The Logic of Anarchy: Neorealism to Structural Realism

Barry Buzan, Charles Jones and Richard Little

New York

Columbia University Press

1993

5. Structural Realism and World History

 

The apparent ahistorical character of Neorealism has been a persistent source of criticism in the literature. By reconstituting and extending Neorealism into Structural Realism we hope that this criticism can be deflected. In this chapter, therefore, we wish to explore the implications of the criticism in a little more detail. It is important to recognize, in the first place, that Neorealists do not deny the relevance of history or the potential for change in international politics. But they do assert that there are important features of international politics, such as imperialism, that have occurred throughout the history of the international system and that need to be accounted for in terms of an unchanging systemic structure. It is this claim that analysts imbued with historicism wish to deny. Imperialism, they accept, is a label that has been attached to state practices throughout history. But the historicists insist that distortion will inevitably occur if it is presupposed that these practices always play an identical role in the international system or that they always carry the same subjective meaning simply because they are identified by a common label. From this perspective, then, the imperialistic practices associated with the Athenians must be distinguished sharply from those associated with the British. Pushed to the limits, this line of argument makes it impossible to engage in comparative analysis across time or cultures. But it is not intended here to engage with this extreme line of argument. Instead, attention will be focused on the work of historians who have accepted the value of comparative historical analysis but who deny that world history can be comprehended within the confines of an unchanging structural framework.

Before looking at this work, it may be helpful, first, to summarize just how extensively Waltz's theory has so far been changed. Against this backdrop we can then examine a number of competing structural frameworks that focus on the way the international system has evolved during the course of world history. Once these have been presented, it becomes easier to identify how further refinements to the Structural Realist framework will help to throw light on the relationship between system continuity and system transformation. Finally, we need to examine in more detail some of the reasons for drawing on the Roman Empire as a case study.

 

Neorealism to Structural Realism

Waltz starts from the underlying premise that since international phenomena like war, alliances, and imperialism have been ever-present features of world politics, some constant factor in the international system must be identified and drawn upon to account for them. Waltz argues that the only permanent feature in international politics is the structure of the system itself. We begin by accepting that the structure of the international system must play an essential role in any general theory in the field of International Relations. The reason for paying such close attention to the agent-structure problem later in this section, therefore, is because the debate surrounding the problem focuses on the role played by the idea of structure in the theory and methodology of the social sciences.

In Neorealist theory, the structure of the international political system is characterized by only two components: (1) the anarchic ordering principle of the international system; and (2) the distribution of power capabilities among the political units in the international system. With the aid of these structural components, Neorealists go on to build a conception of the balance of power, which constitutes the core of their theory of international politics. Little reference was made to the balance of power in the previous section because attention was devoted, instead, to the structural bedrock of Neorealist theory. The reconstructions carried out on the bedrock substantially modify Waltz's conception of structure and also the way he distinguishes between the structure and constituent units of the system. The result is to produce a much richer conception of structure and a more complex relationship between the structure and units of the international system.

Attention must be drawn, first, to the illuminating metaphor of "deep structure," taken from the literature on structuralism, and which Ruggie (1986:135) has applied to the principle of anarchy. Although he does not use the idea, there is no doubt that Waltz also sees the anarchic principle as a more fundamental feature in the structure of the international system than the power component, which, extending the metaphor, can be considered to represent the "surface" or what we call the distributional structure of the system. Ruggie argues that "deeper structural levels have causal priority" with the result that "the structural levels closer to the surface of visible phenomena take effect only within a context that is already `prestructured' by the deeper levels" (1986:150). It is for this reason that deep structures are said to be "generative," because they "generate" observable patterns of behavior in the system. Moreover, Waltz believes that the structure of the system is so powerful that it will generate common patterns of behavior among very different types of unit. He notes that "The logic of anarchy obtains whether the system is composed of tribes, nations, oligopolistic firms or street gangs" (1990:37).

The conceptual utility of deep structure becomes evident only when attention is drawn to the potential for the international system to undergo transformation. As far as Neorealists are concerned, system transformation can happen only if the ordering principle of anarchy gives way to that of hierarchy. In the international context, therefore, for system transformation to occur, independent political units must be taken over by world government or empire. At that juncture, the defining structure of international politics is eliminated. As a consequence, the shift from anarchy to hierarchy constitutes a transformation from decentralized to centralized politics. Since there has never been a global hierarchy in world history, a Neorealist transformation has never occurred. World history, therefore, can be described in terms of a continuous system that has never undergone transformation. Neorealists accept, of course, that dramatic shifts have taken place in the distribution of power among political units in the international system. But these changes at the "surface" of the structure do not constitute a transformation in the system. Indeed, the Neorealist theory of the balance of power is designed to show how the structure of the system inhibits any single political actor from achieving global dominance.

