The long-standing debate over the foundations of international relations scholarship heated up after the Cold War and intensified with the publication of Alexander Wendt's metatheoretically rich Social Theory of International Politics (1999). Proponents of strictly empirical methods have pursued statistical and rational choice approaches. Their critics contend that the social construction of the objects of study must be taken into account, with some arguing that the latter process undercuts anything remotely resembling scientific-style inquiry. And there are many positions in between.

Milja Kurki's book addresses this debate, seeking in particular to show how scholars in IR and other social sciences should understand the concept of “causation.” Kurki examines uses of causation through the history of Western philosophy, their importation into IR scholarship, and the ways that importation has, in her view, distorted IR theorizing and limited IR research. The book has three parts: The first discusses the Humean philosophy of causation and its legacies (four chapters); the second promotes reconceptualizing causation through “deepening” and “broadening” (two chapters); and the third considers the implications of the previous discussions for causal inquiry in IR (two chapters).

Kurki says that the book aims not to develop a substantive theory of international politics but, rather, to improve philosophical understanding of the social world and theories of it. The book is thus fundamentally a work in the philosophy of social science aimed at an IR audience. As such, the author advocates (e.g., pp. 10, n. 26; 183) a particular philosophical doctrine, a version of Roy Bhaskar's “critical realism” (CR), although she also offers emendations of it. Bhaskar's position claims a strong parallel between the natural and social sciences based on the socially grounded nature of each, and develops an ontological distinction (pp. 163–64) between three levels of being—“empirical,” “actual,” and “real.” Although Bhaskar's work has not stirred much reaction in the academic philosophy of science, a number of political scientists and sociologists have enthusiastically embraced and applied it.

There is much that is valuable in Kurki's book, including the explication of CR; the weaving together of the tenets of Bhaskar, Margaret Archer, Andrew Collins, Rom Harré, Edward Madden, and Andrew Sayer; the illustration of their connections to IR authors such as Alexander Wendt, David Dessler, Heiki Patomäki, and Colin Wight; the mapping of Aristotle's familiar fourfold typology of causes; and, above all, the critique of Bhaskar's position by various critical realists, including Kurki herself. Although her dialogue with critics outside of CR is limited, her investigation of debates within CR is thorough and illuminating.

This interesting volume does, however, encounter difficulties at a number of points. Part I, which focuses on the history of philosophy, takes a surprisingly dismissive tone toward philosophers whose views do not fit with CR, which is to say, almost all of them. This severely limits engagement with their ideas. For example, following a long quotation (p. 36) from Hume's Treatise, Kurki identifies “Humean assumptions”—a term used throughout the book. But the statements she cites, so far from being assumptions, are rather key conclusions derived from a powerful, extended argument in the Treatise. However one may disagree with the argument, it is regarded as a major contribution to the history of philosophy and cannot be treated as an assumption. This misstatement is symptomatic of the lack of regard displayed for major philosophers, who are consistently depicted as offering or accepting mere “assumptions,” “taken-for-granted” beliefs (p. 191), and “prejudices” (p. 226). The history of Western philosophy, on Kurki's description, appears to have been largely devoid of logically reasoned argument prior to the emergence of the critical realist philosophy of science.

Furthermore, given Part I's emphasis on the dangers of Humeanism, it is surprising that Kurki never engages Hume's core argument. One would expect an IR book focusing on the influence of Hume's account of “cause” and endorsing Bhaskar's claim that causes are “naturally necessitating” (p. 231) to attempt a serious analysis of Hume's argument against causal necessity (since the sweeping implications of Humeanism take on quite a different aspect if Hume is right). Beyond citing Bhaskar's so-called epistemic fallacy, Kurki adds only that ordinary language usage of “cause” conforms better with Aristotle's far broader conceptualization than with Hume's (p. 189). But the fact that Hume's analysis conflicts so starkly with ordinary usage, and discredits the latter for philosophical purposes, is precisely what made his contribution a landmark in the history of philosophy.

Similarly, Kurki rejects many philosophical views chiefly because they violate tenets of CR. Pragmatism and various other doctrines are discarded once they are shown to entail some form of CR-incompatible relativism. But she shows neither that CR is right nor that the form of relativism under discussion is untenable.

Some of the characterizations of philosophical positions miss their marks, including Kurki's depiction of Hume (pp. 33–40), the major figure of Part I. She mistakenly depicts him as holding that all knowledge comes from experience and that knowledge never achieves certainty. (Kurki paraphrases him as asking, on p. 34, “given that we cannot know anything for certain, how we can justify science and knowledge?”) But Hume's account of mathematical knowledge attributed to it both certainty and a nonobservational basis in “relations among ideas” (Hume's famous “fork,” Treatise IV, 1, 20–21). This is a major error, since the certainty available in the mathematical knowledge allows Hume to contrast the flawed basis for claims about necessary connections attributed to causal relations. Inaccuracies occur throughout the first half of the book.

Part II pursues a broadening and deepening of the notion of causation. The broadening is needed, in Kurki's view, because IR uses an excessively narrow notion of it, confined to what Aristotle termed “efficient causation.” On this point, Kurki especially attacks Kenneth Waltz. However, most IR theorists today offer structural theories, and Waltz, in particular, very clearly defines “system” and “unit”; he argues that systemic influences should properly be designated “causal” (Man, the State and War, pp. 233–34, 1959). Thus, Waltz claims that both active (precipitating) causes and permissive (structural) causes shape outcomes and are properly causal. Kurki dislikes Waltz's skepticism regarding the “reality” of structures. But his attitude there does not turn his theory into one that expunges all but efficient causes.

Kurki's goal in Part II is to unify rationalist and reflectivist (that is, antiscientific, interpretivst) approaches. She notes that many forms of research conducted in IR are, to varying degrees, valuable and legitimate. She thus joins many other metatheorists in endorsing methodological pluralism. She says (p. 196) that her analysis “by no means entails that Humean research or reflectivist theories should be thrown out as inadequate.” Both forms are acceptable. (Even poststructuralists may be surprised to find [p. 19] that they receive “scientific legitimacy,” despite her later criticisms.) But she says that reflectivist and constitutivist theorists should “reclaim” the causal label. The unification thus proceeds essentially by relabeling as “causal” whatever various social theorists have claimed to affect behavior.

Part III advances the argument that CR improves IR theorizing and research. Some authors (myself included) have questioned whether and how it does so. While Kurki does a better job than others on this score, her argument ultimately comes up short. For example, in Part III she says that Robert Cox's IR theory meets her standards—“[d]espite his aversion to causal terminology and lack of engagement with philosophically realist philosophy” (p. 260). This is a big “despite.” Similarly, in her analysis of democratic peace literature, Kurki lauds the work of Tarak Barkawi and Mark Laffey, despite their nonendorsement of CR. If her methodological standards can be met without broadening, deepening, or the highly contestable metaphysics of CR, then it is difficult to see the value of the preceding chapters.

Kurki's wide-ranging volume thus exhibits some serious shortcomings. All the same, this ambitious first book clarifies principles of CR while also demanding that the reader confront important and increasingly prominent questions in IR metatheory. The book highlights the consequences of the ways that social scientists understand “causation” and underscores the relevance of metatheoretical questions for scholars who seek to show the connection between theory and practice, which is often overlooked in metatheory debates. Although there are difficulties with the exposition and argument, as noted, Causation in International Relations displays Kurki's analytical talent, and signals the possibility of much good work to come.