The Logic of Anarchy: Neorealism to Structural Realism

Barry Buzan, Charles Jones and Richard Little

New York

Columbia University Press

1993

7. Agency and Competing Theories of the State

 

A major criticism leveled at Waltz relates to his conception of the units in the international system. As already noted, Waltz has been attacked for being unconcerned about the ontology of the units. He is seen to argue that the rigid structure of the international system will have an unchanging impact on the behavior of states irrespective of their internal characteristics. Critics argue that this reputed lack of interest in the state has fatal consequences for his theory of international politics because it means that he is unable to account for the origins of either the state or the system. According to Wendt, the problem arises because in Neorealist analysis, the state is treated as an ontological given--a primitive concept which does not require further explanation. Wendt, however, also asserts that the state is a concept that requires the sort of substantial explanation that can be provided only by a full-blown theory of the state.

This line of argument then leads Wendt to attack Waltz's conception of the structure of the international system. He argues that Waltz defines the structure in terms of the distribution of power among states in the system. As a consequence, the international system can only be thought of in terms of states, because the structure of the system has been defined by the characteristics of the constituent agents. It follows that it is not possible to think of the system pre-existing the state and helping the state to emerge. So no generative element exists within the system to account for the formation of the state, or, indeed, any new kind of structure for the system. Wendt accepts Ruggie's argument that the theory can account only for how states once in existence endlessly reproduce themselves and the system in which they operate. There is no provision for system transformation.

 

Wallerstein's Theory of the State

Wendt contrasts this approach to the international system with the one established by Wallerstein. He argues that Wallerstein's framework represents an improvement because the structure of the international system is not defined in terms of the attributes of the agents under investigation. It is therefore possible to develop a theory to explain how the state emerged in the first instance. In contrast to the Neorealists, therefore, the world systems theorists have access to a theory of the state rather than to pre-theoretical intuition. The world systems theory of the state is embedded in the theory of the world economy. According to world systems theory, at the end of the Middle Ages an economic surplus began to be produced. As a consequence, trade began to increase and an economic division began to take place, with some regions becoming responsible for the production of primary products and other regions concentrating on more advanced activities--processing primary products, ship-building, and banking. Each economic activity, it is argued, produced a particular type of labor and a specific type of class structure and there was a movement away from the simple feudal relations that had prevailed during the Middle Ages. Complexity required a stronger state to mediate among the conflicting classes and to assert collective interests in the international arena. The strong states could then use their power to establish increasingly beneficial terms of trade, thereby reinforcing their power. It follows that the strong states operated at the core of the international system, while the weak states operated at the periphery.

Wallerstein argues, therefore, that "the world economy develops a pattern where state structures are relatively strong in the core areas and relatively weak in the periphery"(1974:355). The center then pursues policies that have the effect of reproducing the structure of the system, although the world systems theorists also acknowledge that the world economy operates on the basis of its own logic and that states do not stay in the same position. Change is possible and is identified when states move from the center to the periphery of the world economy. The basic structure of the system, however, remains unaltered, although Wallerstein does assume that at some point in the future there will be system transformation and the existing exploitative world economy will give way to a more just socialist world system.

Gourevitch (1978) has identified some very serious weaknesses in the approach. In the first place he points to a certain circularity in the argument. It is unclear if states are powerful because they are at the center of the world economy or if they are at the center of the world economy because they are powerful. The issue becomes more problematic when attention is turned to the empirical correlation between strong states and core economies. Gourevitch analyzed conditions in 1550, when the situation was as follows:

Core economies: Netherlands, northern Italy, southeastern England, parts of Spain, parts of France, parts of Southern Germany and Portugal;
Strong states: Spain, England, France and Portugal.

It follows that some of the most advanced areas had weak states. Conditions had changed by 1700, but from the standpoint of Wallerstein they remain problematic:

Core economies: Britain, France and Netherlands;
Strong states: Britain, France, Prussia, Austria and Sweden.

Here we see that there are more strong states than core economies.

