The Logic of Anarchy: Neorealism to Structural Realism

Barry Buzan, Charles Jones and Richard Little

New York

Columbia University Press

1993

13. Summary and Conclusions

 

The purpose of this chapter is to highlight the main themes from the three sections, and to focus some thoughts on the research agendas to which we think our arguments point. We are fully aware that our pursuit of the logic of anarchy in the preceding pages has opened up more areas than it has in any sense "closed." In moving from a purely structural theory to a more fully systemic one, it is impossible to avoid a very considerable expansion of the subject area.

Our aim has been to sketch out a fairly comprehensive theoretical framework for this larger area, and by doing so to indicate how much of International Relations theory can be integrated. Structural Realism seeks to construct an ordering principle by which the previously dispersed "islands" of International Relations theory can now be viewed as an archipelago. The islands are mostly still separate, but the distance and direction between them is clearer, and the geology and geography that set the conditions for building causeways, bridges, or tunnels (or merely setting up ferry routes) begins to emerge into view. We do not claim that this framework is complete, or that all parts of it are equally well developed.

Like the Neorealists we have focused on the system level, though we hope that we have much expanded what counts at that level. We have ventured more into the rich layers of the unit level than they do, but there is still much there that we have left largely untouched. We have tried to show more clearly than the Neorealists have how the structure and unit levels interplay with each other, first by showing how structural hypotheses can set the framework for unit level analysis, and second by demonstrating how the process of structuration brings system structure and the domestic structure of the units into a strong relationship. Structural Realism is much more open to change than Neorealism, though we hope that we have preserved and enriched the analytical utility of the static power of structure to shape and constrain the behavior of units. Although the main purpose of this exercise has been to expand away from Neorealism we hope along the way to have reduced the level of misunderstanding about it. One advantage of using Waltz's text as a starting point has been to expose just how much misunderstanding of TIP (as opposed to mere disagreement about it) persists despite more than a decade of intense debate.

In the summaries that follow, we take the opportunity that we rejected in the main text of starting with methodology and epistemology, and then going on to elaborate the logic of the theory.

In Section III Waltz was accused of a residual positivism inconsistent with the generally pragmatist tone of a majority of his remarks about theory and knowledge. This positivism was especially evident in his treatment of the pervasive analogy between microeconomics and the balance of power, which was couched in the language of hypothesis and testing. Besides, it relied upon what were felt to be unjustified assumptions about the superiority of economics and about the similarity of competition between firms to competition between states.

An attempt was made to ground Structural Realism on pragmatist epistemological foundations. At the level of methodology, an alternative characterization of the analogy as metaphor was proposed. By this strategy it was hoped that some of the difficulties encountered in TIP might be circumvented and a clear recognition of the rhetorical force of language restored to pride of place within the Realist tradition without sacrificing the gains of a structuralist, systematic theory of international relations.

What was proposed, in short, was the frank acceptance of language as power, rather than as a transparent medium of investigation. Language, and its use both by statesmen and publicists in deliberations about foreign policy, was put forward as one manifestation of power: a structure, like the balance of power or the market. The point of this was to bring back into a systemic theory of international relations those facets of experience, such as interpretation, symbolism, persuasion, motive, and the unconscious, which TIP had seemed to marginalize by its much more sparse definitions of power and state capabilities. It was also suggested that this incorporation of language was bound to strengthen any structuralist Realism in a number of ways, most of all, perhaps, by providing a better understanding of the causes and possibilities of change in the international system than had been offered by Neorealism.

The philosophically trained reader may detect a hankering after scientific realism and a desire to tell truths about things as they are beneath the pragmatism of Section III, frustrated in the end by epistemological resignation. The tension is left unresolved not out of sheer funk, but because this appears to be one of those situations in which some hunches, requiring further empirical work, need to be played before the more abstract issues can be taken much further. In particular, the stress in Section III on language, metaphor, and their spontaneity and power, coupled with the evidently tentative attitude adopted there toward theories of reference and possibilities of absolute rather than historically and socially relative truths, suggests that the way forward may lie in a close scrutiny of foreign policy debates and diplomatic discourse, transtemporal as well as transcultural, employing the techniques of analysis of the literary critic. It would seem quite unsafe to rest too much philosophical weight on the diplomatic form of life before testing it out a little. As things stand, indeed, the reliance on linguistic spontaneity as a source of change in Section III is no less incautious than Waltz's own resort to intuition in accounting for theory generation. This said, a research program comprising close empirical study of the diplomatic and foreign-policy-making community, its practice and discourse, has this great advantage: it leaves open the passes to sociological theories of reference and to the admittedly somewhat diluted forms of scientific realism which they might permit, though those attempting the route should be sure to carry chains in the trunk.

