Martha Finnemore, The Purpose of Intervention

Finnemore's topic is changes in the purposes of international intervention. In the nineteenth century it was not unusual for a great power to send gunboats to a hapless debtor country, particularly in Latin America and the Caribbean, to take over a customs house and coerce repayment. This is no longer done. It was also fairly common for European powers to intervene to protect coethnics and coreligionists in Asia. Today, interventions are intended to vindicate not the rights of Christians or whites, but human rights, or claims inherent to all human beings. Finally, interventions to preserve or restore international order in the early nineteenth century typically suppressed liberal revolutions and reformations; today they typically have something like the opposite purpose.

For constructivists, who regard norms and intentions as central to analysis, the importance of these shifts in purpose is self-evident. The changes are also noteworthy to rationalists, not least because they imply that over time, of the (nearly infinite) set of all possible interventions, great powers have chosen certain types in the past and certain other types more recently. But we need constructivism to explain which types. Consider the interventions to collect debts. In 1838 Mexico owed various French nationals money. French troops landed at Vera Cruz to exact repayment (p. 28); outrage, at least in Europe, was minimal. In 1995 Mexico was on the brink of defaulting on its massive foreign debt; neither France, the United States, nor any creditor country even considered sending troops to collect. Either states are no longer self-interested—a claim few would take seriously—or they secure their interests through different means; in other words, a rationalism that ignores norms cannot say why some potential interventions do not happen:

Here, it seems, social costs matter to actors, and hence a full account of changes in great-power intervention over time must attend to the sources of changes in social costs.

The same is true concerning intervention to help oppressed people. In the early nineteenth century European powers used force to relieve fellow Christians suffering at Muslim hands. Russia helped Greece, a fellow Orthodox Christian people, attain independence from the Muslim Ottoman Empire in the 1820s. In 1860 France led a great-power expedition to aid Maronite Christians in Lebanon and Syria. In 1877 Russia sent troops to Bulgaria to defend Orthodox Christians against the Ottomans. Finnemore also discusses an interesting nonintervention: between 1894 and 1917, when Armenians suffered severe Ottoman persecution, no European state used force on their behalf; Armenians were Christian, but as Monophysites they were not in communion with any European churches, including the Russian Orthodox.

Since 1945, by contrast, great powers have frequently intervened to help religious and ethnic groups other than their own. In addition, it seems that these days humanitarian intervention must be multilateral to be legitimate; unilateral interventions such as that of India in East Pakistan in 1971, Tanzania in Uganda in 1979, and Vietnam in Cambodia in 1979 had humanitarian effects but were justified in terms of security. Recent interventions presented as humanitarian have all been multilateral. These include interventions to protect Kurds and Shia Arabs in Iraq in 1991; to end the Cambodian civil war; to halt starvation in Somalia; and to protect Muslim civilians in Bosnia-Herzegovina in 1995 and Kosovo in 1999. Unless one makes the dubious empirical claim that prior to 1945 Muslims were never oppressed, the conclusion is that the great powers never even considered certain potential humanitarian interventions in the nineteenth century.¹⁰

Finnemore steers clear of the hypothetico-deductive method favored by rationalists in favor of a largely inductive procedure. Indeed, Finnemore's book is the most overtly and unabashedly inductive of the four works considered here.¹¹

Why does Finnemore reject a hypothetico-deductive approach?

In other words, she is in the early stages of a research project; others are free to derive and test hypotheses from the mechanisms she lists in chapter 5. Although “[t]his book does not … offer lawlike statements or if-then predictions about when purpose will change … [t]hese mechanisms may inspire others to construct if-then predictions they may wish to test….” But the rest of this sentence is significant: “I expect that much of the value of the mechanisms lie[s] in their contribution to explanation, not prediction” (p. 14).

The significance lies in Finnemore's particular implied distinction between explanation and prediction. We may agree that one may predict an event without explaining it (for example, that the sun will rise every twenty-four hours), but the standard positivist understanding is that one may not explain an event without providing information that would have allowed one to predict it; hence the need for laws or if-then statements in the positivist account of explanation. Finnemore evidently holds to an alternative view, namely, that explanation is possible without laws or predictions. But just what does “explanation” mean in this context? Finnemore appeals to the distinction between causal and constitutive accounts noted above. However, the line between causality and constitution is blurred in her analysis. She maintains, “Causality in this book has an important constitutive component” (p. 14) in that “beliefs about legitimate intervention constitute certain behavioral possibilities and, in that sense, cause them” (p. 15). This conflation of causal and constitutive analysis pervades the book. Her goal is “to construct constitutive explanations of the type that answer the question, ‘how possible’” (p. 15). Yet she signifies the process of norm change by the term “mechanism,” which suggests more traditional causal analysis. A “mechanistic” connection between two things, after all, is a type of contingent (causal) rather than logical (constitutive) relation.