At the very least, this exercise suggests that it is not possible to derive anything like a complete theory of the state on the basis of a theory of the world economy. Nevertheless, it can be suggested that the world systems theorists have a much more advanced theory than the Neorealists.

 

Waltz's Theory of the State

Wendt's claim that the Neorealists are devoid of a theory of the state is false. It is possible to extract at least a rudimentary theory from Waltz's TIP. In the first instance, Waltz insists that states represent the "primary political units of an era be they city states, empires, or nations." Waltz, therefore, has an undifferentiated theory of the state. His theory also presupposes that the emergence and initial evolution of these institutions is unrelated to the existence of the international system because they precede the existence of such a system. Although he does not elaborate on this point, it can be inferred that he would concur with the view of Thucydides (Pouncey 1980:51) that, in the first instance, states are purely inward looking institutions that lack the necessary power to bring them into contact with each other. It follows that the structure of an international system can only "emerge from the coexistence of states" and is then consolidated by the "coaction" of these units. The international political system, therefore, is deemed to be "individualist in origin," generated "spontaneously and unintentionally" because, as Waltz observes, "No state intends to participate in the formation of a structure by which it and others will be constrained." Because states precede the emergence of an international system Waltz argues that they are "formed and maintained on the basis of self-help." The relevance of self-help may become more important when states start to "coact" but, both before and after that point, the fact remains that whether states "live, prosper or die depends on their own efforts"(1979:91).

Once states are coacting, however, Waltz is quite clear that the structure of the international system does have an important bearing on the development of the state. This is because he believes or theorizes that the ordering principle of anarchy generates a competitive environment. States, it is argued, have no alternative but to compete with each other. This is because the anarchic international arena requires them to formulate self-help solutions to problems. In the absence of a central institution to police cooperative solutions, states have no alternative but to assume that they are operating in an antagonistic or at any rate a competitive system. Their own actions then help to reproduce the system's competitive character.

Waltz therefore acknowledges this competitive logic of anarchy and develops his theory of the state on the assumption that the international system is a "competitive realm" (1979:127). He goes on to argue, as a consequence, that however different states may be when they come into existence, there will be a tendency for certain basic features to converge once they begin to coact in an international system. The reason used to justify this position is very straightforward. In a competitive system, states have no alternative but to try to match any progressive development that occurs in another state. A state that fails to match such developments in other states will inexorably find itself falling behind and becoming vulnerable. As a consequence, any advance in one state will be quickly copied in the others. Waltz observes how states "imitate each other" so that it becomes possible to predict that "states will display characteristics common to competitors"(1979,128). For example, "the weapons of major contenders and even their strategies, begin to look much the same all over the world" (1979:127). The argument need not be restricted to the military dimension. It operates equally well in the economic domain. Failure to adopt successful commercial practices will give rise to uncompetitive and unsuccessful states.

Waltz's theory of the state, therefore represents the internal dimension of his balance of power theory and thus plays an integral part in the process that maintains the balance of power and, in turn, sustains the anarchic system. Waltz admits, however, that there are limits to his theory of the state. He acknowledges that the theory does not lead one to expect that "emulation will proceed to the point where competitors become identical" nor does it give any clear indication about what will be imitated or "how quickly and closely the imitations will occur"(1979:124). The external dimension of the balance of power--alliance formation--compensates for the deficiencies of the internal dimension.

It is apparent from this discussion that Wendt is incorrect when he argues that Waltz's approach is ontologically reductionist because he has no theory to account for the emergence or generation of the state. As Waltz himself admits, the theory is individualist in origin, but this is not from his point of view a methodological weakness. It is a necessary consequence of his theory of the state. States were ontologically prior to the international system. Once states start to coact, however, Waltz's theory presupposes that there will be a growing homogenization because they exist within an anarchic system.

