These trends continued and deepened in the eighth century. In Francia, the aristocracy’s grip on their retinues tightened at the expense of royal command. And while Lombard rulers maintained a greater degree of control over levies and upheld the principle of a nobility of service rather than of wealth, personal followings nevertheless gained in importance. In England, lordly retinues operated alongside royal bodyguards.
Developments diverged from the ninth century onward: whereas the power of Frankish lords eclipsed that of their kings and eventually reduced them to a somewhat marginal position, England experienced a trend reversal: a more compact territory, a weaker nobility, and the strong Danish threat enabled King Alfred and his successors to establish a permanent army. Sustained by men who rotated between their homes and the field army, this created an instrument not just for defense but also for domestic coercion.31
Yet for the time being, this remained the exception that proved the general rule of aristocratic autonomy and declining state capacity. Moreover, armies were small regardless of how they were raised. Alfred’s field army numbered in the low thousands. Entire kingdoms could change hands in engagements of barely more than 10,000 men. Visigothic Spain and Anglo-Saxon England are the best examples. Logistical constraints—and the need for attendants—ruled out operations of armies of more than 5,000 soldiers. Larger forces could only be concentrated for short-term defensive measures. Fast raids, the most common format, involved hundreds rather than thousands of seasoned warriors. Most campaigns focused on plunder, and cities were so unimportant that siege warfare received little attention even though defenses were poor.32
It should go without saying that none of this was conducive to expansion and state-building on a large scale. Sustainable empire was far beyond the reach of these rickety polities. The contrast to the Roman period could hardly be starker. For much of the last two centuries BCE, forces numbering in the tens of thousands had been deployed in the Iberian peninsula to grind down local resistance. Caesar subdued Gaul with ten legions—30,000−40,000 men even if seriously understrength, and bolstered by considerable allied contingents—a display of military might the region would not witness again until the early modern period. In 43 CE, the emperor Claudius invaded Britain with some 40,000 men, four or five times as many as William the Conqueror would marshal in 1066.
THE DEATH OF EMPIRE
In chapter 5, I touched on the dynamics that enfeebled and splintered the Carolingian domain. The power of nobles, enhanced by a phase of lucrative expansion that culminated under Charlemagne, turned against the center once gains from conquest dried up. Dynastic feuding—fueled by the lack of primogeniture in the royal succession—not only weakened rulers but invited intervention by aristocratic factions. The resultant self-sustaining and self-reinforcing process of decentralization played out in different ways depending on local conditions, which favored rural lords in Francia and Germany and city-states and communes in Italy and Lotharingia. The overall outcome—intense fragmentation of power—was similar everywhere.33
As the aristocracy controlled a growing share of the land and the economy began to recover, lords set up their own castles, a practice emulated by lesser nobles and church leaders. Armed retinues turned into knights, whose presence simultaneously broadened the aristocracy and limited military service to a clearly defined in-group. This caused power to disperse more widely in elite circles and reduced any leader’s ability to concentrate resources for the purposes of protection and state-building. The path of least resistance available to kings—further gifts of royal land to church and nobility—proved self-defeating in the longer term: it merely decreased what was left of what the ninth-century Frankish historian Nithard called their “wherewithal to reward followers.”34
But rulers were not the only ones to lose ground. Aristocratic and church power expanded at the expense of the peasantry, which gradually lost its independence. This shift was exacerbated by the growing exclusivity of military service and the concurrent decline of assemblies—armies having served as political gatherings as well—except in England and in Italian communes. Rural dependence further debilitated rulers, who lacked autonomous means of coercion to counterbalance lordly control over the means of production.35
Following an early post-Roman hiatus, aristocratic hegemony had reached unprecedented heights. By the ninth century, hereditary landownership and office-holding by counts had become the norm in West Francia and Germany, and super-magnates of regal or quasi-regal status emerged together with new duchies and kingdoms. In the tenth century, their knightly subordinates established control over the peasantry. By the eleventh century, power had become highly localized: lesser nobles relied more on their own resources than on the support of dukes and counts, and erected large numbers of castles that chopped up territory into autonomous blocks impeding state formation. Rulers had to forge alliances with ducal houses if they hoped to exercise any degree of control.36
By 1000, aristocracies were dominant everywhere in Latin Europe. Public power structures at the supralocal level failed with varying intensity, most notably in France and to some extent in Italy. Germany and Islamic Spain also moved in that direction. England, where governmental structures had collapsed more radically and had also been rebuilt earlier, was the only outlier, yet far too small and peripheral to embark on expansion beyond the Channel.37
In Joseph Strayer’s judgment, these fragmentation processes had become so intense that—with the exception of what remained of the East Roman polity in the Balkans—“by the year 1000 it would have been difficult to find anything like a state anywhere on the continent in Europe.” As far as the political order was concerned, at that point, in Chris Wickham’s words, even the “shadow” of the “inheritance of Rome” had “faded away.”38
