(Page 19 of 22) - Bayesian Decision-Makers Confront the Power Cycle: A Formal Model authored by Doran, Charles F.. and Doran, Kirk.

decision-maker in general does not learn from history and past mistakes that have led to horrible
consequence such as major war but even worse, for some governments, to military defeat and political
subjugation.

Very little true learning seems to occur in international relations over time, and from the

experience of others.

This is especially true because most aggressors believe that the worst will not

happen to them. Does this formalization shed any new light on this failure to learn?

A clear conclusion stems from the formalization of this aspect of power cycle theory. Over time, as the
state moves up or down its power cycle, each succeeding decision density function obtains a sharper
peak. Each forecast made by the government is reinforced by structural change that seems to confirm the
extrapolation of foreign policy role and security into the future. That each subsequent density function is
sharper than its prior indicates that the degree of confidence of the government in the forecast has
increased. This seems to be a form of learning with immediate reinforcement for the validity of each
decision forecast. Where is the problem?

The problem with this form of so-called learning is that the learning is based on an erroneous
understanding of decision reality.

Unfortunately that is also the most salient reality the decision-maker

possesses regarding the foundation of future foreign policy conduct. Power has increased and will
continue to increase. The sharpening peak of the decision density function concerns a perspective that
will continue to appear correct until the critical point is reached. At that point everything changes, the
assumptions underlying the Bayesian estimates are proven wrong, and a wholly new understanding of
structural conditions dawns on the government.

The problem regarding learning on the power cycle is this. As the formalization shows, what is
reinforced, or “learned,” is that which in the end is demonstrated to be utterly wrong. At the same time,
there is no potential to adopt a different perspective that is “right” because no reliable basis exists for any
other kind of knowledge about changing structure and its impact on foreign policy conduct. As this
formalization so strikingly conveys, extrapolation of future foreign policy behavior, and the expectations
underlying that behavior, is only reinforced in terms of Bayesian estimate regarding a set of perceptions
that ultimately proves to be completely fallacious; hence the shock of the critical point on all of foreign
policy behavior.

Although this formalization deals with only a single inflection point, a similar logic can be applied to the
other critical points, and the formalization can be generalized.

The argument made here is that whatever

the actual uncertainty associated with passage through a critical point for governments in world history,
the decision logic outlined by this formalization reveals why such passage, generating an actual
discontinuity in future foreign policy expectations about role and security, is so troubling for the state
undergoing this passage and for the other states in the system as they attempt to cope with the resulting
decision instability.