Because of this homogenization of the state Waltz believes it unnecessary to distinguish between empires, city states or nation states. He assumes the logic associated with strategic interaction will apply to any kind of political unit provided the system is made up of like units, that is, functionally undifferentiated units that emerge as a product of the internal dimension of the balance of power. But Waltz accepts that there will still be differentiation of power between the units in the system. This represents an important feature of the international system. As Waltz sees it, the move from a multipolar to a bipolar system, for example, represents a major change in the system, although not enough to constitute a system transformation. This is because the basic logic of the system remains the same no matter how many units there are in the system or how power is distributed between them. His theory presupposes first that there will no functional differentiation between states because of the internal dimension of the balance of power, and second that power differentiation between the functionally undifferentiated units will not affect their capacity to reproduce themselves because of the external dimension of the balance of power. Waltz's theory of the balance of power thus conceals a theory of the state in which states are differentiated in terms of power but not function.

Waltz's theory of international politics does embrace an embryonic theory of the state, but the theory cannot be incorporated into Structural Realism without considerable development. As it stands, Waltz's theory is unable to account for why functional differentiation can become a feature of the international system. We shall now reconstruct Waltz's theory of the state to account for divergent patterns of adaptation, which generates functional differentiation between states.

 

Theory of the State and the Double Security Dilemma

While in Section I we opened the way to a consideration of type 4 anarchic systems where political units are functionally differentiated, we did not explore the nature of this functional differentiation in detail. As the result of the modifications made in so far in this section it is also not clear whether or how the idea of functional differentiation impinges on Waltz's theory of the state and his theory of the balance of power. Although multinational corporations and colonies are obviously functionally differentiated from states, Waltz insists that they have relatively little impact on the structure of the international system. As a consequence, although Waltz accepts that many international actors "flourish" in the international system, he insists that only states, the "major" actors, play any role in defining the structure of the system (1979:93). But this still leaves open the question of whether tribes, city states, and empires, all of which Waltz treats as states, are nevertheless functionally differentiated units. Waltz fails to explore this issue because his theory of the state commits him to the conclusion that any significant functional differences between political units will disappear once they start to coact. In other words, the anarchic structure of the international system generates functional, and therefore structural, uniformity among states.

Waltz's theory predicts that the anarchic structure will over time eliminate the differences between tribes, empires and city-states, because structural pressures push them in the direction of functional uniformity. States are treated as like units; any residual structural or functional differences can then be ignored. Waltz does not deny, of course, that there will be residual functional differences. But he insists that these can be treated as unit-based characteristics, and therefore irrelevant as far as his structural theory is concerned.

One of the obvious dividends of introducing functional differentiation into Waltz's theory is that it forces us to look more closely and take more seriously the presumption that states can be functionally differentiated across time and space. Unlike Waltz, our analysis is not constrained by a homogenized state-centric model. As a consequence, our approach can accommodate, for example, Strange's idea of imperial political units that are not territorially based.

This line of argument runs counter to Waltz's circumscribed theory of the state. Waltz, of course, is unperturbed by the limitations of his theory. From his perspective, before coaction takes place, the structure of the isolated states in the incipient international system may take many different forms. He does not need to explain why these different forms emerge because he is committed to the theory that the functional variations will disappear when states begin to coact. The absence of functional differentiation between states once coaction occurs is considered, as a consequence, to be purely a product of the logic of anarchy that prevails in the international system. Waltz's assumptions about the impact of anarchy and thus the balance of power on the structure of the state allows him to circumvent the need to develop a more sophisticated theory of the state. This escape route is blocked off in our reformulation, which makes explicit provision for functional differentiation between states. As a consequence, it is necessary to develop a more extensive theory--one that needs to take into account that the agents of the state are constrained not only by the international structure associated with the balance of power, but also by the domestic political structure.

The agents of the state, like Janus, are required to look in two directions simultaneously. They are confronted by two sets of structures: one internal and the other external. Waltz is able to override this argument by assuming that the agents of the state are infinitely malleable and operate under no domestic structural constraints. Here this assumption is relaxed and it is accepted, as a consequence, that the two sets of structures generate a double security dilemma. Waltz argues forcefully that the external security dilemma requires the agents of the state to take actions that reproduce the state. But the internal structures of the state also precipitate a security dilemma that constrains the decisionmakers to reproduce the state in a particular fashion. Agents of the state are aware that the fabric of the state can be destroyed not only by external forces, but also by internal forces. To remain in power, they have to respond to the constraints imposed by both external and internal structures. Once it is accepted that agents of the state are constrained by two sets of structures, it becomes possible to understand why the anarchic structure of the international system does not always generate homogeneous units. Structural Realism thus requires us to look more closely at the reasons why domestic structures sometimes prevail over international structures, with the result that, despite coaction, homogenization does not occur and radically different types of political units may coexist within the same system. This line of argument can be illustrated by looking at the very different modes of reproduction adopted by the Greek city states, the Diadochi empires, and the Roman state--all of which eventually acted within the same subsystem. Once these empirical examples have been explored, it is possible to return to the theoretical implications.