China
DISUNION AND PERSISTENCE
The same year, many of the lands that once been governed by the Han lay united under the rule of the Zhenzong emperor. After severe disturbances that echoed the fall of the Roman empire, East Asian state formation had followed a very different path. The drawn-out demise of the empire of the Eastern Han dynasty in the late second and early third centuries CE led to a three-way split between the states of Wei, Shu, and Wu. Imperial restoration under the Jin dynasty in the late third century proved short-lived. Internal power struggles encouraged ethnic groups from the frontier zone that had been settled on imperial territory to take matters into their own hands. In the 310s, coalitions led by Xiongnu forces sacked the capital cities of Luoyang and Chang’an and forced the Jin court to retreat south.39
Several generations of instability and political disunity followed. In the fourth and early fifth centuries—labeled as the era of the “Sixteen Kingdoms”—northern China was dominated by a series of often brittle conquest regimes. Power kept shifting between military coalitions associated with several peripheral ethnic groups: Xiongnu, Xianbei, Tuoba, and Di. Although the Later Zhao and Former Qin states managed to seize most of northern China from the 320s to the 340s and in the 370s and early 380s, respectively, both of them failed in short order.40
A general pattern is visible behind the chaotic dislocations of this period. The principal northern regimes established dual systems of rule that distinguished between conquerors of steppe or frontier descent and Han Chinese. The former were known as guoren, “compatriots,” whereas the latter were subordinated to them as taxpaying subjects, liable to provide grain and labor services. Ethnic cavalry, obliged to fight for their leaders, formed the backbone of the field armies, whereas Chinese could be drafted as needed to serve as foot soldiers or provide logistical support.41
Flight and other types of attrition had opened up land and made labor scarce. The conquest regimes consequently focused on capturing people rather than land. Forcible mass transfers of subjects to core areas of regime control were meant to secure manpower. Regime change could lead to resettlement to new centers. Rulers’ grip on population outside these core zones was generally much
weaker.
Much of the northern elite had fled to the south to join the Jin. Those who remained—local gentry clans—organized fortified settlements in defensible positions. As cities were ravaged or abandoned by competing military coalitions, these positions became focal points of local government and formed basic blocks of power. Soon numbering in the thousands, these self-defense communities could form leagues: the centralized conquest regimes were left with only two options—to confront or co-opt them—to assert preeminence.42
Northern rulers’ reliance on unstable military coalitions rendered their polities fragile: these constructs were readily dismembered and reconfigured upon defeat by rival forces. The underlying power relations were more enduring, as effective control was balanced between the conquest cores with their reservoirs of ethnic cavalry and native forced labor on the one hand and numerous small-scale but capable local communities on the other.
This environment differed from post-Roman Europe in two crucial respects. One was the presence of strong centrally managed and highly mobile cavalry forces that, even if unable to sweep everything before them, were generally capable of checking local power blocks. This in turn prevented the evolution of feudal relations: military assets remained sufficiently centralized and local power bases were insufficiently autonomous for military and subsequently also economic and political power to disperse and devolve upon small-scale units. As Ray Huang points out, “Had this trend been allowed to continue, a new form of feudalism might have taken root in China.… But mobile warfare over a large region, which favored huge bodies of fighting men, settled the course in a different fashion.” Ethnic cavalry, rooted in steppe traditions, proved vital in first containing local units and later pressuring them into cooperation.43
The second key factor was the persistence of fiscal structures to support these core military forces by taxing agricultural production and mobilizing civilian labor. Evidence for fiscal institutions is exiguous but intriguing. There can be no doubt that the collapse of the Han order drastically curtailed the ability of subsequent central authorities to count and tax their subjects. Han census totals for the second century CE averaged about 10 million households with 50 million residents. By contrast, the “Three Kingdoms” of the mid-third century counted merely 1.7 million households, while the Western Jin mustered 2.5 million after reunification. Notwithstanding potentially significant population losses, most of this decline reflects eroding state capacity rather than demographic change.44
Under the northern conquest regimes, registration quality began to recover from this nadir. In the 370s CE, the Murong state of Former Yan counted 2.5 million households with 10 million people in an area that was only a fraction of the size of that covered by the highly deficient third-century censuses. It thus achieved much better penetration, especially allowing for attrition and the sheltering and concealment of parts of the remaining population in walled villages. The fact that two centuries later, the better-organized state of Northern Qi was able to register 20 million residents in a roughly equivalent territory shows that fourth-century state power, however constrained by local obstruction, was by no means negligible and was capable of capturing a sizable share of the actual population, perhaps close to half.
