While much has been written about deterrence theory, satisfaction with it is very uneven, which suggests we might better leave it alone. However, deterrence remains important and many issues turn on whether it can be a reliable instrument of statecraft today. It helps to consider how useful the theory can be, and the place to start is with rationality because much recent discussion about deterrence, weapons of mass destruction (WMD) proliferation, and other aspects of security has been stimulated by concern about actor rationality. Thus the subject of this chapter (and the next) must be bearded with energy, though enthusiasm comes hard. Deterrence theory has been called one of the most influential products of the social sciences (Achen and Snidal 1989). This is true but misleading. In various ways it is also a case study in how not to design a theory, test it, or apply it.

It is fairly clear how the theory came to be closely associated with rationality. The Cold War put the emphasis in thinking about deterrence on competition and strategy, on making the best moves. The very high stakes reinforced this. We needed instruction on how to do our best, to be rational, to avoid disaster but not lose. Finally, nuclear weapons and the Cold War seemed to simplify matters greatly, to force on actors a set of overriding preferences which could serve as the starting point for a rational decision approach.

With this in mind, the theory has been expected to perform three services.