Jonathan Kirshner’s stimulating book wages a two-front battle against the corrosive effects of hubris in international relations: first against the arrogance of power that has repeatedly led to tragic imperial overstretch as first described in Thucydides’s account of the Peloponnesian Wars, and second against the arrogance of social-scientific overstretch in the vain pursuit of valid predictions about world politics that are based on oversimplified, deterministic theories. Kirshner brings a diverse and well-honed set of skills to this lively contest, drawing on his wide-ranging career as a theorist of international relations and scholar of international political economy, foundational economic theory, and financial history.[1] His penetrating essay, responding here to three reviews by well-chosen interlocutors, breaks additional new ground beyond the many fresh insights of his important book.
H-Diplo | Robert Jervis International Security Studies Forum
Roundtable Review 16-9
Jonathan Kirshner, An Unwritten Future: Realism and Uncertainty in World Politics. Princeton: Princeton University Press, 2022. ISBN: 9780691166773.
25 October 2024 | PDF: https://issforum.org/to/jrt16-9 | Website: rjissf.org | Twitter: @HDiplo
Editor: Diane Labrosse
Commissioning Editors: Jennifer Erickson and Joshua Shifrinson
Production Editor: Christopher Ball
Pre-Production Copy Editor: Katie A. Ryan
Contents
Introduction by Jack Snyder, Columbia University. 2
Review by Stacie Goddard, Wellesley College. 9
Review by Mariya Grinberg, MIT. 14
Review by Barry Posen, MIT. 20
Response by Jonathan Kirshner, Boston College. 29
Introduction by Jack Snyder, Columbia University
Jonathan Kirshner’s stimulating book wages a two-front battle against the corrosive effects of hubris in international relations: first against the arrogance of power that has repeatedly led to tragic imperial overstretch as first described in Thucydides’s account of the Peloponnesian Wars, and second against the arrogance of social-scientific overstretch in the vain pursuit of valid predictions about world politics that are based on oversimplified, deterministic theories. Kirshner brings a diverse and well-honed set of skills to this lively contest, drawing on his wide-ranging career as a theorist of international relations and scholar of international political economy, foundational economic theory, and financial history.[1] His penetrating essay, responding here to three reviews by well-chosen interlocutors, breaks additional new ground beyond the many fresh insights of his important book.
His book and response aim at dual scholarly targets. Most of his fire is trained on “structural realists” who, he argues, follow too closely in the footsteps of Kenneth Waltz, especially “offensive realists” such as John Mearsheimer.[2] Additional targets are the “hyper-rationalists” among international relations scholars, particularly proponents of James Fearon’s widely-employed “bargaining theory of war.”[3] Kirshner says these approaches inevitably fail in their unachievable ambitions to emulate natural science in using overly simplified, deterministic theories to try to understand the elusive moving target of behavior in international politics.
Kirshner prescribes as an antidote the modest, prudent “classical realist” approach to the study of international relations, which accepts that international behavior is shaped by contingency and complexity, and is carried out in conditions of profound uncertainty. Prediction of the behavior of actors in this drama is futile, he says; actors are unable to predict even their own behavior in fluid circumstances that they cannot anticipate or even imagine. In this setting, understanding is more readily achieved through a humble respect for the constraints imposed by limited knowledge in the face of the countervailing power of multiple rivals. The best preparation for understanding these uncertain dynamics, Kirshner argues, is a multifaceted grounding in history, drawn upon flexibly with an eye toward shifting contexts. His recommended exemplary models on this path to understanding are thinkers who highlight history or uncertainty, such as Thucydides, E. H. Carr, Hans Morgenthau, Reinhold Niebuhr, George Kennan, and Carl von Clausewitz.
This academic debate is a battle among rival schools of thought that share much in common in terms of substantive interests, motivating research questions, and intellectual formation. It’s a bit like the rivalry between the Stalinist and the Trotskyite lunch tables at the City College cafeteria in the 1930s. As Stacie Goddard’s review below notes, the difference between the respective camps is more about “temperament” than the content of their core substantive arguments about international politics.
The classical realists as described by Kirshner agree with the structural realists that international politics is shaped by the fact of international anarchy (the absence of an order-imposing sovereign power above states), which makes war an indispensable tool of self-help needed for survival and produces states’ tendency to balance threatening power through expedient alliances. Classical and structural realists (whether “offensive” or “defensive”) agree that “buck-passing,” i.e., letting other states bear the cost of opposing an aggressor, is a factor that can destabilize the equilibrium of the balancing system.[4]
Similarly, rational bargaining theory identifies a major cause of war as the “commitment problem:” how to conclude a stable peace agreement when it is impossible to convince the opponent that you won’t attack later when you are stronger. This structural problem is a key aspect of “the security dilemma,” a situation in which anything one side does to improve its security necessarily reduces the security of others. Offensive and defensive structural realists agree that this security dilemma is a prominent cause of war. The main difference is that offensive realists see it as endemic and remediable only by conquering a self-sufficient continent buffered by oceans on all sides, whereas defensive realists think that lesser geographical barriers and defense-enhancing arms control can mitigate it.[5]
While rational bargaining theory mainly seeks to explain how even rational leaders can get into costly wars that are likely to severely harm both sides, its rationalism is flexible enough to acknowledge that non-rational approaches might also have something to contribute to bargaining failure and war. Fearon starts his article by saying that psychologically based misperceptions might sometimes cause war, too; he just sets that aside to focus on what he sees as a harder puzzle to solve. Likewise, he says that one of his three major causes of war in bargaining—the indivisibility of the stakes—can happen only if some kind of non-rational discourse artificially constructs an issue, such as a holy place, as culturally indivisible.[6]
For the most part, it is not a big stretch for our three reviewers to agree with much of the substantive content of Kirshner’s flexible, modified realism. Barry Posen, a Waltz student, has made seminal contributions on the structural consequences of anarchy, acknowledging inefficiencies in balancing as well as geographical, technological, and demographic conundrums of the security dilemma, processed through the biases of military organizations. He coined the term “restrainers” for those who oppose hubris and extoll prudence in realist grand strategy.[7] Stacie Goddard’s work on indivisible stakes and rising great powers combines structural problems of international security with contextually specific insights from sociological network theory about the role of persuasive brokerage in managing and prevailing in conflict. A Robert Jervis student with a diverse conceptual toolkit, Goddard moves flexibly across theoretical traditions in studying factors that exacerbate or mitigate security dilemmas and bargaining failures. Goddard’s analysis of Chancellor Otto von Bismarck’s clever strategies of persuasion and coalition brokerage shows that adept prudence can serve the offensive goal of neutralizing resistance from potential balancers against expansionism, facilitating the consolidation of dominant German power in the center of Europe.[8] Mariya Grinberg, a John Mearsheimer student, works on the reasons why states trade with their enemies during wartime. Her findings are fully compatible with Kirshner’s plea for realists in security studies to pay more attention to economics as well as his urging political economists to integrate their own arguments with realists’ insights on the structural consequences of anarchic rivalry.[9]
Where these three diverge from Kirshner is less in empirical or even theoretical substance than in their attitudes toward scientific methodology, particularly regarding the implications of uncertainty for strategies of inference, explanation, and research on international power politics. Posen says that Kirshner’s anti-scientific classical realism lacks intellectual “discipline” insofar as it rejects the logic of prediction that makes possible the empirical testing of arguments. Grinberg argues that without testable predictions, there is no useable theory, and “without theory, there is no explanation, no roadmap for the unwritten future.” Making the point in a different way, Goddard says that realists already know that hubris is a problem, prudence is a virtue, and power needs to be respected and accommodated. The real task, she says, is to understand better when and how to accommodate or resist rising power and how to achieve the internal social cohesion that allows states to act prudently. This requires organizing knowledge such that conditional predictions can be tested empirically in some kind of systematic way.
A related methodological dividing line is Posen’s claim that the kind of knowledge that realism can provide is of broadly “recurring behaviors,” which can be understood in a parsimonious way through enduring mechanisms of power politics. Kirshner, in contrast, argues in his book that radical uncertainty means that international politics “will never be a science” because “most questions of international politics, most of the time—and invariably the most important ones—are concerned with the particular behavior of a specific actor at a critical time, not the average reaction of an imagined median actor under everyday circumstances” (215). But the reviewers point out that Kirshner does generalize sweepingly in his warnings about hubris and his endorsements of prudence and accommodation. They want to see these conclusions subjected to a more systematic test and applied with scope conditions and contingencies specified.
The reviewers have a different notion of “prediction” than Kirshner. They consistently discuss prediction in the sense that given certain initial conditions, the scholar’s hypotheses should specify an expected outcome given the logic of the theory. For them, the logic of prediction is identical to the logic of explanation of past events. Kirshner, however, typically portrays “prediction” as an attempt to foretell specific future events or actions, as in a crystal ball. This is connected to his view that the intellectual objective must be understanding particular events rather than generic, recurrent patterns.
The reviewers are not minimizing the uncertainties and contingency of international politics nor the need to layer on economic or psychological features of the situation in order to gain a fuller understanding of complicated interactions. However, they are not convinced that uncertainty is so impenetrable that the usual structure of empirical science needs to be abandoned.
It is worthwhile to recall how some leading international relations scholars have recommended dealing with the endemic problems of complexity, uncertainty, and the inevitability of unintended consequences of action. Robert Jervis’s concluding advice in his book on Perception and Misperception is to keep an open mind, always remembering that evidence that is consistent with your own interpretation may also be consistent with different theories.[10] This is an issue, for example, in evaluating deterrence theory against the spiral model. Jervis’s advice in his System Effects book is to “do two things at once,” one a direct move toward accomplishing your goal and another to anticipate and counteract the contrary feedbacks that your move will propagate through the interconnected system.[11]
Even more directly on point for the debate on uncertainty between Kirshner and the three reviewers is how Glenn Snyder and Paul Diesing integrate the complex structural and psychological dimensions of crises at the brink of great power wars.[12] Working generally within a realist framework, Snyder and Diesing establish the game structure of the crisis interaction—such as symmetric Chicken or Prisoners Dilemma or asymmetric Called Bluff or Bully—through archivally discovered preference orderings for both sides. Then by theory they assume that one or both sides must be underestimating the bargaining power of the other, based on estimated capabilities or intensity of interests, or else there would be no crisis. (Kirshner might call this hubris. Fearon might call it a bargaining failure in the making.) Snyder and Diesing proceed to analyze the strategic interaction of moves in the crisis through Jervis’s theory of interpreting ambiguous new information (Kirshner’s uncertainty) in ways that reinforce incorrect prior beliefs about relative bargaining power. Snyder and Diesing’s advice for the defender is to start the interaction with unambiguously firm resistance to disabuse those false beliefs and then explore terms of compromise only in the end game. Arguably the concerns and conjectures of all of the schools of thought in this debate are accommodated to a considerable degree within this conventionally scientific setup.
But let’s probe more deeply into Kirshner’s understanding of radical uncertainty as undermining the assumptions of conventional social sciences, whether international politics or neoclassical economics. Kirshner notes that his formative understanding of uncertainty came from reading John Maynard Keynes’s General Theory of Employment, Interest and Money.[13] Kirshner argues for the relevance of this to international politics not only because of the centrality of economics to power but also because Waltz’s structural realism is based on an analogy between the balance of military power and the competition among oligopolistic firms.[14]
Before Keynes, insofar as classical economics had a theory of booms and busts in the business cycle, it was hampered by having only a static concept of market equilibrium and even much later by generally assuming perfect information about supply, demand, and price in models of market adjustments. Under these assumptions, there was no convincing general mechanism to explain self-reinforcing downward or upward spirals away from equilibrium in the economy.[15] Keynes’s crucial conceptual innovation pointed to the inherent uncertainty in anticipating the future state of the economy, opening a wide latitude for the role of subjective expectations of consumers about prices, workers about employment and wages, and investors about market opportunities. Consider the situation of investors who have to commit resources to projects that involve the development of new technology which doesn’t yet exist to produce hypothetical products for a completely unknown market demand. Think of AI today or the South Sea Bubble in the early days of global capitalism. Many such decisions by their nature are taken not in response to an existing market equilibrium of supply and demand but in conditions in which any hypothetical future equilibrium is pure speculation. Keynes’s insight was that subjective market confidence in the face of vast uncertainty, which could be influenced by fiscal and monetary policy, was the crucial ingredient affecting the direction taken by the business cycle for good or ill.[16]
What are the implications of this insight about the pervasive impact of radical uncertainty for economics, politics, and international relations? Ironically, Keynes’s otherwise diametrical opponent Friedrich Hayek argued that the inscrutable complexity of the modern economy made centralized government economic planning impossible and consequently concluded that only the decentralized, automatic “invisible hand” of the market could accomplish the task of coordinating economic activity.[17] Neoclassical economists such as Paul Samuelson have tried for decades to construct a scientific model of a dynamic macroeconomic systemic equilibrium. In part the goal was to help democratic governments fine-tune their Keynesian adjustments of inflation and unemployment. In part, this was also an ideological project meant to show how the invisible hand of free market competition, with some limited tinkering by welfare-oriented economists, can produce Pareto-improving outcomes that lift all boats.[18] The 19 editions of Samuelson’s economics textbook from 1948 to 2009 began by invoking Adam Smith’s “invisible hand” as a metaphor, then briefly elevated it under William Nordhaus’s co-editorship at the beginning of the 1970s to a logically air-tight, empirically demonstrated finding, before demoting it once again to an aspirational status when Kenneth Arrow gave up on formally modeling the macro-equilibrium of the free market.[19] On the empirical side, the economic stagflation of the US and global economies in the 1970s was a further setback to the project.[20] Despite this, and supporting Kirshner’s views, the hubristic “rational expectations” theory of competitive markets soon made its comeback as an ideological underpinning of the Washington Consensus for domestic deregulation and global free trade and investment.[21]
Analogously, Karl Deutsch and David Singer advanced the claim that the international balance of power was something like an invisible hand mechanism that tended to produce a peaceful equilibrium under multipolarity.[22] Waltz corrected this by explaining that the balance of power was a result of states’ self-help efforts to use any expedient including war to help them survive in anarchy.