It was argued in the previous section that this position is profoundly unsatisfactory, because historically there have been momentous changes in the international system that transcend modifications in the distribution of power among the political units. The inability of Neorealists to accommodate more extensive systemic transformation represents a familiar line of criticism. In one of the most interesting and sympathetic studies of Waltz, Ruggie sets out to show how Waltz's theory contains "only a reproductive logic but no transformational logic" [1986:152]. Cox, who is more critical of Waltz, also points to the "inability of his theory to account for or to explain transformation" (1986:243). The nature of the criticism is very straightforward: while Neorealism can help to explain continuity in the international system, it lacks the capacity to account for structural transformation. Waltz is unable to describe, let alone explain, what Ruggie sees as the most important system change in the last thousand years: "the shift from the Medieval to the modern international system" (1986:141). It is seen to be essential that a structural approach to international politics should be able to describe and explain such massive transformations in the system. But it is also presupposed that in accommodating the potential for system transformation, it is vital that the important advantage derived from the idea of deep structure--that of continuity--must not be lost.

The desire to take account of both continuity and transformation within a structural framework provides the entry point for our analysis in the previous section and it represents the touchstone for Structural Realism. Like Ruggie, we reintroduce the idea of functionally differentiated units as a structural dimension of the international system. Waltz eliminated this structural component on the grounds that anarchy generates functionally similar units. Although that argument is a persuasive one, we insist that it is not a logical consequence of anarchy. It is possible, after all, to conceive of functionally differentiated units operating in an anarchic system and there are substantial historical precedents for systems with these structural features. Unlike Ruggie, however, we do not see functionally differentiated units as simply mediating the deep structural effects of anarchy (1986:135). Instead, the functional differentiation of political units is seen to dictate the character of the anarchic deep structure of the system.

By making this move, we can argue that the deep structure of the international system, defined in terms of its ordering principle of anarchy, has remained unchanged throughout world history. But by incorporating the additional structural component of functional differentiation we are also able to identify that there can be change in the deep structure of the international system, as, for example, when the complex web of competing authority relations in the Medieval world gave way to the contemporary international system made up of functionally undifferentiated sovereign states. It also allows us to distinguish between the modern international system which is made up of sovereign states and earlier systems where states were autonomous without being sovereign (Wight:1977; Wendt:1990). This line of argument seems to go some way to meeting Cox's objection that Waltz has endeavored to achieve theoretical clarity, but at the cost of relying on an "unconvincing mode of historical understanding" (1986:243).

At first sight, this solution certainly overcomes some of the criticisms directed at Neorealist structural theory. But the solution does more than deflect criticism. By consigning both functional differentiation and anarchy to the deep and most generative structure of the system, we open up the possibility that anarchy is a much more complex phenomenon than Waltz recognizes. The significance of the move becomes apparent when a comparison is made with Ruggie, who argues that the deep structure of the international system is restricted to the ordering principle of anarchy. As a consequence, according to Ruggie, the differentiation of political units can only mediate the deep structural effects of anarchy. Logically, however, if functional differentiation is considered to be part of the deep structure of the international system, it must do more than simply mediate the generative consequences of anarchy. It will be argued in this section that the notion of anarchy is more complex and problematic than is generally acknowledged. It constitutes a differentiated phenomenon and cannot simply be characterized in terms of the absence of central government. The implications of assigning functional differentiation to the deep structure of the international system and its impact on the conception of anarchy will be explored later.

A second line developed in Section I is the provision made for the analysis of different sectors of international activity. Attention is drawn to economic, societal, and strategic sectors, for example, as well as to the political sector that Waltz exclusively focuses upon. Stated baldly in this fashion, the full significance of this development is not immediately apparent. But as will be shown in more detail later, it has profound consequences for the direction theory and method should take in the discipline. The sectoral argument is made, in the first instance, because it is presupposed that the structure of the international system has an impact on a wide range of activities extending beyond international politics. But the deeper significance of this presupposition can become apparent only when the methodology underpinning the idea of structure has been more extensively spelled out. It can then be demonstrated that any attempt to understand the nature and impact of the structure of the international system will be necessarily incomplete if the focus is restricted to the political sector. It follows that a theory of international politics cannot account for the structure of the international system; only a wider theory of international relations can do that.

