Escape From Rome

by Walter Scheidel

  In the summary of my findings at the end of chapter 9, I argue that the most important outcomes—enduring fragmentation in post-Roman Europe and serial empire formation elsewhere—were substantially overdetermined. Notwithstanding Rome’s early success, Europe was always less likely to be ruled by very large empires than other parts of the Old World, most notably East Asia. This holds true regardless of whether we privilege geographic, ecological, or cultural factors.

  STATE FORMATION IN EUROPE AND EAST ASIA: ANCIENT PARALLELS, CONVERGENCES, AND DIFFERENCES

  Parallels

  The eastern and western ends of Eurasia had not always developed in such strikingly different ways. For more than a millennium, state formation proceeded in parallel or even convergent ways. In the early first millennium BCE, political power in Europe and the Mediterranean on the one hand and in East Asia on the other was spatially highly fragmented. Europe was a vast expanse of small and stateless groups lined by Mediterranean city-states. In the Levant, which had long been home to larger and more complex polities, widespread collapse at the end of the Bronze Age had shattered empires from Anatolia to Mesopotamia and Egypt, leaving mostly smaller kingdoms in its wake. In what are now the central-eastern reaches of the People’s Republic of China, the Western Zhou regime gradually lost its grip on its numerous local vassals. The erosion of central authority created a web of more than a hundred smaller polities that were interspersed with ethnically different but similarly modestly sized groups.

  Over time, diversity very gradually diminished as larger political entities emerged and absorbed competitors. This process commenced a little sooner in the West but in both cases yielded comparable results. In the Middle East, the Neo-Assyrian empire expanded from the ninth and especially the eighth centuries BCE onward, as did Kushite and Saite Egypt from the eighth and seventh centuries. Stretching from the Balkans to the Indus valley, the vast Iranian-centered empire of the Achaemenids that arose in the second half of the sixth century BCE eclipsed everything that had come before. After Alexander the Great’s conquest caused it to unravel in the late fourth century BCE, several substantial imperial states took its place. Meanwhile Rome, Carthage, and Syracuse established growing domains in the western Mediterranean. In China, by the fifth century BCE seven major warring states had swallowed the many small polities that had emerged from the wreckage of the traditional Zhou order.3

  In both cases, these processes of consolidation gave rise to increasingly powerful polities at the western margins of each ecumenical zone, Rome and its Italian alliance system in the West and the kingdom of Qin in the East. In building up their military capabilities, they benefited from their respective “marcher” positions that enabled them to expand without being heavily constrained by similarly developed neighbors. Thus, as noted in chapter 3, Rome was protected by Italy’s location outside Western Eurasia’s central “political-military network,” while Qin was shielded by mountains that separated it from the more economically advanced states in the Central Plain and ensured its control over the “land within the passes” (Guanzhong), a peripheral region centered on the Wei River valley.

  These positional advantages facilitated initial expansion away from major established powers: into the western Mediterranean for Rome (in the third century BCE) and southward into Sichuan for Qin (in the late fourth century BCE). Low protection costs boosted the value of these gains. These added resources as well as a particularly strong focus on martial capabilities allowed Rome and Qin to engage and overpower great powers to the east. In both cases, de facto hegemony morphed into direct rule, albeit much more rapidly in China (in the 220s BCE) than in the Mediterranean basin, where Rome drew out this process across the last two centuries BCE and in a few places even beyond that.4

  Quasi-monopolistic authority on a subcontinental scale affected the political structure of both conquest empires. This included a shift from military mass mobilization to more professional and socially as well as geographically peripheral forces. The mature Roman empire sustained a standing army deployed in and increasingly replenished from frontier regions and augmented by provincial auxiliaries, whereas the Han empire (the much longer-lived successor to the Qin regime) gradually came to rely on a mixture of convicts, colonists, mercenaries, and “barbarian” contingents. In both cases, territorial expansion slowed and then ceased altogether. At the same time, rent-seeking local elites increasingly constrained central authority and state control (figure 7.1).5

  FIGURE 7.1   Empires of the Old World, c. 200 CE.

  The two empires even failed in a similar fashion. The third century CE witnessed temporary splits into three subimperial states: however, Rome experienced a more robust recovery (from the 270s to the 390s CE) than Jin China (where unity was chiefly limited to a quarter century after 265 CE). Yet in the end, these differences in detail mattered little. In both cases, the more exposed halves of each empire were taken over by outside (or rather preinfiltrated) conquest regimes: groups from the steppe frontier region in northern China (in the fourth century CE) and Germanic confederations in the western Roman empire (mostly in the fifth century CE). Traditional rump states survived in the more sheltered southern half of China (from 317 to 589 CE) and in the eastern Mediterranean (from 395 CE), which unlike southern China was less well protected by geography but up to the early sixth century benefited from the temporary weakness of Constantinople’s Iranian competitors.