Kirshner’s thought-provoking book—and our lively debate on it here—unpacks all of these issues. Deborah Avant’s blurb on An Unwritten Future hits the nail on its head: “I love this book—even the parts I don’t agree with.” The reviews and Jonathan Kirshner’s value-added reply are just as enjoyable.
Contributors:
Jonathan Kirshner is Professor of Political Science and International Studies at Boston College, and the Stephen and Barbara Friedman Professor of International Political Economy Emeritus in the Department of Government, Cornell University. He is the author of numerous books, including American Power after the Financial Crisis (Cornell University Press, 2014) and Hollywood’s Last Golden Age: Politics, Society and the Seventies Film in America (Cornell University Press, 2012). Kirshner is the co-editor (with Eric Helleiner) of the book series Cornell Studies in Money, and co-editor (with Peter Katzenstein), of The Downfall of the American Order? (Cornell University Press, 2022). His current research projects include “Listening to Thucydides: Contingency, Chance and Catastrophe in Contemporary World Politics.”
Jack Snyder is the Robert and Renée Belfer Professor of International Relations in the Political Science Department and the Saltzman Institute of War and Peace Studies at Columbia University, is the author of Human Rights for Pragmatists: Social Power in Modern Times (Princeton University Press, 2022).
Stacie Goddard is the Betty Freyhof Johnson ‘44 Professor of Political Science and Associate Provost of Wellesley in the World at Wellesley College. Her research and teaching focuses on questions of great power competition and international order. Her latest book, When Right Makes Might: Rising Powers and World Order was published by Cornell University Press in 2018.
Mariya Grinberg is an Assistant Professor of Political Science at the Massachusetts Institute of Technology and a member of the MIT Security Studies Program. Her research examines why states trade with their enemies, investigating the product level and temporal variation in wartime commercial policies of states vis-a-vis enemy belligerents. Her broader research interests center on the question of how time and uncertainty shape the strategic decisions of states, focusing on order formation, military planning, and questions of state sovereignty. Dr. Grinberg earned her PhD in Political Science at the University of Chicago in 2019, and previously served as a Postdoctoral fellow at the Center for International Security and Cooperation at Stanford University, as well as the Dickey Center at Dartmouth College.
Barry Posen is Ford International Professor of Political Science at the Massachusetts Institute of Technology and Director Emeritus of MIT’s Security Studies Program. His research focuses on broad issues of grand strategy, the use of force, and great power politics. The author of three books and dozens of articles, Posen received the 2017 Distinguished Scholar Award from ISA’s International Security Studies section.
Review by Stacie Goddard, Wellesley College
After three decades of US hegemony, we are witnessing the return of great-power competition. The aftershocks of the COVID pandemic, combined with the rise of populist movements, continue to undermine what seemed to be an inexorable march towards economic interdependence. For decades, realist scholars have warned that liberal optimism about the post-Cold War era and the “liberal international order” was misguided, and that it was inevitable that states would rise to challenge US unipolarity.[23] The rules of the liberal world order were only as strong as the power that underpinned them. We are ultimately in a tragic world, one where competition over power will rule the day.
Yet this moment, Jonathan Kirshner instructs us, is not one of inevitability, but of fundamental uncertainty. Yes, it is true that the optimism of the last decades was unwarranted. But so too is the confidence of scholars who contend that today’s competition is the inevitable consequence of anarchic structures and the ensuing competition for power and security. There is no certain future in front of us, be it one of measured cooperation or of tragic competition. For this reason, Kirshner tells us that this is a moment to return to classical realism. Drawing from the works of Thucydides, E.H. Carr, Hans Morgenthau, Reinhold Niebuhr, and George Kennan[24], he argues that to approach this moment properly requires a classical realist’s sense of caution, an appreciation of complexity, a grasp of history, and an understanding that prudence must overcome hubris.
Kirshner is uniquely situated to make a case for classical realism’s continued relevance. His work has long crossed theoretical and disciplinary boundaries, and An Unwritten Future engages both international relations and economic theories to make the case that the field has been unduly obsessed with parsimony and determinist economic models. Kirshner argues forcefully that the political scientists should abandon the futile goal of prediction, and instead embrace classical realism’s aim to “describe, explain, understand, and anticipate” (53, emphasis original). This need for this shift is urgent, as both scholars and policymakers face an increasingly uncertain world, one where—as Kennan put it, “the greatest law of human history is its unpredictability” (55).
The False Promise of Determinacy
Kirshner trains his critical sights on two targets: structural realism, particularly “offensive realist” approaches, which often reduces behavior and outcomes to anarchic structures; and “hyper-rationalist” theories, which build their frameworks on the foundation of utility-maximizing individuals.[25] What these approaches share in common is their determinism, their aim to build theories that not only explain but attempt to predict actors’ behavior and outcomes. This determinism is a problem. As Kirshner explains, the study of international relations is necessarily a science of uncertainty and indeterminacy (68).
To see what is problematic about these deterministic approaches, it is useful to compare Kirshner’s arguments to a straightforward model of Bayesian updating, where actors have incomplete information about which “state of the world” they inhabit. For example, leaders may be uncertain about whether they face a world of status quo or revisionist states. Actors thus begin interacting with a probabilistic assessment about which world they live in, looking for information that allows them to update their prior beliefs. For example, aggressive action by one actor might lead another to update their beliefs and decide they are in a revisionist world.
Kirshner argues that classical realists see at least two problems with this model. First, it is not simply that actors lack information about the future; it is that the future—as the book title tells us—is fundamentally unwritten. Kirshner points to the difference between uncertainty and risk or, putting it another way, the difference between epistemological and ontological uncertainty (77). In the former case, actors are uncertain about what they know about their environment—the truth is out there, but actors lack information about their world. With the latter, we are talking about an environment that is fundamentally uncertain because its very existence is contingent on unfolding processes and interactions. Indeed, the actions leaders take as they negotiate international politics will shape how that future unfolds. How actors interpret information, how they respond to each other and their circumstances is not simply reaction and updating; these are acts of creation. No theory can predict a future that is always in motion.
To make matters even more complicated, there are problems with the assumption—shared by structural realists and rationalists alike—that actors will respond to information in similar ways. This is, of course, why Bayesian updating should produce predictable outcomes: as actors update their beliefs about their environment, they converge on a similar understanding of the world. In contrast, classical realists hold that while actors are rational—insofar as they are reasonable, calculating, and strategic—how actors interpret and respond to information involves a host of factors outside the scope of both structural realism and hyper-rationalist theories. These other factors might include the actor’s own history, which provide an essential lens shaping how information is perceived. Decisionmaking also invokes emotional responses—notably, anger and fear—and hinges on ideology. For this reason, no two actors will respond to their environment in the same way.
These are important critiques. In fact, by focusing only on structural realism and rationalist explanations for war, Kirshner almost undersells the significance of his message. It is not merely realists and rationalists that should be paying attention to classical realist arguments, but anyone who embraces the hope of understanding international politics through invariant causal models of any sort, where we assume shifts in independent variables will produce shifts in dependent variables (even if that causal connection is probabilistic). If Kirshner is correct, the contingency, indeterminacy of interactions, combined with the agency of self-reflective individuals makes this form of “science” impossible.
Is Classical Realism the Answer?
The indeterminacy of international politics is the foundation for Kirshner’s primary argument that both scholars and policymakers would benefit from a return to classical realism. What exactly does this look like in practice? Kirshner explains that classical tools include the “role of historical legacies, uncertainty, contingency, contestation, and exogenous shock.” He devotes a great portion of the first chapter of his book to defining classical realism inductively, using Thucydides’s The Peloponnesian War to draw out core principles. Some of these principles—anarchy and attention to the balance of power—are shared with structural realists, although Kirshner is quick to point out that these factors shape, but do not determine, human action. Added to this are the role of a political community’s “content and purpose”; its national character and identity; and its leadership (25-26). And throughout the chapter, invoking the words of Thucydides as well as the realists of the mid-twentieth century—Carr, Morgenthau, Kennan, and Niebuhr—Kirshner reminds us of classical realists’ warnings about the dangers of hubris, the importance of social cohesion, and the need for scholars and practitioners alike recognize the role of uncertainty and contingency.
This is all persuasive, but I struggled to understand what unique insight classical realism adds to our understanding of politics. The chapter “Why We Need Classical Realism” revisits a few cases—Britain’s appeasement of German leader Adolf Hitler, the US’s interventions in Vietnam and in Iraq—in order to demonstrate the power of classical realism, but this chapter left at least this reader somewhat puzzled. Did I need classical realism to understand that history and ideology shaped Britain’s response to Hitler? While Kirshner’s command of history is on display here—particularly in his account of the London’s financial sector and its opposition to confrontation—I remained uncertain about the specific value added of a classical realism approach (given how much this narrative already exists in the historiography[26]). Likewise, do we need classical realism to see Vietnam and Iraq as catastrophic acts of hubris that were paradoxically driven by the United States’ overwhelming power?
Kirshner argues that there is an “urgent need” (1) for classical realism in theory and practice, yet it seems like these core commitments to history, contingency, and contestation are already present in other theoretical approaches. Kirshner is clear from the outset that he is not interested in providing a “comprehensive overview of theories or approaches to IR.” He is not engaging in “paradigm wars,” not is he attempting to show how classical realism trumps its liberal or constructivist counterparts (2). That is all fair, but if this book is designed to meet an urgent need, then it seems it would have made sense for him to demonstrate that this need is not being met elsewhere in the discipline. It seems, for example, that certain strands of psychological approaches, with their focus on individual leaders’ beliefs and how they shape decision making, fits Kirshner’s model of “classical realist” scholarship.[27] Similarly, certain types of constructivist theorizing embrace contingency and contestation as essential to studying international politics.[28]
If the work Kirshner describes as being essential is being done by scholars, then it is not clear why we need the “return” to classical realism? Is the reason that the work isn’t being done or that published work still misses some of the insights of classical realism. This point should have been addressed. Another possibility is that the work is being done, but that these scholars aren’t overtly embracing their “classical realist” sensibilities. After struggling to interpret what set classical realism apart, I concluded that it seems less a theory or even a paradigm than a temperament that shapes both the analysis and practice of politics (158). This suggests that while classical realism might not exist as a fully-formed theoretical approach, it lives on as a scholarly temperament across a variety of approaches. It is less urgent that we return to or revive classical realism than it is that we continue to recognize its ongoing critiques of rationalist, structuralist, and materialist approaches to international politics.
The Unwritten Future
The future may be unwritten, but classical realists understand that any analysis of international politics still must grapple with the unknown. Kirshner’s final chapters reflect on how classical realism might help us understand the emerging international system. He rightly scolds the discipline for continuing to treat international security and international political economy (IPE) as separate fields of study.[29] He is equally critical of security studies scholars who see the economy as an afterthought, and IPE scholars for embracing overly deterministic economic models and far too often espousing the argument that economic interdependence decreases the political contestation and war.
All of this provides necessary context to his discussion of the tension between US and China, a great power competition that, unlike the Cold War, is necessarily shaped by processes of globalization and dense economic ties. Kirshner excoriates deterministic theories—especially those of John Mearsheimer and Graham Allison[30]—that see conflict between the US and China as all but inevitable. While Kirshner is no optimist about the future of US-China relations, he maintains that the pursuit of regional hegemony, which would likely drive the US and China into a military confrontation, is not simply a biproduct of power transitions and the perils of anarchy. Instead, the real danger for the US and China lies “in hubris and fear: that it is every likely a rising China will become increasingly arrogant” and that the United States “will be first too arrogant and then too anxious to take the measures necessary to address the realities of changes of the international balance of power” (211).