Research of this sort, aimed at the identification of social entities that might act as the subjects of a Structural Realist theory of international relations, cannot simply go after the best available expert consensus on what was meant by "state," or "war," or any other key descriptive term in this or that culture, language, or period. Not can it go hunting innocently, without its own preconceptions and language. The best that it is likely to be able to supply is an account, based on the utterances of policymakers and intellectuals and phrased in the most open, ironic, and provisional manner possible, of the tensions that have formed around contested concepts, the strategies by which usages of key terms have been established and maintained, and the implications such established usages have had for social action. As a conjecture to guide research, one might regard the development of the kinds of highly reflexive, self-conscious, and manipulative systems of symbols, signs and images that suffuse modern political discourse as an instance of change in interaction capacity of the kind identified in Section I.

What is required, moreover, is not simply empirical groundwork to clear the way for a more precise theoretical statement, but also ways of reporting that work which respect the reflexive, tightly folded character of the object of study in their own mode of expression. There is plenty of space between the style of the best current work on international relations and the much more cunning and playful prose of Nietzsche, let alone the less accessible expressions of Derrida. Some attempts have been made to occupy this stylistic space, but it is still thinly populated, remote, and poorly serviced, though probably fertile.

Another face of the attempt made here to reinstate rhetoric at the core of Realism is the way in which it blurs the distinction between practitioner and observer. The implicated and manipulative politician or statesman can no longer so easily be contrasted with the supposedly objective academic once the latter's utterances are allowed a part, albeit sometimes as no more than the merest eddy, in the flow of policy-making and the reproduction of its terms and procedures. On this view each uses language as much to achieve objectives as to express beliefs. Academics cannot have the satisfaction of naming reason as trumps and dealing the cards. Some will find this uncomfortable, but to many it will seem less uncomfortable than the evident positivist lie of scientific detachment and objectivity. Once again, however, the way to deal with this would seem to lie initially in practice, which to academics consists in research, as a way to establish provisional trust. No amount of abstract reasoning is going to help.

All this might seem to amount to no more than a suggestion that careful, theoretically informed scrutiny of the language of public utterances on international relations coupled with an engag;aae approach to the issues of the day are appropriate practices for people who might perhaps better conceive of themselves as intellectuals than academics, and that, diligently pursued, these strategies may pay dividends in clearer theoretical understanding of the practices themselves. The response of many historians and "common sense" Realists, not to mention Marxists, will surely be that they knew this very well already. Maybe. But one defense against this charge of redundancy would be that the trick of maintaining theory and practice in productive tension needs to be constantly re-learned in new circumstances, and, if nothing else has happened since Waltz wrote TIP, circumstances have certainly changed.

Indeed, these changes provided a point of departure for our comparison between TIP and LoA in the Introduction. Section I then provided an exegesis, clarification, and critique of both Waltz and some of his critics. In addition to exploring the logic of Waltz's theory in its own terms, it reconstructed and extended the theory in such a way as to transform it from Waltz's narrow and excessively closed structuralism into the more open theory that we have labeled Structural Realism. The opening of Neorealism into Structural Realism has two consequences. First, it brings the logic of Neorealism into contact with several other areas of international relations theory. In particular, clarifying the boundary between the unit and structure levels of analysis shows how theory at the level of distributional structure can usefully be tied into study of process formations in fields as diverse as Strategic Studies and International Political Economy (IPE). The opening up of the second tier of structure (functional differentiation) under anarchy also points to links with theory of the state, and to possible further links with IPE, some of which were pursued further in Sections II and III.

The other consequence of opening was to undo the excessively static formulation of Neorealism by introducing three elements of change to work alongside its highly durable "shoving and shaping" structures. Waltz's formulation was virtually static except for the distribution of power in which one change from a multipolar to a bipolar system after the Second World War was allowed.

The first element of change is achieved by disaggregating power. Once singular power is disaggregated into plural capabilities the possibilities for change in the distributional structure increase automatically. Economic and ideological polarity can and have changed while military bipolarity has remained more stable. When power is disaggregated, the possibility for change is increased not only by the additional variables in play, but also by the combinations and permutations of how the different types of polarity relate to each other. In turn, the overall pattern of distributional structure has to be related to the shifting contexts (or regimes) of non-state activity within which state capabilities are exercised. This unpacking of power is both supported and made necessary by the use of system sectors as a tool of analysis, a theme explored further in chapter 12.