 

Reproducing the Greek City States

The Greek city states were highly successful political communities and continue to elicit respect and admiration. During the course of their history, these small political units managed to survive in an international system that included massive and sometimes hostile empires. Although the internal political structure of the city states changed over time, attempts to transform the states into a single empire, emulating the powerful neighbors of the Greeks, were not successful. Yet despite the ability of the city states to preserve their independence for several centuries, they failed to retain their autonomy when confronted by the power of Rome. As the Roman Empire spread eastwards, it successfully absorbed the independent Greek states as well as the Diadochi empires with which the Greeks had interacted. The capacity of the Greek city states and the Diadochi Empires to interact and the reasons for their demise will be examined later. The aim here is to reveal how the internal structure of the Greek city states as well as the external structure of the international system constrained the agents of the state in their task of reproducing the city state.

By the sixth century B.C. there were 1,500 Greek cities, most of which had begun life as commercial centers, so that none was more than 25 miles from the sea (Anderson 1974:29). The cities, however, were established not by traders, but by tribal aristocracies and there was intense competition and rivalry among them. From an early stage in Hellenic history, the city states established colonies outside of Greece in Italy and Asia minor. But after the colonialization process came to an end in the late sixth century, Anderson notes that the typical path of expansion was "military conquest and tribute"(1974:37). As Waltz predicts, as the cities began to coact, the military (i.e., the strategic sector), quickly became an essential element of the capacity of states to reproduce themselves.

When the city states were first established, the aristocracy possessed a monopoly on the right to fight and conflicts between states took place on the basis of individual combat. In a highly competitive system, however, it was impossible to maintain this restricted mode of fighting for any length of time. Although it is not known how or when the change came about, the Greek city states began to acknowledge the advantages of fighting in organized groups. These developments made it necessary to extend the number of citizens who could fight for the state. The resulting developments precipitated a military and political revolution in Greece, because as Aristotle argued "the class that does the fighting wields supreme political power" (cited in Cartledge 1977:24). By the seventh century, conflicts between the Greek city states were fought by heavily armored hoplites--infantrymen who carried among other things a round shield called a hoplon--who were organized to fight in a phalanx formation. Although it has been argued that alternative strategies could have developed (Cartledge 1977), when one of the city states developed this strategy, it would have been suicide for the others not to follow (Halladay 1982). Cartledge argues that by the middle of the seventh century nearly all the major Greek city states had adopted phalanx tactics, although he suggests that the Asiatic Greeks and the Western colonists did not follow suit until a little later (1977:21).

From the perspective of the ruling aristocrats, this new mode of fighting had two advantages. In the first place, the expensive equipment required by the hoplite could be afforded, of course, only by wealthy citizens. As Cartledge notes, "the devolution of military responsibility did not obviously imperil the aristocratic structure of society" (1977:24). Functional differentiation within the state was preserved intact. In the second place, it was an extremely efficient mode of warfare. Adcock asserts that it is "hard to conceive of a method of warfare that, in peace, made a more limited call on the time and effort of most citizens of most communities"(1957:4).

