Raise topics of philosophy of science and meta-theory around scholars of international security, and most will begin checking their phones. It is fair to say that a majority of the security studies guild finds questions about research programs and epistemology irrelevant and a distraction. That is a pity. As Fred Chernoff makes clear in this important book, philosophy of science has a great deal to say about a question that all students of security—all social scientists—should care about: Are we making progress? Can we explain more than previous generations could?
The question has arisen before, of course, and most IR scholars have some familiarity with what the works of Thomas Kuhn and Imre Lakatos imply about their field. But Chernoff takes a fresh approach to the question of progress by conducting exhaustive surveys of three important literatures—on nuclear proliferation, alliances, and democratic peace—and teasing out a different kind of answer.
Chernoff reformulates the question, asking: Why are so many decades-old debates in security studies unresolved? The debates he has in mind are not over paradigms. Chernoff has no interest in thawing the frozen conflicts between realism and liberalism, or rationalism and constructivism. He has in mind, rather, the intractability of so many empirical puzzles. After nearly fifty years of study, for example, why can we not agree on why states form one set of alliances rather than another, or ally at all?
In three chapters that could serve as introductions to graduate students on diversity in IR, he selects ten or eleven highly cited authors in each of the literatures, summarizes the arguments and methods of each, points out their similarities and differences, and offers some critiques of individual works. Some of the works are quantitative, others qualitative; some are realist, others liberal; some are rationalist, others constructivist.
The chapter on nuclear proliferation, for example, surveys ten amply cited works of scholarship, starting with George Quester’s 1973 book The Politics of Nuclear Proliferation, running through work by Kenneth Waltz, William Potter, Stephen Meyer, Etel Solingen, Scott Sagan, T. V. Paul, Sonali Singh and Christopher Way, Jacques Hymans, and Dong-Joon Jo and Erik Gartzke. Chernoff finds wide disagreement across these authors over how best to explain proliferation, with some holding to a strategic hypothesis about external threats, and others arguing for bureaucratic variables, domestic politics, national identity, or the characteristics of decision makers. Policy makers could use closure on this academic debate, but none is forthcoming.
One might think that empirical impasses such as this are caused by prior theoretical disagreements, especially between realists and non-realists. Chernoff certainly recognizes the importance of theory, and in his final chapter thoughtfully considers the possibility that scholarly stalemates are produced by lack of agreement over the precise question being asked, which in turn could be tied to theoretical diversity.
His own answer, however, looks to philosophy of science. Where security scholars cannot agree on the best explanation of a phenomenon, it is in large part because they do not agree on what constitutes a good explanation to begin with. Few security scholars explicate criteria for an adequate causal argument. (This seems to irritate Chernoff, but surely economists, geologists, physicists, and others in more progressive sciences are similarly opaque regarding philosophy of science.) In any case, Chernoff performs a service by carefully extracting such criteria from the texts themselves. Some of the authors explicitly or implicitly stress empirical adequacy, others falsifiability, others predictive power, others simplicity, others the explication of causal mechanisms, others the elimination of alternative explanations, and the list goes on. What scholars present as theoretical or empirical debates, then, are to some extent philosophical debates.
Chernoff takes care to follow a sound research design himself. His three cases vary along the dependent variable: He finds more consensus in the literature on alliance formation, and finds near-convergence in the democratic peace literature. That is, Chernoff asserts that most scholars accept the proposition that liberal democracies do not fight one another. He acknowledges that no such consensus exists regarding the causes of the democratic peace, and for many scholars that means that the most important question is still contentious. Chernoff's cases are too few for statistical significance, but the results are suggestive: The more scholarly consensus on an empirical question, the more scholars agree on philosophy of science. The point is made clear in the concluding chapter, which presents some simple quantitative analysis of his data from the three literatures.
Still more interesting, scholars who favor one kind of answer to a puzzle tend to employ similar criteria for what makes an explanation adequate. The nuclear proliferation debate is bedeviled by a wide divergence of criteria. Those who favor realist arguments strongly tend to favor explanations that locate “true causes” and apply to a broad range of cases. Those who favor non-realist arguments favor instead “deep causes” and falsifiable hypotheses. Such a divergence is far less evident in the democratic peace literature: There, realists and liberals alike claim to seek empirical adequacy, “true causes,” falsifiability, and so on.
No book can do everything, and Chernoff’s analysis raises some questions that it does not answer. First, Chernoff’s precise question—why no scholarly consensus on puzzle x?—turns on one particular notion of scientific progress, namely what Chernoff calls “approach to consensus” or the withering away of disagreement over how to explain phenomena of interest to the scholarly community. As he notes, however, some philosophers say instead that science progresses when it is able to explain more facts, make better predictions, or give people greater control over the world. Chernoff’s canvassing of leading positions among philosophers of science on progress—Popper, Kuhn, Lakatos, van Fraassen, et al.—is a tour de force, but he does not give enough reasons why we should favor his own “approach to consensus” view over the alternatives. As Galileo, Newton, and Einstein knew, consensus does not always mean progress.
Nor does Chernoff show that security studies is any more divided as regards empirical puzzles than economics, psychology, or the natural sciences. Consider the heated debate in evolutionary biology between E. O. Wilson, who argues that selection happens at the multiple levels, including groups of organisms, and Richard Dawkins, who counters that selection takes place only at the genetic level. Or consider the interminable (and highly policy-relevant) debates among economists over how best to pull an economy out of recession. Even if these sciences do enjoy more consensus than these cherry-picked examples suggest, one is left wondering if their practitioners agree more than security scholars do as to what makes for a good explanation.
But these are questions that emerge from an ambitious and well-executed study. They do not damage Chernoff’s claims, but rather are evidence that his is an original book pursuing a fruitful line of inquiry. The chief lesson is that, insofar as security scholars want to close the book on persistent empirical puzzles, we should have a sustained collective conversation, not so much about methodology, but about the assumptions that ground our methods.