The second element of change comes from opening the functional differentiation tier of structure within anarchy, which Waltz treats as closed. Once opened, this tier offers opportunities for change not only as between type 2 and type 4 anarchies (respectively, those with like and with unlike units), but also within type 4 as between different classifications of functional differentiation. This opening paves the way for the more comprehensive exploration of Structural Realism in historical application in Section II.

The third element of change came from adding interaction capacity as a new level of analysis enabling Structural Realism to meet the requirements of a full system theory. Because interaction capacity rests on the absolute qualities of attributive power rather than the relative weight of relational power, it is by definition sensitive to a substantial range of ongoing technological, normative, and organizational changes in the international system.

These three elements add considerably to the reach and complexity of structural theory in International Relations. They meet many of the criticisms of Neorealism, but do so without losing touch with its basic ideas about the conditioning effects of international political structure. The theory is nevertheless quite transformed from narrow Neorealism; Structural Realism takes on many qualities not obviously present in Neorealism, and in some cases seemingly opposed to it.

The purpose of Section I was to establish an overview of the transformation of Neorealism to Structural Realism, but there has not been space to work out to the fullest extent all of the implications and consequences of the argument. There is thus a research agenda arising that still needs to be addressed, and it will come as no surprise that this agenda stems largely from the three elements of change introduced into the theory.

Waltz's theory on the distribution of power generated a whole subliterature on the analysis of power polarity. For the most part, this work has reached a dead-end because the variable turned out to be too crude to explain much. The disaggregation of power offers the possibility of reviving this debate on a sounder footing, but it requires systematic rethinking and elaboration of new hypotheses. Several possible lines of development were suggested in chapter 3. It is clear that disaggregation involves not only thinking through the definitions and implications of polarity regarding specific capabilities, but also the new step of thinking through the consequences of different polarities occurring in parallel in different sectors. What happens when military polarity and economic polarity don't line up, as is the case during the 1990s? What is the difference between a situation in which ideological polarity correlates with the location of military power, as during the 1930s, and when it doesn't, as might be one reading of the relationship between Islam and liberal capitalism as state ideologies during the 1990s?

One consequential question arising from these considerations is whether the influence of distributional structure (and thus its explanatory power) declines in relation to unit and interaction capacity explanations when the different sectoral polarities pull in different directions. In other words, is the shaping force of distributional structure stronger when events produce the (Neorealist) condition of aggregated power, and weaker when they do not? A second consequential question: do the prevailing structures of non-state systems in the economic and ideological spheres substantially constrain the shaping force of the distributional structure of state power?

A second area needing further thinking is the clarification of criteria for defining and specifying differentiation of function. Some suggestions are made in chapter 3 and the theme is carried forward in Section II, but this task is filled with hazards as well as opportunities. One tempting, though in the end untenable, reason that Waltz might have given for leaving the second tier of structure closed is that it creates too many problems if opened. Two issues need to be confronted. One is to clarify what counts as functional differentiation and what doesn't. The mistaken desire of some to classify the international roles of states as a function has already been mentioned. Another way of looking at this task is to see it in terms of the need to construct definitions for the structure of states that are as clear as those we now have for the structure of the international system. Once an acceptable range of "function" has been established, the other task is to establish criteria for the extent of change that is to be considered structural in terms of the international system. Setting very broad criteria, as Waltz does for his conception of states as like units, reduces the frequency of structural change. Narrower criteria, such as those suggested in chapter 3 and Section II, make deep structural change more frequent, but have the advantage of linking structural theory clearly and neatly into other areas of theory, particularly theory of the state and IPE. It is conceivable that further work in this still largely virgin territory would expose strong links between differentiation of function and interaction capacity. If so, this might at some point call for a reconsideration of the close linkage between tiers one and two of deep structure as laid out in chapter 3.

The third item on the research agenda of Structural Realism is to develop a fuller conceptual apparatus for handling interaction capacity, and to further integrate this level of analysis into international systems theory. One task is to articulate criteria for differentiating between significant and incidental change in the level of interaction capacity. When does the introduction and spread of a new technology, a new norm or a new organizational network actually produce a step-level change in the functioning of the international system? Another task is to think further about interaction capacity and the definition of system. The arguments in Sections I and II indicate that there are very powerful connections between the level of interaction capacity and the existence and function of structure in the system. Among other things, we need to clarify the criteria for structure in the light of interaction capacity, and to attempt to understand the history of the international system in terms of both the unit-structure and interaction capacity insights of Structural Realist theory. Indeed, one purpose in developing Structural Realism was to equip ourselves with theoretical tools able to address the whole history of the international system.