A third point made in the previous section is to extend the Neorealist conception of power. A distinction is drawn between relative and absolute ("attributive") power and it is demonstrated that Waltz focuses on the former rather than the latter. It is then argued that whereas relative or relational power is inextricably part of the structure of the international system, absolute or attributive power must be separated from the structure of the system, although it also has distinct and important consequences at the system level. It is argued that this attributive dimension of power cannot be pushed into the unit level of analysis. Instead, attributive capabilities are seen to be a distinct systemic level of analysis. Identified as the interaction level of analysis, attributive capabilities are seen to measure the interaction capacity of the system. By opening up this distinction, it becomes possible to differentiate between international systems characterized by system and subsystem dominance. The importance of this distinction becomes more apparent when more detailed attempts are made to describe and explain the evolution of world history.

It will be shown later that these three developments interact and so, when combined, have very important theoretical and methodological consequences for any attempt to analyze world history. An attempt is made to draw out some of these consequences in subsequent chapters in this section. Before doing so, it is worth pausing briefly to look in more detail at the implications of Neorealism for the study of world history, and to examine two alternative frameworks. It is hoped to show how Structural Realism is more effective for analyzing transformation and continuity in the international system than any of the alternatives.

 

Frameworks for Analyzing World History

Few theorists in the contemporary field of international relations have attempted to take account of world history. Attention has been focused instead almost exclusively on the sovereign state system which prevails today. As a consequence, when consideration has occasionally been given to world history, the temptation to think of the past in terms of the present has often proved overwhelming. When this is done, it becomes relatively easy to argue, for example, that the conflict between Athens and Sparta in the Greek city state system, or the conflict between Carthage and Rome can be taken as analogues for the Cold War struggle between the United States and the Soviet Union (Fleiss 1966). This widespread practice can also be used to illustrate the apparent dominance of Realist thinking in international relations.

But there are alternative frameworks for thinking about the past that do not rely on prevailing realist or neorealist assumptions (Alker 1987, 1990). The contrast becomes quite striking when a comparison is made of the way that the various frameworks handle familiar phenomena such as imperialism and empires. It quickly becomes apparent that the Neorealist's lack of interest in system transformation has major consequences for the way that notions of empire and imperialism are conceptualized. For example, it can be shown that frameworks which accommodate the possibility of system transformation can develop a differentiated approach to the concepts of empire and imperialism whereas the Neorealist framework encourages an undifferentiated approach to those concepts.

Empires in the Neorealist's Framework

As already indicated, Neorealists argue that imperialism is a universal phenomenon and it is presupposed, as a consequence, that it is possible to fit imperialism into a general theory of international politics. In making this presupposition, Neorealists assume that all empires take the same structural form. From this perspective, then, no structural distinction can be drawn, for example, between the British and Roman Empires. In both cases, the imperial state is seen to have extended its territory in the context of a broader anarchic system. By formulating imperialism and empire within the context of an anarchic arena, the Neorealists can contemplate a theory of imperialism that will account for the formation of empires in the ancient and modern world. Moreover, Neorealists can accommodate imperialism within their general theory of international politics, which revolves around the conception of the balance of power. It follows, moreover, that as far as the Neorealists are concerned, imperialism can precipitate system change, but it cannot be considered to be a source of system transformation. The deep structure provided by the principle of anarchy is seen to transcend the rise and fall of nonuniversal empires.

Empires in the World Systems' Framework

By contrast, at the other extreme to the Neorealists, lie the World Systems theorists, as exemplified by Wallerstein (1974). Here, a sharp distinction can be drawn between the British and Roman empires. From the World Systems perspective, Rome, unlike Britain, is seen to have effectively eliminated competing centers of power; as a consequence, it constituted a "world" empire, whereas the British empire always operated in an anarchic arena made up of competing states. Wallerstein, therefore, draws a sharp structural distinction between empires which exist within an anarchic system and those which transform the system and eliminate the anarchic structure. Wallerstein is one of the few theorists to explore this distinction and it enables him to view world history in terms of alternating patterns of world systems, with anarchic international systems, identified as world economies, being recurrently overtaken by world empires. Anarchic systems, he argues, "were always transformed into empires" (1974:16) and he gives the examples of China, Persia, and Rome. Wallerstein is quite clear that these are all examples of world empires and need to be distinguished from empires, like the British empire, that existed in the context of a broader anarchic system. A world empire is seen to be a self-contained, autonomous unit existing in a global environment. The British Empire, by contrast, was never a self-contained unit. It was always an integral part of a much larger system with which it interacted continuously.