Over time, however, the emergence of the hoplite mode of fighting inevitably changed the social and political composition of the Greek city states. The hoplite troops were drawn from the farmers and the craftsmen class of the cities. In fifth-century Athens, for example, about a third of the citizens (ten to twelve thousand) were hoplites; the rest, the thete class, were poor peasants, unable to afford the cost of equipping themselves for heavy duty. As Anderson has noted, the precondition for movement away from the aristocratic mode of government was the "self-armed citizen infantry"(1974:33). This development first became apparent in Sparta where eight or nine thousand citizens formed a professional army. They were both economically self-sufficient, relying on the helots to relieve them of any direct involvement in economic production, and politically enfranchised. Sparta as a consequence, was the most egalitarian of the aristocratic city states. The hoplites, moreover, proved their worth when the city states allied in 478 in order to resist and eventually eliminate the threat posed to Greece by the Persian Empire.

The Persian menace proved to be particularly acute for Athens. Doyle notes that "As a commercial society, Athens depended for its prosperity and for its survival (its food supplies) on keeping the sea lanes of the Aegean open and free from an equal naval power" (1986:62). Athenian agriculture was highly specialized. Grain was imported from the Black Sea granaries, while Athenian prosperity was achieved by the export of high value-added goods derived from the olive and the vine and the manufacture of pottery and weapons (Doyle 1986:621). Athens as a consequence became the main focal point for trade in the Aegean. Athens also had access to a major source of wealth in the form of the silver mines in Attica. Silver financed the Athenian navy, which was to defeat the Persian navy at Salamis. The creation of the navy gave the Athenians a very distinctive role in the alliance of the Greek city states and also had a major impact on functional differentiation within Athens itself. The sailors were recruited from the poorer class of thetes who were paid a wage and were required to serve for eight months of the year. They almost equaled the number of hoplites and eventually the thetes were also enfranchised--turning Athens into a democracy (Anderson 1974:401). The agents of the state recognized, in other words, that if the state was to be successfully reproduced, the distribution of power within it had to be extended.

The evolution of the Greek city state demonstrates that the agents of the state are constrained by both internal and external forces. The example therefore does give qualified support to Waltz's theory of the state while providing an inkling as to how political units become and remain differentiated. In the first place, in line with Waltz, the case of Athens demonstrates that once the idea of an organized fighting force crystallized, each state found its only alternative was to make appropriate adjustments to the internal structure of the state. But it would appear that the aristocrats believed that these adjustments could be safely made without threatening their position of power. Although the necessary historical detail is not available to reveal the precise process, it is apparent that, over time, this initial development had the long-term consequence of eroding the power of the aristocrats. But in most states they managed to hang on to the vestiges of power. As a consequence, there was a continuous struggle for power within the state. So, despite the need to accommodate to external structural constraints, the internal structure of power continued to exert a powerful influence on the nature of the state.

The case also demonstrates that because of the Athenian dependence on trade, functional differentiation developed between Athens and Sparta. Athens had no alternative but to develop as a naval power, although this required substantial internal structural readjustment, turning Athens into an unequivocally democratic state. The oligarchs in Sparta failed to follow suit because they had no wish to divest themselves of power. Once this differentiation occurred, it had the effect of rendering all the other states vulnerable to intervention because of the existence of their chronic power struggles. The initial differentiation between Athens and Sparta, moreover, generated further differentiation, because as a democratic, naval state, Athens was ideally suited to develop an informal empire. Ironically, the capacity to consolidate the empire was constrained not by external structural forces, but by internal ones. As we shall see later, the democratic distribution of power within the state, which facilitated the creation of the empire, had the further effect of making its consolidation impossible. The agents of the Greek polis were constrained by its very structure from developing an effective empire. It becomes of interest, therefore, to turn to the reproduction of the Diadochi and Roman empires to see how easily this structural constraint could be circumvented.

 

Reproducing the Diadochi Empires

Diadochi is Greek for successor and the Diadochi empires emerged after the failed attempts to hold together Alexander's massive empire. Alexander was the son of Philip, the King of Macedonia (359-336 B.C.). McNeill portrays Macedonia as a "border state"--a tribal monarchy--which benefited from its position on the "margins" of the Greek civilization (1979:149). So the noblemen from Macedonia were educated in the Greek manner (Alexander, for example, was tutored by Aristotle) with the result that feuding among the noblemen gave way to loyalty to the king. Citizenship was extended to inhabitants of the newly conquered regions who were then required to give their loyalty to the King of Macedonia. According to Anderson, moreover, because Macedonia was "morphologically much more primitive" than the Greek city states, it was able to "overleap" their structural limits (1974:45). The Macedonians, unlike the Greeks, were not constrained by beliefs about the democratic polis. In particular, there were no constraints to limit the size of the state or prevent the formation of a state bureaucracy, which was essential for the administration of a large geographical area.

