[From Bill Powers (980425.0335 MDT)]
Bruce Nevin (980424.1656)--
Rick Marken (980423.1830)--
[interpersonal conflict]
is based on the fact that different individuals want the
_same_ perception in different states.
This is necessarily so for conflict within a control system; it is not
necessarily so and in fact somewhat difficult to prove for conflict between
control systems, such as interpersonal conflict.
It's better, I think, to speak of wanting the same variable to be in
different states, rather than the same perception. When we speak of other
people we can't say anything provable about their perceptions, but we can
observe what they seem to be controlling in environmental terms.
What Colin in east Belfast wants might be revenge for the killing of uncle
Padriac and others, and the nationalist jingoism is a convenient and useful
cover and justification. You can't tell what people are doing just by
listening to what they say they are doing, and here Bruce Gregory's
objection about the names of perceptions is cogent indeed.
But you can observe what they are controlling as manifested in the
environment that you can see, in your own terms. The relationship among
actions, disturbances, and controlled variables is not concealed inside the
other person. The signature of a control system is clear. When things get
complicated -- many controlled variables at many levels -- it may take time
and extensive systematic observations to sort out all the relationships,
but until it's proven otherwise we might as well assume it can be done.
What we observe is that the actions of A interfere with those of B and vice
versa; and we can see that each of them increases the effort of their
respective action even as the other does. From which we conclude that the
interference that B's actions present to A's actions is a disturbance to
some perception or perceptions that A is controlling, and correspondingly
for B. How can we conclude that something is is a disturbance even though
we do not see the disturbance cancelled?
The question here is what model we adopt to explain human behavior. I don't
think it's likely that the human brain switches from one fundamental
organization (control systems) to another (stimulus-response or top-down
cognitive) from one moment to the next. There's a basic physical structure
behind the control organization -- the pathways and neurons have to be
present and functioning. While the brain is adaptable, stuctural
adaptations do not occur in the space between breaths; they take minutes,
hours, days, or weeks to complete. So we really have to settle on one model
and demand that the same model account for all kinds of behavior seen
either from inside or outside the organism.
Perhaps to really verify that we
are observing control (and therefore that the interference is indeed a
disturbance) we have to apply the Test on occasions when the mutual
interference isn't happening, and assure ourselves in some way that control
is continuing after the interference starts. Or perhaps the maximum effort
to resist is sufficient.
Yes, it's perfectly possible to do this. Kent McClelland has started doing
it. You fit a model to the behavior of a single system controlling
something without interference, and then put two of these models together
under conditions that create conflict. In this way you can fit a two-system
model to the behavior of two organisms in a specific conflict. That's just
a sub-category of studies of human interactions done in the same way, as
Tom Bourbon has done with multi-person control. Conflict is only a special
case of interaction; one of the possible things that can happen when
organisms disturb each other.
We see that the actions of A and B interfere with each other because we can
project in imagination what the effects of their escalating efforts would
be without the interference of one another. It is easy to conclude that the
variable we observe at the nexus of their conflict (a variable in the
observer's perceptual universe) is the variable that they are controlling
(a variable in A's perceptual universe, and a variable in B's perceptual
universe), and assert that these three perceptual variables are in some
sense the same thing, or are perceptions of the same underlying reality
(see last point below). But this might not be so.
Real organisms are always controlling multiple variables at the same time.
They can even have multiple conflicts simultaneously. Normally, the whole
collection of control systems will shift its actions until there is minimum
error; it's only in relatively rare cases that the minimum error is too
large to be tolerated. When it is too large, we search for the locus of the
conflict and try to resolve it in some way. Once it's resolved, there might
still be other conflicts that keep the error too high, so we must resolve
them, too.
As always, we are dealing with continuous systems, so "too large" involves
a judgement that becomes fuzzy between the extremes. We can't say there is
or is not a conflict unless it is very obvious or extreme. And we're
dealing with multiple systems; to apply the Test as one day we will have to
apply it, we must record many variables and solve many simultaneous
equations. There are no longer any barriers to handling hundreds of
equations in hundreds of variables, even nonlinear equations; we're not
talking about anything impossible, even if we can't do it right now.
More interesting: it might not have been so when they started, but because
of the conflict the attention of A is focussed on that aspect of the
environment that seems recalcitrant, and similarly for B. So their
awareness of what they are controlling is likely to shift from their
original goal to the point of interference, which is the point of contact
between the two of them, and they may even lose sight entirely of what it
was they originally were controlling.