So what is a good classical realist to do? Here Kirshner points to classical realists’ instinct to dispassionately recognize and accommodate changes in the balance of power. A US defense of the hegemonic status quo is not simply impossible in the face of changing power, it would likely be catastrophic. The US needs to look for reasonable accommodations to manage this changing world. This raises at least two questions that I wish Kirshner’s narrative had pursued in more depth. The first concerns the ethics of accommodation. How do policymakers know when they have gone too far in accommodating power? Throughout the book, Kirshner stresses that there are significant differences among actors. Not every aggressive actor is a Hitler or Napoleon. But some of them are. He points to the disagreement between Carr and Morgenthau over whether to confront or accommodate Hitler. He notes that Kennan (in his view, wrongly) opposed the funding of NATO. Classical realism does not provide a hard and fast guide to identifying those moments where a power must refuse accommodation; that is understandable. But some indication of how policymakers will know when accommodation is the prudent and ethical choice seems essential.
Second, Kirshner raises in these final chapters the question of social cohesion. Kirshner argues that grand strategy—be it one of accommodation or confrontation—does not work if the nation cannot formulate and act upon a shared social purpose, and a defined national interest. The United States managed to do this at the start of the Cold War. The French, in contrast, failed to cohere around a shared social purpose before World War II with disastrous results. This raises significant questions of how and whether the US will weave together a shared social purpose as it attempts to manage its competition with China. Here Kirshner’s response appears particularly pessimistic. Yet, is this warranted? Most if not all nations enter into great power competition with some level of societal fragmentation. Exploring how and when leaders manage to mobilize in the face of contention is a topic worth exploring in more depth.
Perhaps it is fitting that Kirshner leaves us without a clear guide to these questions of accommodation and cohesion. It seems consistent with his classical realist approach to anticipate problems, prescribe prudence and humility in moving forward, but avoid definitive answers. It may very well be the best we can do.
Review by Mariya Grinberg, MIT
In An Unwritten Future: Realism and Uncertainty in World Politics, Jonathan Kirshner takes on the task of reviving classical realism as the viable realist alternative to the two dominant, yet flawed, approaches of structural realism[31] and rational expectations theory[32] (i.e., the bargaining model of war). The book provides both the security and the economic underpinnings of classical realism, distinguishes it from its realist cousins, and, very importantly, takes a chapter to dwell on the limitations of classical realism. Additionally, the approach is used to elucidate several difficult cases in world politics: the British decision to appease German leader Adolf Hitler in 1939, the US involvement in Vietnam and Iraq, and the consequences of China’s rise in contemporary politics.
For Kirshner, classical realism shares several starting assumptions with all the other variants of realism. Anarchy exists between states—or any autonomous political units—such that there is no external restraining force on the behavior of others (14). The primary goal of states is survival. Therefore, states pay great attention to the military capabilities of others, as the distribution of power dictates where states see threats (4). The basic understanding is that economic power underwrites the military power of the state, even if most other variants of realism forget to explicitly mention this factor (159).
To these common foundations, classical realism adds a unique flavor of socially and historically contingent motivations. Beyond survival, the goals of states are “not uniform across actors but shaped [by] the perceived lessons of history and the social-cultural environment in which behavior takes place” (3). There is no predetermined hierarchy of state needs. Moreover, Kirshner notes, the basket of state motivations changes over time based on factors such as domestic interest, relevant recent historical experiences, and shifts in the state’s power. In addition to the multitude of material interests, from which states are free to choose, states are also driven by fear, status and honor concerns, and hubris (15).
As Kirshner argues the national interest is informed by domestic contestation and interpretations of historical experiences; it is greater, however, than the sum of the interests of domestic actors (16). The state has its own interests.[33] And when the interests of private actors diverge from those of the state, the state prioritizes its own interests—those of the nation as a whole—over those of narrow coalitions (160). Precisely because each state’s national interest can be driven by any number of different factors at each specific point, there is constant friction between states in international politics. Conflict between states is ever-present and unceasing (34-5). Also, due to the same factors, Kirshner contends that it is difficult, if not impossible, to predict how states will behave in a particular situation. The world is full of uncertainty; states are even uncertain of their own future actions a few steps into the future (4). If states cannot do this, classical realism contends that scholars, more surely, cannot predict with any specificity what a particular state will do at a crucial historical juncture (73).
And yet, even classical realism recognizes some trends. The ever-changing balance of power—the dynamic, not the static comparison—is the “primal engine of conflict” (14). States’ interests and ambitions “tend to grow in tandem with their capabilities,” which suggests that states with increasing power will find themselves in more conflicts with other states (17). This is so not only because the greater capabilities are frightening in their own right, but also because growing ambition provides greater potential for conflict between the interests of different states. This is only exacerbated by the inevitable changes to the state’s economic power that stem from differential growth rates across countries, which creates yet another destabilizing force (159). The state with the faster growing economy will inevitably rise in the ranks of powers, and with this rise, its ambitions will grow. Short of the complete destruction of state power, a function of war that is not much engaged with in the book, it is unclear what could stop a rising state from being driven to take all in sight, especially if these trend lines are considered alongside the foreign policy prescription that great powers are best served by accommodating rising challengers (198).
Kirshner presents a compelling depiction of classical realism which integrates factors from all three levels of analysis, making moot the typical charge laid against classical realism that it seeks to explain the variable behavior in world politics with the one constant of human nature. The classical realism articulated in this book is broader and richer, charging scholars to take note of the historical experiences that motivate state action, the different levels of social cohesion, different ideologies, and different pursuits of happiness. This version of classical realism requires “dirty hands” but can present a more holistic view of world politics (216).
This resurrection of classical realism does leave one vexing question—is classical realism a paradigm or a theory? At different points in the book, it seems that classical realism has to serve both functions. The introduction and chapter 1 strongly suggest that classical realism is a paradigm. And kudos must be given for Kirshner’s calling out the necessity of paradigms (2). Common assumptions in world views exist whether we chose to acknowledge them or call them irrelevant (all part of classical realism seeing the world as it is). Classical realism, as described in these chapters, closely fits the designation of a paradigm—it provides a set of general assumptions and agreed upon concepts to serve as a solid base for scholars to build their theories around. The question of paradigm or theory remains, however, in chapters 3 and 6, where Kirshner uses classical realism to explain specific foreign policy choices made by decision makers—British Prime Minister Neville Chamberlain’s reliance on ideology and memory of World War I to select appeasement, United States’ hubris leading to Vietnam and Iraq wars—and to suggest best practices to the United States in the face of a rising China. He also directly compares classical realism to other theories (not paradigms) in its explanatory power regarding these events, suggesting that he sees classical realism not as a paradigm but as a theory in its application.
The crux of the issue is that paradigms organize, but cannot explain, specific outcomes.[34] Explanation is the purview of theory. For example, liberalism, the paradigm, does not have an explanation of conflict, or lack thereof.[35] Specific theories under that paradigmatic umbrella, however, do. Democratic peace focuses on regime type of states as an explanatory variable to account for a lack of war or militarized interstate disputes, depending on the specific permutation of the theory.[36] Economic interdependence theory focuses on the level of commercial or financial intercourse between states as the explanatory variable that accounts for lack of conflictual behavior.[37] Theories can explain specific outcomes in world politics. But the best that liberalism can do—as a paradigm—is to acknowledge that cooperation between states is not only plausible, but likely given necessary conditions. Liberalism also does not have an answer to why Chamberlain chose appeasement in 1939; as a paradigm, it doesn’t have to.
Because paradigms and theories serve different roles, they have different requirements. Classical realism, the paradigm, is expertly elucidated in the book, but is also overstretched when used as an explanatory tool of specific foreign policy outcomes. As a theory, classical realism might provide insight into the chosen cases, but it is underspecified if it is meant to be used for analysis. If classical realism is a paradigm, it would be good to know which specific theories fit under its umbrella. Scholars who wrote before the twentieth century certainly influenced classical realist thought,[38] but as the book suggests, they cannot be described as “IR Theorists” (31). The birth of structural realism was in large part a response to the conventional wisdom of the time that the field of International Relations cannot have a unified theory of politics. And neither E. H. Carr,[39] nor George Kennan,[40] nor Hans Morgenthau,[41] (not even Thucydides[42]) sought to provide one. Of the scholars mentioned in the book, Robert Gilpin does provide a theory of world politics,[43] however, Kirshner considers his contribution too focused on structural factors to be classified as classical realism (106-8). It would have been a very useful exercise to populate the paradigmatic umbrella of classical realism.
If classical realism is also meant to serve as a theory, several big questions require an answer. What outcomes does classical realism focus on? “World politics” (1) is the frequently mentioned outcome in the book, but it is too broad for a theory as it encompasses virtually every decision made by a state—not all of which are likely to be comparable. How does it aggregate factors across the levels of analysis? When do structural, domestic, and individual factors matter, to what extent, in what proportion? For example, when do non-material human motivations such as fear, honor, and hubris overwhelm the lessons learned from history (the US in Vietnam, [113-123], versus all bids for hegemony that failed, [187])? Likewise, when do domestic interests destroy society (France before World War II, [238]) or when can the state ignore domestic concerns in favor of the national interest (160). When does ideology prevent sound decision making (Britain in 1939, [99-103]) and when can states ignore ideology for mutual gain (Germany and Russia in 1939, [104])? What history is relevant in specific decision making? When looking backwards in time, scholars have the advantage of archival material to use as a guide. However, even then, different leaders draw on different analogies depending on which outcome they seek to promote. But when looking forward in time, and anticipating how a state might behave in a specific crisis, how would scholars know which historical antecedents are the right ones to focus on?
Since the book strongly suggests that classical realism is a paradigm, these questions can remain unanswered. A paradigm provides the fertile ground of starting assumptions on which scholars build theories to explain behavior in world politics. A theory under the classical realism umbrella, however, would have to grapple with these questions in order to provide an internally consistent causal logic and still remain true to its paradigmatic roots.
This connects to a deeper philosophical question of what is the purpose of inquiry. A central point in the book is to caution scholars against following the natural sciences in their pursuit of prediction. There is much to be admired in this point of view. Point predictions for human behavior, especially those which are based on simplistic models with relatively few explanatory variables, are assuredly difficult. Such predictions can, at best, approximate the natural sciences when surrounded by large confidence intervals making the point predictions largely pointless.
The book suggests that social scientists should focus on anticipation rather than prediction. To predict is to make statements such as “there is a 70 percent chance you will win this war” (53). To anticipate is not to “attribute some probability to that outcome but to call attention to its plausibility and likely consequences” (53). How predictions are made varies from theory to theory. But expectations such as “70 percent chance of winning a war” tend to be expected values, which are, by definition, the likely consequences of an outcome multiplied by its likely chance of occurrence summed over the range of possible outcomes. The 70 percent incorporates both the “likely consequences” and their “plausibility”; prediction incorporates anticipation. This makes it hard to see the difference between anticipation and prediction. Kirshner argues that “anticipation is not about guessing outcomes of political conflicts it is about looking past the resolution of that conflict toward the plausible range of cascading political consequences that will follow” (53). Possibly, it is the specificity of 70 percent that it is the problem, as it assumes an underlying distribution of likely outcomes. However, the outcome of an event is fundamentally important for shaping its consequences. A victory in war will lead to radically different consequences than a defeat; a pyrrhic victory will lead to different consequences than a materially beneficial one. It seems, then, that anticipation must incorporate prediction.
But what if we take prediction off the table completely? As the book suggests, prediction provides “the average behavioral response of a random actor drawn from large population making similar choices” (56). International relations, on the other hand, is concerned with “the behavior of a particular state at some significant moment in time” (56). A general trend cannot account for the individual variations of a specific state at a specific point in time. That is true. And yet, to remove all prediction—all generalization—from the enterprise begs the question: what is the goal of inquiry in political science? Yes, without generalization and with careful attention to all the relevant factors in a specific case (even if the same factors are not relevant in similar appearing cases), it is possible to describe past choices made by states. But what goal does description serve, if not to teach us something about the future?
Unfortunately, since the future is unwritten, the only way to learn something about it is through general trends. As Kirshner quotes the British philosopher Carveth Read, “It is better to be vaguely right than exactly wrong” (5) But having some conception of what a foreign policy could be built on—even if it turns out to be completely wrong—is better than not having any expectations at all, for lack of knowing which specific factors will be relevant to the specific case of an individual state at a significant time. World politics is a difficult business. It is hard to make even vaguely right decisions, because all decisions in politics are about the future, which is unknowable. Generalizations allow some limited, yes predictions, about the potential shape of the future. Without them, statesmen are walking blind. If the purpose of studying human behavior is to improve the quality of human life—that is to avoid the mistakes of the past—then projecting past trends into the future seems to be a necessary evil.
To connect this discussion to the theories under the umbrella of the classical realism paradigm, it seems that their answer to most of the questions posed above of classical realism as a theory would be some variation of “it is contingent on the situation” (58). After all, classical realism requires careful attention to historical, social, and domestic political considerations of the moment. In practice, as a descriptive exercise that is perfectly fine—dirty hands and all. But it is the very opposite of theory. A theory is a simplification of reality, not a carbon copy of it.[44] Theory requires prioritizing some explanatory factors above others, with the explicit acknowledgement that this leads to imprecision and some false positives. That is the cost that theory pays to provide the general trends that are necessary to guide future behavior. If such generalizations are not permitted, and if all factors of the moment must be considered to understand the situation correctly, there can never be a theory under the classical realism paradigm. And without theory, there is no explanation, no roadmap for the unwritten future.
