Models of large-scale state formation fueled by antagonistic interaction between steppe and settled areas work best for empires that emerged in the steppe itself, in the form of entities that were secondary to and in some ways parasitical upon agrarian societies on the other side of the ecological divide. The underlying dynamic has most famously been captured by Thomas Barfield’s concept of steppe-based “shadow empires.” Steppe populations formed confederations out of a need to organize manpower in order to obtain material benefits from adjacent sedentary societies by a variety of means from trade to plunder from raiding and the extraction of tributary payments. Access to such goods buttressed and legitimated centralized authority among nomadic and seminomadic groups by enabling leaders to reward and thus control followers whose mobile lifestyle granted them considerable exit options.
FIGURE 8.7 Potential vegetation cover of Asia. Source: https://nelson.wisc. edu/sage/data-and-models/atlas/maps/potentialveg/atl_potentialveg_asi.jpg, displaying “vegetation that would exist at a given location had human forms of land use never existed” (https://nelson.wisc. .edu/sage/data-and-models/atlas/maps.php?datasetid=25&includerelatedlinks=1&dataset=25). Data used by permission of The Center for Sustainability and the Global Environment, Nelson Institute for Environmental Studies, University of Wisconsin–Madison.
In Barfield’s view, steppe empires were locked into a symbiotic relationship with agrarian polities: “They could not exist except as part of an interaction with an imperial state because they lacked most of the essential characteristics of primary empires,” most notably the ability to tax large populations directly. This warrants their designation as “shadow” or “mirror” empires. In creating spatially extensive empires, steppe groups did not merely borrow from the institutional toolkit of their sedentary neighbors but responded directly to the latter’s development. Thus, the larger and more complex primary empires grew to be, the greater the number of steppe empires that followed suit.39
Simplifying as it is, this scenario is not inherently biased against nomads. Contrary to what is sometimes averred, this perspective does not depend on steppe peoples taking the initiative in aggression: scaling-up among herders could just as readily be triggered by expanding sedentary polities’ encroachment on their pastures. Nor does it denigrate nomads as poor or “barbaric.” Both sides generally sought out resources from the other: horses in particular were prized steppe exports. Yet even if we were to follow Christopher Beckwith in picturing steppe groups as primarily interested in peaceful trade and in casting agrarian empires in the role of aggressors that provoked hostilities by unilaterally restricting the flow of goods, the structural tensions identified by Barfield would remain unchanged: conflict and empire-building arose from a fundamentally unbalanced distribution of resources and capabilities.40
This imbalance was further exacerbated by climatic variation. From the third century BCE, the intensity of steppe pressure on China—proxied by the frequency of battles and southward incursions—was inversely correlated with precipitation levels: drought undermined the fragile ecology of the grasslands and provided an impetus to move on. The migration of the Seljuq Turks may have been similarly motivated. At other times, climatic conditions generated additional resources, as at the beginning of the Arab expansion and again for Genghis Khan’s Mongols. All of this introduced a “natural” element into the relations between the settled and nomadic spheres that superseded specific cultural dispositions.41
No matter how much the literate elites of sedentary societies vilified steppe inhabitants as bloodthirsty savages—or how deeply modern observers have variously been taken in or appalled by such tropes—our task is not to assign or deflect blame but to acknowledge key elements on which all sides in the academic debate consistently agree: that even though inhabitants of the steppe were able to control or produce many vital resources directly, they generally strove to acquire outside goods, whether by trade, war, or military service for others; that agrarian polities either pushed them away or defended themselves, depending on who happened to have the upper hand; and that thanks to their lifestyle, nomadic and seminomadic groups tended to excel at horse-riding and mounted warfare in ways that helped offset their numerical inferiority. These few ingredients were sufficient for sustaining endemic conflict.42
Viable solutions were in short supply. Conquest of sedentary regions by steppe invaders was rare and invited either expulsion by the local population or renewed invasion by the next wave of steppe warriors. Pacification and settlement by overwhelming force or the transformation of the steppe into farmland—as in modern Ukraine—became technologically and logistically feasible only a few centuries ago. For a much longer period before that, competitive friction between agrarian and steppe was structural in nature and therefore highly persistent.43
