European Journal of International Relations 9(4) emerge within 100200(?) years, but nothing below turns on that prediction.

Instead, I am concerned with the macro-structure of all pathways, which channel the international systems development toward an inevitable end-state. In that respect the theory is progressivist, although in an explanatory rather than normative sense. Resistance to progressivist, much less teleological, thinking runs deep within contemporary IR scholarship. Realists are skeptical, arguing that the logic of anarchy is one of endless conict and war (Waltz, 1979). Liberals are more optimistic, arguing that international institutions, interdependence and/or democratic states can lead to cooperation and peace within anarchy (Keohane, 1990). However, liberal progressivism is contingent, not teleological. If institutions are upheld, if interdependence deepens and/or if democracy spreads, then progress is possible. The forecast is based on extrapolating lawlike regularities from the past into the future, assuming certain conditions continue to hold. Since there is no guarantee they will, we cannot say that any given future is inevitable. Indeed, if there is one thing almost all social scientists today agree on, from the most hardened positivists to the most radical postmodernists, it is that teleological explanations are illegitimate. To call a theory teleological is considered a decisive criticism, with no need even to explain why. This may be due to the fact that teleology has been considered unscientic since the triumph of the mechanistic worldview in the 17th century, and is also sometimes thought to deny human agency in the social world. In my view both objections are unfounded, and with them a priori resistance to teleological thinking about world politics. To show this, in the rst section I synthesize recent attempts to rehabilitate teleological explanation. These efforts span many disciplines and indicate that, although the scientic status of teleology remains controversial, it is being taken increasingly seriously. One reason is that much of this literature builds on self-organization theory, which is emerging as an important challenge to the neo-Darwinian theory of evolution.1 Selforganization theory hypothesizes that order in nature emerges not only through the mechanism of mutation-selection-retention, but also spontaneously from the channeling of system dynamics by structural boundary conditions toward particular end-states. With a few exceptions this theory has been little noticed by IR scholars,2 who are just beginning to engage neo-Darwinism.3 But in the social sciences more generally the idea of selforganization has been around since the spontaneous order tradition of the Scottish empiricists, and is getting considerable attention today.4 Much of this work is not teleological, and many self-organization theorists might vigorously reject any such reading of their approach. On the other hand, 492

Why a World State is Inevitable

ALEXANDER WENDT University of Chicago

Long dismissed as unscientic, teleological explanation has been undergoing something of a revival as a result of the emergence of selforganization theory, which combines micro-level dynamics with macrolevel boundary conditions to explain the tendency of systems to develop toward stable end-states. On that methodological basis this article argues that a global monopoly on the legitimate use of organized violence a world state is inevitable. At the micro-level world state formation is driven by the struggle of individuals and groups for recognition of their subjectivity. At the macro-level this struggle is channeled toward a world state by the logic of anarchy, which generates a tendency for military technology and war to become increasingly destructive. The process moves through ve stages, each responding to the instabilities of the one before a system of states, a society of states, world society, collective security, and the world state. Human agency matters all along the way, but is increasingly constrained and enabled by the requirements of universal recognition. KEY WORDS cultures of anarchy downward causation logic of anarchy self-organization struggle for recognition teleology world state

In this article I propose a teleological theory of the logic of anarchy which suggests that a world state is inevitable (cf. Buzan et al., 1993). Like any structural tendency, the speed with which this one will be realized is historically contingent. At the micro-level the process is neither deterministic nor linear, and forward movement may be blocked for periods of time. There are many pathways by which a world state may be achieved, and human agency matters along every one. In that sense anarchy is [still] what states make of it (Wendt, 1992). However, I am not concerned here with historical contingencies or timing. My own guess is that a world state will
European Journal of International Relations Copyright 2003 SAGE Publications and ECPR-European Consortium for Political Research, Vol. 9(4): 491542 [13540661 (200312) 9:4; 491542; 038724]