Finnemore clearly rejects the rationalist's situational analysis. In her account of the changing norms of debt collection, for example, she stresses that “[Luis] Drago, [Elihu] Root, and Root's delegates to the Hague Conference [of 1907] are the primary agents of change” (p. 143). Their actions, in her account, are essentially uncaused causes in the sense that these actors did not respond to the logic of the situation in which they found themselves; rather, they acted in light of novel ideas about the forcible collection of sovereign debt that they brought to an international forum that fostered normative change. The same general story holds for Finnemore's two other cases. In that of humanitarian intervention, “a great many agents worked hard to create … new social structures” (p. 144). In the case of international order, a series of events that changed legitimate uses of force over time was catalyzed by “participants in the Vienna Congress [who] deliberately replaced [existing] social structures in the treaties they created” (p. 145). In all three stories, Finnemore stresses the creativity of significant agents and the novelty of the actions they took.

In sum, Finnemore makes several important empirical contributions in her book. First, she shows how the constitution of international society and practice has changed along several dimensions in the past two centuries. Second, she demonstrates that such changes affected which potential interventions became actual ones. Finally, she makes a strong case against rationalism's ability to account for these changes in purpose and behavior.

Neta Crawford, Argument and Change in World Politics

Crawford devotes her entire book to a proposition briefly considered by Finnemore, namely, that argument and persuasion affect norms and hence practices (pp. 152–54). Pointing out that most scholars of IR agree that our understanding of stasis is superior to our understanding of change, she chooses decolonization, surely one of the seminal changes in international practice over the past few centuries, as her case study. IR theory does not handle decolonization well, Crawford submits, because it focuses on agents and structures to the neglect of process. Argument is a kind of process (pp. 1–2). Thus we cannot understand decolonization without putting argument and persuasion into the picture.

The cases are structured chronologically, describing in great detail European colonial practices across five centuries, arguments for and against those practices, and the gradual end to the practices. Characterized by “political control, economic expropriation … and cultural control, such as forced religious conversion and education in the language of the metropole,” colonialism was practiced by early modern Europeans in Africa and the Americas beginning in the late fifteenth century (p. 135). One of Crawford's impressive achievements is the compilation and presentation of the copious arguments for and against colonialism and associated practices over five hundred years. From the start, colonizers said it was natural for the strong to subjugate the weak; that Europeans made better use of the seized resources than the natives did; and that it was good to convert the colonized to Christianity (p. 142). Yet these arguments were opposed from the beginning. In sixteenth-century Spain dissent was lodged against certain coercive practices in South America; most famously, Francisco de Vitoria and Bartolomé de las Casas condemned brutal treatment of the natives in South America (pp. 148–55). For centuries those who favored colonialism had their way, but in the twentieth century opponents began to win. Crawford's stories of the increasing plausibility and apparent efficacy of anticolonial arguments, and the tactics used by those who made the arguments, are often gripping.

Subsequent chapters describe the end of slavery (or the “colonization of bodies”); the “aggressive humanitarianism” of late-nineteenth-century colonialism; the Mandates system of the League of Nations; and finally the end of colonialism. In each chapter Crawford lists and analyzes arguments for and against change. Among the most striking arguments are those of eighteenth-century Evangelicals and Quakers that slavery would invite divine judgment (pp. 174–75); of nineteenth-century humanitarians that colonialism would “uplift” the colonized in Africa (pp. 234–36); of W. E. B. DuBois in favor of the internationalization of Africa (pp. 257–60); and of post–World War II advocates of decolonization linking colonialism to the racial theories discredited in that war (pp. 307–9).

Argument and Change is not entirely inductive; indeed, its structure gives it the appearance of a standard deductive work. Its first chapter is a lengthy review of literature on “argument, belief, and culture”; its second offers a theory of argument and change, ending in a discussion of method and hypotheses; chapters three through seven present case studies; and chapter eight considers counterarguments. An examination of her argument, method, and evidence, however, reveals that her approach is more inductive than it seems at first.