These competing conceptions of empire raise the difficult question of what might be meant by a boundary that circumscribes a world or international system. Neorealists treat the international system as completely closed; it is not considered to have a boundary. They are unconcerned about the question of how to define the boundary of a world or international system. On the other hand, the question of boundary definition is of central importance for world systems theorists. For them, it is essential to be able to identify the grounds on which the Roman Empire, for example, can be treated as a closed system. It is on the basis of the conception of a bounded system that Wallerstein is able to argue that prior to the modern world system, a number of world systems always coexisted on the earth's surface, taking the form of either a world empire or a world economy. He argues that these world systems coexisted and should not be seen as components of a larger social system because life within each of the systems was "largely self-contained" and the dynamics of development in the system were "largely internal" (1974:347). This formulation accepts, in other words, that there was some contact between the coexisting world systems. But it also presupposes that the contact was insufficient to have any material effect.

Despite the extensive discussions of Wallerstein's work, little attention has been paid to his notion of a bounded international system or to the structural distinction he draws between a world empire and a world economy. The failure to explore the distinction says a good deal about the ethnocentric and state-centric orientation of the contemporary discipline of International Relations. The absence of discussion can be accounted for by the general lack of interest displayed by specialists in international relations in either world history or comparative method (Onuf:1982).

Empires in the World Historian's Framework

When attention is turned to the work of world historians, it becomes apparent that they do not endorse the concept of empire developed by either the Neorealists or the world systems theorists. The theoretical framework developed in world history not only poses problems for the Neorealist conception of empires operating in a closed or unbounded anarchic system but also throws up difficulties for the distinction drawn by the World Systems theorists between world empires, like the Roman Empire, formed within a closed or bounded system and the mercantile empires, like the British Empire, which constitute a type of state operating in an open anarchic system. Bozeman points out, in the first instance, that although the empires of Han China and Rome coexisted they both pretended to hold "unique and undisputed power." Both regarded the frontiers of their respective empires as the limits of the civilized world. Nevertheless, they each knew of the other's existence and they maintained contact indirectly by means of trade. Moreover, Bozeman identifies a "steady flow of intellectually adventurous people" who managed to stimulate a "general awareness in the ancient world that no one empire is ever quite as self-sufficient as it pretends to be." It was certainly the case that "the two world states and the great area of Indian civilization touched at their peripheries" (Bozeman 1960:162-164). Bozeman goes on to assert that Han China aimed to synchronize "explorative diplomacy, economic infiltration, and military pressure, with the ultimate purpose of pressing through to the West." And she argues that Rome used similar tactics when it extended to the East. So both China and Rome must have contemplated the consequences of a direct confrontation. Bozeman goes on to argue that:

Contrary to what one might expect under such circumstances, this prospect does not seem to have fanned fear or rivalry in either of the imperial camps. The indications are to the contrary, namely that their imperialistic policies moved each of these two powers towards an interest in direct, reciprocal, and friendly commercial relations. It is impossible to say whether this amicable attitude would have endured following the consummation of a direct meeting, however, for the Sino-Roman trade was destined to remain indirect (1960:169).

McNeill (1979), has established a general theoretical framework that builds upon this idea of indirect cultural contact. Empires can be viewed as a subordinate component of this broad scheme. According to McNeill, it is possible to explore world history on the basis of competing civilizations. He traces the evolution of four centers of civilization which, he argues, emerged in the period to 500 B.C. the civilizations were located in Greece, India, China and the Middle East. McNeill's central organizing idea indicates that whenever a development took place in a community which was seen to make it more attractive or more powerful, then the techniques or ideas underlying the development would be borrowed and adapted by neighbors and neighbors' neighbors to suit their local conditions. By the processes of borrowing and adaption, the four civilizations managed to retain their distinctive characters while coexisting in a rough equilibrium for two thousand years from 500 B.C. to 1500 A.D.

