ABSTRACT
Human political behavior is patterned: Certain sequences of behavior occur on multiple occasions under predictable circumstances. Computational models and other artificial intelligence techniques provide a rich set of alternative definitions of pattern that may prove more effective in modeling international behavior than the statistical patterns studied to date. The essence of pattern is repetition or regularity: Something is "patterned" when one can infer unknown information about an object on the basis of partial information about it. International relations theory has also begun to reemphasize the importance of nonstatistical patterns. The importance of sequences in determining international political behavior is discussed in general terms, emphasizing that in a complex system such as international politics, pattern recognition is a more efficient use of human cognitive abilities than deductive approaches. Sequences of international events are useful in understanding international politics because they can be employed as analogies.
