Deep within a vast, primeval forest, three hunters silently track a stag. As they follow fresh hoof prints through stands of birch and alder, each hunter worries about the trustworthiness of the others. Armed with rudimentary weapons, they must collaborate to have a reasonable chance of slaying game large enough to feed them all. If one hunter spots a hare and breaks ranks, he could easily catch it and feed himself, but his defection would deprive everyone else of a meal, for they could not bring down the stag without his help. Yet if he faithfully maintains his post, he risks going hungry because another hunter might abandon the group to pursue the hare. Faced with this dilemma, what should each hunter do?
The eighteenth-century Swiss political philosopher Jean-Jacques Rousseau used this allegory to highlight the tension between risky cooperation and individual self-help whenever no higher authority exists to enforce commitments.1 Many international relations theorists believe that the strategic choices facing his fictitious stag hunters parallel those confronting national leaders. In Rousseauâs stag hunt, and, by analogy, the modern interstate system, fear, friction, and fighting arise when egoists, concerned primarily about their own survival, interact without a central arbiter. If one actor expects that another may shirk his responsibility, he will defect from their tacit partnership rather than gamble on being helped during a time of need, thus undermining the prospects for mutually beneficial collaboration.
For stag hunters as well as for national leaders, what is reasonable to do in certain situations depends on beliefs about what others will do. Although the hunters prefer large quarry because it provides more sustenance than small prey, they face a coordination problem: Can any of them trust the others not to pursue the hare, even when they all promised to work together? Itâs best to hunt stag if everyone else does, but if someone in the group is unreliable, the safest choice is to hunt hare and avoid starvation. Given the incentives of self-interested actors to defect when there is no one to enforce commitments, what would cause the hunters to adopt stag-hunting over hare-hunting? Similarly, in an anarchic state system, what would incline national leaders to risk joining forces with potentially unreliable partners rather than engage in defensive noncooperation?
The Problem of Cooperating Under Anarchy
Harmony, a situation in which self-interested behavior by one actor automatically helps other actors achieve their goals, is rare and fleeting in an anarchic environment. As the allegory of the stag hunt suggests, even if everyone desires the same outcome from their interaction, they often have difficulty achieving it. Mistrust nourishes discord. To realize jointly desired outcomes, everyone must make a conscious, deliberate effort to cooperate. Teamwork does not occur spontaneously. Political negotiation, mutual adjustment, policy coordinationâall are required (Keohane 1984, pp. 51â53).
Cooperation is difficult when strategic choices depend on beliefs about the anticipated behavior of prospective partners whose intentions are unknown. The temptation to forsake others and fend for oneself can be overwhelming. Indeed, without clear, convincing assurances that everyone will stay the course during a collective venture, apprehension may cascade through a group: One actorâs worries about opportunism among the others triggers their fear that he will defect, prompting them to consider defecting, which ultimately makes it prudent for him to defect (Jervis 1978, p. 168).
Anarchy can thus hinder non-altruistic cooperation even when working in unison would leave everyone better off. Without a âcommon powerâ to keep everyone in awe, submitted the seventeenth-century English political philosopher Thomas Hobbes (1651), gnawing suspicions and unbridled competition impede collaboration.2 Those living under anarchy, he argued, experience âcontinual jealousies,â which leads them to adopt âthe posture of gladiators,â with âtheir weapons pointing, and their eyes fixedâ on one another. Without a âcoercive Power to tye their hands from rapine, and revenge,â these âmasterlesse menâ suffer âcontinuall feare, and danger of violent death.â Thus, for Hobbes, anarchy invites strife, making life ânasty, brutish, and short.â
Hobbesâ argument raises important questions about the prospects for cooperation under anarchy. Without a supreme authority to regulate their conduct, how might autonomous actors move from self-help to cooperation? Do private appetites and personal brawn define acceptable behavior? Are covenants without the sword meaningful? Is a global Leviathanâan awesome, absolute, and overarching powerâthe price of controlling the use of military force and ameliorating the horrors of interstate war?
A Conjecture
To begin our search for answers to these questions, letâs engage in a brief thought experiment. Imagine that the members of Rousseauâs hunting party belong to a segmentary society.3 This form of political organization was common in precolonial Africa. The Igbo of Nigeria, the Nuer of Sudan, and the Tonga of Zambia are a few examples. Each of these societies was based on the belief that political power should be localized, not concentrated in a central authority. People in such societies understood that they were members of a larger collectivity, but their primary affiliation was with smaller kinship-based segments of that collectivity. Under the principle of exogamy, individuals married outside of their core lineage group, therein forging a complex network of political alliances among the collectivityâs segments (Weiner 2013, p. 58). Sometimes these segments coordinated their activities. For instance, different clans of the Masai of Kenya and Tanzania participated in a common socialization process, and the Teke of Zaire collaborated during certain religious festivals. Notwithstanding these occasional joint activities, local autonomy was maintained, and independence remained a principal value.
Kinship-based segments, the elemental units of these societies, frequently competed with one another, not unlike nation-states, the constituent units of the modern international system. Disputes between competing segments could escalate to war, but most were held in check by socially-agreed-upon rules of behavior. Some rules outlined procedures through which a third party could diffuse contentious situations; others reduced the severity of altercations by prohibiting certain methods of retaliation; and still others helped restore peace by delineating how malefactors could atone for their transgressions. The content of these customary rules varied from one society to another (Middleton and Tait 1958; Bohannan 1957; Fortes 1949; Evans-Prichard 1940). Rules followed by the Anuak of the Sudan, for example, differed from those of the Tallensi of Ghana, even though they both possessed horizontal, decentralized political systems. Segmentary societies that are structurally alike do not necessarily have identical political cultures.
Suppose that Rousseauâs hunting party contains members from different kinship-based units in a segmentary society whose political culture prizes what the ancient Greeks called kerdeaâinitiative, resourcefulness, and guile.4 Within this permissive normative order, self-preservation is the paramount value and an ability to advance oneâs security interests through wit and wile are respected. As reflected in the Bedouin proverb âMe and my brother against my cousin; my cousin and me against the stranger,â loyalty hinges on the immediacy of kinship bonds. Familial and clan interests take precedence over the welfare of people from other segments of the society, and kinfolk have the latitude to do whatever they believe must be done to protect these interests, including reneging on agreements, resorting to force, and intervening into the affairs of other segments.5
Now further suppose that each hunter originally set out alone and later happened on the others while trekking through the forest. Their meeting was coincidental and, hailing from different kinship units, was unlikely to happen again. But on this singular occasion, they realized working together offered the possibility of bagging big game rather than smaller quarry.
To investigate how the members of our imaginary hunting party might act in this situation, let us examine the strategies that they can choose and the outcomes that would result from the combination of their choices. Game theory, a branch of applied mathematics, provides a set of formal tools for modeling strategic interaction. When used for heuristic purposes, it can yield insights into the conduct of interdependent actors facing trust dilemmasâsituations like Rousseauâs stag hunt where recognition of the potential benefits of trusting is offset by a realization that counting on others exposes oneself to the possibility of betrayal. Because Rousseau omitted some details in his description of the hunt, game theorists have modeled it in different ways. Letâs conceptualize the hunt as a symmetric three-person game where the spoils from cooperating are divided equally, each hunter chooses to cooperate or defect without knowing what the others are about to do, and a single defection ruins the hunt.
Table 1.1 depicts the strategic structure of the decision problem facing our three-person hunting party. Recall that the hunters belong to different clans within a segmenta...