In Crawford's account, ethical arguments “denormalize” (render strange) or delegitimize (render morally questionable) a prevailing norm (pp. 100–117). They advance a “reconstruction” or a plausible and desirable alternative set of practices. Powerful actors then reconceive their interests to render them consistent with the new norm; when they institutionalize the norm, general compliance is made more likely. We know, however, that ethical arguments frequently, perhaps usually, fail to alter prevailing practices. Under what conditions do they succeed? Crawford writes that the arguments must be heard; their advocates must be credible; and they must be shown to fit with some preexisting beliefs or values. An example of the third condition is a plea of Guenter Lewy for the United States to intervene more often for humanitarian purposes: “America was founded on a moral purpose, and its people retain a strong commitment to a moral foreign policy…. Given able and forceful leadership, it should be possible to generate a spirit of disinterested altruism that will support humanitarian intervention.”¹²

There is much here that makes good sense. But Crawford's causal story is afflicted by two problems. First, although Crawford offers empirical tests of her argument, she ultimately concedes that it is not falsifiable. Second, the strong role she allows for creative agency undermines her causal claims.

What causes practices to change, in Crawford's account? Suppose that in a given field argument A delegitimizes practice P by telling a plausible story about how P contradicts strongly held belief B. Change in P, then, is more likely than if no such contradiction existed. In this story, A is part of a mechanism linking the contradiction between P and B with the change. The argument, in other words, is the communication of a contradiction. An argument is a causal process that “transmits” or conveys a contradiction between beliefs and practices from one or more actors to others. We can summarize Crawford's causal theory, then, as follows:

How would we know if this theory were wrong? Crawford lists seven “‘tests'” (inverted commas hers) “to see whether the arguments are important causally” (pp. 123–24). Some of the tests are promising, and the example of Britain's violation of material self-interest in opposing the slave trade is impressive (pp. 162–67; 397). But other tests are either consistent with any number of nonconstructivist theories or appear to employ circular logic. Crawford predicts that violators of a norm will justify their violation on ethical grounds; most hard-nosed realists would entirely agree and take such evidence as supporting their cynical interpretation of norms. Crawford writes that after an argument wins the day, behavior should conform to it; but she supplies no way for us to know that an argument “won” without referring to the very behavior it is supposed to explain. Nor is it clear why Crawford does not offer stronger tests of whether arguments really do change minds. For example, if we observed actors saying or writing (especially privately) that an ethical argument changed their minds, or at least gave them pause, Crawford's argument would be supported. But in neither her discussion of method nor any case study do we see this sort of process at work. The closest Crawford comes is in alluding to key actors’ ruminations, as when Emperor Charles V “seems to have been swayed by arguments on both sides” of the question of native South Americans (p. 148). Crawford's case studies primarily describe old practices, arguments for and against change, and whatever changes took place.

Crawford does consider several alternative explanations for decolonization. Among these are liberation movements in the colonies (which she might have incorporated into her theory about arguments); changing economic costs and benefits of empire; and imperial overreach. She answers these ably: liberation movements were constant but only worked after 1945, and the costs and benefits of empire are always ambiguous and were contested before and after 1945. In addition to these alternative accounts, Crawford might have considered the “American hegemony” thesis, namely that the opportunity costs of empire increased after 1945 as the United States offered Western Europe wealth and security through a liberal order.¹³

But Crawford seems less concerned with testing hypotheses than inferring provocative conclusions about arguments and practices. The narratives seldom note any connection to her deductive argument, but instead read like inductive cases. Her inductive method is no surprise, given her explicit skepticism about the hypothetico-deductive method, in particular the possibility of empirically falsifiable hypotheses. “At the most fundamental level,” she writes, “I am not sure that any social science theory, especially rational actor theory, is falsifiable. Rather, theories can be more or less useful for understanding” (p. 129).

Consistent with her skepticism about deduction is Crawford's insistence that her “explanatory aims are both constitutive and causal” (p. 1). Unfortunately, she nowhere explains what she means by constitutive, merely citing Wendt.¹⁴

Thus we see again a rejection of the logic of situational analysis that lies at the heart of rationalist inquiry. Crawford similarly makes explicit her goal of showing that “A focus on argument may also allow us to see room for human agency within the operations of seemingly inexorable political and economic forces” (p. 2). It is these political arguments, she contends, that drive the story of decolonization, and there is nothing in the objective conditions of action that drive the arguments. Again, as in Finnemore's work, agency is an uncaused cause of changes in the dependent variable.