A straightforward model of antagonistic scaling-up works less smoothly for agrarian empires. Barfield’s model treats them as antecedent, which means that steppe power cannot be used to explain their existence. Moreover, traditional empires were not generally established in order to combat nomadic challengers. The earliest large states in particular appeared well before steppe inhabitants mastered cavalry warfare and turned into a force to be reckoned with. All this creates serious problems for overly abstract notions of agrarian responses to nomadic raiding, such as Turchin’s competing options of alliance-building (vulnerable to free-riding) and conquest (to centralize command) within settled areas. It thus will not do to quip, with Kent Gang Deng, that “the irony of history decreed that the Chinese had to kill each other in civil wars to unify the land and cooperate with each other in order to fend off the nomads.” Even if this idea contains a kernel of truth, it cuts too many corners to be of much use to historians.44
The presence of the steppe acted upon the formation of agrarian empires in a more indirect and mediated manner. Geographically and ecologically intermediate contact zones between the agrarian and nomadic spheres—Owen Lattimore’s “marginal zone”—played a critical role. It was there that the most intense and consequential interactions—trading, raiding, co-optation, and acculturation—took place. In these areas, such as Manchuria, settled groups were exposed to and had ready access to steppe assets—to a “reservoir,” in Lattimore’s parlance, of horsemen and horses—that gave them disproportionate influence on state formation in the fully settled zone. The “marginal zone,” rather than the steppe grasslands proper, was the principal source of conquest. This was as true of the Xianbei tribes of the Period of Disunion and the later Manchu takeover as it was of the late antique Arab formations that mediated between Bedouin and the Fertile Crescent.45
But outright conquest was merely the most dramatic manifestation of that influence. More generally, resources from the steppe shaped processes of state formation in the agrarian sphere, most notably access to horses (much sought after in South Asia, for example), cavalry, and military services provided by steppe groups. The history of Jin and Tang China and of the many Middle Eastern regimes that employed and then came to be ruled by Turkic forces would have unfolded very differently without them. The historical record thus favors a somewhat more complex model in which nomad/settled antagonism is supplemented by recourse to steppe-sourced assets to create large agrarian polities. In this scenario, cavalries’ ability to project military power was a powerful driver of empire-building in the cultivated sphere.
Once more, Turchin and associates have taken the lead in generalizing this relationship with the help of a deliberately simple model of state formation in the Old World between 1500 BCE and 1500 CE. In their simulation, which takes account of only a few basic features such as land cultivability, elevation, rivers, and coasts, military technology is seeded at the steppe frontier in order to simulate the impact of chariots and then cavalry. Their presence in turn helps shape the evolution of ultrasocial traits that sustain political scaling-up.
A large number of simulation runs based on these parsimonious premises predicts patterns of s
tate formation that fairly closely approximate historical outcomes. In projecting the probability that any particular cell on the model’s grid was part of a state of at least 100,000 square kilometers at certain intervals during this three-millennium period, the simulation captures about two-thirds of the variance observed in actual history. This finding suggests that equine warfare was critical to processes of large-scale state-building. Given that the spatial distribution of states of 100,000 square kilometers or more closely resembles that of empires of a million square kilometers or more (as shown in figure 8.4), the same conclusion applies to the formation of many of the largest traditional empires in history.46
It thus appears that the steppe precipitated imperial state formation both directly—by encouraging cooperation among agriculturalists and among pastoralists and conflict between these two—and indirectly, in the first instance by supplying horses (and horsemen) for military purposes. One big caveat applies. Broadly generalizing statements derived from parsimonious premises are bound to be greeted with skepticism by most historians, who tend to focus on the trees at the expense of the forest.47
This is not unreasonable: on their own, statistical correlations and spatial patterns do not necessarily make a compelling case. We also need to consider in greater depth the proximate mechanisms that translated what we might call the “steppe effect” to specific outcomes in state formation. I do so by delving a little more deeply into the dynamics of regional political and military history as they unfolded over the long term. As we will see, this historical evidence from China, Europe, and other parts of the Old World provides ample support for a steppe-focused account of premodern imperiogenesis.
