![]()
1
Both fox and hedgehog
The art of nesting structural and perceptual perspectives
Jack Snyder
Scholars of international relations are put in a quandary by Philip Tetlock’s research showing that foxes who know many little things are better at forecasting future events than are hedgehogs who know one big thing.1 After all, the main characteristic that differentiates scholars from journalists and practitioners is that scholars try to integrate their knowledge into unified, generalized intellectual systems called theories. Even scholars who are put off by grandiose paradigms like realism and liberalism nonetheless usually try to connect their insights into integrated, abstract arguments. It is the job of academics to be hedgehogs, yet we know that this risks making us more cartoonish in our judgments than less theory-bound experts.
Scholars justify this, correctly, by arguing that foxes need to bring a list of hedgehog-like insights to the complex, evolving situations that they are trying to understand and manage. Foxes whose heads are filled only with disorderly facts and unprocessed impressions would not be good foxes. Hedgehogs help to furnish the minds of the foxes, who then rearrange or dispose of the furniture as they please.
Some hedgehogs are happy with this division of labor. Kenneth Waltz, for example, makes no pretence of having a “theory of foreign policy” that can predict whether North Korea will test a nuclear weapon next year, the kind of task that Tetlock set for his subjects. He sees his job as providing a background theory that will help orient people to enduring patterns that are likely to be relevant to their more specific intellectual tasks. But many scholars seek to combine the benefits of the hedgehog’s depth and unity of insight with the fox’s openness, flexibility, and specificity.
Robert Jervis offers an example of how to have it both ways. On the one hand, he has made seminal contributions to the canon of hedgehog-like simple, unified, general theory, while on the other hand being a master of open-minded, multifaceted analysis of specific political problems and processes. His work meets the approval of international relations theorists as well as historians and intelligence professionals.
How exactly does he do this? One view might be that his trick is to rigorously pursue one hedgehog insight at a time, but then step back and admit that this is only a partial truth, and that other insights, including ones that may fundamentally contradict it, could also be true.2 As an exasperated undergraduate once exclaimed to me, “He keeps arguing AGAINST himself! It’s SOOOO FRUSTRATING!” Another view might be that his insights are nested, such that the central hedgehog assumptions provide the anchor for the second-order elaborations, the third-order qualifications, and the radical exceptions. I think the nesting view comes closer to capturing Jervis’s method.
Nesting theories of international structure, strategic interaction, and perception
Robert Jervis stands out for his highly original, multifaceted contributions to understanding the dilemmas of international conflict and cooperation. He has shown in a compelling fashion how the efforts of political leaders and strategists to escape from the insecurities of international competition often serve only to compound these dangers. While remaining alert to the role that greedy aggressors play as a source of conflict among states, Jervis’s research has stressed the more surprising and tragic sources of conflict that can embroil even states that mainly seek security from each other. He points out the misperceptions, unintended consequences, and perverse situational dilemmas that trip up statesmen, in the hope that these pitfalls, if better understood, can be guarded against.
In studying these problems of international security, Jervis has made contributions to basic social science in a broad range of fields. He originated the distinction between signals and indices in strategic bargaining, which are similar to the now ubiquitous concepts of “cheap talk” and “costly signals.”3 He pioneered the sophisticated application and elaboration of the laboratory findings of cognitive psychology to problems of real-world political decision-making.4 He has been a central figure in theoretical dialogues between historians and political scientists on the nature of the international system. From this base, he has explored the implications of feedback effects in complex systems of all kinds, ranging from supertanker safety systems to the balance of power.
Jervis’s work has been driven not by an a priori commitment to any particular theoretical paradigm, but rather by the problem of international cooperation in the face of the danger of war, especially nuclear war, and by a series of intellectually challenging puzzles posed by that substantive agenda. Thus, his work constitutes a conundrum for the many international relations specialists who like to categorize thinkers exclusively as realists, liberals, rationalists, or constructivists. Jervis is a realist in that he argues that anarchical circumstances pose distinctive problems, yet he diverges from many realists in arguing that it is the perception and mis-perception of these circumstances and of the intentions of the adversary that crucially shape choices about how to compete or cooperate. Jervis is a rationalist who has made seminal contributions to the conceptual underpinnings of game theory and bargaining, yet the bulk of his work shows how rationality is confounded by cognitive shortcuts and by unpredictable consequences in complex systems. Jervis is a methodological positivist who has laid out a scrupulously rigorous method for assessing the impact of ideas on behavior, yet he has also explored the limits of conventional positivism in grappling with nuclear symbolic politics and complex feedback effects.
Because Jervis’s insights and style of argument cannot be reduced to a simple “ism,” it is superficially tempting to view his various contributions to knowledge as eclectic and piecemeal. His subtle, non-dogmatic arguments draw on diverse intellectual sources, cut across standard categories, and characteristically depend on the integration of opposing considerations. But his approach is by no means merely eclectic. The hard core of Jervis’s theoretical project has remained coherent and consistent as he has explored its implications in varied domains. He consistently portrays international politics as structured by complex systems of strategic interaction, by the often myopic striving of actors to perceive the nature of the system they are operating in, and by these wary actors’ use or misuse of guile to deal with the problems of insecurity, opportunism, and uncertainty that are inherent in the workings of the system.5
I will review five areas in which Jervis has made central theoretical contributions and then explain how I see them as constituting a tightly coherent research program.