Mlada Bukovansky, Legitimacy and Power Politics

Bukovansky states her purpose at the outset:

The brief answer to her first how-question—“how this transformation of political legitimacy came about”—is that the American and French revolutions accomplished it. These politically seismic events, she argues, delegitimized the old sovereignty upon which the international system was built and replaced it with a new sovereignty. Her second how-question—“how [that transformation] influenced the conduct of international politics”—has many answers, including some related to wars, alliances, and interventions. Bukovansky's book may be placed alongside other works that explore the roots and results of shifts in the grounding of state sovereignty and hence of the international system itself.¹⁵

In addressing her first how-question, Bukovansky borrows from the theory of cultural change proffered by sociologist Margaret Archer. Archer posits two properties of cultural systems, contradictory and complementary. A contradictory culture is grounded in propositions that are inconsistent. For example, Archer writes that early Christians held both that there was value in pagan culture and that pagan culture must be destroyed. A contradictory situation causes agents to exit or else to reinterpret or “correct” one or both propositions. Thus J. H. Newman, the nineteenth-century churchman, argued that pagan knowledge was fully valid for the Christian but, unlike Christian knowledge, was amoral, that is, could not teach virtue.¹⁶

In adopting Archer's approach, Bukovansky redefines complementarity as, roughly, newly emerging compatible interests among actors. An example is the common interest of the philosophes and ambitious monarchs in eighteenth-century France in applying Enlightenment principles to society (p. 33). It is not clear that Bukovansky's version of cultural complementarity is anything other than a special case of contradiction, in which two or more groups agree on a strategy for eliminating the contradiction. In any case, here Bukovansky offers a causal theory that implies lawlike statements applicable across cases: contradictions and complementarities make change more likely. Thus, for example, the early Enlightenment made two contradictory assertions: that right reason was universal and hence accessible to all, regardless of heredity or other status; and that monarchy should be absolute. This contradiction was part of what produced the French Revolution and the move from monarchical to popular sovereignty as the grounding of states and the states system (for example, pp. 104–5).

Her second how-question Bukovansky begins to address with a highly general assertion that culture emerges from the interactions of units in a given system and exerts downward causation upon those units, much like Kenneth Waltz's familiar conception of international structure, which emerges from state interaction and socializes states into seeking security.¹⁸

Thus far the book appears to hew a deductive line of inquiry. In fact, however, its overall cast is inductive. Most of Bukovansky's empirical arguments about the two revolutions and their effects are not the results of hypothesis tests, but rather conclusions she draws by generalizing from the facts of her case studies. Her study of “Old Regime Political Culture” (chap. 3) does identify complementarities, chiefly that between the philosophical Enlightenment and absolute monarchy, and contradictions, such as conflicts between monarchs and nobles and mercantilism and economic liberalism, and it does make the case that these helped produce the revolutions. Enlightenment discourse was crucial, inasmuch as it “indirectly exacerbated and heightened the level of international competition, on the one hand,” by supporting monarchical (“enlightened”) absolutism, “while constituting a continuous public commentary on international affairs, on the other,” by supporting the free international exchange of ideas (p. 108). However, in Bukovansky's treatment of the revolutions themselves (chaps. 4–5), the potential deductive arguments fade from view. Chapter 4 has little to say directly about the causes of the American Revolution or its effects on international politics. Instead, it describes a number of contradictions in American political culture at the time of the founding, such as that between elites and popular authority; secular republicanism and Christianity; economic liberalism and colonialism; independence and the persistence of colonial dependency; Hamiltonian and Jeffersonian political economy; commercial discrimination and free trade; cosmopolitanism and nationalism; and neutral rights and entanglement. Along the way Bukovansky attempts to settle longstanding historiographical debates, such as whether the American founding was liberal or republican. The connection with the book's central arguments here seems to be that the American Revolution was a crucial event in the evolution of legitimacy of a state by virtue of its effects on republican discourse: it made that discourse more democratic, commercial, and nationalistic.

