The Complexity of Cooperation: Agent-Based Models of Competition and Collaboration; System Effects: Complexity in Political and Social Life
These two books represent recent efforts to apply complexity theory to international relations. Axelrod, who is famous in game theory circles for his iterated solution to the Prisoner's Dilemma, has been at it the longest. This volume, which reprints previously published articles with new introductions, shows how so-called genetic algorithms can be applied to everything from sexual selection in biology to the evolution of international cooperative norms. The Jervis book sees the international system as a complex one in which positive and negative feedback loops tend to produce unexpected second- and third-order effects. The international system, in this view, is much more than the sum of its nation-state parts, and cannot be adequately described by highly reductionist systems-level theories like that of Kenneth Waltz.
The application of concepts like complex adaptive systems, arising from quantitative fields like biology and information theory, to the study of international relations is overdue. The assertion of realist theory that anarchy prevails in the absence of an international sovereign is much too simple: many forms of order, both natural and social, evolve in decentralized systems lacking any form of hierarchical authority. Axelrod's book gives a number of examples of this, from international business and military alliances to the emergence of new actors over time. What is much less clear is whether the complex adaptive systems approach can offer much in the way of a positive theory of international relations. Such models tend to work best in situations with large numbers of relatively simple agents, as with biological populations. International politics consists of a small number of very complex agents whose behavior tends to be more chaotic. And indeed, when Jervis leaves theory for the real world, his book reverts to a more familiar historical, case-by-case approach.