Review by Barry Posen, MIT
Jonathan Kirshner’s An Unwritten Future seeks to affirm the explanatory power and policy-analytic utility of “classical realism” in contrast to its modern progeny, “structural realism.”[45] He defends the former extensively and subjects the latter to sustained, at times biting, criticism. Integrated with the theoretical debate are substantive discussions of globalization, the history of appeasement in France and Britain, and the emerging US-China strategic competition.
Different Theories; Different Aims
It is hard to assess the competition of structural with classical realism without reviewing the theoretical purposes of the two schools. At several junctures in the text, Kirshner correctly explains their different purposes. This is ultimately frustrating for the reader, for Kirshner’s actual complaint is not with the structural realist enterprise itself. Rather, he believes its exponents have exaggerated structural realism’s explanatory power and its ability to provide useful real-world guidance.
Both structural realism and classical realism aspire to help us understand international politics, but they disagree on what “understanding” international politics means.[46] To use an analogy given us by the history profession, structural realists are akin to the “lumpers” looking for big and recurring causes, and classical realists are more like the “splitters,” who see history as one damn thing after another. Structural realists look across time and space to explain broad patterns of repeated behavior with a single persistent cause—the anarchical condition—whose effects are shaped by the distribution and sometimes the nature of the material elements of state power. By contrast, classical realism is committed to a holistic understanding of international politics at any given time and place, privileging whatever causes seem most important to explaining particular “world-events” (e.g., world wars, global economic crises, etc.) whether the pertinent causes are to be found in human nature, the character and beliefs of key leaders, the nature of individual states, and also anarchy, the same structural variable that structural realism privileges (41, 44-45).
Kirshner claims greater explanatory power for classical realism, but his exposition reveals that it achieves that power by proliferating a bevy of additional variables. This move turns the theory into a shape shifter; there is no way to assess its explanatory power across time and space, because enough variables can always be added to the theory to explain whatever event is under scrutiny and the variables will change according to each theorist’s taste. But as he often notes in the text, Kirshner is a “classical realist” in part because he fully understands “the ominous consequences of anarchy as the fundamental point of departure for understanding international relations.” This, of course, is precisely the central concern of structural realism (13-14).
Classical realism got its start as advice to statespersons on how both to understand and to navigate their world.[47] Its principal theorists were all deeply steeped in the study of military and diplomatic history and their theories do not fall far from this tree. They were all reacting to various scourges of their times: the Peloponnesian war, the Thirty Years War, the Napoleonic Wars, the Great Depression, the First and Second World Wars, the Cold War and the accompanying nuclear age. All blend, according to their own lights, observations about international politics, leaders, economics, societies, and statecraft. But as Kirshner notes repeatedly, the foundation on which classical realism rests turns out to be the condition of anarchy, even if the ways in which states respond to that condition can be affected by a host of other factors.[48]
Structural realists, by contrast, are trying to understand the most with the least, a fundamental scientific value. This value can be fetishized of course, as Kirshner warns. Although they all acknowledge that domestic politics, leadership, and other “unit-level” factors can be important, structural realists are trying to focus on the “international,” not what one country wants, but on how multiple countries influence each other to produce results that tend to recur. This is why, in part, they use the term structural. They are trying to milk as much explanatory power as possible out of the fact that states cohabit an anarchical system, a world without a sovereign. As John Mearsheimer likes to say, “there is no higher authority to come to their rescue when [states] dial 911.”[49] Structural theories are theories that privilege as a cause the things that actors cannot change, even if they wanted to do so. They can either adapt or not, and if they do not they will pay a price. In anarchy you either look after yourself, or the fact that others are not constrained will allow them to prey upon you. Anarchy gives others an incentive to prey on their neighbors because the main answer to insecurity is power, and if you can grab more abroad you will be tempted to do so. Anarchy shapes the behavior of all actors. Thus, great states that get grabby often encounter balancing behavior, or outright war.
For structural realists, the nature of inter-state competition is also influenced by the number of great powers. Kenneth Waltz observed that a two-actor structure produces patterns of interaction different from three or more. In multipolar systems the potential to aggregate power through alliances permits rapid changes in a state’s security, and thus all are incentivized to play that game. In bipolar worlds, allies do not matter much—the real balancing is internal to the state and mostly consists of military competition.
Structural realists have been somewhat desultory in figuring out whether there are or are not other elements of structure that matter aside from polarity. Are nuclear weapons a structural phenomenon? Many theorists, including me, would say yes, but other realists disagree.[50] Prior to the invention of nuclear weapons, was there such a structural property as the “offense-defense balance”? Despite much useful scholarly criticism of this concept, I would still say yes.[51] Is geography an attribute of structure? Actors can change it only with difficulty, and it strongly affects the projection of military power so I think it is, though over time its influence has changed due to gigantic construction projects such as the Suez Canal, and changing technology such as the ballistic missile.
What does structure explain? Structural realists should be more disciplined than they have been when they answer this question. Structure explains recurring behaviors, not a specific action or event. Anarchy encourages competition. Competition produces emulation and socialization. These terms mean that states that are living in an anarchical system tend to eye one another warily, act in certain ways, and often start to resemble on another. States that do too much will often meet opposing power and states that do too little will be exploited. In multipolar systems, alliance-making (and breaking) is a more or less constant practice. States which are concerned about their security have little choice but to play this game, because of the possible speed and magnitude of changes in the balance of power that coalitions produce. Another typical practice is the pursuit of spheres of influence and also their recognition by others. Likewise, preventive war is a behavior that recurs, as Morgenthau also notes somewhat cryptically, deep in Politics among Nations.[52] States are sometimes so concerned about their relative power that they occasionally become fixated on a competitor’s rate of economic or military growth, or the number of allies it seems to be collecting, and they “balance” in advance by striking now rather than later.[53]
States will also come to look alike. War made the state and the state made war, as Charles Tilly observed in the Formation of National States in Western Europe, exploring insights from Max Weber and Otto Hintze.[54] Tilly and his co-authors showed us how in the cockpit of Europe states emulated successful practices—they all became experts in the extraction of resources from their societies and the mobilization of those resources for war.[55] My favorite emulated domestic structure was the “general staff system” of military command, which spread across Europe and the world in the late-nineteenth and early twentieth century. To say that states become similar is not to say that they become identical. The “general staff” system worked somewhat differently in every major state.
Again, structural realism makes predictions about patterns and recurring tendencies emerging from the problem that anarchy creates. As Waltz succinctly put it, “Structures shape and shove; they do not determine the actions of states.”[56] Those of us who wish to work with structural theories empirically therefore face a problem. Once you have catalogued the predicted behaviors, and shown that they are quite common across time and space, what is next? There will always be a scholar grounded in a particular place and time who will argue that the observed behavior is the product of unique qualities and that it is merely a coincidence that structure seems to be an explanatory factor. The fact that the same behavior turns up elsewhere is often explained not by the common problem states face, but by inside-out theories of social, political, or economic development. Perhaps for this reason, structural realists have not done a very good job of cataloguing and explaining the behaviors they predict. I once argued that nationalism and the mass army spread across Europe in a fashion that was “mutually causative.”[57] If other powers had a mass army, you had better have one too. If you needed a mass army, you had to fire up the soldiers who would fill it before their sacrifices on the battlefield were needed, and you could not wait for a crisis to do so. A structural theory cannot explain the existence of the preconditions for nationalist politics, but it can explain the propensity of different states to embrace nationalist ideologies, purvey them through the national educational system, and connect the individual’s welfare to the group, and the group’s welfare to the army and to war.
Structural realists like to dabble in explaining particular national decisions, which is not the theory’s main purpose. Structural realism cannot be mobilized to explain 100 percent of the particular actions of particular states. Indeed, it is often hard to tell what share of any given behavior should be attributed to structural variables. That task requires exactly the kind of research that Kirshner commends. Even so, structural realism can be mobilized in particular cases to illuminate the way that anarchy influenced a decision or practice. Careful exploration of the role of structural factors in particular state decisions can strengthen or weaken our confidence in the “shaping and shoving” power of anarchy, and may uncover patterns of interaction with other variables of which we may have been previously unaware. Jack Snyder and Stephen Van Evera studied the influence of military organizational and institutional biases on perceptions of the offense-defense balance in Europe before the first World War, and observed a strong propensity to see offensive advantages where more dispassionate observers would not.[58] Anarchy tells us why perceptions of offensive advantage would be particularly problematical. Focused attention on professional militaries and their relationships with their civilian masters, domestic structures that were forged and sustained by inter-state competition, produced valuable insights into one cause of the First World War. We should be wary of similar problems that we might face in the future. A classical realist might claim this work for their school, because it looks at variables beyond international structure. But the work was guided in the first instance by insights that were grounded in structural realism, and also disciplined by its focus on a small number of state level variables, the kinetic tools of the state to which anarchy gave birth and which anarchy sustains.
Structural theories can also help us understand why states with very different domestic characters sometimes act in strikingly similar ways—as the United States and Soviet Union did during the Cold War—and why a state’s grand strategy and foreign policy can change rapidly while its internal character remains constant. The United States offers a telling example: it went from being a lightly armed buck-passer and offshore balancer in the 1930s to being a global superpower committed to anti-Soviet containment in the late 1940s. Its domestic structure, political ideology, and governing elite did not change in this period, but its position on the global structure of power did.
Since classical realists are trying to understand the most, they don’t care much about “with the least.” This leads them to privilege additional causes beyond structure; indeed, it leads some to throw in whatever important influence they happen to have observed. Hans Morgenthau devotes many pages in the middle of Politics Among Nations to nationalism and what he perceived to be its pernicious effects on statecraft.[59] Morgenthau and Kennan are both skeptical that democracies will be very good at statecraft, because mass politics constrains the wise practitioners of classical diplomacy.[60] Kirshner encourages us to think about “hubris” as an independent variable, though its close connection to unbalanced power inclines one to credit “structural realism” with its parentage. Morgenthau’s guide to measuring the power of a state includes both material factors such as wealth and military capability that are the main concern of structural realists, but also non-material factors such as national morale and diplomatic skill, which are hard to measure, but which would be a concern for any experienced statesperson.
One could try to assess the relative merits of these two forms of realism by asking what each predicts and comparing those predictions against the historical record, but Kirshner in effect rejects the discipline of testing because he rejects prediction. [61] In his view, politics is too fluid an enterprise to permit prediction—almost any political outcome can happen. Scholars’ testing methods for both of these theories are a subject of endless debate, as are the results of our tests. Thus, everything we think we have discovered has some uncertainty attached to it, but discipline and repetition and diversity of methods can give us more confidence in our insights. Because classical realism aspires to utility as a policy analysis tool, this problem cannot be wished away. A policy is a bet, and realists of every stripe, if they wish to advise, should try to assess systematically the explanatory power of their theory or theories.
Appeasement Re-Revisited
Despite his ambivalence about testing rival theories, Kirshner revisits the historical battlefield of appeasement in the 1930s in order to advance his case that classical realism is “better” than structural realism at explaining the policies that helped produce the second world war. In a way this “test” is too easy, for who can study the 1930s without wanting to scream through the pages at the actors, “don’t you know where this is going?” Of course we do know where it was going—now—and though some did at the time, many others did not, most notably the statespersons who were in charge. Kirshner works hard to locate the explanation for their strategic mistakes in non-structural variables of the kind that structural realism leaves out, including the memory of the hideous bloodletting of the First World War; the deep social, political, and economic divisions in France; and the political naivete and arrogance of British Prime Minister Neville Chamberlain. These features are all important to a full understanding of appeasement, but they do not, as Kirshner’s narrative suggests, somehow telling us everything we need to know.
In particular, Kirshner’s discussion of the period omits facts limned by structural realism that classical realism would acknowledge as equally relevant to explaining the tragic events of the 1930s.[62] For starters, France was materially and demographically much weaker than Germany, with perhaps two thirds the population and a third of Germany’s economic potential. Balancing Germany was bound to be a fraught business for its leaders, and they wanted to be damn sure they had a powerful ally on side if they were going to fight the Germans. That meant Britain, though one can argue about whether the disqualification of the USSR for this role was a “structural” fact produced by the political-geography of eastern Europe (which gave the USSR and Germany a strategic buffer), or a “political” fact that resulted from ideological divisions. British leaders, especially Chamberlain, could experiment with appeasement because Britain had a margin of safety in the English Channel and they knew that in the worst case, France was to some extent doomed by its geographical position to fight Germany on Britain’s behalf. They knew that France needed them and thus they had enormous leverage over French diplomacy. The distribution of power tells us a lot about the policies of these two states.