The four civilizations, according to McNeill, occupied but did not cover the Eurasian land mass. They were separated by bands of territory inhabited by nomadic peoples. They were characterized by the Greeks and Romans as barbarians. Inevitably, as the four civilizations developed, they also expanded into the territory occupied by the barbarians, thereby reducing what McNeill refers to as the "barbarian zones," which served to insulate one civilization from another. Although contact between the civilizations increased, McNeill insists that their autonomy and the overall equilibrium between them was maintained.

There is obviously a degree of similarity here between McNeill and the Neorealists, with both pointing to the existence of a balance maintained on the basis of a process of diffusion between autonomous entities. Moreover, in each case, the balance is viewed as an unintended consequence of action taken within the entities. But there are also substantial differences between the two analyses. The entities identified by Waltz are hierarchically organized political units he refers to as states. The term is seen to embrace political actors as different as tribes, city states, and empires. By contrast, McNeill makes reference to civilizations where the identifying link is cultural. Within a civilization, he suggests, individuals conform to a "loose yet coherent life style" (McNeill 1979:v). Moreover, despite the equilibrium between the civilizations, these life-styles are seen to have diverged very substantially. The variations were certainly sufficient to ensure that the political configurations in each civilization were quite distinct. So, for example, although the Chinese civilization fragmented politically on several occasions when it came under pressure from the barbarians, political unity was always reestablished. For most of its history, China operated, in Wallerstein's terms, as a world empire. By contrast, the civilizations that evolved around Greece took a very different form. At one point, it was transformed, by the Romans, into a world empire. But the empire eventually fragmented. And in contrast to China, attempts to restore the empire failed so that the civilization remained politically fragmented.

There are other obvious differences between McNeill and Waltz. The Neorealist framework focuses on an international system that consists of interacting states. These states are considered to be mutually vulnerable, because they all represent a potential threat to each other. But in the case of McNeill, the emphasis is on civilizations that tend to coexist rather than coact. McNeill observes some interaction but identifies it as a "disturbance" or "shock" which upsets the equilibrium between the civilizations. It is possible, he argues, to identify four major disturbances in the period between 500 B.C. and 1500 A.D. These represent major "bench marks" in the course of world history (McNeill 1979:129). The first two disturbances, the Greek penetration of the Middle East and the extension of Indian influence into China and Japan, had few long-term consequences because there was a reassertion of the indigenous civilization. But other "bench marks" represented by the spread of Islam from the Middle East and the capitalist expansion of the West were both to have major and long-term consequences for other civilizations.

But these "shocks" to McNeill's cultural equilibrium are very different from the disturbances associated with Waltz's political balance of power. McNeill's framework embraces both the Roman Empire and the Han Chinese Empire which were linked by a process of cultural diffusion. But the two empires cannot both be included within the Neorealist framework, where balance of power theory presupposes that states pose each other with a direct threat and, by the same token, provide each other with the possibility of direct assistance. The Roman and Chinese empires were not linked in this way. McNeill's framework suggests that the boundary drawn by Wallerstein is too tight. It ignores the complex nonpolitical interactions that took place across the frontiers of the Roman Empire. In comparing these three approaches it quickly becomes apparent how dependent each of them is on certain assumptions about the types and levels of interaction capacity in the system. It also becomes clear how vital it is to achieving any understanding of international systems to develop a clearly specified interaction level of analysis.

World Empires in the Structural Realist Framework

The impossibility of dealing with the Roman Empire as a closed system becomes immediately apparent once it is acknowledged that the frontiers were never static. Throughout the history of the empire the boundaries were either expanding or contracting. Nor is it possible to suggest that a process of expansion gave way eventually to a process of contraction. The evolution of the Roman Empire was much more complex than such a formulation allows. But it can be suggested that during the period of major expansion Rome impinged upon and then overtook a number of independent or relatively autonomous regional subsystems.