At the same time, the Macedonians were also able to graft their cavalry skills onto the Greek military innovations. Macedonian peasants began to be trained as hoplites; but they were also armed with a long lance and were given lighter armor to make them more mobile. The Macedonian phalanx was then flanked by the cavalry. As a result of these changes, all linked to the process of Hellenization, Macedonia--a Hellenistic rather than a Hellenic state--was soon in a position to pose a considerable threat to its neighbors. Under Philip, the Macedonians proved capable of overwhelming local barbarians and then the Greeks themselves. Under Alexander, Macedonia subdued the Persian Empire and his rule extended as far as India.

Alexander died in 323 and conflict among his successors persisted until 311, when there was an agreement about how to divide up the empire. It was only after thirty more years of warfare that three stable great powers were established in the Hellenistic world. They were ruled by descendants of a Macedonian general: the Ptolemies in Egypt, the Seleucids in Asia, and the Antigonids in Macedonia. These Hellenistic kingdoms were hybrid political creations that combined in a complex fashion Hellenic and Oriental features. In the first place, there was a proliferation of cities throughout the Near East, making it the most densely urbanized area of the ancient world. But although modeled on the Greek cities, the new Hellenistic cities were much larger (Anderson 1974:47) and developed in a very different political setting. In the Hellenic world, strong walls were impregnable and the city could survive as a defensible political unit. With the development of the torsion catapult and the siege tower, however, Philip and his successors established a means of overpowering the city. Walls were strengthened, but this merely led to improvements in the catapults, so that the security of the city was always in jeopardy. As Price notes this "crucial fact underlies the dominance of kings over cities"(Price 1986:330).

The scale of conflict increased dramatically as a consequence. In Greece, conflict was restricted to fights over disputed territory between city states. But in the Hellenistic kingdoms there were vast tracts of land at stake and the kings could muster 60,000 to 80,000 men against each other. The size of armies would never be larger until the end of the eighteenth century. There was a concomitant increase in destruction and whole cities could be razed to the ground. Nevertheless, the Hellenistic kings established only a hegemonial status over the cities, with their power often being exercised informally or indirectly. The kings used a range of diplomatic devices to ensure "harmony between and within cities without involving invidious direct interventions"(Price 1986:332). Yet the threat of force always persisted because whereas in Greece legitimacy came from tradition, the Hellenistic kings derived their legitimacy from military prestige (Price 1986:326).

The Hellenic influence, however, was pervasive. It was reflected in the monetary standard, which was generalized throughout the Hellenistic Kingdoms, and the banking system. These developments encouraged commerce and trade (Anderson 1974, 47-48). But slave labor, although of great importance in Greece, failed to develop in the Hellenistic East because of the strength of established social and political traditions. These traditions also encouraged a veneration of the Hellenistic kings unknown in Greece (Anderson 1974:48-50). As we shall see, these factors were to prove very significant when an attempt is made to account for the survival of the Eastern half of the Roman Empire.

What this case study demonstrates, therefore, is that the process of emulation does not necessarily lead to the increasing homogenization of political units, as Waltz's theory predicts. On the contrary, when the Macedonians copied Hellenic techniques and grafted them onto the structure of their state, a new and very much more powerful political entity emerged, which looked nothing like the Greek city states. The Greeks, moreover, with their fixed belief in the virtues of the polis, were structurally constrained from emulating the Macedonians. It is important to note, moreover, that the Macedonians did not emulate the Greek city states out of strategic necessity. The Macedonian state underwent its transformation as the result of cultural transfusion rather than as a consequence of the pressure of operating in an anarchic arena. Waltz's theory, therefore, performs much more effectively in the context of the Greek city states than it does in the context of the Macedonian Empire. We shall examine the reasons for this, but first let us look at the way the Roman Empire managed to sustain itself for so long, when compared to the almost instant "death" of Alexander's empire.