East Asia
In China, empire emanated almost exclusively from the northern frontier. Over the course of 3,600 years, all but one of a dozen unification events originated in the north. Seven of them were rooted in the northwest, especially the Wei River valley: Western Zhou (twelfth century BCE), Qin and Han (third century BCE), Sui (sixth century CE), Tang (seventh century CE), Yuan (thirteenth century CE), and the communist takeover out of Shaanxi (twentieth century CE). The Manchu Qing came from the northeast (seventeenth century CE), and the Shang (sixteenth [?] century BCE), Western Jin (third century CE), and Northern Song (tenth century CE) from the north-central area. Two further unifications merely of northern China—Northern Wei (fourth century CE) and Jin (twelfth century CE)—originated from the northwest and northeast, respectively. The Ming regime (fourteenth century CE), centered on the Yangzi basin, was the sole outlier.48
The origin of the leadership reflects the same long-term pattern. Victor Mair’s survey of the ethnic and regional affiliation of ruling houses from the Shang dynasty to the People’s Republic finds that most unifying dynasties were founded by individuals from the northwest: Shang, Western Zhou, Qin, Eastern Han, Sui, Tang, Northern Song, Yuan, and Qing, nine overall, compared to only three—Western Han, Western Jin, and Ming—that were not, alongside the Republic of 1911 and the current regime. Much the same is true of dynasties that controlled less than the whole of China but more than half of it. Moreover, dynasties from the northwest were on average more durable. Thus, in Mair’s words, northwestern actors “were nearly always responsible for the recurrent creation of an Extended East Asian Heartland”—“for the building of vast empires” in that part of the world.49
This focus was not simply a function of ethnicity: both early Chinese dynasties (Western Zhou, Qin, Han, and Western Jin) and later regimes that were either of foreign origin (Yuan and Qing, alongside Wei and Jin) or of a more hybrid nature (Sui, Tang, and to some extent even the Northern Song) shared this characteristic. Geography was key.
Northern predominance went back a long time, well before the first steppe conquest regimes appeared in the fourth century CE. The emergence of the Shang coincided with the appearance of domesticated horses and war chariots. In a striking reflection of steppe influence, Shang and Western Zhou royal burials were replete with horses and chariots. More tantalizingly, accounts of the First Emperor of Qin and the founders of the Western and Eastern Han dynasties emphasize somatic features—long noses and facial hair—that were coded as markers of ethnic identity pointing to admixture from the steppe and only much later came to be associated with European foreigners.50
In the grasslands northwest of China, the transition to pastoral nomadism occurred during the Western Zhou period, and classic nomadic steppe culture appeared in the seventh and sixth centuries BCE, at a time when the states began to coalesce in the agrarian zone. It is true that this process was primarily driven by interstate warfare. But this does not mean that proximity to the steppe had no significant impact. The ultimate victor, the kingdom of Qin, was located in China’s northwest and appears to have owed much of its strikingly martial disposition to its frontier position. It was there that state-strengthening reforms—which had been pioneered elsewhere—were taken the farthest and were implemented with the greatest success, even though Qin, centered on the Wei River valley, was less exposed to rival states than others.51
Yet Qin’s belligerence exceeded that of its rivals: by one count, it launched several times as many offensives during the Warring States period as its principal competitors. The runner-up, Zhao, was even more exposed to the steppe frontier. Some version of the “steppe effect” may well have made itself felt. The Qin were later said to have started out as herders. They were locked into prolonged conflict with surrounding chiefdoms (“Rong,” some of which may have been pastoralists) over control of the Wei River basin, alongside wars with the state of Jin and its successor Wei to the northeast. These efforts helped produce a precociously strong state with a relatively weak aristocracy, and muted cultural attainment.52