Section II provided a preliminary foray into this terrain, and ensures that the Structural Realist approach developed in this book is not subject to the charge of ahistoricism. The rise and fall of the Roman Empire was chosen as a historical case study in part because these events are difficult to accommodate within the Neorealist framework and in part because they occur in a very different context to the modern international system. The charge of ahistoricism has often been leveled at Waltz and, indeed, at many social scientists who appear to be working within the confines of positivism. But as was demonstrated in Section III, the claim that Waltz is a positivist is suspect; on closer inspection his epistemology is seen to contain more than a dash of pragmatism. By the same token, the superficial signs of ahistoricism which undoubtedly litter the text of TIP mask a clear acceptance on Waltz's part that there is a need to study historical change as well as continuity in the international system. Because Waltz asserts so vigorously that he is interested in continuity rather than transformation, most critics have failed to observe just how much allowance he makes for historical change in his theory. Section II identifies the dynamic features of the theory and then goes on to extend them.

There are two major lines of development and both build on ideas raised in Section I. The first carries forward the inclusion of functional differentiation in the Structural Realist approach. It begins this task by exposing an incipient theory of the state embedded in TIP. Critics of Waltz have argued that he treats the state as an ontologically primitive concept. By this they mean that there is no attempt to explain the origin or evolution of the state. In making this claim, the critics have been misled by Waltz's rhetoric. A more sympathetic reading of TIP reveals that it operates at the interface of the structure of the state and the structure of the international system. As far as Waltz is concerned, international anarchy presupposes state autonomy: the two are inextricably interconnected. In Waltz's balance of power theory, as a consequence, the structure of the state and the structure of the international system are essentially fused. To ensure that the state survives, its agents have not only to take account of shifts in the pattern of alliances, but also to be prepared to make adjustments to the structure of the state to preserve its relative power position. Waltz is clear, moreover, that international moves to form alliances cannot offset, in the long term, the need to sustain domestic sources of power. The implications of this simple "truth" are examined in Kennedy's controversial account of the rise and fall of the Great Powers in the modern international system. States that fail to emulate successful modes of domestic power formation steadily slip down the power hierarchy.

In Section II we develop Structural Realism by defusing the structure of the state and the structure of the international system. Waltz's fusion of these structures opens the way to the Neorealist prediction about the homogenization of states over time. The defusion by contrast opens the way to our Structural Realist prediction about the potential for functional differentiation within the international system. Agents of the state, it is argued, are sometimes unable to respond to pressure imposed upon them by the structure of the international system because of the strength of the constraints generated by the structure of the state. So while Neorealism presupposes that all states will undergo restructuring, whenever necessary, thereby ensuring that the anarchic structure of the international system is reproduced, Structural Realism presupposes that the internal structure of states in the international system can diverge, and the resulting functional differentiation can be preserved by the external dimension of the balance of power. This defusion of domestic and international structures helps to account for the divergence between type 2 and type 4 anarchies discussed in Section I.

Section II goes on to develop the idea that anarchy can also be conceived as a differentiated phenomenon which is not governed as Waltz indicates by a single logic. The nature of the logic is dictated by the structure of the units in the system. Examples of semi-autonomous units reveal that anarchy does not necessarily generate balance of power behavior. This extension has particularly important ramifications if it is accepted that, until quite recently, the international system consisted of relatively autonomous subsystems that have operated on the basis of different logics of anarchy. Only with the rise of interaction capacity at the global level have these subsystems finally coalesced.

The second development of Structural Realism is designed to build on the idea of an interaction capacity. Reference is made in Section I to the role of societal norms in the process of interaction. Section II further explores the reasons why, contrary to Waltz, cooperation can occur within an anarchy, and give rise to norms and rules within the international system. By opening up this line of thought, contact can be made with a growing body of literature that defines the international arena in societal rather than systemic terms. Ironically, this school of thought also fuses the structure of the state and the system, but in very different terms to those favored by Waltz. Its adherents argue that the sovereign state and the international system are mutually constituted, and it is therefore not possible to conceive of the state as an autonomous entity within an independent international system. Membership of international society is a defining feature of the state with the structure of the state and that of the international society being mutually constituted. So whereas Neorealists argue that there are mutual constraints set up by the structure of the state and the system, the societal school of thought argues that the state and system are mutually constituted.