Structural realism predicts that France and Britain should have balanced against a resurgent Germany and they did.[63] But they did not quite do it at a pace that matched the rate of growth of German power, especially a Germany that was led by the likes of Germany’s Adolf Hitler, who is a historical personage that structural realism would not pretend to explain and that Kirshner does not mobilize classical realism to explain either. But how far behind were France and Britain? Even today, students of military history and strategy do not have a settled explanation for the disaster the allies suffered in the Battle of the Low Countries and Northern France in 1940.[64] It is fair to say, however, that the battle was not lost due to a lack of equipment or combat units or military personnel. Enough balancing occurred in the 1930s for the allies to field a large and reasonably well-armed force, and to do it together as a coalition on the battlefield. Kirshner dismisses one reasonable explanation for their defeat in 1940: the Germans had done a better job understanding technological change and assimilating those changes into their tactical and operational doctrine than did the French and British. (More effective learning is not uncommon for those who lost the last war.) France and Britain also chose a risky military plan for the campaign, one which was at variance with their-prewar operational thinking and the structure and skills of their militaries, but consistent with the logic of “buckpassing,” the propensity of states in a competitive system to conserve their own power at one another’s expense, even when they may share a threat perception. Finally, their soldiers fought hard if not well. 27,000 German soldiers were killed: someone did the killing.[65]
Structural realism is obviously not sufficient to explain either appeasement as a policy or the defeat France and Britain suffered in 1940, and Kirshner correctly highlights the negative impact of France’s internal politics. Civil-military relations in France were poisonous, and the hearts of France’s most senior commanders do not seem to have been in the battle. None of them showed the steely resolve, careful calculation, and mastery of their subordinates that Chief of the Army Staff General Joseph Joffre showed in 1914. Indeed, during the 1940 campaign, it seems that most senior commanders excused themselves from responsibility for conduct of the battle. Why was that? The question is rendered more vexing because the Dreyfus affair had revealed an equally poisonous pattern of civil-military relations in pre-World-War I France, yet the French officer corps rose to the occasion in 1914. Structural realism does not pretend to explain puzzles of this kind.
If we are honest, it is both easier and probably more fun to be a classical realist that a structural realist. And this is not always a bad analytic strategy, so long as it is practiced with discipline. For example, when structural realists take off their theory hats and put on their policy analysis and advising hats, most of them begin to morph magically into classical realists and include non-structural variables into their analysis.
It is hard to find a realist of any stripe who does not assign an important role to nationalism in international politics and especially in war. And as Kirshner observes, a state pursuing security at time t probably has both an elite and a society with a historical memory of what happened before, which sometimes involves quite horrible things that happened before. The Polish experience of Soviet conquest and occupation in 1939, and then again from 1944 to the end of the Cold War, produces what seem to be a very ambitious set of war goals in the Ukraine-Russia war, goals that seem either unreachable or downright suicidal to some realists.
The flip side is also true: some states remember nothing. Americans can barely recall that they waged a preventive war in 2003 against a chimerical Iraqi weapons of mass destruction program. Nor do they recall that during the Vietnam war they dumped ten million tons of bombs and shells onto three agrarian societies, which is several times what they threw against vastly more developed industrialized adversaries during the global Second World War. We now condemn as criminal the practices that US policymakers often found expedient.
Kirshner’s criticism is that even when they turn they hands to policy analysis, structural realists still privilege anarchy and material power. He is correct, but his comments here are misplaced. Structural realism’s emphasis on anarchy and the distribution of power is valuable precisely because there is too little of that kind of thinking in Western foreign and security policy establishments. Those of us who have taken up the effort to argue a bit of “restraint” into US national security policy find ourselves constantly up against liberal premises that are so deeply engrained in the foreign policy establishment that they have become undebatable.[66] Realists believe that spheres of influence are normal. Liberals believe that they are sinful. Realists believe that states of all kinds will worry about their security and sometimes discard their own principles in order to defeat a foe; liberals believe that democracies pursue more virtuous foreign policies than autocracies do and that most international conflicts are caused by the latter.[67] By reminding policymakers that even dictatorships may face real military threats and that the “security dilemma” applies to all states in anarchy, structural realism helps guard against the sort of Manichaen, “us versus them” hubris that classical realism also warns against.
Readers should not be left with the misapprehension that Kirshner’s open preference for classical realism and criticism of structural realism is a reason to give this book a miss. Kirshner’s account of the theoretical origins and evolution of classical realism is as good as theory-writing gets. Those who have read the classics can profit from the review; those who have not can profit even more. It is lucid, comprehensive, and sprightly. (Graduate students preparing for qualifying exams, take note.)
What about China?
The book ultimately takes up the central strategic question of our time: are China and the US destined for an intense security competition in Asia and elsewhere? Structural realism says yes. One structural realist, John Mearsheimer, has even advised both parties to embrace the struggle.[68] Kirshner rejects this argument in part because that affirmative advice does seem somewhat beyond the boundaries of structural realism’s theoretical enterprise.[69] Kirshner agrees that some level of competition is baked in the cake, but he questions whether intense competition is either wise or inevitable and he explores the plausible range of choice for each party. Donning my “classical realist” hat, I agree with him that there is room for human agency in this competition. But as we witness the evolution of the US-China relationship, it is hard to be optimistic that cooler heads will prevail. How much of the deterioration of their relationship we should ascribe to their structural situation and how much to their peculiar natures and histories is good ground to compare the two theories and we are already beginning to see the experiment unfold. That much of our future is written.
Response by Jonathan Kirshner, Boston College
Let me begin by offering my gratitude and appreciation to the editors and organizers of this forum, to Jack Snyder for his characteristically learned and thoughtful introduction, and, most especially, to distinguished scholars Stacie Goddard, Mariya Grinberg, and Barry Posen, each of whom offers serious, thoughtful engagements with and well-argued constructive criticisms of my book, An Unwritten Future: Realism and Uncertainty in World Politics. An author can ask for little more than to have his work be read with care and attention, and it is clear from their contributions that each reviewer understood what I was trying to accomplish as they honed lines of inquiry shaped by their particular interests and expertise.
Like the other contributors, Stacie Goddard efficiently and very well summarizes the main themes of the book; I am especially appreciative of her attentiveness to the crucial notion of (analytical) uncertainty, which she characterizes extremely well in an outstanding early section of her essay. She then raises one global criticism, which I will attempt to rebut, and offers three more specific challenges, which I think are spot on and should be acknowledged and engaged.
In her most potentially subversive section, Goddard asks, “Is Classical Realism the Answer?” and writes that she “struggled to understand what unique insight classical realism adds to our understanding of politics.” She notes, correctly, that its “core commitments to history, contingency, and contestation are already present in other theoretical approaches.”
I would answer this challenge with a question of my own: do we need realism? My answer is yes. I find realism to be a valuable paradigm for understanding and explaining world politics. As a paradigm, all subbranches of realism take as points of departure the consequences of the presence of anarchy, attentiveness to the capabilities of others, and an expectation of inevitable conflicts of interests between actors. Some of this may be resolved by the resort to force, and in that context there are no assurances that behaviors will be restrained, including horrifying acts of barbarism. Although realism, inevitably, has its blind spots, flaws and limitations, when I look out the window this worldview is consistent with much of what I see in contemporary world politics. Thus I do think we are well-served by having realism around—and in that context, I do think we very much need classical realism. Indeed, the purpose of my book is to reclaim realism, arguing that the modern embrace of structural realism, and, distinctly, of hyper-rationalist schools of thought such as the influential “Rationalist Explanations for War” perspective and those that would integrate its core assumptions, has led us down blind analytical alleys.[70]
Goddard acknowledges that I am opposed to paradigm wars, which are intrinsically unproductive. Moreover, regarding the discipline of international relations most generally, my project is not evangelical. Everyone should not be a classical realist; that would be a recipe for intellectual stasis and even disaster. Intellectual monocultures are unhealthy—indeed, in macroeconomics, an intellectual monoculture contributed to the Global Financial Crisis.[71] Yet even as paradigm wars are to be shunned, we must recognize that paradigms themselves are inescapable. Many scholars like to assert that their work is non-paradigmatic, but it is not—all theories begin with a bundle of assumptions, dispositions, and worldviews that select certain variables, events, measurement styles and even questions in, and others out. This is necessary and not a bad thing. But paradigmatic dispositions should be recognized, so that we can understand those crucial analytical choices that are made, often implicitly.
Thus classical realism is not the answer, if the question is whether everyone should be a classical realist. But if the question is “should we have realism?” then my project of reclaiming realism from the wrong turn in took it the wake of the structuralist revolution comes more clearly into focus. In that context, although Goddard is correct to observe that some other approaches might share an emphasis on the influence of historical legacies with classical realism, that does not mean that we ought to drop realism as redundant. It is not. It simply means there are elements of paradigmatic approaches that can overlap. For example, Marxist theories of international politics, which can be similar to those of realists in general and classical realists in particular, place great emphasis on shifts in the distribution of power as an engine of international conflict, rooted, from a Marxist perspective, in the pathologies of capitalism and the “law of uneven development.”[72] Yet there are also vast differences between the two paradigms, and there is no reason why one should subsume the other.
In Goddard’s more specific criticisms, I want to call attention to three issues, which are very much on target, and, for me at least, the source of enormous, ongoing, intellectual stimulation. Goddard notes my emphasis in chapter 7 on the role of domestic social cohesion in explaining the ability of states to meet foreign policy challenges. She queries, fairly, whether my own pessimism about the contemporary United States is adequately grounded—how we know how much social fragmentation is consequential. As she notes, “Exploring how and when leaders manage to mobilize in the face of contention is a topic worth exploring in more depth.” I fully agree that this topic is not discussed in An Unwritten Future; it is a vital question for my own future research agenda. Another critique—which is a perennial challenge to realism in general and classical realism in particular—has to do with the realist respect for power, the power of others, the limits of one’s own power, and how changes to the balance of power must be understood in that context. As Goddard observes, the classical realist instinct is “to dispassionately recognize and accommodate changes in the balance of power.” But she raises two enormous and enormously consequential reservations about that instinct. The first is the relatively familiar question, “How do policymakers know when they have gone too far in accommodating power?”
There is no easy answer, and no off-the-shelf universal law that can confidently be applied to all such cases. This is why classical realists tend to focus more on changes to the balance of power than to its static distribution as the engine of international conflict; it raises questions that are perennial, perplexing, and profound. As I stress in the book, managing the emergence of a new great power in the international system is perhaps the single greatest challenge to stability in world politics.
Finally, in this context, another issue arises, as Goddard slips in a crucial concept. She asks about “the ethics of accommodation,” and what classical realism has to tell us about “when accommodation is the prudent and ethical choice” (emphasis added). This is no small query. As I discuss in chapter 4, “The Limits of Classical Realism,” (which, as Mariya Grinberg astutely observes in her commentary, is an important and integral part of the book), too many realists imagine that they can escape ethical questions, conflating the realist observation that world politics is often a sphere in which actors behave amorally with the position that therefore one can check their ethics at the door, that is, the national border, and practice their own foreign policy without regard for moral concerns. But the latter does not follow from the former.
As Arnold Wolfers observed, states may have a moral obligation to assure their survival, but “nations engaged in international politics are faced with the problem of survival only on rare occasions.” Most of the time, most foreign policy practices—especially those of great powers—are directed at the pursuit of a much broader range of national interests. This also considerably broadens their discretion regarding what policies and objectives they will choose to pursue, which is a crucial point I will revisit when engaging Barry Posen’s contribution. As Wolfers notes, “attempts to evade, silence or ignore moral judgment merely play into the hands of those who relish the uncriticized use or abuse of their power.”[73] Thus, although states do have a responsibility to protect their citizens and their interests, their foreign policies should nevertheless reflect their national values, even though those policies are necessarily circumscribed by the realities of relative power. Accommodation may thus mean ceding political terrain to actors whose conduct we abhor. That may at times be necessary, but Goddard is right to hold realists’ feet to this fire, and suggest that we cannot simply dodge these vital, even existential, issues by hiding behind the skirts of prudence.
In her commentary, Mariya Grinberg also paints a clear picture of what I am trying to accomplish with An Unwritten Future. And I very much appreciate her appropriate, summary rejection of the all-too-common trope that dismisses classical realism as attributing “behavior in world politics” to “human nature,” a charge which is leveled by those unwilling or unable to seriously engage the perspective. In what strikes me as a modest miscommunication, Grinberg suggests some confusion in my use of classical realism as a paradigm as opposed to a theory. I think this particular concern is overstated, although perhaps I could have been more explicit in drawing the distinction. I agree that classical realism is a paradigm or a branch of a paradigm, and not a theory. Nevertheless, theories can be generated from a paradigmatic tradition, and, as Grinberg notes, in the illustrative examples I do draw on such theories. An important distinction here, and returning to the folly of paradigm wars, is that those theories can be evaluated by criteria that are established a-priori, and thus found wanting or valuable—but the paradigm, or any paradigm is not a theory itself and cannot be falsified, although it surely can and should be critiqued.