 

Reproducing the Roman Empire

Like the Greek city states, Rome emerged in an intensely competitive international environment. But its horizons were initially local. What is perhaps most remarkable about Rome is that its agents eventually should have come to think in such universal terms, and that they could have contemplated ruling the entire orbis terrarum (Brunt 1978:168). While they failed to achieve this goal they did successfully unify into a larger subsystem what had previously been a number of independent international subsystems. To be able to do this, the agents of Rome had to overcome or circumvent the structural restraints experienced by the agents of the Greek city states and the Hellenistic empires.

The methodology of structuration suggests that to understand how the expansion of the Roman state came about, it is necessary to look more closely at the process of reproduction implemented by the agents of the state. Although very little is known about the early history of Rome, it is clear that from the very start its evolution must be considered in the context of a broader system. Rome began as a village on the Tiber at a time when the peninsula of Italy was fragmented by competing tribes. To the North of Rome were the Etruscans, who operated within a loose confederation of twelve cities. They represented the dominant power in central Italy in the middle of the first millennium B.C. From an early stage, the Romans must have recognized that they had no alternative but to develop a military capacity in order to maintain their independence against armed neighbors and thereby reproduce their city state. During the early phase of Rome's development, the main rival was the Etruscan city of Veii, which lay ten miles north of Rome. Although Veii and Rome were both were much bigger than other cities in this region of Italy, Keppie insists that the wars between Rome and her neighbors "were little more than scuffles between armed raiding bands of a few hundred men at the most"(1984:14).

As Fox has observed, in primitive combat superior numbers are likely to prevail, so there was always an "inherent logic that led inexorably to larger and larger political units" (1971:21). The competition between the Romans and Etruscans, according to this logic, encouraged them to extend their borders. But as Mann has argued, in line with Waltz, the competition also meant that both the Romans and the Etruscans were constantly looking for new strategic ways of achieving military superiority. In the seventh century, for example, the Etruscans began to copy the military tactics associated with the Greek hoplites. However, once the Romans had copied the Etruscan tactics, to ensure the reproduction of the state, then the independence of their neighbors to the North was simultaneously threatened. Moreover, while Rome was consolidating its strength, the Etruscans' power was being steadily eroded as new actors entered the system. The Etruscans had to face the Celtic tribes, in particular the Gauls, who migrated across the Alps during the fifth century. Keppie notes that with hindsight it can be seen that the Etruscans unintentionally provided "a buffer" for Rome and in doing so whittled away much of their own strength (1984:18). Ironically, therefore, when Rome eventually captured Veii in 396 they were almost immediately confronted by the Gauls, who defeated the Roman forces and looted Rome in 390 before being repelled. The experience was salutary and the Roman army underwent major changes over the next fifty years as they endeavored to adapt the tactics of the phalanx to the more mobile tactics used by the Gauls (Keppie 1984:19).

Because of the double security dilemma, the agents of the state seek to reproduce the state in such a way as to preserve both the internal and the external power structures. As already observed, the Greek aristocracy introduced the hoplites only to find, in the long term, that this method of fighting eroded their own power base. By contrast, the aristocratic agents of Rome managed to develop an effective fighting force without democratizing the state. Anderson (1974:57) notes, for example, how "a hereditary nobility kept unbroken power" despite substantial constitutional changes for two centuries after the emergence of the Roman Republic. It has been argued that continuous expansion proved to be a vital feature in the reproduction of the aristocratic state (Harris 1979). Closely associated with this development was the decision to promote a conception of citizenship. Such a conception did not initially exist. During the early history of Rome, for example, there was "no rigid conception of citizenship to tie a man to a community of his birth" (Crawford 1986:393). As Rome expanded its area of control the importance of possessing citizenship increased dramatically. Land acquired in the course of expansion was given to peasants who in turn were used as soldiers in the process of acquiring more territory. New communities absorbed by the Romans were granted citizenship, given land, and they too became liable to serve in the military operations that saw the further expansion of Rome's area of control. Land, however, was never equally shared. There was always a segment of the community more generously treated, providing the basis for a social elite and a governing class.

