Middle East Conflict

Greetings:

I've drafted a short document on a control-theory approach to international
conflict, and posted it on the web at
http://www.indiana.edu/~socpsy/papers/Terrorism.htm .
In some ways, this is very different from recent discussions on CSGnet.
However, it is a PCT approach with some extra understandings, such as: 1.
We perceive and control our feelings as well as sense stimulations. 2. Many
things have to be controlled simultaneously - e.g., our feelings about
ourselves and about our interaction partners, too. A full introduction to
this particular approach to PCT is available at
http://www.indiana.edu/~socpsy/ACT/acttutorial/sentiments.htm .

My international conflict document will be the basis of a talk this summer,
so I would appreciate helpful comments you can offer on improving the
paper. (It seems to me that recently some members are acting as trolls,
just trying to get a rise out of others, and I won't respond to that kind
of message.)

Dave Heise

[From Rick Marken (2002.04.18.1640)]

Dave Heise wrote:

Greetings:

I've drafted a short document on a control-theory approach to international
conflict, and posted it on the web at
http://www.indiana.edu/~socpsy/papers/Terrorism.htm .

My international conflict document will be the basis of a talk this summer,
so I would appreciate helpful comments you can offer on improving the
paper.

Hi Dave --

Thanks for the reference. I think your work has great possibilities. The only
suggestion I would make pertains to the last paragraph on p. 3. You find a
correlation of only .45 between "focal events'" (what I would call
disturbances) and levels of cooperation-conflict (what seem to be the presumed
response to the disturbance). The very low correlation between these variable
suggests that your hypothetical controlled variable (the EPA score) is not
what is being controlled. It may be _related_ to what is being controlled, but
your results show that it is not the variable that leaders control. So I would
change the conclusion in that paragraph. The .45 correlation doesn't support
(or reject) affect control theory but it does reject EPA as a possible
controlled variable. This would be a good way to bring the PCT "test for the
controlled variable" into the discussion.

Best regards

Rick

···

--
Richard S. Marken, Ph.D.
The RAND Corporation
PO Box 2138
1700 Main Street
Santa Monica, CA 90407-2138
Tel: 310-393-0411 x7971
Fax: 310-451-7018
E-mail: rmarken@rand.org

Rick,

Thanks for the response.

First, everyone needs to understand that measurement problems in analyses
of macrosociological dynamics are *severe* so no one is going to obtain the
high correlations that are possible in computerized studies of an
individual's movement of a mouse. Accounting for twenty percent of the
variance typically is considered a definite success in social science.

Second, alternative theories of international relations are not anywhere
near that successful. Social scientists ordinarily find it to be extremely
hard to predict a nation's next event - even in fairly generalized terms.

Third, I took your remark as a challenge, and I re-did the twenty-year-old
study I was reporting, using better statistics and the more powerful
simulation software that is now available. The result is a correlation of
0.69 in the new study. That is, affect control theory predictions account
for almost half the variance in behaviors of nations with each other. I
think it will be a long time before anyone with a different theory matches
this level of explanatory power, achieved by assuming that rulers are
trying to control the evaluation, potency, and activity impressions of
their own and others' nations.

I uploaded a revised manuscript reporting these matters to the same address
on the web, http://www.indiana.edu/~socpsy/papers/Terrorism.htm .

Dave Heise

···

At 04:44 PM 4/18/2002 -0500, you wrote:

[From Rick Marken (2002.04.18.1640)]

Dave Heise wrote:

> Greetings:
>
> I've drafted a short document on a control-theory approach to international
> conflict, and posted it on the web at
> http://www.indiana.edu/~socpsy/papers/Terrorism.htm .
>
> My international conflict document will be the basis of a talk this summer,
> so I would appreciate helpful comments you can offer on improving the
> paper.

Hi Dave --

Thanks for the reference. I think your work has great possibilities. The only
suggestion I would make pertains to the last paragraph on p. 3. You find a
correlation of only .45 between "focal events'" (what I would call
disturbances) and levels of cooperation-conflict (what seem to be the presumed
response to the disturbance). The very low correlation between these variable
suggests that your hypothetical controlled variable (the EPA score) is not
what is being controlled. It may be _related_ to what is being controlled, but
your results show that it is not the variable that leaders control. So I would
change the conclusion in that paragraph. The .45 correlation doesn't support
(or reject) affect control theory but it does reject EPA as a possible
controlled variable. This would be a good way to bring the PCT "test for the
controlled variable" into the discussion.

Best regards

Rick

[From Rick Marken (2002.04.28.0915)]

Dave Heise wrote:

First, everyone needs to understand that measurement problems in analyses
of macrosociological dynamics are *severe* so no one is going to obtain the
high correlations that are possible in computerized studies of an
individual's movement of a mouse. Accounting for twenty percent of the
variance typically is considered a definite success in social science.

But wasn't the theory about the perceptions controlled by individual rulers? I
thought the theory is that rulers use macro actions to control EPA perceptions
relative to their references for those perceptions. There macro actions should
precisely counter the effects of disturbances (what I think you are calling "focal
events") to the EPA perceptions. So, for example, disturbances to Sharon's
controlled EPA perceptions of the PLA (say) should be precisely countered by macro
actions. If you put the right variables into the analysis, the correlation between
disturbances and macro actions should be on the order of .99 (if you have the
right hypothesis about the perception controlled by Sharon).

Second, alternative theories of international relations are not anywhere
near that successful. Social scientists ordinarily find it to be extremely
hard to predict a nation's next event - even in fairly generalized terms.

I bet that if you get a better idea of what these rulers (or ruling parties) are
controlling you will be able to account for well over 90% of the variance on a
regular basis, unless, of course, these rulers are not controlling and are just
reacting to events statistically.

Best regards

Rick