To my eyes, the more important challenges raised by Grinberg’s review have to do with prediction, generalization, and the much-dreaded “kitchen sink” argument that classical realism throws every conceivable explanatory variable into mix. These are found in her assertion that “anticipation must incorporate prediction” and that classical realism renounces the search for generalization, and without at least the aspiration to generalization, all is lost. As she writes: “If such generalizations are not permitted, and if all factors of the moment must be considered to understand the situation correctly, there can never be a theory under the classical realism paradigm. And without theory, there is no explanation, no roadmap for the unwritten future.”
These are very important and well-articulated potential objections. That I have heard them before (often in vigorous and extensive private correspondence) is not a critique; it reflects on my own apparently unclear exposition. Since I do not think these criticisms are on target, the fact they are often raised plainly suggests that I have not been adequately clear. My positions are: that the ability to make generalizable claims is important, and classical realism aspires to do so to the extent possible; that classical realism does not imply that all factors must be considered at all times, rather, the craft of classical realism lies in the wisdom to know which factors are likely central to understanding a given situation; and that anticipation is fundamentally different from prediction.
It remains the case that as implemented, these three attributes diverge from much contemporary IR scholarship. These differences are rooted in the disposition of classical realism which sees an unbridgeable gap between the social sciences and the natural sciences, with one consequence of that yielding a profound skepticism about predictive acuity in IR, and from there, the embrace of what can be called a “partial equilibrium” rather than a “general equilibrium” approach to the study of world politics. On these themes, and on uncertainty, which I will revisit in my discussion of Posen’s contribution, classical realism draws its inspiration from seminal figures in economics—none of whom as a result of their perspectives simply shrugged his shoulders and announced “well, everything matters.”
Another wrong turn in IR that I bemoan, here especially in the company of Hans Morgenthau, is the aspiration towards “scientism” in our endeavors.[74] Too many political scientists worship neoclassical economists, too many mainstream economists worship Newtonian physics. But this aspiration is misguided; the gap between the social world and the physical world is unbridgeable. A volcano does not care whether or not it erupts. Geologists can make only the vaguest of “predictions” about earthquakes, despite the fact that tectonic plates shift unburdened by the complexities of politics and purpose. They cannot be inhibited, or not, by deterrent threats. The moon is indifferent as to whether or not spacecraft will land on it; it will not throw a head fake to try and prevent that from happening, or be influenced by debates between well-intentioned domestic communities with distinct interpretations regarding the implications of the approaching others.
John Maynard Keynes drew this distinction forcefully and repeatedly. Emphasizing the “sharp difference” between the natural and the social sciences, he warned that “The pseudo-analogy with the physical sciences leads directly counter to the habit of mind which is most important for an economist to acquire.” Social science deals with “motives, expectations [and] psychological uncertainties,” which do not exist in the natural sciences, requiring a fundamentally different approach to studying one as compared with the other. “One has to be constantly on guard against treating the material as constant and homogenous,” he admonished. “It is as though the fall of the apple to the ground, depended on the apple’s motives” as well as “on mistaken calculations on the part of the apple as to how far it was from the centre of the earth.”[75]
Otherwise virulent anti-Keynesians Frank Knight and Friedrich von Hayek also emphasized this difference between the natural and social world, and it was at the root of why they were so profoundly skeptical of chasing prediction in economics. Knight stressed “the inherent, absolute unpredictability of things, out of the sheer brute fact that the results of human activity cannot be anticipated.” One important source of this “is the variation in the power of reading human nature, of forecasting the conduct of other men, as contrasted with the scientific judgment in regard to natural phenomena.” Hayek insisted that “in the study of such complex phenomena as market,” economists could expect to offer no more than “only very general predictions about the kind of events which we must expect in a given situation.”[76]
Note that Keynes, Knight, and Hayek each engaged in elaborate and sophisticated theorizing, and they were not shy about proffering (typically divergent) policy advice that was rooted in their view of economics. Nor did they say that because the world is complicated and humans are unpredictable that all variables matter and we have to abandon the aspiration to generalizability. I stress this point because in his review Posen joins Grinberg (and other constructive critics of classical realism) in pressing it, first with his lovely turn of phrase about how classical realism achieves its “power by proliferating a bevy of additional variables,” and, more sharply, with the claim that some within this school “throw in whatever important influence they happen to have observed.”
Again, I don’t think that this is what classical realism is about. Rather, in the spirit of analytical modesty, it embraces an approach associated with Charles Kindleberger, and his work in historical economics. Kindleberger, who was very much not a realist, but an enormous intellectual influence on my work, argued that “there is not one all-purpose economic theory or model that illuminates economic history.”[77] Instead—and here classical realism could not agree more—he reached for a “toolbox” analogy, in which experienced economists could draw on a set of theories and analytical devices to be applied judiciously as appropriate to the particular case. Generalizability is to be aspired to—but in this context. As he wrote, Historical Economics “believes in partial equilibrium…and looks for patterns of uniformity but is wary of insisting on identity.”[78] Classical realism similarly reaches for a trusty tool kit, which is full of analytical devices that are capable of widespread application but are introduced as appropriate for the demands of each specific, idiosyncratic, and unique situation. In a similar spirit, Kindleberger, as a specialist in international economics who often participated in debates about IR, with regard to world politics again stressed a partial equilibrium approach: “it is futile to spend time at the over-all level,” because “the total system is infinitely complex with everything interacting. One can discuss it intelligently, therefore, only bit by bit.”[79]
Classical realism, then, does not insist that “everything matters,” but strives for the wisdom to judiciously apply the appropriate tools to a given situation—which, yes, will vary. But general patterns can be observed, and they can inform anticipation, which is indeed distinct from prediction. Anticipation stresses notions like “this commonly follows from that,” which is very different than assigning, even loosely, probabilities to all possible outcomes. As I will discuss below, frequently in IR we don’t even know what all the outcomes might be.
Drawing on that tool kit to anticipate what might happen next is perfectly compatible with classical realism. At the height of the early military successes of the Iraq war, and at the peak of that war’s domestic popularity, I wrote that the war “was very unlikely to achieve, and in fact would probably undermine, the broader political objectives for which it was fought,” and that eventually “a fatigued and impatient America,” would finally distance itself from “the chaos that ensues” (121). Classical realists understand that other outcomes were possible—a respect for the roles of contingency and chance that goes all the way back to Thucydides. For Thucydides, the Sicilian campaign was a catastrophic blunder, and the cause of Athenian ruin. If there is such a thing as a “Thucydides Trap,” it is found there, in the war’s sixteenth year—and plainly not at the initial commencement of hostilities. Yet Thucydides also tantalizingly suggests that the Athenians could have been successful in that effort. What was clear about Sicily, and Iraq, were that they were invitations to disaster. But they each could have gone better. I could have been wrong about the outcome—but not about the invitation.[80] And that is the craft of classical realism: not to make predictions, even probabilistic ones, but to offer important insights into the likely plausible range of consequences that will follow from one choice or another.[81]
Barry Posen offers an incisive and wide ranging engagement with An Unwritten Future as befits his status as one of the eminent patricians and renowned teacher-scholars of our discipline. And it is not surprising that in his defense of structural realism, Posen finds much to agree with in classical realism—these are, after all, sub branches of the same paradigm. And as he generously and wisely notes, classical realism “is not always a bad analytic strategy, so long as it is practiced with discipline.” Nevertheless, the thrust of his commentary appropriately focuses on areas where our perspectives diverge; similarly, in my rejoinder I will dwell on areas of disagreement between us, which should not be misunderstood as a lack of appreciation for, and agreement, with many of the arguments he expresses. To take one example among many, alertness to the perceived offense-defense balance is an indispensable analytical arrow in the classical realist’s quiver, which is not to try and wrest that variable from the structuralists, should they wish to lay claim to it.
This observation speaks to a modest dissent I would register with Posen’s commentary. I think it leaves the impression that classical realism does not take seriously structural variables. This is not the case, and is well illustrated by our treatments of the China challenge, where we are not as far apart as he might suggest. As Goddard summarizes, “Kirshner is no optimist about the future of US-China relations.” Moreover, that pessimism is rooted in a systemic level variable: changes to the balance of power. It is true that I am a vehement critic of John Mearsheimer’s structuralist “theory of offensive realism,” but that is because the theory, as I elaborate in chapter 6, is logically incoherent on the basis of its own premises, and proffers misguided and catastrophically self-defeating policy advice to both China and the US.[82]
That said, classical realism, which appropriately reaches for a systemic level variable to apply in this instance, is even more pessimistic than its structuralists cousins. This is because both “defensive” and “offensive” structural realists see “security seeking” as the primal engine of behavior.[83] Problematically, by that logic China should present less of a disruptive challenge to world politics, not more, as China today is orders of magnitude more secure than it was decades ago. In contrast, one of the most oft-deployed tools in the classical realist kit is the expectation that states will tend to get more ambitious as they grow more powerful. Notably, classical realists then add a non-structural variable into this mix, arguing that this will be an especially difficult problem to resolve, due to the role of hubris on both sides. Which is also why I agree with Posen that it is “hard to be optimistic that cooler heads will prevail”—but taking that temperature is more of a classical than a structural move (to say nothing of the hyper-rationalists).[84]
The most basic point of contention between my argument and that of Posen regards how much structural variables can tell us on their own. For classical realism, structure is an essential part of the story, but on its own, it is irretrievably indeterminate, and thus needs help. That indeterminacy derives from the market metaphor (described below) that systemic theory draws inspiration from, and, as suggested by the title of my book, the consequences of analytical uncertainty.
Structural realism was built on a market analogy: the international system creates pressures and imposes constraints on its actors—and as with market forces, these pressures derive from the collective behavior of all actors, but they are beyond anyone’s control. But that metaphor describes what economists call “perfect competition”—which does not well describe international politics, and, especially, great power politics, which is the explicit focus of structural realism. As Kenneth Waltz argued, “A general theory of international politics is necessarily based on the great powers.”[85] Rather, as Raymond Aron observed “the structure of the international system is always oligopolistic.” This means that great powers (like oligopolists) are not simply subject to systemic pressures, but the choices they make can shape the nature of those pressures (thus, for Aron, throughout history “the principal actors have determined the system more than they have been determined by it.”)[86] As noted above, via Wolfers, great powers can safely choose their preferred policies from a menu of plausible possibilities. Those choices will shape the incentives faced by others. Moreover, to the extent that “systemic pressures” are analogous to a “budget constraint,” different states will also make distinct choices about what costs they are willing to bear to pursue certain policies. Systemic pressures can help us understand the costs—but they tell us nothing about the willingness of states to bear them.
But the analytical problem for structural realism is even greater than that. Even if we knew, exactly, the goals that a particular great power harbored, we still could not know how they would choose to pursue them. A key consequence of uncertainty is that individual decisionmakers, advocates, and scholars will disagree as to the best means to pursue a given goal. Knight is best remembered for his emphasis on “true uncertainty,” which is “unmeasurable” and which “must be taken in a sense radically different” from risk. Uncertainty brings about the “necessity of acting upon opinion rather than knowledge,” and following one’s own instincts, while trying to gauge the opinions of others for additional clues and insights.[87]
Keynes held similar views. “The orthodox theory assumes that we have a knowledge of the future of a kind quite different from that which we actually possess,” the great economist explained. “This hypothesis of a calculable future leads to a wrong interpretation of the principles of behavior which the need for action compels us to adopt, and to an underestimation of the concealed factors of utter doubt, precariousness, hope and fear.” As Keynes emphasized explicitly, one of the two pillars of his new economics—what would become known as the Keynesian revolution—was that actors in the economy made decisions in an environment characterized by uncertainty, not risk.[88]
This is not the same as noting that “we don’t know what is going to happen in the future”; it rather means that we do not—and cannot—have access to a single, shared unified theory of how the social world works. Thus actors, drawing on different (often implicit) models, will disagree on the causal consequences of different choices, and what will be the best policies to advance the national interest. For classical realism, international relations take place in the context both of anarchy and of Knightian uncertainty. The latter, which is often disregarded in IR, voids hyper-rationalist models (because rational, dispassionate experts can look at exactly the same information yet come to radically different conclusions), and exposes the keenly circumscribed limits of structural realism, left to its own devices.
Under uncertainty there will endure multiple, often competing causal theories, which means that systemic pressures will yield indeterminate outcomes. Different experts and actors will come up with very distinct policies and postures designed to ensure state security, even when faced with exactly the same external pressure, and even when seeking to achieve precisely the same goals. Ironically, Posen’s own work well illustrates this. As the most prominent advocate of a grand strategy of “restraint,” Posen argues—forcefully and cogently—that this posture will best advance the American national interest.[89] Yet some of his intellectual opponents, who were often trained in the very same traditions, and are also motivated to craft a grand strategy that will best serve the American national interest, advocate a markedly different approach. This is often colloquially termed “deep engagement.”[90] That dispassionate experts who are drawn from a relatively narrow intellectual community, observe exactly the same structural setting, and seek the same overarching objectives reach for such disparate policy recommendations well illustrates the fact that in the real world states and statesmen might choose wildly divergent courses of action, that is, international political behavior, to achieve those goals. This is the case even if we can know exactly what their goals are, which, it should be added, is not always the case when assessing potential adversaries.