Chapter 5, on the French Revolution, relates more clearly to the book's central concerns. Bukovansky argues that the revolution was caused in part by some of the old regime contradictions described in chapter 3, including that between privilege and equality. She also devotes close attention to her second how-question: how the shift from monarchical to popular sovereignty altered international practice. Thus she seeks to undermine realist interpretations of the consequences of the revolution. Stephen Walt argues that the wars of the French Revolution vindicate his balance-of-threat theory, which sees ideas as of secondary importance;¹⁹

Bukovansky's claims about 1789 and international relations are not new, but she argues them with admirable lucidity and brevity. Her more general and central claim, that some time between the eighteenth century and today the locus of legitimate sovereignty changed from dynastic monarchs to the people, is not especially controversial, although it needs some qualification (the Gulf monarchies of today, which enjoy wide diplomatic recognition if not robust domestic stability, are obvious exceptions). More contestable is Bukovansky's claim that the shift came with the American and French revolutions. No doubt those political events were seismic inasmuch as they sparked transnational movements and increased the credibility of republicanism as a stable form of government. But political strife in Europe in the nineteenth century suggests that sovereignty norms were contested for many decades thereafter. As Finnemore reminds us, the absolute monarchies of Europe through 1849 acted according to a norm of monarchical sovereignty, intervening to suppress popular revolutions in Germany, Italy, and Iberia. It might even be said that popular sovereignty became normative in the West only in 1917–18, with the disappearance of Imperial Russia, Germany, and Austria-Hungary and their norms of monarchism and hierarchy.

Bukovansky's inductive method deserves special attention. “Although studying the political culture of the international system in the manner suggested here does not in and of itself generate definitive, deductive hypotheses about the conditions under which systemic change will occur,” she writes, “it does lend insight into necessary if not sufficient preconditions for systemic change” (p. 53). But if one's goal is to refute Walt and Posen, then why not do as they did and follow the hypothetico-deductive method? Because, she asserts,

In other words, no book can do everything; like Finnemore, Bukovansky implicitly invites subsequent researchers to test her causal claims on new evidence.

But also like Finnemore, Bukovansky seems to have more fundamental reasons to choose inductive analysis over deductive. Her causal language notwithstanding, Bukovansky says she also treats culture as constitutive, inasmuch as it “actually helps make the actors who they are” (p. 36). Even actors that adopt a norm for instrumental reasons are embedded within a larger culture (p. 55). Bukovansky sees constructivism as necessary to her analysis because it better captures “broad shifts in the parameters of political legitimacy”; for example, the French revolutionaries sought to rid themselves not only of Louis XVI and other oppressive elites, but “the whole system of rule they thought despotic” (p. 37).

Bukovansky would achieve her dual approach, treating norms or culture as both causal and constitutive, by decomposing questions temporally.²¹

Bukovansky's case studies would not be helpful were they simply tests of her theory of cultural change or of her general claim that culture is causal. But as excavations of the constitution of actors and acts, they are exemplary. They describe in depth predominant norms and practices in old-regime Europe and the competing norms—contradictions and complementarities—of America and France during the revolutionary period. One leaves them with a surer understanding of the life-world of these people and their institutions.

It is worth emphasizing in this context that Bukovansky too bases her analysis on a repudiation of the situational logic characteristic of rationalist research. She wrestles with the status of contradictions in the cultural system in world politics. These contradictions might be objective, she notes (which would set up the possibility of generalizing research through the study of a situational logic), or they might simply be the constructions of actors (which would undercut such a possibility). In the end she adopts a “dualistic approach,” essentially accepting both the objectivity and subjectivity of the cultural contradictions at the heart of her story—but tellingly she shades her analysis toward the subjective pole. “The cultural system appears to actors as an objective reality, but one that they are more or less free to interpret” (p. 55). This makes her story more agent-driven than structure-driven, in a manner similar to the arguments of Finnemore and Crawford.

Ted Hopf, Social Construction of International Politics

Hopf's book differs from the other three in a striking way: It seeks to explain changes not in international norms or behavior, but rather foreign policy attitudes or understandings. The dependent variable, in other words, is the outlook of the actors under study, not their actions. “I aim to show how a state's collection of identities, how it understands itself, can affect how that state, or more precisely its decision makers, understands other states in world affairs” (p. xiv). Both the independent and dependent variables in this analysis consist of decision makers' understandings. In this sense Hopf's departure from rationalistic social science is more dramatic than that of the other three authors.²²

Hopf's main focus is the ways that foreign policy thinking is structured within one city, Moscow, at two distinct times, 1955 and 1999. His central argument is that identities are linked, causally and constitutively, to foreign policy attitudes. He never defines “identity,” but he clearly means a type of cognitive heuristic: “[A]n individual needs her own identity in order to make sense of herself and others and needs the identities of others to make sense of them and herself” (pp. 4–5); “[a]n individual's identity acts like an axis of interpretation, implying that she will find in the external world what is relevant to that identity” (p. 5). Hopf believes identities are multiple, and this notion is at the core of his approach: “Every individual in society has many identities” (p. 1).

