Inspired by a seminal essay of Albert O. Hirschman, as well as by the ongoing debate on the empirical foundations of social science, this article “revisits” (1) the paradigm concept popularized by T. S. Kuhn in the 1960s and (2) the relationship between probabilistic and “possibilistic” modes of theorizing that has acquired renewed relevance in comparative politics mainly with respect to recent theories of democratization and development. It does so by reviewing three major paradigm crises in modern political science: the shift from the Aristotelian polls to the social “system,” the refocusing of political explanations from the social to the global environment, and the contemporary attempts to reevaluate the role of technology in political change. The review takes stock of the record of the discipline of comparative politics, of opportunities provided by paradigm shifts, seized upon or missed by the discipline. It also allows one to seek a more even balance between the potential utility and limitations of the paradigm concept, while at the same time pointing to the perils of divorcing the art of the possible from the laws of probability.

Stephen D. Krasner's article in this journal in 1976, “State Power and the Structure of International Trade,” defined the agenda for years of scholarship by being both lucid and problematic. Krasner presented a clear puzzle but manifestly failed adequately to answer the questions that he raised. His key proposition, that strong international economic regimes depend on hegemonic power, was supported by only half of the six cases that he discussed. Yet the cogency of Krasner's formulation of the argument, and the pungency of his rhetoric, led “State Power” to serve as a focal point in a coordination game among three major constituencies in the international political economy field. Liberal transnationalists, statist realists, and their audiences all benefited from Krasner's lucid specification of the issues. As a result of research prompted by Krasner's article, we understand the relationship between international political structure and economic openness much better than we did before it appeared.

Robert Jervis's article “Cooperation under the Security Dilemma” is among the most important works in international relations of the past few decades. In it, Jervis develops two essential arguments. First, he explains that the security dilemma is the key to understanding how in an anarchic international system states with fundamentally compatible goals still end up in competition and at war. The security dilemma exists when “many of the means by which a state tries1 to increase its security decrease the security of others.” It provides the rational foundation for what Jervis termed the “spiral model,” which describes how the interaction between states that are seeking only security can fuel competition and strain political relations.2 Second, Jervis explains that the magnitude and nature of the security dilemma depend on two variables: the offense-defense balance and offense-defense differentiation.3 As a result, the security dilemma can vary across space and time. Although states exist in a condition of international anarchy that does not vary, there can be significant variation in the attractiveness of cooperative or competitive means, the prospects for achieving a high level of security, and the probability of war.