This impression is consistent with accounts of mass mobilization during a campaign against the south in 383 CE: even though the reported totals of 270,000 cavalry and 600,000 infantry are hardly credible, even a reduction by an entire order of magnitude would leave forces much larger than any medieval European state could hope to muster. The first post-Roman polities in Europe to encompass 10 million people—leaving aside a brief blip under Charlemagne and Louis the Pious—were Germany and France in the High Middle Ages. But European rulers would have been ignorant of this fact and could not have hoped to impose regular taxes on most of their subjects.45
Manuscript evidence for civil service examinations dating from 408 CE has been unearthed from a tomb in Gansu, an area then ruled by the short-lived and peripheral state of Western Liang. This tantalizing tidbit speaks to the robustness of bureaucratic traditions in unexpected places. More generally, northern conquest regimes recruited Chinese literati to perform administrative services. The aforementioned Murong state was dominated by the Xianbei, a frontier group that had long enjoyed deep contacts with China. This provided a basis for sustained cooperation, and prompted rulers to rely more on taxation than on outright pillage, as earlier Xiongnu leaders had done.46
In sum, despite repeated regime change, the central state managed to hold its own in its struggle against local interests. Its administrative capacity was sufficient to maintain adequate registration and taxation systems, which in turn enabled it to field enough forces to curb local autonomy. Instead of accommodating a hybrid elite of tax-exempt landowner-soldiers, the state retained control over revenues and military compensation. As a result, the tributary state as a means of managing people and resources did not dwindle nearly as much as it did in post-Roman Europe.
The root causes of this divergence are a matter of conjecture. While it is true that the Han empire had fostered more ambitious traditions of bureaucracy and granted local communities less autonomy than the Roman empire ever had, the enormous deficiencies of Chinese censuses in the third century CE cast doubt on the notion that the continuous persistence of imperial institutions had ensured higher levels of state capacity under the northern conquest regimes than under the Germanic successor regimes, which had inherited an arguably somewhat less dysfunctional fiscal infrastructure.
Instead, the difference between individual and group land allotments to Germanic warriors that invited privatization on the one hand and greater emphasis on taxation and centralized provision among the conquest regimes of northern China on the other may well have been of critical importance. As a corollary of that emphasis, the latter were more invested in directly managing core areas and their residents, which provided them with the means required to prevent runaway decentralization.
Modes of subsistence and combat help account for these divergent priorities. The dominance of cavalry forces in the Chinese case was more conducive to the concentration of military power—as they could easily be gathered to project power across the Central Plain region—and to social separation between (foreign) horsemen and herders and (local) farmers. Both factors encouraged centralized provisioning by a sedentary subject population.
In the Germanic successor states, military forces came to be dispersed to make use of desirable farmland. Even as armored cavalry gained in importance from the eighth and ninth centuries onward, this shift unfolded in the context of intensifying localism. In northern China, by contrast, cavalry forces tended to be not only more concentrated but also larger. Ecology played a role: access to the steppe ensured a greater supply of horses, effectively “democratizing” their use.
Finally, this proximity added another factor to the mix that was mostly missing from post-Roman Europe. The steppe did not merely produce horses but also new challengers to those conquest groups that had already established themselves in northern China: the Tuoba and Rouran, discussed below, are the most prominent examples. This ongoing pressure placed a premium on the scaling-up of military capabilities. As we saw in chapter 5, the political fracturing of the Islamic world and the logistical limitations of Avars and Magyars saved the remaining Germanic states of the Early Middle Ages from comparably intense competition. In this sheltered environment, unlike in northern China, state deformation driven by localized self-interest was a viable option.
EMPIRE REDUX
During the fifth and sixth centuries, centralized state power in northern China continued to recover. The collapse of the Murong state in the 380s enabled the Xianbei Tuoba clan to regain independence and to take over much of the former’s territory. By the late 430s, the Northern Wei dynasty of the Tuoba had expanded their rule across the whole of northern China. For most of the fifth century, they operated the traditional dual status system: military garrisons that were
mostly made up of Tuoba—classified as guoren—maintained control over the local population. Over time, these units became more mixed, admitting Xiongnu as well as some Han Chinese. However, even though Chinese could be conscripted and served in support roles, they did not routinely participate in military activities: yet, unlike military commands, administrative positions were open to them. That many Tuoba continued to practice herding highlighted their detachment from the Chinese agricultural population.47
These farmers were subject to direct taxation that sustained the conquest regime. A variety of land taxes are attested, and from the 420s onward civilian officials were in charge of fiscal affairs. This strengthened the hand of the central state at the expense of local autonomy. Large numbers of clan leaders were pressured into surrendering their often sizable communities, which could comprise thousands of households. In return, the Tuoba court bestowed official titles on these fortress chiefs to incorporate them into formal state hierarchies. This process of co-optation arrested and reversed earlier trends toward localization that had counterbalanced central power.48
Tuoba military capabilities critically relied on access to horses from the northern grasslands, where a new steppe confederation, the Rouran, had arisen concurrently and in interaction with the Tuoba state. Ongoing conflict between these two powers provided strong impetus for centralization. Adopting a strategy common among earlier Chinese empires, the Tuoba set up garrisons along the steppe frontier.49
A series of integrative reforms accelerated Tuoba state-building. In the mid-480s, their Northern Wei state launched an “equal-fields” program in which plots of land were assigned to households in exchange for tax and service obligations. Different versions of this arrangement survived for the next three centuries. The land remained state property, keeping elites from taking over and subordinating the peasantry. This measure was accompanied by the appointment of prominent villagers to verify census registers and supervise tax collection. Improved registration density shored up the center’s position by ensuring access to manpower and material resources, which allowed a uniform salary schedule for state officials to be introduced.50
Escape From Rome Page 25