The difference is not simply one of terminology. Neorealists argue that the autonomy of states is preserved by the balance of power while the societal school argue that the independence of sovereign states is preserved because these states collectively implement a body of rules that ensures the reproduction of a system of states. From a structural realist perspective, however, both of these schools of thought are thinking in terms of the contemporary global system which is made up of units that can be defined in either sovereign or autonomous terms. Structural Realism, as defined in this book, does not seek to adjudicate between these two schools of thought. Instead, it accepts that there have been international systems where power predominates and others where societal norms play a role in the constitution of both the state and the system. But it is also accepted that even in systems where norms play an integral role, the significance of power persists. The quality of anarchy will be very different as between international systems and international societies. Among other things, international society might support type 4 systems, as was the case during the nineteenth century. Developing these ideas further requires the clarificatory work on functional differentiation and the structure of the state sketched above.

In maintaining this position, a number of difficult methodological questions about the relationship between explanations based on causal factors and on the process of reasoning have been glossed over. Scientific realism provides one possible route for reconciling these two different modes of explanation. But as is made clear in Section III, there is, as yet, no solution capable of overcoming all the difficulties. It would be a mistake to presuppose that it is not possible to make any progress until the problems of methodology and epistemology have been overcome. On the contrary, it can be argued that methodologists and epistemologists necessarily follow in the wake of social scientists who are engaged in the practice of trying to understand social reality (Little 1991). This section leads to a conclusion very similar to that of Section III: that progress in the task of understanding international relations will only come about by moving beyond the narrow confines of the modern international system and drawing much more systematically on world history.

Structural Realism, as mapped out in sections I and II, generates an extensive research program. In the first place, the theoretical ideas associated with Structural Realism can be used to trace the broad outlines of how the international system has evolved. It is already clear that these outlines will diverge sharply from what Clark (1989) has called Whig and Tory views of world history. The former depicts world history in terms of the steady advance of civilization, while the latter sees only historical cycles, and sometimes regressive cycles at that (Farrenkopf 1991). Structural Realism, by contrast, focuses on the initial diversity of autonomous subsystems within the international system. Once the range has been identified, it becomes possible to look in more detail at the different structures that maintain these autonomous subsystems. In looking at the detail of these empirical cases, it also becomes both necessary and possible to pay more attention to the kinds of theoretical debates that are examined in sections I and II. Although an attempt has been made to reconcile some of the major theoretical disputes in the literature, detailed comparative research is undoubtedly necessary to make further progress.

Thinking about the meaning of structure in a full historical and systemic context points to the disturbing conclusion that the effects of structure are not constant. Significant variations occur caused by important links between both unit and structure and interaction capacity and structure. Between unit and structure it becomes impossible to avoid the conclusion that units and structures are mutually constitutive, and that the nature of the units therefore affects the consequences of structure. This emerged from arguments in both sections II and III. It is perhaps most visible in International Political Economy in the difference between a system of liberal states and a system of mercantilist ones. The openness of liberal states is what enables an international market structure to emerge, the closure of mercantilist ones directly suppresses market structure. Between interaction capacity and structure it becomes impossible to avoid the conclusion that both the capacity and type of interaction that define the system also affect the consequences of structure. Neorealist theory focuses on a middle level of capacity, and produces a logic of anarchy centered on strategic interaction. At higher and lower levels of capacity the system may be defined principally by types of interaction other than strategic. Such systems will display a logic of anarchy quite different from that sketched by Waltz.

Structure is therefore only one component of a more complex systemic equation. There is not one logic of anarchy but many. This is not a cause for abandoning the theoretical enterprise in despair. It does wreck the fences around Waltz's tidy universe in which a single logic of anarchy prevails. But it leaves intact the powerful theoretical apparatus of structural analysis. Since anarchy remains the great constant in the international system this is no mean legacy.

Structural Realism provides the tools with which one can assess continuity and change in the international system. It offers the possibility of thinking systematically about both the whole history of the international system, and some aspects of its future. To do this, Structural Realists need to equip themselves with a set of lenses through which they can examine all of the sectors of the system. They need to train themselves to interpret the partial picture of the whole that each lens gives, and to be able to see how the different perspectives relate to each other. They need to avoid the illusion that sectors are self-contained universes, and to cultivate an awareness that language is part of the political process. They need to understand that a systems approach necessitates awareness not only of the constraints of structure, but also of the structuration process between units and structure, and of the pervasive influence of interaction capacity on the meaning and significance of system. They need, in other words, to take a very comprehensive view of the traditional wisdom that Realism is about understanding what is. Only with that foundation can one think usefully and responsibly about the potential for purposive change.

When more fully developed, and if used with care and precision, these tools could take us a long way down the road toward an integrative and coherent theory of international relations. There still remains a great deal to do.