Once again, although Keynes embraced uncertainty as central to his understanding of how the world worked, he did not just throw up his hands and announce “well it’s uncertainty out there and everything matters and we have no idea what’s going to happen.” Rather, he tailored his theoretical apparatus, and his extensive policy advice, to take account for that crucial, formative context. Admittedly, this is not the easiest path to follow, which was an important reason why the post-war Keynesians, not unlike structural realists, found they could build more elegant, “sophisticated” and more parsimonious theories if they shed the complicating, if for Keynes, essential, factor of uncertainty. Keynes himself died in 1946. Joan Robinson, who was one of Keynes’ most eminent students and participated in the “Cambridge Circus” of the early 1930s that was the cauldron of the Keynesian revolution, noted that dropping uncertainty in favor of more model-friendly abstractions meant that “all these pretty, polite techniques, made for a well-paneled board room and a nicely regulated market, are liable to collapse.”[91]
This returns us to the question of parsimony, the ground on which Posen makes a strong stand: “Since classical realists are trying to understand the most, they don’t care much about ‘with the least.’” I would disagree with this characterization, and argue, following Kindleberger, that classical realists care passionately about “with the least”—we just have a very different view of what “the least” can productively be reduced to. This leaves us with a potentially unproductive food fight on our hands, because my perspective sees the converse: structural realism has nailed “with the least” but at the expense of not being able to tell us, on its own, very much at all.
What does it explain? Posen argues that “Structure explains recurring behaviors, not a specific action or event.” But when will those behaviors recur? While structure surely can help explain a general, but not universal, pattern of recurring behaviors—a valuable tool of analysis which no classical realist would fail to deploy—I am reminded again of Keynes, and his admonition that “this long run is a misleading guide to current affairs.”[92] IR is commonly concerned with the particular behavior of a powerful state at a critical juncture; very rarely is there much to be gained by gleaning the median behavioral response of an imagined average state.
Posen twice invokes the notion that structure “shapes and shoves” state behavior. And again, although I agree that structure matters, its inherent indeterminacy remains problematic. I think that it does not “shape and shove” so much as it presents “costs and constraints.” The difference is that shoving implies directionality; the consequences of imposed costs, on the other hand, are incomplete without understanding the willingness of particular actors to bear those costs (or not) in a given setting.
This can be seen in studies of the interwar years, where Posen makes some very formidable arguments about state choices. But France was a front-line state, facing an obvious, proximate, and mortal threat. That is a pretty hard systemic shove. Yet in Posen’s landmark study of interwar grand strategy, he concludes that the French failed “to insure themselves against the worst.”[93] This is no small swing and miss for structural realism, which surely must aspire to say more than that great powers will take appropriate measures to defend themselves from mortal threats, except when they do not.
Let me close with enormous thanks for the generous engagement of all of the participants in this forum, and conclude with a short summary of my core claims. International politics will never be well understood if it is envisioned as social physics. Instead, following the admonition of George F. Kennan, “we must be gardeners and not mechanics in our approach to world affairs.”[94] This does not mean that “everything matters”—but it does describe a discipline where the savvy scholar needs to know what to look for in each particular case, by drawing on a well deployed analytical tool kit, and by understanding which tools are most appropriate for a given setting. The distinctiveness of classical realism rests on three pillars: (1) In IR we live in a world of anarchy, but we also live in a world of analytical uncertainty, which is just as defining for the study of IR; (2) Great powers have a range of plausible, contestable options available to them as to how best to assure their security and advance their interests, and the choices that they make matter for shaping world politics; (3) We will not be able to understand those choices in ignorance of the relevant historical context.
Ultimately, my twin responses to this most stimulating forum are that realism is a valuable paradigm for understanding world politics, and that realist approaches that do not acknowledge these three things will lead us astray.
[1] Jonathan Kirshner, Appeasing Bankers: Financial Caution on the Road to War (Princeton University Press, 2007); Kirshner, Currency and Coercion: The Political Economy of International Monetary Power (Princeton University Press, 1995); Kirshner, “Rationalist Explanations for War?” Security Studies 10:1 (Autumn 2000): 143-50.
[2] Kenneth N. Waltz’s Theory of International Politics (Addison-Wesley Pub. Co, 1979); John J. Mearsheimer. The Tragedy of Great Power Politics, (W.W. Norton, 2001).
[3] James Fearon, “Rationalist Explanations for War,” International Organization 49:3 (Summer 1995): 379-414.
[4] See the extensive footnotes to the literature in the review by Barry Posen and in Kirshner’s response.
[5] Robert Jervis, “Cooperation Under the Security Dilemma,” World Politics 30:2 (January 1978): 167-214.
[6] Fearon, “Rationalist Explanations for War,” 389-390.
[7] Barry Posen, Restraint: A New Foundation for U.S. Grand Strategy (Cornell University Press, 2014).
[8] Stacie Goddard, “When Right Makes Might: How Prussia Overturned the European Balance of Power,” International Security 33:3 (Winter 2008/09): 110-42.
[9] Mariya Grinberg, “Wartime Commercial Policy and Trade between Enemies,” International Security 46:1 (Summer 2021): 9–52.
[10] Jervis, Perception and Misperception in International Politics, second edition (Princeton University Press, 2017), 409-424.
[11] Jervis, System Effects: Complexity in Political and Social Life (Princeton University Press, 1997).
[12] Glenn H. Snyder and Paul Diesing, Conflict among Nations: Bargaining, Decision Making, and System Structure in International Crises (Princeton University Press, 1977).
[13] John Maynard Keynes, General Theory of Employment, Interest and Money (Modern Library, 2000, original ed., 1936).
[14] Waltz draws on William Fellner, Competition among the Few: Oligopoly and Similar Market Structures (Knopf, 1949).
[15] Donald A. Walker, Walrasian Economics (Cambridge University Press, 2006), 147-52; Jonathan Schlefer, The Assumptions Economists Make (Harvard Belknap, 2012), 21-38; Maurice Obstfeld, “International Macroeconomics: Beyond Mundell-Fleming,” IMF Staff Papers Special Issue (2001), 215.
[16] Keynes, General Theory, 46-51, 144-148; Schlefer, Assumptions, 137-214.
[17] F.A. Hayek, “Economics and Knowledge,” Economica 4 (February 1937): 33-54.
[18] Gavin Kennedy, “Paul Samuelson and the Invention of the Modern Economics of the Invisible Hand,” History of Economic Ideas 18:3 (2010): 105-119; Adam Smith, The Wealth of Nations (Modern Library, 2000; orig. ed. London, 1776), 484-5, on the invisible hand; Ben Fine and Ourania Dimakou, Macroeconomics: A Critical Companion (Pluto, 2016), 73; 81; Schlefer, The Assumptions Economists Make, 7-8, 142, 151-2, 259; John Gerard Ruggie, “International Regimes, Transactions, and Change: Embedded Liberalism in the Postwar Economic Order,” International Organization 36:2 (Spring 1982): 379-415.
[19] Kennedy, “Paul Samuelson and the Intervention of the Modern Economics of the Invisible Hand,” 105-119; Paul Samuelson, Economics (McGraw Hill, 1948 first edition; 2009 nineteenth edition with William Nordhaus); Andrew Schotter, Free Market Economics: A Critical Appraisal (St. Martin’s, 1985); Fine and Dimakou, Macroeconomics, 12-17; Schlefer, Assumptions, 37, 193.
[20] Andrew Glyn, Alan Hughes, Alain Lipietz, and Ajit Singh, “The Rise and Fall of the Golden Age,” in Stephen A. Marglin and Juliet B. Schor, eds., The Golden Age of Capitalism (Clarendon, 1990), 59-125 at 72-93, and v-vii; Schlefer, Assumptions, 195-214; Fine and Dimakou, Macroeconomics, 138.
[21] Andrew Schotter, Free Market Economics: A Critical Appraisal (St. Martin’s, 1985), 101-7; Mark Blyth, Great Transformations: Economic Ideas and Institutional Change in the Twentieth Century (Cambridge University Press, 2002), 6, 35-45, 180-2, especially 254, 259, on ideas reducing uncertainty in ways that make collective action possible; Schlefer, Assumptions, 201-207, 215-234, 268-73.
[22] Karl Deutsch and J. David Singer, “Multipolar Systems and International Stability,” World Politics 16:3 (April 1964): 390-406.
[23] See e.g., Christopher Layne, “The Unipolar Illusion: Why New Great Powers Will Arise,” International Security 14 (1993): 5-51; John Mearsheimer, Tragedy of Great Power Politics (W.W. Norton, 2014), 401-402.
[24] Thucydides, History of the Peloponnesian War (Penguin Classics, 1972) E. H. Carr, The Twenty Years Crisis (Harper Perennial, 1964); Hans J. Morgenthau, Politics Among Nations: the Struggle for Power and Peace (Knopf, 1948); Reinhold Niehbuhr, Moral Man and Immoral Society (Westminster John Knox Press, 2021); George Kennan, American Diplomacy, (University of Chicago Press, 2012).
[25] On offensive realism, see e.g. Mearsheimer 2014; for an example of what Kirshner sees as hyper-rationalist approaches, see James Fearon, “Rationalist Explanations for War,” International Organization 49(3), 379-414.
[26] For a succinct overview of the historiography, see Patrick Finney, “Introduction,” in The Origins of the Second World War: a Reader, ed. Patrick Finney (Bloomsbury, 1997), 12-17.
[27] See e.g., Elizabeth Saunders, Leaders at War: How Presidents Shape Military Interventions (Cornell University Press, 2011); Rachel E. Whitlark, All Options on the Table: Leaders, Preventive War, and Nuclear Proliferation (Cornell University Press, 2021).
[28] See e.g. Antje Wiener, Contestation and Constitution of Norms in Global International Relations (Cambridge University Press, 2018).
[29] Seen, for example, in the organization of international relations around subfield journals, such as International Security, Security Studies, and the Review of International Political Economy.
[30] Mearsheimer, “The Gathering Storm: China’s Challenge to US Power in Asia,” The Chinese Journal of International Politics 3, no. 4 (December 1, 2010): 381-396; Graham Allison, Destined for War: Can America and China Escape the Thucydides Trap? (Houghton Mifflin Harcourt, 2017).
[31] Kenneth N. Waltz, Theory of International Politics (Waveland Press, 1979).
[32] James D. Fearon, “Rationalist Explanations for War,” International Organization 49, no. 3 (July 1, 1995): 379-414.
[33] This puts Classical Realism on the “bring the state back in” side of the debate: See Peter Evans, Dietrich Rueschemeyer, and Theda Skocpol, eds., Bringing the State Back In (Cambridge University Press, 1985).
[34] Thomas S. Kuhn, The Structure of Scientific Revolutions (University of Chicago Press, 2012).
[35] Andrew Moravcsik, “Taking Preferences Seriously: A Liberal Theory of International Politics,” International Organization 51, no. 4 (1997): 513–53. Ironically, this article also conflates the use of the terms “theory” and “paradigm,” yet it describes the liberal paradigm, along with several theories built on the foundations of that paradigm.
[36] Michael E. Brown, Sean M. Lynn-Jones, and Steven E. Miller, eds., Debating the Democratic Peace (MIT Press, 1996); Michael Mousseau, “Democracy and Compromise in Militarized Interstate Conflicts, 1816–1992,” The Journal of Conflict Resolution 42, no. 2 (1998): 210-30.
[37] Solomon William Polachek, “Conflict and Trade,” The Journal of Conflict Resolution 24, no. 1 (March 1, 1980): 55-78; Jonathan Kirshner, Appeasing Bankers: Financial Caution on the Road to War (Princeton University Press, 2007).
[38] Niccolo Machiavelli, The Prince, trans. George Bull, Reissue edition (Penguin Classics, 2003); Thomas Hobbes, Leviathan: Or, The Matter, Forme & Power of a Commonwealth, Ecclesiasticall and Civill (University Press, 1904).
[39] Edward Hallett Carr, The Twenty Years’ Crisis, 1919-1939: An Introduction to the Study of International Relations (Harper Perennial, 1964).
[40] George Kennan, “The Sources of Soviet Conduct,” Foreign Affairs 25, no. 4 (July 1, 1947), https://www.foreignaffairs.com/russian-federation/george-kennan-sources-soviet-conduct.
[41] Hans J. Morgenthau, Politics Among Nations: The Struggle for Power and Peace, 3rd edition (Alfred A Knopf, 1963).
[42] Thucydides, History of the Peloponnesian War, ed. M. I. Finley, trans. Rex Warner (Penguin Classics, 1972).
[43] Robert Gilpin, War and Change in World Politics (Cambridge University Press, 1981).
[44] John J. Mearsheimer and Stephen M. Walt, “Leaving Theory behind: Why Simplistic Hypothesis Testing Is Bad for International Relations,” European Journal of International Relations 19, no. 3 (September 1, 2013): 427-57, https://doi.org/10.1177/1354066113494320.
[45] The book sets out to subject game theory to a similar treatment, but Kirshner seems to lose interest in this effort, which wanes as the book unfolds. As game theory is not my specialty, I will say no more about it.