Cooperation in anarchy
Jervis’s most cited article, “Cooperation under the Security Dilemma,”6 draws on Rousseau’s parable of the Stag Hunt and the Prisoner’s Dilemma game to show how status quo states can wind up arms-racing and fighting due to the fears engendered by the situation of anarchy. Jervis defines a security dilemma as a situation in which any state’s efforts to increase its security necessarily decrease the security of others. In this situation, one side’s efforts to escape its insecurity through an arms build-up or through the conquest of strategic territory will inevitably trigger similar behavior by other security-conscious actors.
Many realists have written as if any anarchical situation creates this kind of dilemma. Jervis showed, however, that this is true mainly when the side that takes the offensive thereby gains an advantage. If defense is easier or cheaper than offense, as it often is in international politics and in military strategy, then the security dilemma is relaxed or even eliminated, because both sides can be secure simultaneously by adopting non-threatening measures of self-defense. This works best when offensive weapons and strategies are easily distinguishable from defensive ones. Since nuclear threats are issued more credibly by the side that is defending encroachments on its core interests, a nuclear-armed stalemate eases the security dilemma, Jervis argues. Thus, the anchor for Jervis’s theorizing is the objective strategic circumstances that actors find themselves in.
However, in the typical Jervis two-step, it is not just the situation of anarchy, offensive technology, or barrier-free geography that heightens vulnerability to attack and thus intensifies the security dilemma. As he notes, strategists and political leaders have often misestimated the ease of attacking or defending. What counts is their perception of the balance between offensive and defensive incentives, as well as their perception of the likelihood that the other “stag hunters” will defect from cooperation. Thus, both strategic circumstances and perceptual biases as discussed in Jervis’s other work shape behavior under the security dilemma.
This seminal article spawned an immense literature on the security dilemma: testing its propositions against the historical record, debating ways to measure the causal variables in the theory, studying the origins of offensive and defensive strategies, applying the theory to grand strategy as well as ethnic conflict, and using the theory to devise more stable strategic postures.7
Communication in strategic bargaining
Jervis has been a central contributor to research on communication in strategic bargaining, bridging from Thomas Schelling’s seminal studies of the credibility of threats in the 1960s to contemporary game-theory examinations of bargaining under conditions of imperfect information. The Logic of Images in International Relations,8 adapting the dramaturgical insights of Erving Goffman’s The Presentation of Self in Everyday Life,9 introduced the distinction between what Jervis called signals “on the cheap” of an actor’s intentions and indices of the actor’s expected behavior, which are too costly or too integral to the actor’s fundamental make-up to fake. Applying these ideas to problems of deterrence and bargaining in the nuclear era, he examined precisely the kind of subtleties in the use of signaling and the manipulation of indices that have become staples of the burgeoning literature on bargaining in conditions of imperfect information, such as the dynamics of reputation and commitment in the face of incentives for opportunism and deception. These themes provided the conceptual grounding for much of Jervis’s later work on the role of perception in bargaining, on nuclear deterrence, and on conflict spirals and the security dilemma – all of which hinge on such problems as the mis-reading of an adversary’s intentions or the difficulty of credibly committing not to attack an adversary.
Jervis has continued to write on questions of game theory, signaling, and bargaining.10 Typically, his work has stressed that the most important factors driving behavior in games, such as the preferences of the actors and their mode of perceiving each other, are precisely the elements that most game theory simply takes as givens. His work in this area shows the importance of embedding formal game theory in a broad conceptual context, drawing on psychology and other sources of theoretical inspiration.
All of this work has a nested character insofar as the analysis starts with the given characteristics of the bargaining situation in anarchy, proceeds to the logic of making threats and signaling commitments under uncertainty, and then adds the complications that stem from perceptual deviations from a rational baseline.11 This is not just a list of disconnected factors, but a nested causal hierarchy in which the structural features of the situation (such as the security dilemma) shape the bargaining problems (such as credible commitments) and perceptual issues (such as the likelihood that the other will cooperate) that play out according to the “logic of images” and the rules of cognition.
Perception and misperception in international politics
Since the appearance of his seminal article, “Hypotheses on Misperception,”12 Jervis has practically owned the subject of perception in international relations. His comprehensive study, Perception and Misperception in International Politics,13 creatively assessed the applicability of a wide range of psychological propositions to the study of deterrence failures, conflict spirals, intelligence failures, strategic assessments, and the full panoply of diplomatic judgments and misjudgments. The central message of the book is that perception is profoundly theory-driven, that decisionmakers tend to see what they expect to see, and that these expectations are often driven by stereotyped lessons of history, analogies, or routine scripts that provide shortcuts in making assessments under uncertainty. As James Wirtz’s chapter in this volume shows, Jervis is remarkably consistent in showing how erroneous expectations drove US intelligence failures in the overthrow of the Shah of Iran and in the absence of weapons of mass destruction in Iraq.
Unlike much of the more recent work on the role of ideas in international politics, Jervis does not see the causal role of “ideas”...
