Progress in International Relations Theory: Appraising the Field (review)

Is there a rational method for judging theoretical and empirical progress in the recent scholarly work on international relations (IR), conceived of as a subfield of political science? If so, what does it tell us? This book was designed as a collective articulation, application, and appraisal of Imre Lakatos's "methodology of scientific research programmes," his explicit, analytical "metric for theory appraisal" (p. 11). The contributors offer assessments of international institutionalism (by Robert Keohane and Lisa Martin), power transition theory (Jonathan DiCicco and Jack Levy), liberal IR theory (Andrew Moravcsik), democratic peace theory (James Lee Ray), operational code analysis (Stephen Walker), the contest between realist and neoliberal theories of international cooperation (Robert Jervis), neoclassical realism (Randall Schweller), and normative-cum-empirical theorizing about ethnic conflict and minority rights (Jack Snyder). Initiated with a foreword by Kenneth Waltz explaining why he assigned Lakatos to his IR students—"Lakatos's assaults crush the crassly positivist ideas about how to evaluate theories that are accepted by most political scientists" (p. xii)—the volume has an introductory overview and a careful, richly footnoted exposition by the Elmans of Lakatos's and other Lakatosians' "lessons" for IR theorists, as well as somewhat broader concluding commentaries on "Lakatos, and Beyond," by David Dessler, Roslyn Simowitz, John Vasquez, and Andrew Bennett. With the sponsorship of the Belfer Center for Science and International Affairs at Harvard University, and with contributions from three past presidents of the American Political Science Association, this book aspires to be an authoritative statement on its subject.

Does it succeed ? To answer this question, I first need to summarize the Elmans' version of Lakatos's argument, which their contributors are supposed to apply. A boxed "Brief Guide" beginning on page 19 and figure 2–1 on page 33 are helpful in this regard. The Elmans maintain that Lakatos is often cited in an inaccurate, perfunctory manner. They take pains to elaborate and summarize his tolerant, innovation- respecting, logically consistent, methodological criteria for deciding whether a series of theories and theoretical amendments produce "progressive" or "degenerative" problem shifts. Lakatos's "scientific research programs (SRPs)," they say, consist of a number of theories constituted and guided by "hard core" fundamental premises shielded from direct attack by a "negative heuristic," plus a "protective belt" of testable and adjustable "auxiliary hypotheses" derived from the "positive heuristic" of the SRPs. These auxiliary hypotheses, they write, "guide the development of specific theories within the program." Degenerative problem shifts are theoretical amendments of three possible types—those that generate no predictions of "novel facts" (ad hoc₁), those that generate predictions of "novel facts" that are not eventually corroborated (ad hoc₂), and those that generate predictions that are eventually corroborated in accordance with the SRPs' negative heuristic but not with the positive heuristic (ad hoc₃). Progressive problem shifts are either "intra-program," when the theoretical amendment follows both the negative and the positive heuristics of the SRPs, or "inter-program," when the hard core of the theory is amended. Like most of the other key concepts in Lakatos's methodology, "novel facts" are open to several interpretations. The Elmans prefer the Zahar/Lakatos concept of "heuristic novelty," which involves facts that played no heuristic role in the theory's construction (p. 36). They are dismissive of Musgrave's concept of "background theory novelty," which refers to facts that "could not have been expected from the best rival theory available" (pp. 38ff). Musgrave's notion, they believe, is too "permissive."

Beyond such articulations and some reasonably faithful attempts to apply them (resulting in at least partly favorable judgments by all of the direct RSP appraisers), the volume makes a strong contribution to IR in the detailed richness of the sequential accounts of novel facts and theory reformulations offered by some of the best scholars in the field. Too many graduate students read a bit of Waltz, Keohane, Martin, Jervis, and Moravcsik (on realism, institutionalism, or liberalism) or of Alexander George...