The expansion of the Sinosphere into Rong and Di areas in this period brought the Warring States into closer contact with pastoralists. Subjugation of Di tribes in the Ordos region from the eighth through the sixth centuries BCE removed earlier buffers and for the first time led to conflict between the northern states and actual nomads. The states of Yan, Zhao, and Qin abutted the grasslands of Inner Mongolia and southwestern Manchuria. From the late fourth century BCE, the first two of them set up long border walls to fence off land they had seized from pastoralists. Aggression was stepped up upon the Qin conquest of China: existing walls were connected and ambitious campaigns targeted herders. The displacement of the pastoralist Xiongnu from the Ordos in 215 BCE encouraged them to pool their forces under centralized leadership.53
This consolidation process enabled them to put growing pressure on the Han empire, which in the late second and early first centuries BCE responded by adopting an aggressive policy of massive incursions into the steppe. Even though these extremely costly operations could not be sustained, they provided strong impetus for domestic state-building measures from crackdowns on nobles and merchants to greater government intervention in economy, and also pushed Han power into Central Asia all the way to Fergana, a rich source of horses.54
The “steppe effect” on Chinese state formation increased tremendously during the following millennium and a half, when groups from the frontier zone set up a series of conquest regimes. From the fourth century CE, the proverbial “Five Barbarians,” mostly herders—the Xiongnu, Xianbei, and Jie from the steppe and Manchuria and the Qiang from the western highlands, alongside the sedentary Di—dominated the political landscape of northern China. More than half of the “Sixteen Kingdoms” from 304 to 439 CE were established by these groups: Former Zhao, Northern Liang, and Xia by the Xiongnu; Southern Liang, Former, Later, and Southern Yan, and Western Qin by the Xianbei; Later Qin by the Qiang; Later Zhao by Jie; and Later Liang by the Di.55
The Tuoba clan that restored unity in northern China as the Northern Wei was of Xianbei extraction, as were various successor regimes from Eastern and Western Wei to Northern Zhou and Northern Qi. Tuoba rule depended in no small measure on access to horses from the steppe. Throughout the fifth century CE, they launched attacks o
n the Rouran steppe federation to keep it at bay and ensure a steady flow of equine imports: thirteen major conflicts are attested between 402 and 522 CE.56
Sui and Tang rulers were of mixed Han and Turkic-Xianbei descent and relied on the state apparatus that had been built up by the previous northern regimes. In the seventh century CE, the Tang’s Turkic generals commanded largely non-Chinese cavalry in far-flung campaigns across the steppe. Just as under the Han empire, expansionary efforts were concentrated on the grasslands in the northwest of China. Between the 620s and 650s, Tang forces destroyed both Turk khaganates and claimed authority over Xinjiang and areas to the west all the way to what are now Afghanistan and Uzbekistan. By the 660s, Tang cavalry were said to have owned 700,000 horses.57
In the 750s, Uighur forces helped the Tang dynasty survive a massive rebellion. After the Tang’s eventual collapse, renewed fission opened up space for northern military regimes. In the tenth century CE, the Shatuo Turks set up the Later Tang and Later Jin dynasties, which were beholden to the Qidan state of Liao. The subsequent Later Zhou and Northern Song dynasties built on this Turkic governmental apparatus to unify first northern and then all of China. The Northern Song empire, unable to overcome the Liao Qidan in Manchuria and Inner Mongolia by force, invested massively in military capabilities, an effort that led to the creation of a powerful centralized fiscal-military state.58
In the early twelfth century, the Jurchen, Tungusic tribes from the woodlands of eastern Manchuria that had once been subordinate to the Liao, established the Jin dynasty that quickly took over northern China and put even greater pressure on what remained of the Song state until both succumbed to the Mongols. Combining his growing populous and wealthy Chinese territories with extensive steppe cavalry forces, in the late thirteenth century Kublai Khan claimed the largest empire the world had ever seen, even if much of it only nominally acknowledged his suzerainty.59
Escape From Rome Page 29