But it is likely that they go on controlling whatever it was, whether
awaredly or not, at a higher level. As the systems interact, the array of
reference signals shifts as some get larger and others get smaller, so to
an outside observer it may seem that whole control systems are appearing
and disappearing. But in terms of higher-level actions of many systems, the
same things are still being controlled relative to the same reference
levels; only the details are changing as the control systems disturb each
other and the state of the complex environment changes.
An action that would be means for ordinary control of a perception ...
(person A pushing from one side of the road to the other a cart that was
stuck and now at last he's got it rolling)
can be done as though the other were not present
(... across B's direction of walking)
can be construed as a slight to B at which B objects
(Watch where you're going there!)
at which A protests
(Watch out your own self [as if I weren't vexed enough already])
which B takes as an insult, since there the cart stands square in his way
still
(And sure it looks deliberate to me)
and so on, A's reason for pushing the cart and B's for walking in such a
hurry toward the apothecary now almost forgot. There is no "same
perception" in this business of cart-rolling and of walking though there is
conflict where they cross. (Sure ye'll have to be puttin' on an appropriate
way of talkin' now to get this story right.)
You might say, to rescue the matter, that there was no conflict until they
had brought it into focus around the same perception. But even at this
focal juncture you cannot say that they want the same perception in
different states, nor even that they want their individual perceptions of
the same thing (see last point below) in different states, since A is
surely not controlling a perception of his cart staying in the middle of
the road.
That's what's wrong with stating the substance of a conflict in terms of
conjectured perceptions. Much better to say that B's body and A's cart
can't simultaneously occupy the same volume. We can say that A's preferred
perception of the situation is physically incompatible with B's preferred
perception because a physical impossibility is entailed. We don't have to
know how either A or B internally represents the situation, although we can
sometimes make reasonable guesses. We can say that both A and B will
experience abnormally large error, and that (if the output functions are
integrators) their opposing efforts (in terms of the commonly affect
environmental variable) will escalate if the conflict isn't immediately
resolved. We can say that without knowing the internal details of either
system. All we have to do is assume that we're dealing with control systems.
So OK, by this time maybe what they each want is the other to back down and
apologize and neither will do it. So they each control a perception of an
assymmetrical relationship in which one is contrite and the other
righteous, and the conflict is in who is in which role. That would be a
nice observer's-eye view of it. Is it what A and B are severally
controlling? What if A is doing "I'll not take it from these Ulstermen any
more" and B is doing "This will get me off the hook getting that expensive
medicine for old Molly, I've got witnesses how the IRA prevented me, and
how that's why I had to go get a drink with the money."
Conflicts at higher levels can explain why conflicts at lower levels are
not immediately resolved. And there are multiple systems acting at the same
time and influencing each other even when there is no direct conflict. So
all that you say is comprehensible within the HPCT model.
I think the point that Bruce Gregory is obscurely driving after is that
your formulation of interpersonal conflict identically with intra-system
conflict seems to entail that we observers know something about the Real
world behind these perceptions. This is why I have objected to saying that
it is the same perception necessarily. When we have three parties, the two
in conflict and the third observing, none of them can know for sure what
aspect of the interference-laden environment is being perceived by the
other two, however hotly they conclude it.
So, to reach agreement on all this, let's speak of conflicts in terms of
the environmental variable(s) that are being controlled toward different
states at the same time, and speak of the perceptions involved only as
conjectures. It is agreed that we must have some way to estimate how the
variable(s) would be controlled without the conflict, and this can often be
arranged. The Test can be applied to a pair of conflicted systems as well
as to a single system, and we can observe the relationships among the
disturbance, the variable, and the _two_ (or more) actions. Disturbing the
common controlled variable one way will increase the action of one system
and decrease the other, and disturbing it the other way will reverse that
effect. I expect that a satisfactory match of a two-system model to the
observations can be achieved at least in broad terms (there are lots of
possible cases to consider if you want to get into the details).
The same model applies to internal and external conflicts because the same
principle applies (no variable can be in two different states at the same
time) and the conflicting entities are control systems in both cases. All
the apparent differences are only circumstantial; nothing fundamentally
different is happening.
Best,
Bill P.