-1-

-2-

Wendt: Why a World State is Inevitable many others do see a connection, arguing that self-organization theory provides a scientic basis for teleological explanation.5 Assuming that is possible, toward what end-state does the international system move, and by what mechanism does it get there? Three end-states suggest themselves a pacic federation of republican states, a realist world of nation-states in which war remains legitimate, and a world state. The rst is associated with Kant (1991a, 1991b) and the second with Hegel (1977), both of whom based their projections on explicitly teleological arguments.6 In rejecting the possibility of a world state, therefore, they agreed that, strictly speaking, anarchy would remain the organizing principle of the system, albeit different kinds of anarchy. As to the mechanism of progress, in different ways Kant and Hegel also both emphasized the role of conict Kant in mans unsociable sociability, and Hegel in the struggle for recognition. I am in no position here to engage in an exegesis and critique of Kant and Hegels arguments. However, since I share their emphasis on conict as a mechanism of development but reach a different conclusion, it may be useful to introduce my argument by highlighting two ways in which it departs from theirs. The rst concerns the effects of conict on state identity. While envisioning a tendency for conict to create republican states, Kant did not expect them to develop a collective identity. His states remain egoists who retain their sovereignty. Hegel provides the basis for a different conclusion, since the effect of the struggle for recognition is precisely to transform egoistic identity into collective identity, and eventually a state. But Hegel expects this outcome only in the struggle between individuals. States too seek recognition, but in his view they remain self-sufcient totalities. Their struggle for recognition does not produce supranational solidarity, leaving us at the end of history with a world of multiple states (also see Fukuyama, 1992). Some critics have suggested that Hegels reasoning here is inconsistent, and that he should have argued for a world state.7 Be that as it may, I argue that the struggle for recognition between states will have the same outcome as that between individuals, collective identity formation and eventually a state (cf. Walzer, 1986). One reason for this concerns the second difference in my approach, which concerns the role of technology. Kant rejected the possibility of a world state in part because the technology of his day precluded it (Carson, 1988: 177; Guyer, 2000: 41617), and in positing an end-state in which war remained legitimate Hegel did not think its costs would become intolerable. Neither anticipated the dramatic technological changes of the past century, which are in part caused by the security dilemma and thus endogenous to anarchy. As Daniel Deudney (1999, 2000) convincingly argues, these changes have greatly increased the costs of war and also the scale on which it is possible to organize a state. 493

European Journal of International Relations 9(4) With these material changes the struggle for recognition among states undermines their self-sufciency and makes a world state inevitable. Via the struggle for recognition, in short, the logic of anarchy leads to its own demise. Two caveats. First, it is impossible within the constraints of an article to do justice to both the methodological and substantive issues in this discussion. That might have counseled two separate articles (at least), but given the modern hostility to teleology the argument might be dismissed a priori without a defense of its methodology, and the latter might seem unmotivated without a plausibility probe. So at the risk of doing both inadequately I have joined them. However, on the assumption that the logic of teleological explanation will be less familiar and of broader interest than a theory of world state formation, when faced with trade-offs I have opted for preserving discussion of the former, and left the substantive theory more at the level of a sketch, to be eshed out in the future. Second, again for reasons of space, I shall not address the relationship between the logic of anarchy and the logic of capital, which forms a distinct developmental dynamic in the system. The logic of capital generates distributional struggles that cannot be reduced to the struggle for recognition (Fraser, 2000) and, as such, would signicantly complicate the latter, but in the long run it too points toward a world state (Chase-Dunn, 1990; Shaw, 2000). So I bracket its role here, on the assumption that it only makes a world state more inevitable. The teleological explanation is defended in the next section. I then explain what I shall mean by the state, including a world state. In the third section I discuss the struggle for recognition, and the fourth show why such a struggle within anarchy should culminate in a world state. The role of agency in this process is addressed in the conclusion.

Causal Pluralism and Teleological Explanation

In recent years there has been much debate within IR scholarship about what might be called causal pluralism (Asma, 1996) whether explanations of world politics can take different forms. The orthodox positivist position, rooted in a Newtonian worldview, is that an explanation always depicts a mechanical relationship between prior conditions and later effects.8 Other forms of inquiry might be valuable as descriptive inference (King et al., 1994), but they do not explain. Explanations must be causal, and causation must be mechanical. From this perspective, causal pluralism is either confused about what explanation means, or a threat to science itself. Following interpretivist philosophers of social science, constructivists and postmodernists have argued against causal monism in favor of constitutive 494

-3-

-4-