Hopf's argument is novel and derives from a deep and sophisticated reading of social theory. Its general claims are straightforward: identities are constructed “at home,” in everyday life. These identities are always relational—always built upon a distinction between Self and Other—but are not necessarily oppositional. An identity produces particular habits of thought and action and thereby makes certain practices, including actions toward other people, seem self-evidently right and others unthinkable. Identity thus shapes notions of interest. Hence, a Soviet or Russian foreign policy maker's understanding of particular foreign states cannot be understood apart from the complex of his identities. Hopf's constructivism is thus “liberal” in the sense that it is about preference formation at the domestic level.²³

Hopf is not interested in the reasons that people overtly give for their actions, but in the “reasons” that, so to speak, never come up. Eschewing both the logics of consequences and of appropriateness, he builds on the work of sociologists of knowledge and symbolic interactionists to posit a “logic of the everyday” (pp. 12–14). He analyzes the discourse of ordinary life in Moscow in his two periods in order to retrieve identities. The method involves “phenomenology,” or “letting subjects speak … through texts” (p. 23); Hopf reads the daily press, textbooks, popular novels, and the like. He then organizes the identities into “discursive formations,” evidently meaning sets, each of which is constituted by some property that seems important. He finally posits “that any Soviet or Russian foreign policy decision maker is part of a social cognitive structure that comprises the identities and discourses and that they constitute any Soviet decision maker's understanding of himself. His … understandings of other states, therefore, must necessarily involve the interaction between this complicated social Self and those external Others” (p. 20).

In the Moscow of 1955 Hopf finds four “primary identities:” “class, modernity, nation, and the New Soviet Man (NSM)” (p. 41). That is, the works he reads suggest that Muscovites saw themselves as workers within a class hierarchy, as moderns, as Russians—and as at least aspiring toward an ideal construct of the best of the other identities: “[The NSM] transcended national and ethnic identities, was an avowed atheist, and had a working-class consciousness” (p. 69). The properties of the NSM, however, were contested. Some conceived of him very precisely and rigidly, while others, in the post-Stalin “thaw,” allowed some deviance. Based on these identities, which he sees as composing a single discursive formation, Hopf offers several predictions about the understandings of Soviet foreign policy decision makers. For example, he argues that “[t]he more strictly class analysis was pursued, the more likely the Soviet Union understood its allies in a more constrained and binarized manner,” and that the “Soviet modern identity should lead to its potential allies being understood through Soviet understandings of its own modernization” (pp. 81–82). In chapter 3 Hopf finds the evidence to be consistent with his predictions—for example, a presumption that fellow Marxist-Leninist states, being ostensibly ruled by their working classes, would be loyal allies—but also finds an extra identity not anticipated by his discursive analysis, namely, the Soviet Union as a great power (p. 83).²⁴

In the Moscow of 1999 Hopf finds three discursive structures relevant to foreign policy, which he labels New Western Russian (NWR), New Soviet Russian (NSR), and Liberal Essentialist (LE). The discourses all comprise identities such as democracy and culture. Distinguishing them are such identities as nationalism and the Orthodox Church (NWR rejects them), free markets (NWR embraces, NSR and LE are ambivalent toward them), and the Russian countryside (NWR ignores, NSR embraces, LE is ambivalent toward it). The NWR is also characterized by its embrace of the U.S. model; the LE prefers Europe; and the NSR sets store by neither (table 2, 157–58). From these discourses Hopf predicts preferences: the NWR will prefer an alliance with Europe or the United States; the NSR, with China; and the LE, with Europe and China. The NSR and LE will feel most threatened by America. He predicts Russian interests and understandings contingent upon which discourses were predominant. For example, if the NWR and LE dominated, “Europe would supplant the United States as the most desirable Other” (pp. 209–10). In the event, as Hopf describes in chapter 5, Russian policy in 1999 was ambiguous in many ways, reflecting the failure of any discourse to dominate.