  In both the former western Roman Europe and in northern China, conquerors and local elites assimilated into new ruling classes, and transcendent religions that claimed autonomy from the state—Christianity and Buddhism—advanced. Determined attempts to reestablish empire on the previous scale were made in the sixth century CE. This was the final parallel experience: from then on, state formation lastingly diverged as imperial reconstruction succeeded in late sixth-century China (and would do so repeatedly on later occasions) but not (and in fact never again) in Europe.

  Differences and Convergences

  These parallel trends in state formation up to the middle of the first millennium CE masked underlying differences. Chinese polities of the Spring and Autumn and Warring States periods operated in a framework of elite interaction based on the at least loosely unifying traditions of Shang and Zhou (around the sixteenth through the eighth centuries BCE). Rome, by contrast, in most of its European territories created political unity where it had never existed before. It was only in the central and eastern Mediterranean that it could draw on antecedent networks and governmental infrastructure. This slowed but did not lastingly obstruct Roman expansion to the west and north.6

  In terms of political and administrative structures, a sizable gap separated the monarchical systems and (by ancient standards) strongly centralized regimes of the Warring States period from the oligarchic system of the Roman Republic and the decentralized mode of Rome’s governance of its provinces. This gap was widest during the last few centuries of the first millennium BCE but narrowed over time, even if it never completely disappeared. This reflects convergent trends over the long run.

  During their expansion phases, both Rome and Qin (alongside the other Warring States) relied on mass mobilization of peasant infantry, and created conscription districts and conducted censuses for this purpose. From the fifth through the third centuries BCE, the Warring States were locked into prolonged inconclusive warfare whose competitive pressures prompted intrusive centralizing and homogenizing reforms. The state of Qin may have gone the farthest in this regard: its rulers strove to break aristocratic power, subjected the entire population to a ranking system, divided it into small groups for mutual surveillance and collective liability, and instituted formalized rewards for military prowess. Taxes—in money, kind, and military as well as civilian labor—were relatively high. Qin’s overarching ambition, to the extent that it could be realized, was the creation of a centralized territorial state that was fully controlled by employees of the ruler and left no political space for rival groups such as nobles
or the wealthy.7

  As we saw in chapter 2, Republican Rome achieved intense mass mobilization with the help of a much slimmer governmental apparatus and overt streamlining of local arrangements. Its state was highly spatially centralized in that the entire top tier of its leadership was concentrated in the city of Rome. Autocracy had long been blocked by an assertive anti-regal aristocratic oligarchy. In the absence of a monarch, competition within the aristocracy was constrained by tightly regimented popular political participation. This system was structurally opposed to formal bureaucracy and onerous taxes on insiders. A modest number of aristocratic houses relied on patrimonial resources to fulfill their public functions and on patronage relations and ritual performances to exercise power. Fiscal operations were largely farmed out to private contractors, and taxes were relatively low, especially in the Italian core, where military service represented the principal civic contribution for elite and commoners alike.

  Overall, the Roman domain came to be hierarchically stratified into an Italian core and a growing provincial periphery, a feature that was largely missing from imperial China. Urban autonomy and effective self-governance were preserved, and no salaried state agents were imposed on or created in the numerous constituent communities. The main reason that this more loosely structured system managed to prevail lies in the fact that unlike the Warring States, the maturing Roman Republic engaged mostly in asymmetric competition with differently organized challengers: somewhat lopsided in Rome’s favor, this competition did not spur invasive restructuring. It was only early on, as Rome struggled for supremacy over the Italian peninsula, that conflict with more opponents that were more similar encouraged—rather limited—organizational adjustments.8

  Yet once the Roman and Qin-Han empires had grown into behemoths that between them claimed close to two-thirds of the world’s population at the time, their institutions changed in ways that made them look more alike. Undermined by logistical challenges and growing stratification and erosion of consensus among the elite, oligarchy failed in Rome and was replaced by monarchy: although different in style from the Han version—in granting greater autonomy to elite groups and leaning more heavily on a bloated military sector—this was a big step toward the global historical default model of kingship. In China, the imperial center, while maintaining a numerically large bureaucracy of around 150,000 salaried state employees, gradually lost ground as local elites expanded their influence at the expense of these agents. A period of dynastic instability in the early first century CE in particular bolstered the power of magnates who supported the ruling dynasty. The concurrent abolition of conscription surrendered greater control over people and resources to wealthy landlords and patrons.

  In the Roman empire, meanwhile, a patrimonial bureaucracy that had grown only very slowly received a major boost from serious dislocations in the third century CE that encouraged more determined centralizing reforms. As a result, the mature imperial state of the fourth century CE resembled that of the Han more closely than before, featuring a sizable bureaucracy of 20,000-plus men, an overhaul of the fragmented tax system, the separation of military and civilian commands, and halfhearted encroachment on urban autonomy. Ministries, powerful court eunuchs, and child emperors, long common in Han China, likewise appeared on the scene.9

  It is true that this convergence went only so far: the Han administrative apparatus remained larger and Roman urban autonomy stronger. Reduced to essentials, these abiding differences reflect the resilience of the Greco-Roman city-state tradition and the more intense coercion-extraction cycle in China that had shored up centralized ambition and authority at the expense of local and especially urban self-governance.