It is not helpful, therefore, to depict the expansion of Rome in the context of a closed Neorealist system of states. It makes more sense to identify the growth of Rome taking place initially within an expanding subsystem that coexisted with and then eventually overtook other subsystems made up of political units that could be sharply differentiated from Rome. The agents of these units perceived themselves to be operating within an international subsystem that was relatively autonomous. The Greeks, for example, as Herodotus (1987) makes clear, were conscious of two international systems: the system of Greek city states and the Persian empire. According to Herodotus, these two systems naturally balanced each other and had to be kept separate but equal. Herodotus "deals with a period which constitutes a disturbance of that order by the unlimited expansionism of the Persians" (Immerwahr 1966:306). Waltz would argue, of course, that the Persians and the Greek city states were operating within a single international system. But this is not how the agents of the state saw the situation. They clearly acknowledged the need to establish boundaries between independent international systems. They were also aware when these boundaries were becoming vulnerable. World history reveals, moreover, that the Romans, in contrast to the Persians, were able to eliminate the boundaries that separated divergent subsystems of states. It is clear that the concept of interaction capacity can help to give some purchase on the approach developed by the world historians. One of our major aims here is to extend our framework so that it can further clarify the relationship between these subsystems and the overarching international system.

 

Structure, Transformation and the Roman Empire

From a Neorealist perspective, the Roman Empire is an inappropriate choice to illustrate the potential for structural transformation and continuity in the international system. For the Neorealists, the case simply provides an example of a successful unit operating in an anarchic arena of states. But as we have seen, neither Wallerstein nor McNeill would endorse this assessment. Within the context of their frameworks, the Roman Empire poses Neorealist theory with some tricky problems. Nevertheless, most theorists in the field of International Relations have, until recently, largely ignored the enormous and controversial literature associated with its rise and fall. It could be argued that this evasion has occurred primarily because the evolution of the empire does not fit neatly into the familiar interstate framework used to analyze more recent history. As a consequence, with only a few exceptions, there has been no exchange of ideas between the study of international relations and the study of the Roman Empire. Indeed, it is the difficulties associated with any attempt to analyze the history of the Roman Empire in the state-centric framework that makes it an appropriate case study to draw upon to test the strengths of Structural Realism.

The emergence of the Roman empire provides, in the first place, an important instance when the balance of power apparently failed to function. According to Neorealism, competing political units, observing the expansion of Rome, should have been willing to collaborate in order to keep Rome in check. But this failed to happen, thereby falsifying the theory. Morrow has noted, for example, that as the Roman Empire expanded, the Diadochi Empires (Macedonia, the Seleucid Empire, and Ptolemaic Egypt) preferred to fight among themselves rather than form an alliance against the ascendant power. "Why" asks Morrow "does Waltz's argument about the stability of anarchic systems fail in this case?" (1988:87). He could have asked the same question in the context of the Italian city states at the end of the fifteenth century. Instead of consolidating their strength against the emerging nation-states, the French were encouraged by Milan in 1494 to intervene in Italy, thereby fueling an established imperial trait in France. In both these cases, Waltz's theory provides no obvious explanation as to why the necessary collaboration among the weaker powers failed to occur. Morrow is also unable to offer a solution; the best he can do is to suggest that the most effective way forward is to examine the situation on the basis of a methodology which seeks to synthesize structure and agency. The basic features of this methodology will be looked at in more detail in the next chapter.

Structural Realism provides a possible explanation for the failure of the balance of power theory. It can point, for example, to the subsystemic character of the arena within which Rome expanded, as well as to the functional differentiation of the units within the subsystem. When Rome expanded, it confronted a range of radically different political entities, including the barbarian tribes, the Greek city states, and the sprawling Diadochi empires, operating within semiautonomous subsystems. Waltz's theory assumes that the anarchic arena is made up of identical interacting units operating within a closed system. His framework is, as a consequence, ill-equipped to handle the arena within which Rome operated. By contrast, the Structural Realist framework developed in the previous section offers some possible clues that may account for evolution of the Roman Empire.

The consolidation of the Roman empire took place over a relatively long period of time. The Romans first absorbed the Italian peninsula, then they moved west, gaining control of the Western Mediterranean and eliminating the Carthaginian empire in the process. Only after their rule was secure in the West did the Romans turn to challenge the well-established and civilized empires and city states in the East. Once these areas had been engulfed by the empire, it becomes necessary to raise the question as to whether the Romans had managed to transform the structure of the international system in which they were operating. The question raises an interesting theoretical issue which the Waltzian framework must necessarily push to one side.