These results are plausible, but in the end it is not always clear what to make of them, inasmuch as Hopf mingles causal and constitutive language. His adducing of predictions, his effort to describe identities and construct discourses prior to describing understandings of foreign policy, and much of his language suggest that he sees his analysis as causal. His stated aim, as summarized above, is “to show how a state's collection of identities … can affect how that state, or more precisely its decision makers, understands other states in world affairs” (p. xiv). The verb “affect” implies that identities and understandings of other states are separate. Identities “enable and constrain Soviet alliance choices in 1955,” suggesting the identities are separate from the choices (p. 83). His allusions to a psychological approach (chap. 1), in which identities function like heuristics, would also seem to imply causality. Yet in other places Hopf understands his work as constitutive, or an analysis of what makes actors who they are. For him, “identities constitute [discursive] formations” (p. 3), and “an individual's identity implies his interests” (p. 16). The relation between identity and interest, in other words, is logical rather than contingent; it is not that identities cause interests, but that actors of a particular constitution (identity) will by definition have certain interests.

And like our other authors, Hopf adopts a research strategy that is heavily inductive. Indeed, Hopf announces in his preface that he has chosen an approach that is as inductivist as possible:

Hopf does use a hypothetico-deductive method, but one that assumes very little and is hence what many rationalists would call “weak.” As we showed above, the primary causal claim is that identities cause foreign policy preferences; the case studies are intensive efforts to discover and organize the number and content of identities (that is, values of his causal variable) by analyzing discourses. The deductive effort falls short, primarily because like our other authors Hopf conflates causal with constitutive analysis.

Accordingly, the 1999 case study seems constitutive, its findings tautological from a causal perspective. Hopf uses attitudes toward various external Others, such as the United States and Europe, to delineate Muscovite identities and discourses; he then uses those same discourses to explain the attitudes from which he derived them. Since we know NWR people precisely by their identification with the United States, is it any surprise that they favor close relations with the United States? Since we know LR actors in part by their alienation from America, would it not be astonishing if they did favor a close alignment with America? Had Hopf excluded attitudes toward foreign countries from his identities—arguing, for example, that actors who favor free markets would also favor good relations with the United States—then his argument could be causal insofar as free markets and the United States are not logically connected. One can infer, however, that in Muscovite minds in 1999 free markets and America were in fact so connected. Perhaps, then, causal analysis of this case is not possible.

The 1955 study, by contrast, appears causal, insofar as Hopf keeps the Muscovite identities in that year separate from attitudes toward foreign countries. One knows whether an actor's New Soviet Man identity is tolerant by his attitude toward deviance within the USSR, not toward Tito or Nehru. Even here, however, it might be difficult to separate attitudes toward deviance at home and abroad, especially since communists in 1955 were, on their own understanding, internationalists. They retained the Marxist-Leninist deprecation of nationalism and national boundaries.

Perhaps the difficulty of doing deductive analysis concerning identities is what makes Hopf's approach so much more inductive than that of the standard rationalist work. Hopf refuses to assume much about his subjects except that their identities imply their interests. He begs no questions concerning how many identities existed in Moscow in 1955 and 1999, their content, whether they presupposed a hostile Other, and so on. Identities are taken as given; Hopf neither theorizes about them nor tries to trace empirically how agents transform them. His inductivism seems to be driven solely by the constitutive nature of his work. If actors' identities entail their preferences, then the analyst's task is primarily to articulate who they are.

As a constitutive analysis of foreign policy discourse and preferences, Hopf's work will set a standard for years to come, both for Soviet-Russian studies and for foreign policy studies more generally. Still, here again rationalists will be discontented. Both of Hopf's case studies leave us with the question of how one discourse or combination comes to dominate. His prototypical decision maker has multiple identities, each of which has some influence; both preferences and policy choices are thus indeterminate. Hopf thus leaves structural IR theories an opening to explain Soviet and Russian foreign policy. In the 1999 case, neorealists (at least, those interested in explaining foreign policy as opposed to international outcomes) will argue that the distribution of power among states will dictate which Muscovite discourse wins; for example, under unipolarity those discourses wishing to counterbalance the United States (the NSR and LE) will win. Hopf might reply that we are still waiting for this to happen,²⁵