[46] Kenneth N. Waltz’s Theory of International Politics (Addison-Wesley Pub. Co, 1979) is generally seen as the original structural realist work. John J. Mearsheimer’s The Tragedy of Great Power Politics, (W.W. Norton, 2001) both extends and critiques Waltz’s conception and predictions. Though there are many classical realists, the most important US work is Hans J. Morgenthau, Politics Among Nations; the Struggle for Power and Peace, 2d ed. (Knopf, 1954). See also E. H. Carr, The Twenty Year’s Crisis, 1919-1939, 2nd ed. (St. Martin Press, 1946, which was often taught in American universities. As discussed below, Kirshner’s own review of the evolution of classical realism, in chapter 1, “What is Classical Realism? Thucydides and His Descendants,” is superb.
[47] See, for example, Morgenthau, Politics Among Nations; the Struggle for Power and Peace.
[48] It might be possible to construct a “classical realist” theory without it, but such an effort would depend on verbal circumlocutions that would add up to a narrative about anarchy.
[49] Mearsheimer, The Tragedy of Great Power Politics, 32-33.
[50] Viewing nuclear weapons, and indeed the “offense-defense” balance as a structural factor, see Charles L. Glaser, Rational Theory of International Politics, The Logic of Competition and Cooperation,( Princeton University Press, 2010):41-46. Though they do not employ the term “structure,” the standard challenge is Keir A. Lieber and Daryl G. Press, The Myth of the Nuclear Revolution, Power Politics in the Atomic Age, (Cornell University Press).
[51] The best known critique is Leiber, War and the Engineers, The Primacy of Politics over Technology (Cornell University Press, 2005).
[52] Morgenthau, Politics Among Nations, 190. He writes, “Since in a balance of power system all nations live in constant fear lest their rivals deprive them, at the first opportune moment, of their power position, all nations have a vital interest in anticipating such a development and doing unto the others what they do not want the others to do unto them.”
[53] Jack S. Levy, “Declining Power and the Preventive Motivation for War,” World Politics 40:1 (October 1987): 82-107.
[54] See also H. H. Gerth and C. Wright Mills, From Max Weber: Essays in Sociology (Oxford University Press, 1946): 212-214; Otto Hintze, The Historical Essays of Otto Hintze (Oxford University Press, 1975); Chapter 5, “Military Organization and the Organization of the State;” and chapter 7, “The Commissary and His Significance in General Administrative History: A Comparative Study.”
[55] Also see Joao Resende-Santos, “Anarchy and the Emulation of Military Systems: Military Organization and Technology in South America, 1870–1930,” Security Studies 5:3 (Spring 1996): 193-260.
[56] Waltz, “Structural Realism after the Cold War,” International Security 25:1 (Summer 2000), 24.
[57] Barry R. Posen, “Nationalism, the Mass Army, and Military Power,” International Security 18:2 (Autumn 1993): 80-124.
[58] Stephen Van Evera, “The Cult of the Offensive and the Origins of the First World War,” International Security 9:1 (Summer 1984): 58-107; Jack Snyder, “Civil Military Relations and the Cult of the Offensive, 1914 and 1984,” International Security 9:1 (Summer 1984): 108-146.
[59] Morgenthau, Politics Among Nations, 96-102, 220-248.
[60] George F. Kennan, American Diplomacy, 60th anniversary edition (University of Chicago Press, 1951, 2012), 186-192.
[61] I infer this from the discussion, “IR Theory and the Predictive Fallacy,” 51-58, though he inserts in the middle of the discussion an escape clause, stating that classical realism is “committed to rationalism, causality, the search for generalizable claims, productive hypothesis testing, and establishing the criteria by which arguments can be evaluated,” but it “redirects effort away from prediction….” I don’t see how one can have the former without the latter, 52.
[62] Posen, The Sources of Military Doctrine: France, Britain, and Germany Between the World Wars (Cornell University Press, 1984).
[63] See Stephen M. Walt, “Alliances, Threats, and U.S. Grand Strategy: A Reply to Kaufman and Labs,” Security Studies 1:3 (1992): 448-482.
[64] Martin S. Alexander, The Republic in Danger: General Maurice Gamelin and Politics of French Defense, 1933–1940 (Cambridge University Press, 1992); Robert Allan Doughty, The Seeds of Disaster, The Development of French Army Doctrine 1919–1939 (Archon, 1985); Elizabeth Kier, Imagining War, French and British Military Doctrine Between the Wars (Princeton University Press, 1997); Eugenia Kiesling, Arming Against Hitler: France and the Limits of Military Planning. (University Press of Kansas, 1996). There are more recent treatments that I have not reviewed.
[65] This is the accepted estimate. See Stephanie Trouillard, “Why did France lose to Germany in 1940?” France 24, May 16, 2020 https://www.france24.com/en/20200516-why-did-france-lose-to-germany-in-1940.
[66] Posen, Restraint: A New Foundation for U.S. Grand Strategy (Cornell University Press, 2014).
[67] For representative critiques of Restraint see Robert Lieber’s book review in H-Diplo|ISSF Roundtable VIII, No. 16 (11 July 2016): https://issforum.org/ISSF/PDF/ISSF-Roundtable-8-16.pdf. For a structured debate, see Barry R. Posen, “Pull Back,” and Stephen G. Brooks, G. John Ikenberry, William C. Wohlforth, “Lean Forward: In Defense of American Engagement,” Foreign Affairs 92:1 (January/February 2013): 130-142. The three co-authors have been active critics of a grand strategy of restraint for the last decade.
[68] John J Mearsheimer, “The Inevitable Rivalry, America, China, and the Tragedy of Great-Power Politics,” Foreign Affairs, November-December 2021.
[69] Structural realists disagree about the diversity and intensity of behaviors explained by structure. For example, Mearsheimer’s theory of “offensive realism” highlights how the security motive incentivizes power maximization. “Defensive realism” agrees that that is one reasonable prediction, but following Waltz, also places a great deal of weight on how the same security motive incentivizes “balancing behavior” by others, which serves to raise the costs of expansion, and sometimes to nudge the acquisitive to more cautious behavior. Here too, for Waltz’s relevant work Structural realists do not agree among themselves on the amount of explanatory weight to attribute to structural forces. Mearsheimer’s theory of “offensive realism” sees structure as a more powerful causal force than the theory of “defensive realism” developed by Kenneth Waltz and many of his students.
[70] The locus classicus of structural realism remains Kenneth Waltz, Theory of International Politics (Addison-Wesley, 1979). On hyper-rationalism, see James Fearon, “Rationalist Explanations for War,” International Organization 49:3 (1995): 379-414, and also David Lake, “Two Cheers for Bargaining Theory: Assessing Rationalist Explanations of the Iraq War,” International Security 35:3 (2010/11): 7-52. For some of the landmarks of classical realism, see the citations to Hans Morgenthau, Raymond Aron, and George F. Kennan, below. An Unwritten Future seeks to revive, articulate, and build on this tradition.
[71] Michael Woodford, “Convergence in Macroeconomics: Elements if the New Synthesis,” American Economic Journal: Macroeconomics 1:1 (2009): 267-79; David Colander, et al., “The Financial Crisis and the Systemic Failure of the Economics Profession,” Critical Review 21:2-3 (2009): 249-67; Stephen C. Nelson and Peter J. Katzenstein, Uncertainty, Risk, and the Financial Crisis of 2008,” International Organization 68:2 (2014): 361-92.
[72] See Anthony Brewer, Marxist Theories of Imperialism: A Critical Survey (Routledge & Kegan Paul, 1980) and V. I. Lenin, Imperialism: the Highest Stage of Capitalism (International Publishers, 1939/1917). Realist Robert Gilpin emphasizes (and embraces as a crucial explanatory variable) what he dubs Lenin’s “law of uneven development”; realist Stephen Krasner emphasizes that at times distinguishing a realist perspective from some strands of Marxist analysis can be a “difficult task.” Robert Gilpin, War and Change in World Politics (Cambridge University Press, 1981), 76; Stephen D. Krasner, Defending the National Interest: Raw Materials Investments and U.S. Foreign Policy (Princeton University Press, 1978), 8.
[73] Arnold Wolfers, “Statesmanship and Moral Choice,” World Politics 1:2 (1949): 175-95, 189, 193-4.
[74] Hans Morgenthau, Scientific Man versus Power Politics (University of Chicago Press, 1946). See also Raymond Aron, Introduction to the Philosophy of History: An Essay on the Limits of Historical Objectivity (Weidenfeld and Nicolson, 1981).
[75] John Maynard Keynes, letter to Roy Harrod, July 16, 1938, reprinted in The Collected Writings of John Maynard Keynes, Volume XIV (Cambridge University Press, 1973), 299, 300. On this unbridgeable gap between the natural and the social worlds, see also Hans Morgenthau, “The Limitations of Science and the Problem of Social Planning,” Ethics 54:3 (1944): 174-85.
[76] Frank Knight, Risk, Uncertainty and Profit (University of Chicago Press, 1971/1921), 241, 311; Friedrich Hayek, “The Pretence of Knowledge,” (Nobel Memorial Lecture, 11 December 1974), reprinted in The American Economic Review, 79:6 (1989): 3-7, 3, 5.
[77] Charles P. Kindleberger, Economic Laws and Economic History (Cambridge University Press, 1989), ix (quote), x, 127, 193.
[78] Charles P. Kindleberger, Historical Economics: Art or Science? (University of California Press, 1990), 4 (quote), 7, 9.
[79] Charles P. Kindleberger, “Scientific International Politics,” World Politics 11:1 (1958): 83-88, 86.
[80] Emily Greenwood, “Thucydides on the Sicilian Expedition,” in Ryan K. Balot, Sara Forsdyke, and Edith Foster, eds., The Oxford Handbook of Thucydides (Oxford University Press, 2017): 161-77. See also Jonathan Kirshner, “Prevent Defense: Why the Bush Doctrine will Hurt U.S. Interests,” in Iraq and Beyond: The New U.S. National Security Strategy, Peace Studies Program, Cornell University, Occasional Paper #27 (January 2003): 1-10.
[81] Alfred Marshall, one of the founding fathers of neoclassical economics, illustrated his profound skepticism of prediction with an example drawn from a complete information game, a setting (generally absent in economics, impossible to even imagine in IR), where every possible move available to every possible player is known to every observer at all times: “Show an uninterrupted game at chess to an expert and he will be bold indeed if he prophesies its future stages. If either side make one move ever so little different from what he expected, all the following moves will be altered; and after two or three moves more the whole face of the game will have become different.” Alfred Marshall, “Fragments,” in A.C. Pigou, ed., Memorials of Alfred Marshall (Macmillan, 1925), 360; see also Marshall, Principles of Economics (8th Edition) (MacMillan, 1920), esp. book I, chapter iii, “Economic Generalizations or Laws,” 30-33.
[82] John Mearsheimer, The Tragedy of Great Power Politics (Norton, 2001).
[83] On offensive realism, see Mearsheimer, Tragedy; on defensive realism, Stephen Walt, “The Enduring Relevance of the Realist Tradition,” in Ira Katznelson and Helen Milner, eds., Political Science: The State of the Discipline (Norton, 2002): 197-230.
[84] On hubris, Robert Colquhoun, Raymond Aron: Volume 2: The Sociologist in Society 1955–1983 (Sage Publications, 1986), 422.
[85] Waltz, Theory of International Politics, 73.
[86] Aron quoted (approvingly) in Robert Gilpin, War and Change in World Politics (Cambridge University Press, 1981), 29.
[87] Knight, Risk, Uncertainty and Profit, 19 (radically different), 20 (true uncertainty, unmeasurable), 198, 232-3, 268 (opinion), 287-8, 293.
[88] Keynes, “The General Theory of Employment,” Quarterly Journal of Economics 51:2 (1937): 209-233 quotation on 222; see also 213-114. Under risk, although we can’t know for sure what will happen next, we do know with certainty the underlying probability distribution of all possible outcomes. (The classic example is rolling two dice—we don’t know what number will come up, but we know exactly the probability of every possible outcome.) Under uncertainty, we do not have access to that underlying probability distribution.
[89] Barry Posen, Restraint: A New Foundation for U.S. Grand Strategy (Cornell University Press, 2014).
[90] Stephen G. Brooks, G. John Ikenberry and William C. Wohlforth, “Don’t Come Home, America: The Case against Retrenchment,” International Security 37:3 (2012/13): 7-51.
[91] Joan Robinson, “What has become of the Keynesian Revolution?” in Milo Keynes, ed., Essays on John Maynard Keynes (Cambridge University Press, 1975): 123-31,125, 126.
[92] Keynes, A Tract on Monetary Reform (Cambridge University Press, 1971/1923), 80.
[93] Posen, The Sources of Military Doctrine (Cornell University Press, 1984), 224.
[94] George F. Kennan, American Diplomacy 1900–1950 (Mentor Books, 1951), 88; see also Kennan, Realities of American Foreign Policy (Princeton University Press, 1954).