  Yet this contrast, while noteworthy, must not be overrated. Roman city officials and Han provincial agents hailed from comparable socioeconomic backgrounds, and the formally bureaucratic features of Han administration barely masked rampant patronage and simony, which were similarly common in the Roman empire. Roman self-governing plutocrats and Han salaried state agents were equally adept at siphoning off resources claimed by the center, and landlords shielded their own assets and those of their clients, slowly but surely eroding the foundations of the imperial edifice. Whatever differences remained in terms of the relative weight of the military and civilian spheres, of center and periphery, and of bureaucracy and local self-rule, they were very much a matter of degree. Driven by the internal logic of traditional empire, the two systems had become about as similar as their discrepant starting conditions permitted them to be.10

  MID-FIRST-MILLENNIUM EUROPE: THE FIRST GREAT DIVERGENCE

  This gradual if imperfect convergence makes the following divergence in state formation seem even more striking. That process spanned roughly the second half of the first millennium of the Common Era. By about 500 CE, the Roman empire had split into an eastern half, ruled from Constantinople, and five major kingdoms under Germanic successor regimes in the west, a number that fell to only two within the next few decades. In China, the collapse of the Jin empire ushered in the period of the “Sixteen Kingdoms,” a series of often ephemeral polities. Yet by the early fifth century, just two states controlled the northern and southern halves of China. This number fluctuated between two and three until the late sixth century when the north conquered the south. With only a brief interruption, China was then at least formally unified until around 900. The Song restoration ended another period of fragmentation in the first half of the tenth century.

  By then, eleven major states, alongside a number of smaller entities, occupied the area once held by the Roman empire. Moreover, the larger polities that were located in the western and southern European parts of the former empire suffered from intense internal fragmentation, most notably the Frankish and German kingdoms and, before long, Islamic Spain. Geographically peripheral England was perhaps the single reasonably coherent polity in this zone, albeit modest in size and beset by aggressive Scandinavian neighbors.

  The growing divergence between eastern and western Eurasia manifested in three different ways. First, during the second half of the first millennium CE, China had been either politically unified or split two or three ways for more than four-fifths of the time. (These latter years include about a century and a half of effective if informal interprovincial fragmentation under the later Tang dynasty.) The area once ruled by Rome, in contrast, was never unified and just barely met the criteria for a (short-lived) three-way split: its shift toward greater diversity was persistent, from only three major states in the mid-sixth century to close to a dozen by the year 1000.

  Second, as we will see below, eleventh-century China was ruled by the most formidable empire in the world: it contained at least 100 million people and employed over a million soldiers. Nothing even remotely like that existed anywhere in what had been the Roman empire, and most decidedly not in its former European provinces.

  Third, and most important, both the overall trajectories from 500 to 1000 and conditions around 1000 set the tone for the following millennium. Leaving aside short periods of regime change and allowing for intermittent spells of internal decentralization, China never again divided into more than two or three major states. Europe moved in the opposite direction, experiencing intense fragmentation in the High and Late Middle Ages (figure 7.2).

  Thus, the number of independent polities in Latin Europe grew from late antiquity onward, at a moderate pace at first but faster after the fracturing of the Carolingian empire in the ninth century, which had imposed a short-lived semblance—or rather mirage—of Roman-scale unity on substantial parts of Western Europe. By the end of the Middle Ages, Latin Europe was contested among at least a hundred effectively independent polities, and the number of largely autonomous ones was even much greater than that. Consolidation only slowly gathered steam in the early modern period but remained incomplete. In 1900, what had once been the Roman empire was—in a conservative count—divided among twenty-three states, all but two of them located on European soil. The current tal
ly is at least forty, depending on how we define Roman rule and what qualifies as a state.11

  FIGURE 7.2   The Han, Tang, Northern Song, Yuan, Ming, and Qing empires.

  The First Great Divergence brought about an enduring contrast between serial reconstitution of empire and the resultant absence of a stable state system in China and the lack of any comparable scaling-up and the resultant formation of a highly resilient ecology of political polycentrism in Europe. This does not mean that China was always unified: the figures in chapter 1 graph the extent to which this was not the case. Metrics vary, depending on how we define “China” or “unity.” By one count, “core China”—defined as the territory held by the Qin state at its peak in 214 BCE—was unified under one ruler for 947 of the past 2,231 years, or 42 percent of the time (figure 7.2).12

  Imperial persistence in China was thus relative—that is, compared to conditions in other parts of the world—not absolute. A different calculation shows that East Asia was characterized by a unipolar or hegemonic political system for 68 percent of the years between 220 BCE and 1875. This pattern presents a stark contrast to the prevalence of a balanced system in Europe for 98 percent of the years from 1500 to 2000, or indeed at any time after the demise of the mature Roman empire.13

  It is true that imperial persistence in China was at times more formal than real: periods of effective decentralization can be observed in the late Western Han, Western Jin, mid- to late Tang, and late Ming periods. Yet even weakened empire prevented the emergence of anything like an actual multistate system, which had been present in China only in the Warring States period up to the 220s BCE.