Rome did not rely simply on military superiority and territorial conquest to extend its influence. It also developed an elaborate system of alliances designed not only to aid expansion, but also to underpin the aristocratic mode of government in the territories on the border of Rome. The alliances, therefore, were very different from those postulated in Waltz's balance of power theory. The Roman alliances involved "horizontal penetration." When Rome offered to form an alliance with a neighbor, it extended privileges to the upper classes in the communities they allied with, and there were frequent intermarriages. Rome invariably supported this class whenever there were internal or external threats. These allies were therefore soon tied by very tight bonds to Rome. These bonds ensured, moreover, that Rome's major demand on its allies--the supply of troops--was invariably fulfilled. Allies tolerated this exercise of power because these troops did not return empty handed. Although not as generously treated as the citizens of Rome, they did share in the fruits of victory, at the expense of the newly conquered territory. Crawford (1986:399) argues that most ancient empires demanded tribute from the areas they subjugated. The demand for troops represents the distinctive and very effective feature of Roman imperialism permitting not only the reproduction but also the expansion of the system. This form of client relationship was to play a crucial role in the future running of the Roman Empire.

As the Roman Republic extended across the peninsula, the aristocracy acquired huge tracts of land from defeated tribes. These tribes also initially provided the slaves who worked the land. As the empire extended overseas, the slaves were brought from further afield. Anderson (1974:62) argues that the development of a slave economy released the necessary personnel to fight further wars and precipitated a "gigantic social upheaval" in Italy. According to estimates made by Brunt (1971:121-25), in 225 B.C. there were about 4,400,000 free persons in Italy to 600,000 slaves. By 43 B.C. he suggests that there were 4,500,000 free persons and 3,000,000 slaves. Anderson also observes that although the Greeks had slaves, they were used only on small plots of land. The Romans, by contrast, were the first people to introduce large-scale slave latifundia. The consequences of this development proved to be very important, because when the Roman Empire moved West "for the first time, classical Antiquity was confronted with great interior land-masses, devoid of previous urban civilization. It was the Roman city-state, that had developed the rural slave-latifundium, that proved capable of mastering them. . . The successful organization of large-scale agrarian production by slave labor was the precondition of the permanent conquest and colonization of the great Western and Northern hinterlands" (Anderson 1974:63).

The capacity of the Roman state to expand, in contrast to the Greek city state, can therefore be partially explained by reference to a structurationist theory of the state. According to this theory, agents of the state are constrained not only by external structures of the international system, but also by internal structures of the state itself. The agents of the Athenian state were encouraged to engage in imperialism because of the need to combat the oriental threat posed by the Persian Empire, which possessed a quite different structure from that of the Greek city states. But because of internal constraints the Greek city states were never able to consolidate their imperial possessions. By contrast, although external threats also encouraged the Roman city state to engage in imperialistic ventures, because of its distinctive mode of reproduction it was able to expand successfully. The expansion, of course, was not world wide and this raises questions about the factors that constrained the continuous development of Rome.

What we have demonstrated in this chapter is that because of the constraints generated by the internal structure of the state, it does not always follow that the units in a system eliminate any signs of functional differentiation, which can persist because of the external dimension of the balance of power. Although the Diadochi empires were functionally differentiated from the Greek city states, they nevertheless pursued policies under balance of power constraints which secured the independence of the Greek city states. Whereas Waltz presupposes that the balance of power will generate like units in the first instance, and generate alliances to deal with the residual problem of power differentiation only in the second instance, the case studies looked at here indicate that this logic needs to be reversed. Anarchy does not necessarily eliminate functional differentiation. Instead, the external workings of the balance of power can ensure that states continue to be reproduced in very different ways. The logic developed here suggests that functional differentiation will be eliminated only if it cannot be accommodated by the workings of the external balance of power. Reformulating Neorealism in this way makes it possible to account for both the differentiation and the nondifferentiation of units in an anarchic arena.