The competing structural assessments of the Roman Empire, therefore, have important theoretical repercussions. Waltz, for example, expresses an interest in imperialism, but his theory requires him to deny that imperialism can precipitate system transformation. Of course, Waltz can accept in principle that the movement from an international system to a world empire represents a structural transformation. But he works on the assumption that there has never been a world empire and that all empires in the past have existed in the context of a wider international system. This line of argument is certainly valid for the Diadochi Empires, but as already noted, the argument has been questioned in the context of the Roman Empire. Although it is true that the empire did not extend to the limits of the known world, a Roman goal which remained unrealized, it is possible to argue that the empire was effectively an autonomous system.

A number of recent theorists have begun to acknowledge that the Roman Empire provides a useful heuristic for breaking away from the established state-centric view of the world. The empire is now being recognized as a distinctive type of international actor. It is beginning to be argued that by understanding how the empire operated, it may be possible to come to terms with the process of system transformation in the contemporary international system. Strange has argued, for example, that to understand the role of the United States in the modern world it is essential to "escape the corset-like intellectual constraints of the conventional study of international relations" [1989:11]. These constraints are seen to be perpetuated by Neorealists. From Strange's perspective, Waltz, in keeping with other Realists, is bound by a spatial image of the international system which is seen to be neatly divided into territorial plots. Only by erasing this image, argues Strange, can we come to understand that from a contemporary perspective the nature of American power and the character of the American Empire is not territorial in character. Once the constraints of the prevailing perspective are escaped then it becomes possible to see, first, that it is necessary to draw a sharp distinction between the contemporary American Empire and the now dissolved European Empires and, second, that there are a number of interesting parallels between the American Empire and the Roman Empire that also need to be understood in nonterritorial terms. Strange calls for a careful study of the various forms empires have taken in the past and she is convinced that such a study will require scholars to transcend the structural barriers imposed by Waltz.

In making this point, she is confirming an argument made by Luttwak, who considers that the nature of security in the Roman Empire has more in common with the security problems that confronted the Western world in the aftermath of the Second World War than is generally realized. He asserts, furthermore, that scholars of the Roman Empire, like modern strategic analysts, are conditioned by a set of assumptions about the international system that reflect nineteenth-century prejudices. In attacking the judgments of contemporary scholars of the Roman Empire, he argues that they reflect "a seemingly ineradicable Clausewizian prejudice against defensive strategies and defensive construction--a prejudice as common among historians writing of Hadrian and his policies as among contemporary military analysts discussing today's ballistic missile defences" (1977:61). Luttwak believes that, paradoxically, the effect of the revolutionary changes in modern war "has been to bring the strategic predicament of the Romans much closer to our own" (1977:xii). This is an intriguing line of argument that raises important questions about the nature of structural transformation. Like Strange, therefore, Luttwak draws on the example of the Roman Empire to challenge the universality of the framework putatively advanced by the Neorealists.

There is one further point to be made about the value of using the Roman Empire as a case study to illustrate a theoretical discussion about continuity and transformation in the international system. It concerns the difficulty confronted by the Neorealist framework in dealing with the collapse of the Roman Empire. The difficulty arises because the reasons underpinning the demise of a state have not been explored by the Neorealists. Waltz argues that the "death" of a state in the international system, in contrast to the "death" of a firm in the marketplace is an unusual event. As Halliday (1989) argues, however, this is not a proposition that can be sustained when the broad sweep of history is considered. It becomes an extremely interesting theoretical question, therefore, as to why some states fail to survive. The initial success of the Roman Empire draws attention to the importance of the issue. Rome initially brought about the demise of a large number of apparently successful political actors. The question of survival under conditions of anarchy becomes even more interesting in the context of the fall of the Roman Empire because it is far from clear why the Western half of the empire, confronted by a series of weak and divided tribes, should have collapsed so completely and so rapidly while the Eastern half of the empire, confronted not only by the same tribes, but also the massive and powerful Persian empire, should have been able to survive and continue to flourish for another thousand years.

It is hoped that by using the Roman Empire as a case study, it will provide at least a crude test for some of the theoretical ideas that are discussed in this section. Attention is focused on the formation, expansion, and collapse of the Roman Empire not only because, like the Medieval period, it has been very generally ignored by theorists in International Relations, but also because it does throw up a series of difficulties for the Waltzian framework. Inevitably, it is not possible to do more than to provide a brief sketch of some of the titanic developments associated with the history of Rome. But the case study material helps to illuminate the theoretical propositions and justifies the need to think about transformation of the international system at a structural and a systemic level.