Re:

bob denemark asks about what to do about different theories that predict
the same thing.
he says:
" I highlight predictions of war-pronness
during the period from 2030 to 2050 by Modelski and Thompson, Arrighi,
Wallerstein, and Joshua Goldstein.  Each has different reasons for
making
this prediction, and some of them are contradictory."

first point. two of the most important formulations that predict this
are missing.
Karen Rasler and Bill Thompson _Great Powers and Global Struggle
(Kentucky 1994) have a formulation that differs from Modelski and
Thompson in its focus on the interaction between global-level and
regional-level interstate competition. They are also much more explicit
than Modelski and Thompson in predicting a coming vultnerability to core
war.

The other formulation is Chase-Dunn and Podobnik
http://csf.colorado.edu/wsystems/jwsr/vol1/v1_n6.htm
which is forthcoming in Volker Bornschier and Christopher Chase-Dunn
(eds.) _The Future of  Global Conflict (London, Sage 1999).
the C-D and Podobnik model is a synthesis, but the structure of the
argument is clearer than in many of the other formulation (if i may be
so bold).

What to do about different theories that predict the same thing.  One
possibility is that the theories are not really different. Perhaps they
are using different words to mean the same thing.  Wallerstein and
Arrighi call it hegemony while Modelski calls it world leadership.
Modelski and Thompson have now included economic power in there
conceptualization of leadership and Arrighi has included ideological
power. So the differences have diminished.  And does it really matter
for the problem of future conflict?

After examining conceptual differences, we need to look at proposed
causal mechanisms. Here a propositional inventory is required.  The
formulations do indeed differ importantly in this regard, and this is
perhaps more important than the similarity of prediction. It is quite
possible to be right for the wrong reasons.

It is not possible to wait 50 years to find out which theory does
better. And we have not yet figured out how to study the future. So the
best thing would be to test these different theories against the past.
or to evaluate them with tests that have already been done. Brian
Pollins (APSR 90,1 March 1996) looks at how several of these stack up in
predicting changes in the number of all kinds of war.  Pollins finds
that treating hegemony and the K-wave separately improves the prediction
of warfare.

My unfunded NSF proposal used a different approach.  see
http://csf.colorado.edu/wsystems/archive/papers/c-d&hall/warprop.htm

hope this is helpful.
chriscd