one-way),
but the chances of repeating the facilitation are then very low.

Not at all. You are talking about 1 to 1, almost "contractual,"
facilitation, which probably _is_ rare on an accidental basis. But
every time I act according to social norms, I am facilitating the
actions of thousands or millions of other people, as compared with
the situation if I didn't exist, or if I chose the most direct way of
controlling my perceptions--such as by getting money at the point of
a gun, or barging through a crowded sidewalk, or taking the last
goody off a table.

At a much lower level of perceptual control, every cell in my body
controls a myriad of perceptions that have side-effects of producing
chemicals that facilitate the control capability of many other cells,
and I'm not about to concede that each of them has the intention of
cooperating with the others, any more than I will concede that bees
or ants have the intention of cooperating (even thoguh they do
co--together--operate.)

We are talking here about why we have social structures, at all
levels from intra-cellular through fruiting bodies, packs and tribes,
to nations and religions. I don't think you can evoke "intentional
cooperation" to explain why the entities that operate together to
constitute such structures do in fact co--together--operate.

Martin

2

[From Bill Powers (2000.07.03.1149 MDT)]

Martin Taylor 2000.07.03 09:45--

I don't think either category is arbitrary at all. There aren't "many
ways to do this". There are three for "facilitation", four for
"interference". "Interference" has one possibility that "facilitation
does not:

With your diagram, I now see what you're talking about. I have to agree.
(That's the best kind of agreement, meaning the most sincere: I've been
backed into a corner and the only way out is to accept the truth of what
you say).

We act in a world in which many living control systems (not only
human LCSs) sense and could act on the same entities in the
environment. In the absence of facilitation, we would be endlessly
disturbing one another, and in many cases conflicting with one
another. We don't, not to nearly the extent that a randomly organized
set of LCSs would do in that environment.

But neither do we _intentionally_ refrain from actions whose main
effects would control our perceptions but whose side effects would
disturb other LCSs perceptions. We have reorganized so that not only
do we refrain from many of the possible disturbing actions, we act so
that the side effects of our actions are often (unintentionally)
facilitatory. Our reorganizations have created social structures in
which unintended facilitation is the norm.

Very clear, and probably true of most interactions with unidentified
"others." However, this does support my argument that cooperation, or
intentional mutual facilitation, is not likely to happen by accident. It is
a specific mode of control in which a person deliberatly tries to maintain
a state of mutual facilitation with specific, known, other people: for
example, by discussing the rules of a game or of parliamentary procedure
prior to engaging in collective action -- one of Clark McPhail's conditions
for collection action.

Behind your argument is another that probably should be made explicit to
give this thesis full force: we reorganize not in order to facilitate other
people's control process but to facilitate our own. We come to facilitate
others because of what happens to our own control processes if we don't.
This process is not directed by consciousness, of course: reorganization,
as I see it, is a wholly automatic process driven only by the built-in need
to satisfy intrinsic reference signals.

You are talking about 1 to 1, almost "contractual,"
facilitation, which probably _is_ rare on an accidental basis.

Exactly. Being explicitly controlled for, cooperation as I define it is far
more likely to occur than an outcome that is only a statistical outcome
rather than a controlled one.

But
every time I act according to social norms, I am facilitating the
actions of thousands or millions of other people, as compared with
the situation if I didn't exist, or if I chose the most direct way of
controlling my perceptions--such as by getting money at the point of
a gun, or barging through a crowded sidewalk, or taking the last
goody off a table.

This is the general, statistical, population trend, I can agree to that.
But because it is statistical (being the outcome of a random process of
reorganization) we must expect to see many exceptions. Many of a person's
interactions with other people will interfere with their control rather
than facilitating it, even though the overall effect, taking all others
into account, is to minimize their interference with that person's control
processes. In a completely determinate social system, with reorganization
always being based on correct rational decisions and with a background
noise level of zero, we would expect the whole social system to converge to
a state of maximum possible facilition. But reorganization is random, not
optimally rational; in addition, the world is full of variations that are
not directly tied to people's immediate control problems, so even with the
best possible control strategies, there will be an irreducible level of
error (both intrinsic error and systematic, hierarchical, error). And even
in the fully determinate system, the maximum achievable mutual facilitation
will not be the maximum conceivable, because there are probably more
variables than equations and thus no single determinate solution.

There is a possible theorem that I proposed long ago (in that paper on
social systems I did for Don Campbell). It is that the greater the number
of people with whom an individual interacts, the more the disturbances from
those other people will tend to average out to zero. This is because on
many scales, people choose their reference conditions at random (as far as
anyone else knows). No, that's not the whole story -- we have to remember
that each person is in the middle of controlling a lot of different
variables by different means (different means even when controlling similar
variables), while occupying different positions relative to the environment
and the forces in it. We are disturbed only by the _actions_ of others, not
by their intentions. Thus there is a large amount of variability in the
effects the actions of others have on us -- even when they're nominally
controlling for similar things. It's that large component of variability
that would tend to make the sum of all disturbances from a large number of
other people add up, if not to zero, to some vector sum much smaller than
the average absolute disturbance. If you want to be left effectively alone,
move to a crowded city.

We are talking here about why we have social structures, at all
levels from intra-cellular through fruiting bodies, packs and tribes,
to nations and religions. I don't think you can evoke "intentional
cooperation" to explain why the entities that operate together to
constitute such structures do in fact co--together--operate.

No, but neither can you rule out intentional facilitation -- that is,
cooperation -- in some social structures. Bees probably have no Bill of
Rights, but some groups of human beings do (in one or another explicit
form). An explicit system of controlling social variables is probably, in
principle, more effective at promoting social harmony than one arrived at
by random variation and selective retention, to bring Campbell in once
again. How the random approach fares in comparison with the systematic
approach depends, of course, on the quality of the systematic approach you
have in mind. And we have to keep in mind that the systematic control of
social variables must begin with the processes of random reorganization.
But the outcome of the random process is to supplant the random process
with a systematic one that doesn't allow the intrinsic errors to occur again.

Sometimes, Martin, your systematic classifications drive me nuts. This
time, they have taught the old man something. Thanks.

Best,

Bill P.

3

[From Bruce Gregory (2000.0703.1725)]

[From Bill Powers (2000.07.03.1149 MDT)]

With your diagram, I now see what you're talking about. I have to agree.
(That's the best kind of agreement, meaning the most sincere: I've been
backed into a corner and the only way out is to accept the truth of what
you say).

Duly noted.

BG

"Either the wallpaper goes, or I do."
--Oscar Wilde, last words

4

[Martin Taylor 2000.07.04 11:09] Happy Independence Day, Americans!

[From Bill Powers (2000.07.03.1149 MDT)]

Martin Taylor 2000.07.03 09:45--

>I don't think either category is arbitrary at all. There aren't "many
>ways to do this". There are three for "facilitation", four for
>"interference". "Interference" has one possibility that "facilitation
>does not:

With your diagram, I now see what you're talking about. I have to agree.
...

I'm glad. And I appreciate your extensions and elaborations.

Martin